Dark Basic Pro Vol 2

Hands On DarkBASIC Pro
Volume 2
A Self-Study Guide to Games Programming
Alistair Stewart
Digital Skills
Milton
Barr
Girvan
Ayrshire
KA26 9TY
www.digital-skills.co.uk
Copyright © Alistair Stewart 2006
All rights reserved.
No part of this work may be reproduced or used in any form

without the written permission of the author.
Although every effort has been made to ensure accuracy, the

author and publisher accept neither liability nor responsibility
for any loss or damage arising from the information in this book.
All brand names and product names are trademarks of their respective
companies and have been capitalised throughout the text.
DarkBASIC Professional is produced by The Game Creators Ltd
Printed September 2006

2nd Printing November 2006
3rd Printing July 2007
4th Printing February 2008
5th Printing September 2009 (with corrections)
6th Printing July 2010
Title : Hands On DarkBASIC Pro Volume 2
ISBN-10 : 1-874107-09-2
ISBN-13 : 978-1-874107-09-5
Other Titles Available:
Hands On DarkBASIC Pro Volume 1

Hands On Milkshape
Hands On Pascal
Hands On C++
Hands On Java
Hands On XHTML
TABLE OF CONTENTS
Chapter 30 3D Concepts and Terminology
The 3D World . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
The Coordinate System . . . . . . . . . . . . . . . . . . . . . . . 744
Axes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746
World Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Local Axes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
3D Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
Object Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 749
Textures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 750
Images with an Alpha Channel . . . . . . . . . . . . . . . . . . 750
Cameras . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
Lights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
Chapter 31 3D Primitives
3D Primitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Creating a Cube . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
The MAKE OBJECT CUBE Statement . . . . . . . . . . . . . 756
Creating Other Primitives . . . . . . . . . . . . . . . . . . . . . . . 757
The MAKE OBJECT BOX Statement . . . . . . . . . . . . . . 757
The MAKE OBJECT SPHERE Statement . . . . . . . . . . . . 758
The MAKE OBJECT CYLINDER Statement . . . . . . . . . . . 759
The MAKE OBJECT CONE Statement . . . . . . . . . . . . . 760
The MAKE OBJECT PLAIN Statement . . . . . . . . . . . . . 760
The MAKE OBJECT TRIANGLE Statement . . . . . . . . . . . 761
Positioning an Object . . . . . . . . . . . . . . . . . . . . . . . . . 762
The POSITION OBJECT Statement . . . . . . . . . . . . . . . 762
The MOVE OBJECT Statement . . . . . . . . . . . . . . . . . 764
Rotating Objects - Absolute Rotation . . . . . . . . . . . . . . . . . 765
The XROTATE OBJECT Statement . . . . . . . . . . . . . . . 766
The YROTATE OBJECT Statement . . . . . . . . . . . . . . . 767
The ZROTATE OBJECT Statement . . . . . . . . . . . . . . . 767
The ROTATE OBJECT Statement . . . . . . . . . . . . . . . . 768
The SET OBJECT ROTATION Statement . . . . . . . . . . . 769
Rotating Objects - Relative Rotation . . . . . . . . . . . . . . . . . 769
The PITCH OBJECT Statement . . . . . . . . . . . . . . . . . 770
The TURN OBJECT Statement . . . . . . . . . . . . . . . . . 770
The ROLL OBJECT Statement . . . . . . . . . . . . . . . . . 771
The POINT OBJECT Statement . . . . . . . . . . . . . . . . . 771
The MOVE OBJECT distance Statement . . . . . . . . . . . . 772
The FIX OBJECT PIVOT Statement . . . . . . . . . . . . . . . 773
Resizing Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
The SCALE OBJECT Statement . . . . . . . . . . . . . . . . . 775
Showing and Hiding Objects . . . . . . . . . . . . . . . . . . . . . 776
The HIDE OBJECT Statement . . . . . . . . . . . . . . . . . . 776
The SHOW OBJECT Statement . . . . . . . . . . . . . . . . . 776
The DELETE OBJECT Statement . . . . . . . . . . . . . . . . 777
The DELETE OBJECTS Statement . . . . . . . . . . . . . . . 777
Copying a 3D Object . . . . . . . . . . . . . . . . . . . . . . . . . 778
The CLONE OBJECT Statement . . . . . . . . . . . . . . . . 778
The INSTANCE OBJECT Statement . . . . . . . . . . . . . . 779
Retrieving Data on 3D Objects . . . . . . . . . . . . . . . . . . . . 779
The OBJECT EXIST Statement . . . . . . . . . . . . . . . . . 779
The OBJECT POSITION Statement . . . . . . . . . . . . . . 780
The OBJECT VISIBLE Statement . . . . . . . . . . . . . . . . 780
The OBJECT SIZE Statement . . . . . . . . . . . . . . . . . . 781
The OBJECT ANGLE Statement . . . . . . . . . . . . . . . . 781
Controlling an Object's Rotation Using the Mouse . . . . . . . . . . 782
Wireframe and Culling . . . . . . . . . . . . . . . . . . . . . . . . 783
The SET OBJECT WIREFRAME Statement . . . . . . . . . . 783
The SET OBJECT CULL Statement . . . . . . . . . . . . . . . 784
Storage Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
The SET GLOBAL OBJECT CREATION Statement . . . . . . 785
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786
Merging Primitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
The Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
The PERFORM CSG UNION Statement . . . . . . . . . . . . 788
The PERFORM CSG DIFFERENCE Statement . . . . . . . . . 790
The PERFORM CSG INTERSECTION Statement . . . . . . . 790
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792
Chapter 32 Texturing
Adding Texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
Loading a Texture Image . . . . . . . . . . . . . . . . . . . . . . . 798
Using the Image as a Texture . . . . . . . . . . . . . . . . . . . . 798
The TEXTURE OBJECT Statement . . . . . . . . . . . . . . . 798
Mipmaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799
The LOAD IMAGE Statement Again . . . . . . . . . . . . . . . 800
Tiling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801
The SCALE OBJECT TEXTURE Statement . . . . . . . . . . . 802
Seamless Tiling . . . . . . . . . . . . . . . . . . . . . . . . . . . 804
Video Texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805
The PLAY ANIMATION TO IMAGE Statement . . . . . . . . . 805
Other Texture Effects . . . . . . . . . . . . . . . . . . . . . . . . . 807
The SET OBJECT TEXTURE Statement . . . . . . . . . . . . 807
The SCROLL OBJECT TEXTURE Statement . . . . . . . . . . 808
The SET OBJECT TRANSPARENCY Statement . . . . . . . . 810
The SET DETAIL MAPPING ON Statement . . . . . . . . . . . 811
The SET OBJECT FILTER Statement . . . . . . . . . . . . . . 813
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
Other Visual Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815
Changing Colour and Transparency . . . . . . . . . . . . . . . . . 815
The COLOR OBJECT Statement . . . . . . . . . . . . . . . . 815
The GHOST OBJECT ON Statement . . . . . . . . . . . . . . 816
The GHOST OBJECT OFF Statement . . . . . . . . . . . . . 817
The FADE OBJECT Statement . . . . . . . . . . . . . . . . . 817
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 819
Images with an Alpha Channel . . . . . . . . . . . . . . . . . . . . . . 820
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820
Using Images with an Alpha Channel . . . . . . . . . . . . . . . . 820
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821
Creating a Complex 3D Shape . . . . . . . . . . . . . . . . . . . . . . 822
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 822
Designing the Castle . . . . . . . . . . . . . . . . . . . . . . . . . 822
Gathering the Components . . . . . . . . . . . . . . . . . . . 823
Creating the Code . . . . . . . . . . . . . . . . . . . . . . . . . . 823
The Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824
Sky Spheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
Chapter 33 Cameras
Camera Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Positioning the Camera . . . . . . . . . . . . . . . . . . . . . . . 836
The POSITION CAMERA Statement . . . . . . . . . . . . . . 836
The MOVE CAMERA Statement . . . . . . . . . . . . . . . . . 837
Changing the Viewpoint . . . . . . . . . . . . . . . . . . . . . . . 838
The POINT CAMERA Statement . . . . . . . . . . . . . . . . 838
The ROTATE CAMERA Statement . . . . . . . . . . . . . . . 838
The SET CAMERA ROTATION Statement . . . . . . . . . . . 840
The XROTATE CAMERA Statement . . . . . . . . . . . . . . 840
The YROTATE CAMERA Statement . . . . . . . . . . . . . . 841
The ZROTATE CAMERA Statement . . . . . . . . . . . . . . 841
The PITCH CAMERA Statement . . . . . . . . . . . . . . . . 842
The TURN CAMERA Statement . . . . . . . . . . . . . . . . . 842
The ROLL CAMERA Statement . . . . . . . . . . . . . . . . . 843
Retrieving Camera Data . . . . . . . . . . . . . . . . . . . . . . . 844
The CAMERA POSITION Statement . . . . . . . . . . . . . . 844
The CAMERA ANGLE Statement . . . . . . . . . . . . . . . . 844
Modifying Camera Characteristics . . . . . . . . . . . . . . . . . . 845
The SET CAMERA VIEW Statement . . . . . . . . . . . . . . 845
The SET CAMERA ASPECT Statement . . . . . . . . . . . . . 846
The SET CAMERA FOV Statement . . . . . . . . . . . . . . . 847
The SET CAMERA RANGE Statement . . . . . . . . . . . . . 848
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849
Controlling Camera Movement . . . . . . . . . . . . . . . . . . . . . . 851
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
Automatic Camera Placement . . . . . . . . . . . . . . . . . . . . 851
The AUTOCAM Statement . . . . . . . . . . . . . . . . . . . . 851
Following the Action . . . . . . . . . . . . . . . . . . . . . . . . . 852
The SET CAMERA TO FOLLOW Statement . . . . . . . . . . 853
Giving the Player Control of the Camera . . . . . . . . . . . . . . . 856
The CONTROL CAMERA USING ARROWKEYS Statement . . 856
The AUTOMATIC CAMERA COLLISION Statement . . . . . . 858
Controlling the Camera with the Mouse . . . . . . . . . . . . . 859
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
Multiple Cameras . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
Using Additional Cameras . . . . . . . . . . . . . . . . . . . . . . 863
The MAKE CAMERA Statement . . . . . . . . . . . . . . . . . 863
The COLOR BACKDROP Statement . . . . . . . . . . . . . . 864
The BACKDROP Statement . . . . . . . . . . . . . . . . . . . 864
The SET CURRENT CAMERA Statement . . . . . . . . . . . . 865
The DELETE CAMERA Statement . . . . . . . . . . . . . . . 866
Switching Between Cameras . . . . . . . . . . . . . . . . . . 866
Multiple Camera Output . . . . . . . . . . . . . . . . . . . . . 868
The CLEAR CAMERA VIEW Statement . . . . . . . . . . . . . 869
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870
Advanced Camera Techniques . . . . . . . . . . . . . . . . . . . . . . 871
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
The Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
The SET CAMERA TO IMAGE Statement . . . . . . . . . . . . 871
The SET CAMERA TO OBJECT ORIENTATION Statement . . 873
The SET OBJECT TO CAMERA ORIENTATION Statement . . 873
The LOCK OBJECT Statement . . . . . . . . . . . . . . . . . 874
The SET VECTOR3 TO CAMERA POSITION Statement . . . . 875
The SET VECTOR3 TO CAMERA ROTATION Statement . . . 876
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 878
Chapter 34 Lighting
Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Types of Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Ambient Lighting . . . . . . . . . . . . . . . . . . . . . . . . . 886
Point Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Spot Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Directional Lighting . . . . . . . . . . . . . . . . . . . . . . . . 886
Lighting in DarkBASIC Pro . . . . . . . . . . . . . . . . . . . . . . 887
The HIDE LIGHT Statement . . . . . . . . . . . . . . . . . . . 887
The SHOW LIGHT Statement . . . . . . . . . . . . . . . . . . 888
The SET AMBIENT LIGHT Statement . . . . . . . . . . . . . . 888
The COLOR AMBIENT LIGHT Statement . . . . . . . . . . . . 889
The MAKE LIGHT Statement . . . . . . . . . . . . . . . . . . 889
The DELETE LIGHT Statement . . . . . . . . . . . . . . . . . 890
The COLOR LIGHT Statement . . . . . . . . . . . . . . . . . 890
The POSITION LIGHT Statement . . . . . . . . . . . . . . . . 891
The SET LIGHT RANGE Statement . . . . . . . . . . . . . . . 891
The SET SPOT LIGHT Statement . . . . . . . . . . . . . . . . 892
The SET DIRECTIONAL LIGHT Statement . . . . . . . . . . . 892
The SET POINT LIGHT Statement . . . . . . . . . . . . . . . 893
The POINT LIGHT Statement . . . . . . . . . . . . . . . . . . 893
The ROTATE LIGHT Statement . . . . . . . . . . . . . . . . . 895
The SET LIGHT TO OBJECT POSITION Statement . . . . . . 895
The SET LIGHT TO OBJECT ORIENTATION Statement . . . . 897
Retrieving Light Data . . . . . . . . . . . . . . . . . . . . . . . . . 898
The LIGHT EXIST Statement . . . . . . . . . . . . . . . . . . 898
The LIGHT VISIBLE Statement . . . . . . . . . . . . . . . . . 899
The LIGHT RANGE Statement . . . . . . . . . . . . . . . . . 899
The LIGHT TYPE Statement . . . . . . . . . . . . . . . . . . . 899
The LIGHT POSITION Statement . . . . . . . . . . . . . . . . 900
The LIGHT DIRECTION Statement . . . . . . . . . . . . . . . 900
Fog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
The FOG Statement . . . . . . . . . . . . . . . . . . . . . . . 901
The FOG COLOR Statement . . . . . . . . . . . . . . . . . . 902
The FOG DISTANCE Statement . . . . . . . . . . . . . . . . . 902
The SET OBJECT FOG Statement . . . . . . . . . . . . . . . 903
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907
Chapter 35 Meshes and Limbs
Meshes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 912
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 912
Handling Meshes . . . . . . . . . . . . . . . . . . . . . . . . . . . 912
The MAKE MESH FROM OBJECT Statement . . . . . . . . . 912
The SAVE MESH Statement . . . . . . . . . . . . . . . . . . . 913
The LOAD MESH Statement . . . . . . . . . . . . . . . . . . 914
The MAKE OBJECT Statement . . . . . . . . . . . . . . . . . 914
The DELETE MESH Statement . . . . . . . . . . . . . . . . . 915
The MESH EXIST Statement . . . . . . . . . . . . . . . . . . 915
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916
Limbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917
Getting Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917
The ADD LIMB Statement . . . . . . . . . . . . . . . . . . . . 917
The MAKE OBJECT FROM LIMB Statement . . . . . . . . . . 919
The OFFSET LIMB Statement . . . . . . . . . . . . . . . . . . 920
The ROTATE LIMB Statement . . . . . . . . . . . . . . . . . . 920
The SCALE LIMB Statement . . . . . . . . . . . . . . . . . . 922
The COLOR LIMB Statement . . . . . . . . . . . . . . . . . . 922
The TEXTURE LIMB Statement . . . . . . . . . . . . . . . . . 923
The SCALE LIMB TEXTURE Statement . . . . . . . . . . . . . 925
The SCROLL LIMB TEXTURE Statement . . . . . . . . . . . . 927
The HIDE LIMB Statement . . . . . . . . . . . . . . . . . . . . 927
The SHOW LIMB Statement . . . . . . . . . . . . . . . . . . . 928
The REMOVE LIMB Statement . . . . . . . . . . . . . . . . . 928
The LINK LIMB Statement . . . . . . . . . . . . . . . . . . . . 928
The CHANGE MESH Statement . . . . . . . . . . . . . . . . . 931
The GLUE OBJECT TO LIMB Statement . . . . . . . . . . . . 931
The UNGLUE OBJECT Statement . . . . . . . . . . . . . . . 934
The SET LIMB SMOOTHING Statement . . . . . . . . . . . . 934
Creating Doors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935
Retrieving Limb Data . . . . . . . . . . . . . . . . . . . . . . . . . 936
The LIMB EXIST Statement . . . . . . . . . . . . . . . . . . . 936
The LIMB VISIBLE Statement . . . . . . . . . . . . . . . . . . 937
The LIMB OFFSET Statement . . . . . . . . . . . . . . . . . . 937
The LIMB SCALE Statement . . . . . . . . . . . . . . . . . . 938
The LIMB ANGLE Statement . . . . . . . . . . . . . . . . . . 939
The LIMB POSITION Statement . . . . . . . . . . . . . . . . . 939
The LIMB DIRECTION Statement . . . . . . . . . . . . . . . . 940
The PERFORM CHECKLIST FOR OBJECT LIMBS Statement . 944
The LIMB NAME$ Statement . . . . . . . . . . . . . . . . . . 945
The LIMB TEXTURE Statement . . . . . . . . . . . . . . . . . 946
The LIMB TEXTURE NAME Statement . . . . . . . . . . . . . 946
The CHECK LIMB LINK Statement . . . . . . . . . . . . . . . 947
Saving a Model in DBO Format . . . . . . . . . . . . . . . . . . . 947
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947
The DBO File Format . . . . . . . . . . . . . . . . . . . . . . 948
Creating an Elevator Model . . . . . . . . . . . . . . . . . . . 948
The SAVE OBJECT Statement . . . . . . . . . . . . . . . . . 949
The LOAD OBJECT Statement . . . . . . . . . . . . . . . . . 950
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 953
Chapter 36 Importing 3D Objects

Importing 3D Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . 962
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 962
File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963
Statements for Loading and Using 3D Objects . . . . . . . . . . . . 963
The LOAD OBJECT Statement Again . . . . . . . . . . . . . . 963
The PLAY OBJECT Statement . . . . . . . . . . . . . . . . . 965
The LOOP OBJECT Statement . . . . . . . . . . . . . . . . . 966
The TOTAL OBJECT FRAMES Statement . . . . . . . . . . . 966
Moving the Alien . . . . . . . . . . . . . . . . . . . . . . . . . 967
The SET OBJECT SPEED Statement . . . . . . . . . . . . . . 967
The STOP OBJECT Statement . . . . . . . . . . . . . . . . . 968
The SET OBJECT FRAME Statement . . . . . . . . . . . . . . 968
The SET OBJECT INTERPOLATION Statement . . . . . . . . 969
The APPEND OBJECT Statement . . . . . . . . . . . . . . . . 970
Retrieving Animation Object Information . . . . . . . . . . . . . . . 971
The OBJECT PLAYING Statement . . . . . . . . . . . . . . . 971
The OBJECT LOOPING Statement . . . . . . . . . . . . . . . 971
The OBJECT FRAME Statement . . . . . . . . . . . . . . . . 972
The OBJECT SPEED Statement . . . . . . . . . . . . . . . . 972
The OBJECT INTERPOLATION Statement . . . . . . . . . . . 972
The OBJECT SIZE Statement . . . . . . . . . . . . . . . . . . 973
Limbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977
Chapter 37 Screen Control

User Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 980
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 980
Selecting an Object . . . . . . . . . . . . . . . . . . . . . . . . . . 980
The OBJECT SCREEN Statement . . . . . . . . . . . . . . . 982
The PICK OBJECT Statement . . . . . . . . . . . . . . . . . . 983
The GET PICK DISTANCE Statement . . . . . . . . . . . . . . 984
The PICK VECTOR Statement . . . . . . . . . . . . . . . . . 985
The PICK SCREEN Statement . . . . . . . . . . . . . . . . . 986
The OBJECT IN SCREEN Statement . . . . . . . . . . . . . . 987
Selecting Objects using the Mouse . . . . . . . . . . . . . . . . . 988
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 990
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991
Chapter 38 Solitaire
Solitaire - The Board Game . . . . . . . . . . . . . . . . . . . . . . . . 994
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
The Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . 994
The Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
The Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
Creating a Computer Version of the Game . . . . . . . . . . . . . 994
User Controls . . . . . . . . . . . . . . . . . . . . . . . . . . 994
Game Responses . . . . . . . . . . . . . . . . . . . . . . . . 995
Screen Layout . . . . . . . . . . . . . . . . . . . . . . . . . . 995
Media Used . . . . . . . . . . . . . . . . . . . . . . . . . . . 995
Data Structures . . . . . . . . . . . . . . . . . . . . . . . . . 996
Adding SetUpScreen() . . . . . . . . . . . . . . . . . . . . . . 999
Adding SetUpGame() . . . . . . . . . . . . . . . . . . . . . . 1000
Adding CreateBoard() . . . . . . . . . . . . . . . . . . . . . . 1001
Adding CreateInternalBoard() . . . . . . . . . . . . . . . . . . 1001
Adding CreateMarbles() . . . . . . . . . . . . . . . . . . . . . 1002
Adding CreateSelector() . . . . . . . . . . . . . . . . . . . . . 1002
Adding SetUpHelp() . . . . . . . . . . . . . . . . . . . . . . . 1003
Adding GetPlayerMove() . . . . . . . . . . . . . . . . . . . . . 1004
Adding MoveSelector() . . . . . . . . . . . . . . . . . . . . . . 1006
Adding SelectMarble() . . . . . . . . . . . . . . . . . . . . . . 1007
Adding SelectPit() . . . . . . . . . . . . . . . . . . . . . . . . 1008
Adding IsValidMove() . . . . . . . . . . . . . . . . . . . . . . 1008
Adding MoveMarble() . . . . . . . . . . . . . . . . . . . . . . 1008
Adding SelectHelpPage() . . . . . . . . . . . . . . . . . . . . 1009
Using the Mouse . . . . . . . . . . . . . . . . . . . . . . . . . . . 1009
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1009
Updating the Program . . . . . . . . . . . . . . . . . . . . . . 1010
Suggested Enhancements . . . . . . . . . . . . . . . . . . . . . . 1013
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1015
Chapter 39 Advanced Lighting and Texturing

Advanced Lighting and Texturing . . . . . . . . . . . . . . . . . . . . . 1028
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Surface Reflection . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
The SET OBJECT AMBIENT Statement . . . . . . . . . . . . 1029
The SET OBJECT DIFFUSE Statement . . . . . . . . . . . . . 1030
The SET OBJECT SPECULAR Statement . . . . . . . . . . . 1030
The SET OBJECT SPECULAR POWER Statement . . . . . . 1031
The SET OBJECT EMISSIVE Statement . . . . . . . . . . . . 1031
The SET OBJECT LIGHT Statement . . . . . . . . . . . . . . 1034
Mappings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035
The SET LIGHT MAPPING ON Statement . . . . . . . . . . . 1035
The SET BUMP MAPPING ON Statement . . . . . . . . . . . 1038
The SET SPHERE MAPPING ON Statement . . . . . . . . . . 1039
The SET BLEND MAPPING ON Statement . . . . . . . . . . . 1041
The SET CUBE MAPPING ON Statement . . . . . . . . . . . . 1042
The SET ALPHA MAPPING ON Statement . . . . . . . . . . . 1044
Shadows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1045
The SET SHADOW SHADING ON Statement . . . . . . . . . . 1045
The SET SHADOW SHADING OFF Statement . . . . . . . . . 1048
The SET GLOBAL SHADOWS Statement . . . . . . . . . . . . 1048
The SET GLOBAL SHADOW COLOR Statement . . . . . . . . 1050
The SET GLOBAL SHADOW SHADES Statement . . . . . . . 1050
Positioning Shadows . . . . . . . . . . . . . . . . . . . . . . . . . 1051
The SET SHADOW POSITION Statement . . . . . . . . . . . 1051
Shadows and Models . . . . . . . . . . . . . . . . . . . . . . . . 1052
Other Shading Methods . . . . . . . . . . . . . . . . . . . . . . . 1054
The SET CARTOON SHADING ON Statement . . . . . . . . . 1054
The SET RAINBOW SHADING ON Statement . . . . . . . . . 1056
The SET REFLECTION SHADING ON Statement . . . . . . . 1057
The SET SHADING OFF Statement . . . . . . . . . . . . . . . 1058
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1061
Chapter 40 Collisions
Object Collisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068
Object Collision . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068
The OBJECT HIT Statement . . . . . . . . . . . . . . . . . . . 1069
The OBJECT COLLISION Statement . . . . . . . . . . . . . . 1070
The SET OBJECT COLLISION Statement . . . . . . . . . . . 1070
The SET GLOBAL COLLISION Statement . . . . . . . . . . . 1071
How Collision Detection Works . . . . . . . . . . . . . . . . . . . . 1071
The SHOW OBJECT BOUNDS Statement . . . . . . . . . . . 1072
The HIDE OBJECT BOUNDS statement . . . . . . . . . . . . 1072
Modifying Collision Detection . . . . . . . . . . . . . . . . . . . . . 1074
The SET OBJECT COLLISION TO SPHERES Statement . . . 1074
The SET OBJECT RADIUS Statement . . . . . . . . . . . . . 1074
The OBJECT COLLISION RADIUS Statement . . . . . . . . . 1075
The OBJECT COLLISION CENTER Statement . . . . . . . . . 1075
The SET OBJECT COLLISION TO BOXES Statement . . . . . 1076
The SET OBJECT COLLISION TO POLYGONS Statement . . 1076
The MAKE OBJECT COLLISION BOX Statement . . . . . . . 1077
The GET OBJECT COLLISION Statement . . . . . . . . . . . 1080
The DELETE OBJECT COLLISION BOX Statement . . . . . . 1082
The AUTOMATIC OBJECT COLLISION Statement . . . . . . . 1082
The INTERSECT OBJECT Statement . . . . . . . . . . . . . . 1083
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085
Static Collisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1087
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1087
Creating and Using Static Collision Boxes . . . . . . . . . . . . . . 1087
The MAKE STATIC COLLISION BOX Statement . . . . . . . . 1087
The GET STATIC COLLISION HIT Statement . . . . . . . . . 1087
The GET STATIC COLLISION Statement . . . . . . . . . . . 1089
The STATIC LINE OF SIGHT Statement . . . . . . . . . . . . 1093
The STATIC LINE OF SIGHT Coordinates Statement . . . . . 1095
Static Collision Boxes and the Camera . . . . . . . . . . . . . . . 1096
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1098
Chapter 41 Particles
Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1102
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1102
Creating Particles . . . . . . . . . . . . . . . . . . . . . . . . . . 1102
The MAKE PARTICLES Statement . . . . . . . . . . . . . . . 1102
The HIDE PARTICLES Statement . . . . . . . . . . . . . . . . 1103
The SHOW PARTICLES Statement . . . . . . . . . . . . . . . 1104
The DELETE PARTICLES Statement . . . . . . . . . . . . . . 1104
The POSITION PARTICLES Statement . . . . . . . . . . . . . 1104
The POSITION PARTICLE EMISSIONS Statement . . . . . . . 1105
The ROTATE PARTICLES Statement . . . . . . . . . . . . . . 1106
The COLOR PARTICLES Statement . . . . . . . . . . . . . . 1107
The SET PARTICLE EMISSIONS Statement . . . . . . . . . . 1108
The SET PARTICLE VELOCITY Statement . . . . . . . . . . . 1109
The SET PARTICLE GRAVITY Statement . . . . . . . . . . . 1110
The SET PARTICLE CHAOS Statement . . . . . . . . . . . . 1110
The SET PARTICLE SPEED Statement . . . . . . . . . . . . . 1111
The SET PARTICLE FLOOR Statement . . . . . . . . . . . . . 1112
The SET PARTICLE LIFE Statement . . . . . . . . . . . . . . 1113
The GHOST PARTICLES ON Statement . . . . . . . . . . . . 1113
The GHOST PARTICLES OFF Statement . . . . . . . . . . . . 1114
Retrieving Data on a Particles Object . . . . . . . . . . . . . . . . 1114
The PARTICLES EXIST Statement . . . . . . . . . . . . . . . 1114
The PARTICLES POSITION Statement . . . . . . . . . . . . 1115
Particles Statements that use Vectors . . . . . . . . . . . . . . . . 1116
The SET VECTOR3 TO PARTICLES POSITION Statement . . 1116
The SET VECTOR3 TO PARTICLES ROTATION Statement . . 1116
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1116
Other Types of Particles . . . . . . . . . . . . . . . . . . . . . . . . . 1118
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1118
The Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . 1118
The MAKE SNOW PARTICLES Statement . . . . . . . . . . . 1118
The MAKE FIRE PARTICLES Statement . . . . . . . . . . . . 1119
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1120
Examples of Using Particles . . . . . . . . . . . . . . . . . . . . . . . 1121
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1121
A Roman Candle . . . . . . . . . . . . . . . . . . . . . . . . . . . 1121
A Spaceship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122
A Dungeon Torch . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1124
Chapter 42 The Elevators Game

Elevators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
The Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
The Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
The Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
Creating a Computer version of the Game . . . . . . . . . . . . . . 1128
User Controls . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
Game Responses . . . . . . . . . . . . . . . . . . . . . . . . 1128
Screen Layout . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
The Board Design . . . . . . . . . . . . . . . . . . . . . . . . 1129
The Media Used . . . . . . . . . . . . . . . . . . . . . . . . . 1129
Data Structures . . . . . . . . . . . . . . . . . . . . . . . . . 1130
Game Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131
Adding SetUpGame() . . . . . . . . . . . . . . . . . . . . . . 1132
Adding InitialiseData() . . . . . . . . . . . . . . . . . . . . . . 1134
Adding InitialiseLifts() . . . . . . . . . . . . . . . . . . . . . . 1134
Adding InitialiseBoard() . . . . . . . . . . . . . . . . . . . . . 1135
Adding InitialiseVisuals() . . . . . . . . . . . . . . . . . . . . . 1136
Loading Models and Texture Files . . . . . . . . . . . . . . . . 1136
Adding LoadBoard() . . . . . . . . . . . . . . . . . . . . . . . 1137
Adding AddElevators() . . . . . . . . . . . . . . . . . . . . . . 1137
Adding LoadPlayerCharacter() . . . . . . . . . . . . . . . . . . 1138
Adding LoadDice() . . . . . . . . . . . . . . . . . . . . . . . . 1138
Adding PositionCameras() . . . . . . . . . . . . . . . . . . . . 1138
Adding RollDice() . . . . . . . . . . . . . . . . . . . . . . . . 1142
Adding MovePlayer() . . . . . . . . . . . . . . . . . . . . . . . 1143
Adding UseElevator() . . . . . . . . . . . . . . . . . . . . . . 1146
Adding MovePlayerToElevator() . . . . . . . . . . . . . . . . . 1147
Adding TurnPlayer() . . . . . . . . . . . . . . . . . . . . . . . 1148
Adding MoveOntoPlatform() . . . . . . . . . . . . . . . . . . . 1148
Adding MoveElevator() . . . . . . . . . . . . . . . . . . . . . . 1148
Adding MoveOffPlatform() . . . . . . . . . . . . . . . . . . . . 1149
Adding ReturnElevator() . . . . . . . . . . . . . . . . . . . . . 1150
Adding RepositionCamera() . . . . . . . . . . . . . . . . . . . 1150
Fixing the Shortcomings . . . . . . . . . . . . . . . . . . . . . . . 1151
Fixing RepositionCamera() . . . . . . . . . . . . . . . . . . . . 1151
Fixing MovePlayer() . . . . . . . . . . . . . . . . . . . . . . . 1152
Fixing UseElevator() . . . . . . . . . . . . . . . . . . . . . . . 1153
Fixing MovePlayerToElevator() . . . . . . . . . . . . . . . . . 1153
Fixing MoveElevator() . . . . . . . . . . . . . . . . . . . . . . 1153
Adding EndGame() . . . . . . . . . . . . . . . . . . . . . . . 1154
Game Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1154
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1155
Chapter 43 Handling BSP Models

Binary Space Partitioning . . . . . . . . . . . . . . . . . . . . . . . . . 1164
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1164
Creating a BSP File . . . . . . . . . . . . . . . . . . . . . . . . . 1165
Using BSP Files . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165
The LOAD BSP Statement . . . . . . . . . . . . . . . . . . . . 1165
The SET BSP CAMERA COLLISION Statement . . . . . . . . 1167
The SET BSP OBJECT COLLISION Statement . . . . . . . . . 1167
The SET BSP CAMERA COLLISION RADIUS Statement . . . 1169
The SET BSP OBJECT COLLISION RADIUS Statement . . . . 1169
The SET BSP COLLISION HEIGHT ADJUSTMENT Statement 1170
The SET BSP COLLISION THRESHOLD Statement . . . . . . 1171
The PROCESS BSP COLLISION Statement . . . . . . . . . . 1171
The SET BSP COLLISION OFF Statement . . . . . . . . . . . 1171
The BSP COLLISION HIT Statement . . . . . . . . . . . . . . 1172
The BSP COLLISION Statement . . . . . . . . . . . . . . . . 1172
The SET BSP CAMERA Statement . . . . . . . . . . . . . . . 1173
The DELETE BSP Statement . . . . . . . . . . . . . . . . . . 1173
The SET BSP MULTITEXTURING Statement . . . . . . . . . . 1173
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1173
Using a BSP Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1175
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1175
The Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1175
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1178
Chapter 44 Creating Terrain

Creating Terrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1180
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1180
Documented Terrain Statements . . . . . . . . . . . . . . . . . . . 1180
The MAKE TERRAIN Statement . . . . . . . . . . . . . . . . . 1180
The DELETE TERRAIN Statement . . . . . . . . . . . . . . . 1181
The POSITION TERRAIN Statement . . . . . . . . . . . . . . 1182
The TERRAIN POSITION Statement . . . . . . . . . . . . . . 1183
The TEXTURE TERRAIN Statement . . . . . . . . . . . . . . 1183
The GET TERRAIN HEIGHT Statement . . . . . . . . . . . . . 1184
The GET TOTAL TERRAIN HEIGHT Statement . . . . . . . . 1186
The Advanced Terrain Statements . . . . . . . . . . . . . . . . . . 1186
The MAKE OBJECT TERRAIN Statement . . . . . . . . . . . 1186
The SET TERRAIN HEIGHTMAP Statement . . . . . . . . . . 1187
The SET TERRAIN SCALE Statement . . . . . . . . . . . . . 1187
The SET TERRAIN TEXTURE Statement . . . . . . . . . . . . 1188
The BUILD TERRAIN Statement . . . . . . . . . . . . . . . . 1188
The SET TERRAIN TILING Statement . . . . . . . . . . . . . 1189
The SET TERRAIN LIGHT Statement . . . . . . . . . . . . . . 1190
The SET TERRAIN SPLIT Statement . . . . . . . . . . . . . . 1191
The GET TERRAIN GROUND HEIGHT Statement . . . . . . . 1191
The GET TERRAIN SIZE Statement . . . . . . . . . . . . . . 1193
The SAVE TERRAIN Statement . . . . . . . . . . . . . . . . . 1193
The LOAD TERRAIN Statement . . . . . . . . . . . . . . . . . 1194
Terrains as Objects . . . . . . . . . . . . . . . . . . . . . . . 1195
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1195
Documented Statements . . . . . . . . . . . . . . . . . . . . . 1195
Undocumented (Advanced Terrain) Statements . . . . . . . . . 1196
Terrain Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1197
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1197
Creating the Game . . . . . . . . . . . . . . . . . . . . . . . . . . 1197
Constants and Global Variables . . . . . . . . . . . . . . . . . 1198
Adding StartUpGame() . . . . . . . . . . . . . . . . . . . . . . 1198
Adding PositionCamera() . . . . . . . . . . . . . . . . . . . . 1199
Adding CreateScene() . . . . . . . . . . . . . . . . . . . . . . 1199
Adding LoadTerrain() . . . . . . . . . . . . . . . . . . . . . . 1199
Adding CreateSkyBox() . . . . . . . . . . . . . . . . . . . . . 1200
Adding LoadOcean() . . . . . . . . . . . . . . . . . . . . . . . 1200
Adding PlaceOrb() . . . . . . . . . . . . . . . . . . . . . . . . 1201
Adding StartGame() . . . . . . . . . . . . . . . . . . . . . . . 1202
Adding ControlPlayer() . . . . . . . . . . . . . . . . . . . . . . 1202
Adding EndGame() . . . . . . . . . . . . . . . . . . . . . . . 1203
Adding Testing Features . . . . . . . . . . . . . . . . . . . . . 1204
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1206
Chapter 45 Using Matrices

Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212
Creating a Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213
The MAKE MATRIX Statement . . . . . . . . . . . . . . . . . 1213
The RANDOMIZE MATRIX Statement . . . . . . . . . . . . . 1214
The UPDATE MATRIX Statement . . . . . . . . . . . . . . . . 1214
The SET MATRIX HEIGHT Statement . . . . . . . . . . . . . 1215
The GET MATRIX HEIGHT Statement . . . . . . . . . . . . . 1217
The GET GROUND HEIGHT Statement . . . . . . . . . . . . . 1218
The SET MATRIX WIREFRAME Statement . . . . . . . . . . . 1219
The MATRIX WIREFRAME STATE Statement . . . . . . . . . 1220
Adding Texture to the Matrix . . . . . . . . . . . . . . . . . . . . . 1220
The PREPARE MATRIX TEXTURE Statement . . . . . . . . . 1220
The FILL MATRIX Statement . . . . . . . . . . . . . . . . . . 1222
The SET MATRIX TILE Statement . . . . . . . . . . . . . . . 1223
The SET TEXTURE TRIM Statement . . . . . . . . . . . . . . 1226
The SHIFT MATRIX Statement . . . . . . . . . . . . . . . . . 1227
The MATRIX TILE COUNT Statement . . . . . . . . . . . . . . 1228
The MATRIX TILES EXIST Statement . . . . . . . . . . . . . . 1228
Positioning the Matrix in 3D Space . . . . . . . . . . . . . . . . . . 1229
The POSITION MATRIX Statement . . . . . . . . . . . . . . . 1229
The MATRIX POSITION Statement . . . . . . . . . . . . . . . 1230
Matrix Transparency . . . . . . . . . . . . . . . . . . . . . . . . . 1231
The GHOST MATRIX ON Statement . . . . . . . . . . . . . . 1231
The GHOST MATRIX OFF Statement . . . . . . . . . . . . . . 1232
The SET MATRIX PRIORITY Statement . . . . . . . . . . . . 1232
Lighting the Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . 1234
The SET MATRIX NORMAL Statement . . . . . . . . . . . . . 1234
The SET MATRIX Statement . . . . . . . . . . . . . . . . . . 1235
The MATRIX EXIST Statement . . . . . . . . . . . . . . . . . 1237
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1238
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1240
Chapter 46 Manipulating Vertices

Manipulating Vertices . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
The Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
The LOCK VERTEXDATA FOR MESH Statement . . . . . . . 1246
The GET VERTEXDATA VERTEX COUNT Statement . . . . . 1247
The GET VERTEXDATA POSITION Statement . . . . . . . . 1248
The SET VERTEXDATA POSITION Statement . . . . . . . . . 1250
The UNLOCK VERTEXDATA Statement . . . . . . . . . . . . 1250
The LOCK VERTEXDATA FOR LIMB Statement . . . . . . . . 1251
The GET VERTEXDATA NORMALS Statement . . . . . . . . 1253
The SET VERTEXDATA NORMALS Statement . . . . . . . . . 1254
The GET VERTEXDATA Statement . . . . . . . . . . . . . . . 1255
The SET VERTEXDATA UV Statement . . . . . . . . . . . . . 1256
The SET VERTEXDATA DIFFUSE Statement . . . . . . . . . 1257
The GET VERTEXDATA DIFFUSE Statement . . . . . . . . . 1258
Handling More Complex Shapes . . . . . . . . . . . . . . . . . . . 1258
The ADD MESH TO VERTEXDATA Statement . . . . . . . . . 1265
More About the Vertex Data Buffer's Structure . . . . . . . . . . . 1266
The GET VERTEXDATA INDEX COUNT Statement . . . . . . 1267
The GET INDEXDATA Statement . . . . . . . . . . . . . . . . 1268
The SET INDEXDATA Statement . . . . . . . . . . . . . . . . 1270
The DELETE MESH FROM VERTEXDATA Statement . . . . . 1271
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1272
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1274
Chapter 47 Accessing Memory

Accessing Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1280
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1280
Pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1280
Creating Pointers in DarkBASIC Pro . . . . . . . . . . . . . . . 1281
Assigning a Value to a Pointer . . . . . . . . . . . . . . . . . . 1281
The MAKE MEMBLOCK Statement . . . . . . . . . . . . . . . 1281
The GET MEMBLOCK PTR Statement . . . . . . . . . . . . . 1281
Using a Pointer . . . . . . . . . . . . . . . . . . . . . . . . . . 1282
Using a Pointer to Return Values from a Function . . . . . . . . 1283
Larger Memory Blocks . . . . . . . . . . . . . . . . . . . . . . . . 1284
The WRITE MEMBLOCK Statement . . . . . . . . . . . . . . 1284
The MEMBLOCK Statement . . . . . . . . . . . . . . . . . . . 1285
The GET MEMBLOCK SIZE Statement . . . . . . . . . . . . . 1286
The DELETE MEMBLOCK Statement . . . . . . . . . . . . . . 1286
The MEMBLOCK EXIST Statement . . . . . . . . . . . . . . . 1286
The COPY MEMBLOCK Statement . . . . . . . . . . . . . . . 1287
Strings and Memory Blocks . . . . . . . . . . . . . . . . . . . 1288
The WRITE MEMBLOCK (to file) Statement . . . . . . . . . . 1290
The MAKE FILE FROM MEMBLOCK Statement . . . . . . . . 1292
The READ MEMBLOCK (from file) Statement . . . . . . . . . . 1292
The MAKE MEMBLOCK FROM FILE Statement . . . . . . . . 1294
Adding a New Top Score to our List . . . . . . . . . . . . . . . 1294
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1295
Media Contents and Memory Blocks . . . . . . . . . . . . . . . . . . . 1297
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1297
Bitmaps and Memory Blocks . . . . . . . . . . . . . . . . . . . . . 1297
The MAKE MEMBLOCK FROM BITMAP Statement . . . . . . 1297
The MAKE BITMAP FROM MEMBLOCK Statement . . . . . . 1399
Mapping a Screen Position to a Memory Block Location . . . . 1300
Mapping the Mouse Position to a Memory Block Location . . . . 1301
Images and Memory Blocks . . . . . . . . . . . . . . . . . . . . . 1302
The MAKE MEMBLOCK FROM IMAGE Statement . . . . . . . 1302
The MAKE IMAGE FROM MEMBLOCK Statement . . . . . . . 1302
Sounds and Memory Blocks . . . . . . . . . . . . . . . . . . . . . 1303
The MAKE MEMBLOCK FROM SOUND Statement . . . . . . 1303
The MAKE SOUND FROM MEMBLOCK Statement . . . . . . 1305
3D Objects and Memory Blocks . . . . . . . . . . . . . . . . . . . 1306
The MAKE MEMBLOCK FROM MESH Statement . . . . . . . 1306
The MAKE MESH FROM MEMBLOCK Statement . . . . . . . 1309
The CHANGE MESH FROM MEMBLOCK Statement . . . . . . 1310
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1311
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1312
Chapter 48 Open Dynamics Engine

Using ODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1318
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1318
Basic ODE Statements . . . . . . . . . . . . . . . . . . . . . . . . 1318
The ODE CREATE DYNAMIC BOX Statement . . . . . . . . . 1318
The ODE START Statement . . . . . . . . . . . . . . . . . . . 1319
The ODE END Statement . . . . . . . . . . . . . . . . . . . . 1319
The ODE UPDATE Statement . . . . . . . . . . . . . . . . . . 1319
The ODE SET WORLD GRAVITY Statement . . . . . . . . . . 1320
The ODE CREATE STATIC BOX Statement . . . . . . . . . . 1321
The ODE CREATE DYNAMIC SPHERE Statement . . . . . . . 1322
The ODE CREATE DYNAMIC CYLINDER Statement . . . . . 1322
The ODE CREATE DYNAMIC TRIANGLE MESH Statement . . 1324
The ODE SET WORLD STEP . . . . . . . . . . . . . . . . . . 1325
The ODE CREATE STATIC TRIANGLE MESH Statement . . . 1325
The ODE SET WORLD ERP Statement . . . . . . . . . . . . . 1326
The ODE SET WORLD CFM Statement . . . . . . . . . . . . . 1327
The ODE SET CONTACT FDIR1 Statement . . . . . . . . . . 1328
The ODE SET LINEAR VELOCITY Statement . . . . . . . . . 1328
The ODE SET ANGULAR VELOCITY Statement . . . . . . . . 1331
The ODE SET BODY ROTATION Statement . . . . . . . . . . 1332
The ODE SET BODY MASS Statement . . . . . . . . . . . . . 1332
The ODE DESTROY OBJECT Statement . . . . . . . . . . . . 1334
The ODE GET BODY LINEAR VELOCITY Statement . . . . . 1335
The ODE GET BODY HEIGHT Statement . . . . . . . . . . . . 1335
The ODE COLLISION MESSAGE EXISTS Statement . . . . . 1336
The ODE COLLISION GET MESSAGE Statement . . . . . . . 1336
The ODE GET OBJECT Statement . . . . . . . . . . . . . . . 1336
The ODE GET OBJECT VELOCITY Statement . . . . . . . . . 1337
The ODE GET OBJECT ANGULAR VELOCITY Statement . . . 1338
The ODE ADD FORCE Statement . . . . . . . . . . . . . . . . 1338
Surface Contact Statements . . . . . . . . . . . . . . . . . . . . . 1340
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1343
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1345
Chapter 49 Vectors and Matrices
3D Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1350
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1350
A Mathematical Description of 3D Vectors . . . . . . . . . . . . 1350
What is a 3D Vector in DarkBASIC Pro? . . . . . . . . . . . . 1351
Why do we need 3D Vectors? . . . . . . . . . . . . . . . . . . 1351
3D Vector Statements . . . . . . . . . . . . . . . . . . . . . . . . 1351
The MAKE VECTOR3 Statement . . . . . . . . . . . . . . . . 1351
The SET VECTOR3 Statement . . . . . . . . . . . . . . . . . 1352
Retrieving Data from a 3D Vector . . . . . . . . . . . . . . . . 1352
The DELETE VECTOR3 Statement . . . . . . . . . . . . . . . 1353
The COPY VECTOR3 Statement . . . . . . . . . . . . . . . . 1353
The MULTIPLY VECTOR3 Statement . . . . . . . . . . . . . . 1354
The SCALE VECTOR3 Statement . . . . . . . . . . . . . . . . 1354
The DIVIDE VECTOR3 Statement . . . . . . . . . . . . . . . . 1355
The LENGTH VECTOR3 Statement . . . . . . . . . . . . . . . 1355
The SQUARED LENGTH VECTOR3 Statement . . . . . . . . 1356
The ADD VECTOR3 Statement . . . . . . . . . . . . . . . . . 1356
The SUBTRACT VECTOR3 Statement . . . . . . . . . . . . . 1357
The DOT PRODUCT VECTOR3 Statement . . . . . . . . . . . 1357
The NORMALIZE VECTOR3 Statement . . . . . . . . . . . . . 1358
The IS EQUAL VECTOR3 Statement . . . . . . . . . . . . . . 1359
The MAXIMIZE VECTOR3 Statement . . . . . . . . . . . . . . 1359
The MINIMIZE VECTOR3 Statement . . . . . . . . . . . . . . 1360
The CROSS PRODUCT VECTOR3 Statement . . . . . . . . . 1360
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1361
4D Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363
Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1365
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1365
Matrix Statements . . . . . . . . . . . . . . . . . . . . . . . . . . 1365
The MAKE MATRIX4 Statement . . . . . . . . . . . . . . . . . 1365
The SET IDENTITY MATRIX4 Statement . . . . . . . . . . . . 1366
The IS IDENTITY MATRIX4 Statement . . . . . . . . . . . . . 1366
Other Matrix Assignment Statements . . . . . . . . . . . . . . 1367
The COPY MATRIX4 Statement . . . . . . . . . . . . . . . . . 1367
The IS EQUAL MATRIX4 Statement . . . . . . . . . . . . . . 1367
The ADD MATRIX4 Statement . . . . . . . . . . . . . . . . . 1368
The SUBTRACT MATRIX4 Statement . . . . . . . . . . . . . . 1368
The DIVIDE MATRIX4 Statement . . . . . . . . . . . . . . . . 1368
The MULTIPLY MATRIX4 Statement . . . . . . . . . . . . . . 1369
The INVERSE MATRIX4 Statement . . . . . . . . . . . . . . . 1370
The SCALE MATRIX4 Statement . . . . . . . . . . . . . . . . 1370
The TRANSLATE MATRIX4 Statement . . . . . . . . . . . . . 1371
The ROTATE MATRIX4 Statement . . . . . . . . . . . . . . . 1371
The TRANSPOSE MATRIX4 Statement . . . . . . . . . . . . . 1372
The DELETE MATRIX4 Statement . . . . . . . . . . . . . . . 1372
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1372
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374
Chapter 50 Shaders
Shaders and FX Files . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
Vertex Shader . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
Pixel Shader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
FX Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1377
Graphics Card Check Statements . . . . . . . . . . . . . . . . . . 1377
The GET MAXIMUM VERTEX SHADER VERSION Statement . 1377
The GET MAXIMUM PIXEL SHADER VERSION Statement . . 1377
FX Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1378
The LOAD EFFECT Statement . . . . . . . . . . . . . . . . . 1378
The EFFECT EXIST Statement . . . . . . . . . . . . . . . . . 1378
The PERFORM CHECKLIST FOR EFFECT ERRORS Statement . . . 1379
The SET OBJECT EFFECT Statement . . . . . . . . . . . . . 1379

The SET EFFECT ON Statement . . . . . . . . . . . . . . . . 1380
The DELETE EFFECT Statement . . . . . . . . . . . . . . . . 1381
The SET LIMB EFFECT Statement . . . . . . . . . . . . . . . 1381
The PERFORM CHECKLIST FOR EFFECT VALUES Statement1382
The SET EFFECT CONSTANT Statement . . . . . . . . . . . 1383
The SET EFFECT TECHNIQUE Statement . . . . . . . . . . . 1383
The SET EFFECT TRANSPOSE Statement . . . . . . . . . . 1384
Vertex Shader Statements . . . . . . . . . . . . . . . . . . . . . . 1383
The CREATE VERTEX SHADER FROM FILE Statement . . . 1383
The SET VERTEX SHADER ON Statement . . . . . . . . . . . 1385
The SET VERTEX SHADER OFF Statement . . . . . . . . . . 1385
The DELETE VERTEX SHADER Statement . . . . . . . . . . 1385
Other Vertex Shader Statements . . . . . . . . . . . . . . . . 1386
Pixel Shader Statements . . . . . . . . . . . . . . . . . . . . . . . 1386
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
FX Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
Shader Files . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1388
Chapter 51 Network Programming

Networked Games . . . . . . . . . . . . . .
. 1390. . . . . . . . . . . . .
Introduction . . . . . . . . . . . . . . . .
. 1390. . . . . . . . . . . . .
Hardware Requirements . . . . . . . . .
. 1390. . . . . . . . . . . . .
Getting Started . . . . . . . . . . . . . .
. 1390. . . . . . . . . . . . .
The PERFORM CHECKLIST FOR NET CONNECTIONS Statement . 1391
TCP/IP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1392
The SET NET CONNECTION Statement . . . . . . . . . . . . 1392
The CREATE NET GAME Statement . . . . . . . . . . . . . . 1394
Writing Code for the Client Machine . . . . . . . . . . . . . . . . . 1395
The PERFORM CHECKLIST FOR NET SESSIONS Statement 1395
The JOIN NET GAME Statement . . . . . . . . . . . . . . . . 1396
The PERFORM CHECKLIST FOR NET PLAYERS Statement . 1397
Using a Single Machine as Both Host and Client . . . . . . . . . . 1498
Combining the Host/Client Requirements . . . . . . . . . . . . . . 1499
Communicating . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1401
The SEND NET MESSAGE Statement (Version 1) . . . . . . . 1401
The GET NET MESSAGE Statement . . . . . . . . . . . . . . 1401
The NET MESSAGE EXISTS Statement . . . . . . . . . . . . 1402
The NET MESSAGE Statement (Version 1) . . . . . . . . . . . 1402
The NET MESSAGE PLAYER FROM Statement . . . . . . . . 1403
The NET MESSAGE PLAYER TO Statement . . . . . . . . . . 1403
The SEND NET MESSAGE Statement (Version 2) . . . . . . . 1404
The NET MESSAGE Statement (Version 2) . . . . . . . . . . . 1405
The NET MESSAGE TYPE Statement . . . . . . . . . . . . . 1406
The NET BUFFER SIZE Statement . . . . . . . . . . . . . . . 1408
Session Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 1409
The NET PLAYER CREATED Statement . . . . . . . . . . . . 1409
The NET PLAYER DESTROYED Statement . . . . . . . . . . 1409
The NET GAME NOW HOSTING Statement . . . . . . . . . . 1411
The FREE NET GAME Statement . . . . . . . . . . . . . . . . 1411
The CREATE NET PLAYER Statement . . . . . . . . . . . . . 1412
The FREE NET PLAYER Statement . . . . . . . . . . . . . . . 1412
The NET GAME EXISTS Statement . . . . . . . . . . . . . . . 1413
The NET GAME LOST Statement . . . . . . . . . . . . . . . . 1413
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1413
A Networked Game . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1415
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1415
A Non-Networked Version . . . . . . . . . . . . . . . . . . . . . . 1415
Program Data . . . . . . . . . . . . . . . . . . . . . . . . . . 1415
Game Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . 1416
Adding SetUpPlayerDetails() . . . . . . . . . . . . . . . . . . . 1417
Adding SetUpScreen() . . . . . . . . . . . . . . . . . . . . . . 1417
Adding SetUpBoard() . . . . . . . . . . . . . . . . . . . . . . 1417
Adding GetMove() . . . . . . . . . . . . . . . . . . . . . . . . 1417
Adding GetMyMove() . . . . . . . . . . . . . . . . . . . . . . 1418
Adding GetSquare() . . . . . . . . . . . . . . . . . . . . . . . 1418
Adding InRange() . . . . . . . . . . . . . . . . . . . . . . . . 1418
Adding GetOpponentsMove() . . . . . . . . . . . . . . . . . . 1419
Adding CheckForWin() . . . . . . . . . . . . . . . . . . . . . . 1419
Adding the Other Search Routines . . . . . . . . . . . . . . . 1420
Adding EndGame() . . . . . . . . . . . . . . . . . . . . . . . 1422
Networking the Game . . . . . . . . . . . . . . . . . . . . . . . . 1423
Updating the main section . . . . . . . . . . . . . . . . . . . . 1423
Adding WaitForSecondPlayer() . . . . . . . . . . . . . . . . . 1423
Adding NumberOfPlayers() . . . . . . . . . . . . . . . . . . . 1423
Modifying the Call to SetUpPlayerDetails() . . . . . . . . . . . 1424
Modifying GetMyMove() . . . . . . . . . . . . . . . . . . . . . 1424
Modifying GetOpponentsMove() . . . . . . . . . . . . . . . . . 1424
Modifying EndGame() . . . . . . . . . . . . . . . . . . . . . . 1425
A Complete Listing . . . . . . . . . . . . . . . . . . . . . . . . 1425
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1431
Chapter 52 Using File Transfer Protocol

Internet File Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . 1436
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1436
The Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . 1436
The FTP CONNECT Statement . . . . . . . . . . . . . . . . . 1436
The GET FTP FAILURE Statement . . . . . . . . . . . . . . . 1436
The GET FTP ERROR$ Statement . . . . . . . . . . . . . . . 1437
The GET FTP STATUS Statement . . . . . . . . . . . . . . . 1437
The FTP SET DIR Statement . . . . . . . . . . . . . . . . . . 1438
The GET FTP DIR$ Statement . . . . . . . . . . . . . . . . . 1438
The FTP FIND FIRST Statement . . . . . . . . . . . . . . . . 1438
The FTP FIND NEXT Statement . . . . . . . . . . . . . . . . . 1439
The GET FTP FILE TYPE Statement . . . . . . . . . . . . . . 1439
The GET FTP FILE NAME$ Statement . . . . . . . . . . . . . 1439
The GET FTP FILE SIZE Statement . . . . . . . . . . . . . . . 1439
The FTP DISCONNECT Statement . . . . . . . . . . . . . . . 1440
The FTP GET FILE Statement . . . . . . . . . . . . . . . . . . 1440
The FTP PROCEED Statement . . . . . . . . . . . . . . . . . 1441
The GET FTP PROGRESS Statement . . . . . . . . . . . . . 1442
The FTP TERMINATE Statement . . . . . . . . . . . . . . . . 1442
The FTP DELETE FILE statement . . . . . . . . . . . . . . . . 1442
The FTP PUT FILE Statement . . . . . . . . . . . . . . . . . . 1443
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1443
Chapter 53 Dynamic Link Libraries

Creating New DBPro Statements . . . . . . . . . . . . . . . . . . . . . 1446
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1446
A Dynamic Link Library (DLL) . . . . . . . . . . . . . . . . . . . . 1446
Creating a DLL . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1446
Starting Up Visual Studio . . . . . . . . . . . . . . . . . . . . 1446
Adding the Code for New Statements . . . . . . . . . . . . . . 1448
Adding a String Table . . . . . . . . . . . . . . . . . . . . . . 1449
Constructing the Caption . . . . . . . . . . . . . . . . . . . . . 1450
Adding the New Statements to DarkBASIC Pro . . . . . . . . . 1451
Adding Help . . . . . . . . . . . . . . . . . . . . . . . . . . . 1452
Adding More New Commands . . . . . . . . . . . . . . . . . . . . 1456
Functions that Return Real Values . . . . . . . . . . . . . . . . 1456
Functions that Return Strings . . . . . . . . . . . . . . . . . . 1456
More String Handling Functions . . . . . . . . . . . . . . . . . 1459
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1460
Using Standard DLLs . . . . . . . . . . . . . . . . . . . . . . . . . . . 1462
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1462
The LOAD DLL Statement . . . . . . . . . . . . . . . . . . . . 1463
The DLL EXIST Statement . . . . . . . . . . . . . . . . . . . . 1463
The CALL DLL Statement . . . . . . . . . . . . . . . . . . . . 1463
The DLL CALL EXIST Statement . . . . . . . . . . . . . . . . 1464
The DELETE DLL Statement . . . . . . . . . . . . . . . . . . 1464
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1465
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1467
Acknowledgements
I would like to thank all those who helped me prepare the final draft of this book.
In particular, Virginia Marshall who proof-read the original script and Michael Kerr
who did an excellent job of checking the technical contents. Mark Armstrong
researched all the difficult bits for me and produced almost as much in the way of
notes as is in this final text.
Any errors that remain are probably due to the usual extra paragraphs I added after
all the proof-reading was complete!
Thanks also to The Game Creators Ltd for producing an excellent piece of software
- DarkBASIC Professional.
Many of the 3D models and textures are from The Game Creators Dark Matter 1
package and used with their kind permission.
Finally, thank you to every one of you who has bought this book. Any constructive
comments would be most welcome.
Email me at alistair@digital-skills.co.uk.
Introduction
Welcome to the second volume of a book that I hope is a little different from any
other you've come across before. Instead of just telling you about software design
and programming, it makes you get involved. There's plenty of work for you to do
since the book is full of exercises - most of them programming exercises - but you
also get a full set of solutions, just in case you get stuck!
If you've worked your way through Volume 1, then you should have gained a good
grounding in, not only DarkBASIC Pro, but also professional programming skills.
Most of Volume 2 is dedicted to 3D graphics but there a few other interesting topics
such as network programming and how to create your own DarkBASIC commands.
Learn by Doing
The only way to become a programming expert is to practice. No one ever learned
any skill by just reading about it! Hence, this is not a text book where you can just
sit back in a passive way and read from cover to cover whilst sitting in your favourite
chair. Rather it is designed as a teaching package in which you will do most of the
work.
The tasks embedded in the text are included to test your understanding of what has
gone before and as a method of helping you retain the knowledge you have gained.
It is therefore important that you tackle each task as you come to it. Also, many of
the programming exercises are referred to, or expanded, in later pages so it is
important that you are familar with the code concerned.
What You Need

You'll obviously need a PC and a copy of DarkBASIC Pro.
At this stage you'll also need some programming skills and a basic knowledge of
DarkBASIC Pro.
How to Get the Most out of this Text

Experience has shown that readers derive most benefit from this material by
approaching its study in an organised way. The following strategy for study is highly
recommended:
1. Read a chapter or section through without taking notes or worrying

too much about topics that are not immediately clear to you. This will
give you an overview of the contents of that chapter/section.
2. Re-read the chapter. This time take things slowly; make notes and
summaries of the material you are reading (even if you understand the
material, making notes helps to retain the facts in your long-term
memory); re-read any parts you are unclear about.
3. Embedded in the material are a series of activities. Do each task as you

reach it (on the second reading). These activities are designed to test
your knowledge and understanding of what has gone before. Do not be
tempted to skip over them, promise to come back to them later, or to
make only a half-hearted attempt at tackling them before looking up
the answer (there are solutions at the end of each chapter). Once you
have attempted a task, look at the solution given. Often there will be
important points emphasised in the solution which will aid higher
understanding.
4. As you progress through the book, go back and re-read earlier chapters,
since you will often get something new from them as your knowledge
increases.
Syntax Diagrams
The format of each statement is explained using a syntax diagram. Raised tiles
represent keywords of the language while sunken tiles are parts of the statement for
which you are free to create your own values. Parts within square brackets are
optional while braces represent a choice of options. Statements that return a value
show this using an arrowed line and the type of value returned.
Braces represent choice
Raised tiles Sunken tiles

represent keywords represent programmer-assigned
X values
LIMB OFFSET Y ( objno , limbno )
Z An arrowed line
indicates that the
statement returns a value
real
Line Continuation Symbol

Occasionally, a single programming instruction has to be split over two or more
lines because of limited page width. In such cases the second line(and subsequent
Ä
lines) begins with the symbol. For example, the instruction
POSITION OBJECT 2,OBJECT POSITION X(2),OBJECT POSITION Y(2)-0.1,OBJECT POSITION Z(2)+0.1
might appear as
POSITION OBJECT 2,OBJECT POSITION X(2),

ÄOBJECT POSITION Y(2)-0.1,OBJECT POSITION Z(2)+0.1
In such cases you should enter the code as a single line when creating a DarkBASIC
Pro program.
30
3D Coordinate System
3D Primitives
3D Vectors
Cameras
Lights
Vertex and Surface Normals

Rotation
Textures
The Major Planes in 3D
Vertices, Edges and Polygons
Wireframe Models
World Units
DarkBASIC Pro: 3D - Concepts and Terminology 743

The 3D World
Introduction
Welcome to the world of 3D. Of course, we can create great games in 2D - many
people still consider 2D games like Space Invaders and Pac-Man to be some of the
best games ever invented - but for sheer eye candy you really can't beat 3D.
In this chapter we'll get a broad view of the 3D world created by computers. We'll
cover the basic concepts and define some of the terms. Many of these concepts will
be explained in greater detail in later chapters as we discover how DarkBASIC Pro
implements many of these ideas.
The Coordinate System

Axes
In a 3D world, just as in a 2D one, we need to identify the position of any point
within that world. This we do using three axes (known as world axes) for reference.
As before, we need x and y axes for width and height, but this time we also need a
z axis to measure depth (see FIG-30.1).
FIG-30.1
+y
The Axes used in 3D
+z
-x
+x
-z
-y
In the figure above, the axes have been skewed slightly to give a better perspective
In reality the x-axis runs across the screen, the y-axis runs up and down, and the
z-axis points directly out of the screen (-z) and into the screen (+z) (see FIG-30.2).
744 DarkBASIC Pro: 3D - Concepts and Terminology

FIG-30.2 +y
3D Axes and the Viewer
The computer
screen
The viewer
+z
-x
+x
-z
-y
Planes
In mathematics, a plane is a flat surface with only two dimensions. 3D space has
three main planes: the X-Y plane, the X-Z plane and the Y-Z plane (see FIG-30.3).
+y
FIG-30.3
The
The Main 3D Planes
X-Y Plane
+z
-x
+x
-z
-y
+y
The
X-Z Plane
+z
-x
+x
-z
-y
+y
The
Y-Z Plane
+z
-x
+x
-z
-y

The X-Y plane has the x and y axes passing through its centre and, like every plane,
expands to infinity in all directions. The X-Z plane has the x and z axes at its centre,
and the Y-Z plane has the y and z axes at its centre.
These three planes are important since each divides space into two equally sized
areas. The X-Y plane splits space with one half to the front, the other half to the
back. The X-Z planes splits space into above and below sections, and the Y-Z plane
splits space into left and right sections.
With all three planes in place, space is split into eight equally-sized sections. Each
of these sections is known as an octant.
Of course, not all planes lie on axes; there are an infinite number of planes, some
parallel to the main planes, others at angles to those planes, but it is the main planes
that will be useful in many of the calculations required when determining the
position of an object in 3D space.
Points
To specify the position of a point in 3D space we state its distance from the origin
along all three axes in the order, x, y, z (see FIG-30.4).
FIG-30.4 +y
Determining the Position of a
Point in 3D Space A point in
space
+z
-x
+x
1
-z Measure the point’s
distance out from the
origin along the x-axis
-y
+y
+z
2
-xMeasure the point’s
origin along the y-axis
+x
-z
-y

FIG-30.4
(continued) +y
Determining the Position of a

Point in 3D Space
P
+z
-x
3 +x
-xMeasure the point’s
origin-z
along the z-axis
-y
We might, for example, state that point p is at the position (8,12,5) meaning that
point p is 8 units along the x-axis, 12 units along the y-axis and 5 units along the
z-axis.
World Units
Distances are measured in units. These units have no relationship to real-life
measurements such as centimetres or inches. Instead, objects are constructed in such
a way as to be the correct size relative to other objects. For example, if we make a
human character 6 units high, then a simple house might be 18 to 25 units high. Of
course, if you wish, you can think of 1 unit being the equivalent of a real distance.
The scale you choose will depend on the context; when creating a world with an
ant as the main character, 1 unit might be equivalent to a millimetre, while a truly
interstellar game might make 1 unit equivalent to 1 light year.
Local Axes
Every 3D object we create has its own local axes. These axes are (initially, at least)
aligned to the world axes. FIG-30.5 shows a cuboid and its local axes.
FIG-30.5 +y
Each Object has its
Own Local Axes
+z
-x
+x
-z
Each 3D object
has its own local
axes parallel to the
-y world axes

Rotation
An object can be made to rotate about its own, local, axes. In DarkBASIC Pro
rotation is measured in degrees. For example, we might rotate an object 30 degrees
about it's x-axis as shown in FIG-30.6.
Initial Position Rotated 30o about
Y-axis Y-axis
the x-axis
FIG-30.6
A Cuboid is Rotated
30o about the x-axis
X-axis Z-axis X-axis

Z-axis
Rotation is performed in a clockwise direction when viewed down the positive end
of an axis (see FIG-30.7).
FIG-30.7 Rotation is in a
clockwise direction
Clockwise Rotation
Axi when viewed from

s the positive end of
an axis
By specifying a negative angle of rotation, an object will rotate anti-clockwise.
3D Vectors
Although the purpose of this chapter is to describe basic 3D concepts, it's worth
mentioning that DarkBASIC Pro allows the creation of a 3-element vector
specifically for storing the coordinates of a point in 3D space. The vector is created
using the MAKE VECTOR3 statement which has the format shown in FIG-30.8.
FIG-30.8
The MAKE VECTOR3 MAKE VECTOR3 ( vectno )
Statement
integer
In the diagram:
vectno is an integer value giving the ID to be assigned

to the 3D vector being created.
The statement returns 1 if the vector is created successfully; otherwise zero is

returned. Usually we won't worry about the value returned and can create a 3D
vector with a statement such as:
result = MAKE VECTOR3(1)

We can visualise a 3D vector object as shown in FIG-30.9.
FIG-30.9 A 3D Vector
A 3D Vector 12.3 4.6 1.7

...an x value,
x y z a y value
and a z value
The 3D vector
holds three values...
Many of the DarkBASIC Pro statements we'll encounter later make use of 3D
vectors for storing results, so it's useful to give you this quick grounding in them at
this early stage. We'll learn more on this subject in a later chapter.
Object Terminology
Just as 2D has a few basic shapes such as a line, a circle, a triangle and a rectangle,
so we have a set of basic shapes (known as primitives) in 3D. These include the
sphere, cylinder, cone, and cube. In FIG-30.10 we see and example of a cube.
FIG-30.10
A Cube - An Example of
a Primitive
solid wireframe
The cube is shown in two ways: solid, with shading caused by the light falling on
its surface, and wireframe showing how the cube is constructed.
Polygon is the term used for a many-sided enclosed area. The simplest polygon
(that is, the one with the least sides) is the triangle. The point where two lines of a
polygon meet is known as a vertex. A triangle has three vertices (see FIG-20.11).
Triangle
FIG-30.11
The Vertices of a
Triangle Vetrices
The line between two vertices is known as an edge (see FIG-30.12).
FIG-30.12
Edges
Edges
Every 3D shape in a game is constructed from polygons (normally triangles), as

you can see from the wireframe version of the cube shown in FIG-30.10.
The greater the number of polygons used to create an object, the more detailed and
realistic it will appear (see FIG-30.13). But there is a price to pay for greater detail

- higher processing requirements. As you increase the number of polygons that go
to make up the objects in your scene, the harder your processor and video card need
to work. Ask too much of your hardware, and screen updating will slow down. The
number of times the screen is redrawn in one second is known as the frame rate
and is quoted in frames per second (fps). If the frame rate falls much below about
20 fps, then your eyes will become aware of the screen refreshing and the picture
will become jerky.
FIG-30.13
Varying the Polygons in
a Sphere
A Sphere with Few Polygons A Sphere with Many Polygons
Textures
In solid mode (as opposed to wireframe), a 3D object has a bland grey surface, but
we can use an image wrapped around that object to give it a greater reality. By
wrapping the image of riveted steel plate round a sphere, we can create the illusion
of a metal ball. Wrap an image of wooden planks round the same sphere and we
create a wooden ball (see FIG-30.14).
FIG-30.14
Adding Texture to a 3D
Object
Images with an Alpha Channel

The image used to texture an object can be one of many different formats. For
example, JPG and BMP files are often used, but sometimes we will see images
stored in the PNG or TGA format.
PNG and TGA files are amongst those formats capable of embedding an alpha
channel within the image. An alpha channel affects how visible an image is and is
probably best explained with an analogy.
Imagine you've just painted an image on a piece of glass and that the light
illuminating the picture comes from behind the glass (see FIG-30.15) - like looking

out through a church's stained-glass window.
FIG-30.15 Back of Glass
Front of Glass
Perceived Image Light Source
Depends on the Backing
The image seen at

the front of the glass
depends on the paint
used on the back
If we were to paint the back of the glass black, no light would get through and we
wouldn't see the picture. If we used grey paint rather than black, then some light
would get through. If we painted a pattern on the back of the glass using a mixture
of black, dark grey, and light grey paint, the image would appear to have bright,
dull and black areas depending on the paint on the back of the image.
This is how the alpha channel of an image works. As well as the basic red, green
and blue elements (or channels) that go to make up the image, a fourth, alpha,
channel is added. This is just another layer to the image which can only be shaded
using greyscale colours (white through to black). Where black is used, the original
image is unseen; where white is used the image appears at normal brightness (this
is where the glass analogy falls down since it would be at its brightest with no paint
on the back of the glass). Shades of grey give varying degrees of image brightness.
FIG-30.16 shows original images, alpha channels, and the overall effects created.
FIG-30.16
Using an Alpha Channel
Cameras
The real world is a vast place, but with the help of television we can view any part
of it - all we need is a camera. What the TV camera broadcasts we see on our screens.
Move the camera and we see a different part of the world.
This is exactly how the 3D world we create within the computer works; what we
see on the computer screen is the output from a virtual camera. The camera can be
moved, just like a real camera, revealing different parts of our new 3D world. We
can zoom the camera in or out allowing us to enlarge a distant object or show
everything within a small space.

We can even use several cameras, switching between each to change what the user
is seeing on the screen. Unlike real life, there's never any chance of seeing a camera
in the view produced by a second camera - all virtual cameras are invisible!
DarkBASIC Pro creates and positions a single camera automatically at the start of
every program that uses 3D objects. The exact position of the camera depends on
the positioning of the 3D objects, since the camera normally places itself in order
to see the objects that have been created. However, as the programmer, you can take
complete control of the camera and thereby determine just exactly what appears on
the screen.
Lights
We can even set up the lights we want to use to illuminate our new world - just like
placing lights on a movie set. By positioning various types of lights in just the correct
positions, we can create any type of atmosphere we want - from dark and mysterious
to bright and sunny. Like cameras in the 3D world, the lights are invisible, but the
effects they create are not!
To help calculate the effect of lights on the individual polygons of a 3D object, a

set of normals are maintained by the object. A surface normal is a vector coming
from the centre of a polygon, perpendicular to the surface of the polygon. Every
polygon within an object has an associated surface normal (see FIG-30.17).
FIG-30.17
Surface Normals
surface
normal
Normals are stored as mathematical expressions and are not part of the visible
structure of the model.
In principal every polygon can have two surface normals: one on the top side and
one on the bottom. However, often models only use a single normal - on the side
facing outwards.
When using surface normals to calculate how an object should be lit, we sometimes
get a rather faceted appearance, with an obvious jump in shading from one polygon
to the next (see FIG-30.18).
FIG-30.18
Visible Polygons

To solve this, vertex normals may be used. A vertex normal is created at every
vertex of a polygon (see FIG-30.19).
FIG-30.19 Vertex Normals
Vertex normal Vertex normal
Vertex normal
Vertex normal

Vertex normal
Every vertex in a polygon has a When two or more polygons have common vertices, that vertex has a separate
vertex normal. vertex normal for each of the polygons that share the vertex.
These vertex normals are calculated from the values of the two edges which meet
at that vertex.
Using vertex normals creates a smoother lighting effect, but requires more
calculations. You can see the effect produced in FIG-30.20.
FIG-30.20
Polygon Smoothing
Activity 30.1
Load and run the program basic3D.exe. This will demonstrate some of the
basic concepts covered in this chapter.
(You can download this program, and all other files used in this text from
www.digital-skills.co.uk)
As we'll see in the chapters that follow, DarkBASIC Pro has literally hundreds of
commands designed to help us create a 3D world and manipulate the objects in that
world.
Summary
l The 3D world uses three axes: x, y and z.
l 3D space is split into eight octants by the X-Y, X-Z and Y-Z planes.

l Space within the 3D world is measured in world units. These do not relate to real
world units.
l A point in 3D space is defined by its distance along each of the axes.
l 3D objects have their own local axes.
l 3D objects can be rotated about their own local axes.
l Rotations are measured in degrees.
l Rotation is in a clockwise direction (as viewed from the positive end of the axis
of rotation).
l DarkBASIC Pro provides 3D vector objects in which the coordinates of a point

in 3D space can be stored.
l 3D objects are constructed from polygons.
l The simplest polygon is the triangle.
l The end of a line within a polygon is known as a vertex.
l The line between two vertices is known as an edge.
l More detailed objects require more polygons.
l Increasing the number of polygons used in a scene increases the load on the
computer.
l When faced by a heavy load, the computer will output at a reduced frame rate.
l Images can be used to texture a 3D shape to increase realism.
l Some images can contain alpha channels which effect lightness when the image
is used to texture a surface.
l Virtual cameras determine which parts of the 3D world are shown on the screen.
l Lights can be added to a scene to help create the desired atmosphere.
l The effects of lights on a surface are calculated using surface normals or vertex
normals.
l Every polygon has an associated surface normal.
l A surface normal is a vector at right angles to its polygon.
l Using surface normals to calculate shading can result in a patchy effect.
l Every vertex of a polygon has an associated vertex normal.
l Vertex normals may be used to create smoother shading effects, but at the cost
of more complex calculations.

31
Absolute and Relative Object Movement

Global and Local Axes
Creating 3D Primitives
Culling
Deleting 3D Primitives
Duplicating 3D Objects
Merging Objects
Pointing an Object in a Specific Direction
Positioning 3D Objects
Retrieving 3D Object Data
Rotating 3D Objects
Resetting Local Axes
Resizing 3D Objects
Showing and Hiding 3D Objects
Wireframe Mode
DarkBASIC Pro: 3D Primitives 755

3D Primitives
Introduction
DarkBASIC Pro contains several statements for creating and manipulating 3D
primitives such as spheres, cones and cubes. These statements are explained in detail
below. A sample of the possible shapes is shown in FIG-31.1.
FIG-31.1
The 3D Shapes that can
be Created in
DarkBASIC Pro
Creating a Cube
The MAKE OBJECT CUBE Statement
To create a cube on the screen, we use the MAKE OBJECT CUBE statement. Like
sprites, every 3D object created must be given an identifying integer value (its ID).
No two 3D objects within a program can be assigned the same ID. The size of the
cube is also defined in this statement, which has the format shown in FIG-31.2.
FIG-31.2
The MAKE OBJECT
MAKE OBJECT CUBE objno , size
CUBE Statement
In the diagram:
objno is an integer value giving the ID to be assigned to the cube.
size is a real value specifying the width, height and depth of the
cube. This value is given in world units.
A typical usage of this statement might be:
MAKE OBJECT CUBE 1, 10
This would create a cube (with ID 1) which is 10 units wide, by 10 units high, by
10 units deep. FIG-31.3 shows a screen shot of the resulting cube.
FIG-31.3
A Cube in DarkBASIC Pro
This may not look too impressive as a 3D object, but that’s because we’re looking
at the cube straight on and therefore can only see the front face of the object.
756 DarkBASIC Pro: 3D Primitives

The cube shown above was created using the program given in LISTING-31.1.
LISTING-31.1 REM *** Set display resolution and backdrop ***

SET DISPLAY MODE 1280,1024,32
Creating a Cube COLOR BACKDROP 0
BACKDROP ON
Statements such as
COLOR BACKDROP REM *** Make the cube ***
and BACKDROP ON MAKE OBJECT CUBE 1,10
were covered in Volume
1. REM *** End program ***
WAIT KEY
END
Activity 31.1
Type in the program in LISTING-31.1 (object3D01.dbpro) and check that you

get the same display as shown above.
When any of the 3D primitives is first created, its centre is positioned at the origin.
FIG-31.4 shows a model of what has been created by the program in LISTING-31.1.
The 3 axes and parts of the XZ and YZ planes have been included to give a clearer
picture of how the cube is positioned.
FIG-31.4
How the Cube is y-axis
Positioned by the
Program
z-axis
s
10 unit
un
its 10
10 units
The centre of the cube

is at the origin
x-axis
Creating Other Primitives

DarkBASIC Pro has a set of similar MAKE statements to create other basic 3D
shapes. Like the cube, all of these objects are initially positioned with their centres
at the origin. These statements are described below.
The MAKE OBJECT BOX Statement

The MAKE OBJECT BOX statement is similar to the MAKE OBJECT CUBE
statement, but allows the three dimensions of the object to be set separately. The
statement has the format shown in FIG-31.5.

FIG-31.5
The MAKE OBJECT
MAKE OBJECT BOX objno , w , h , d
BOX Statement
In the diagram:
objno is an integer value giving the ID to be assigned to the box

being created. No other 3D object in the program can be
assigned the same value.
w is a real value giving the width (x-dimension) of the box.
h is a real value giving the height (y-dimension) of the box.
d is a real value giving the depth (z-dimension) of the

box.
For example, the line

MAKE OBJECT BOX 2, 10, 3.7, 12
would create a box with ID 2 which is 10 units wide, by 3.7 units high, by 12
units deep.
The MAKE OBJECT SPHERE Statement

The MAKE OBJECT SPHERE statement creates a sphere of a specified diameter
FIG-31.6 but offers extra options. The statement has the format shown in FIG-31.6.
The MAKE OBJECT SPHERE Statement
MAKE OBJECT SPHERE objno , diameter , rows , columns
In the diagram:
objno is an integer value giving the ID assigned to the

sphere being created.
diameter is a real number representing the diameter of the

sphere.
rows is an integer value specifying the number of lines

of latitude making up the sphere.
columns is an integer value specifying the number of lines

of longitude making up the sphere.
The statement
MAKE OBJECT SPHERE 3,40.0
would create a sphere with a diameter of 40 units and assign it the ID number 3.
However, the sphere produced is constructed from a relatively small number of
polygons and hence its curve is not particularly smooth. By using the rows and
columns values, we can control the number of polygons used to construct the sphere
and thereby produce a more realistic effect. For example, the line

MAKE OBJECT SPHERE 3,40.0,100,100
would create a much smoother sphere. FIG-31.7 shows the difference between the
default sphere and the more detailed one.
FIG-31.7
Creating a Smoother
Sphere
The number of polygons

used has greatly increased
CREATE OBJECT SPHERE 3, 40.0 CREATE OBJECT SPHERE 3, 40.0,100,100
However, there's a price to be paid for the more detailed sphere; the more polygons
we use when creating any 3D shape, the more work the processor/video card needs
to do and this reduces the frames per second that can be achieved.
Activity 31.2
Modify your previous program so that a standard sphere (diameter 10) is

created instead of a cube.
Modify the sphere to have 40 columns by 40 rows.
The MAKE OBJECT CYLINDER Statement

A cylinder of a specified height can be created using the MAKE CYLINDER
OBJECT statement. The diameter of the cylinder’s base automatically matches the
height. The statement has the format shown in FIG-31.8.
FIG-31.8
The MAKE OBJECT MAKE OBJECT CYLINDER objno , h
CYLINDER Statement
In the diagram:
objno is the integer value assigned to the cylinder being

created.
h is a real value giving the height and diameter of

the cylinder.

For example, we could make a cylinder of height 31.5 units using the statement:
MAKE OBJECT CYLINDER 4,31.5
The MAKE OBJECT CONE Statement

The MAKE OBJECT CONE statement creates a cone of a specified height. The
diameter of the base automatically matches the height. This statement has the format
shown in FIG-31.9.
FIG-31.9
The MAKE OBJECT MAKE OBJECT CONE objno , h
CONE Statement
In the diagram:
objno is the integer value assigned to the cone being

created.
h is a real value giving the height of the cone and

the diameter of its base.
For example, we could make a cone of height 10.1 units using the statement:
MAKE OBJECT CONE 5,10.1
Activity 31.3
Modify your previous program to display a cylinder of diameter 5.
Modify the program again to show a cone of the same height as the cylinder.
The MAKE OBJECT PLAIN Statement

A flat plane standing on the XY plane can be constructed using the MAKE OBJECT
PLAIN statement which has the format shown in FIG-31.10.
FIG-31.10
The MAKE OBJECT MAKE OBJECT PLAIN objno , w , h
PLAIN Statement
Note the spelling used in In the diagram:

the instruction!
objno is the integer value assigned to the plane being
created.
w is a real value giving the width of the plane.
h is a real value giving the height of the plane.
For example, we could create a plane which is 1000 units wide by 500 high using
the line:
MAKE OBJECT PLAIN 6,1000.0,500.0
The centre of the plane will be located at the origin (see FIG-31.11).

FIG-31.11
y-axis
How a Plane is Positioned
when First Created
z-axis
x-axis
The MAKE OBJECT TRIANGLE Statement

The simplest of all polygons, the triangle, can be constructed using the MAKE
OBJECT TRIANGLE statement. The statement requires the positions of all three
vertices to be supplied, so this statement contains a significant number of values,
as shown in FIG-31.12.
FIG-31.12
The MAKE OBJECT
TRIANGLE Statement
MAKE OBJECT TRIANGLE objno , x1 , y1 , z1
, x2 , y2 , z2 , x3 , y3 , z3
In the diagram:
objno is the integer value assigned to the triangle being

created.
x1,y1,z1 are real numbers representing the position of the

first vertex.

second vertex.

third vertex.
The line
The additional spacing
used within the MAKE OBJECT TRIANGLE 7, 2,0,3, 5,0,3, 3.5,6,7
instruction is used to
highlight the various
parameter groupings.
would create the triangle shown in FIG-31.13.

FIG-31.13 y-axis
Creating a Triangle Object
(3.5,6,7)
z-axis
(2,0,3)
(5,0,3)
x-axis
Notice that, unlike any of the other objects, a triangle can be placed anywhere in
3D space.
Positioning an Object
Other than the triangle, every object is created with it’s centre at the point (0,0,0).
However, once an object has been created, DarkBASIC Pro offers several ways of
moving an object to another position.
The POSITION OBJECT Statement

One way to move an object is to use the POSITION OBJECT statement. The object
is moved so that its centre is at the position specified. The statement has the format
shown in FIG-31.14.
FIG-31.14
The POSITION POSITION OBJECT objno , x , y , z
OBJECT Statement
In the diagram:
objno is the integer value previously assigned to the object.
x,y,z are real values representing the position to which the

object is to be moved. It is the centre of the object that is
placed at this position.
For example, if we wanted the centre of the cube we had created previously to be
moved to position (9,0,0), then we would use the statement:
POSITION OBJECT 1,9,0,0
The result of executing this statement is shown in FIG-31.15.

FIG-31.15
The Result of Moving the
Cube to (9,0,0)
y-axis
z-axis
The centre of the cube

is at (9,0,0) x-axis
LISTING-31.2 is a modification of the previous listing which moves the cube to

position (9,0,0) after the user presses a key. The new lines have been highlighted.
LISTING-31.2
Moving the Cube REM *** Set display resolution and backdrop ***
COLOR BACKDROP 0
BACKDROP ON
REM *** Make the cube **

REM *** Move cube to (9,0,0) after key press ***

WAIT KEY
REM *** End program ***

WAIT KEY
END
Activity 31.4
Modify your previous program to match that given in LISTING-31.2.
Add the lines
REM *** Move the cube backwards ***

WAIT KEY
POSITION OBJECT 1, 9,0,30
so that the object is moved for a second time.
Notice that, in its final position, the cube looks smaller since it has now moved
further away from our viewing position.

Activity 31.5
Write a program (object3D02.dbpro) to create the following objects, and then

position each as specified:
Object Number Object Type Dimensions Final Position

1 CUBE 4 9,0,0
2 BOX 10,15,5 -60,0,100
3 SPHERE 7 -30,-40,50
4 CYLINDER 12 25,0,120
5 CONE 12 0,25,100
Add a WAIT KEY statement between each move.
The screen should appear as shown below.
Screen at Start Screen at End
The MOVE OBJECT Statement

The MOVE OBJECT statement can be used to move an object a specified distance
from its current location.
There are four possible directions available: RIGHT, LEFT, UP and DOWN. The
FIG-31.16
{}
The MOVE OBJECT UP
Statement
DOWN
MOVE OBJECT objno , dist
LEFT
RIGHT
In the diagram:
UP, DOWN, LEFT, RIGHT

One of these keywords must be used to indicate in
which direction the object is to be moved.
objno is an integer value giving the ID of the object to be

moved.

dist is a real value giving the number of units the
object is to moved.
The direction of movement for each option is shown in FIG-31.17.
FIG-31.17
Using the MOVE
OBJECT Statement MOVE OBJECT UP
MOVE OBJECT LEFT
MOVE OBJECT RIGHT
MOVE OBJECT DOWN
Notice that there is no option to move the object backwards or forwards (i.e. along
the z-axis). For example, object 1 could be moved 10.5 units to the right using the
statement:
MOVE OBJECT RIGHT 1,10.5
Activity 31.6
Create a new program (object3D03.dbpro) containing a cube of size 10.
Use POSITION OBJECT to place the cube at (9,0,100).
Now move the cube 31.3 units to the right.
Place a WAIT KEY statement before each action.
Rotating Objects - Absolute Rotation

It is possible to rotate an object about one of its own local axes. FIG-31.18
emphasises the difference between the main (or world) axes and local axes.
FIG-31.18
y-axis
y-axis
An Object's Local Axes
z-axis z-axis
World Axes The Cube’s Local Axes
x-axis x-axis
Possible rotations are shown in FIG-31.19.

FIG-31.19 local z-axis
Possible Rotations about

Local Axes Rotation about
the z-axis
local x-axis
Rotation about Rotation about
the x-axis the y-axis
local y-axis
The XROTATE OBJECT Statement

This command causes an identified object to rotate to a specific angle about the
object’s local x-axis. Rotation is towards the viewer. The statement has the format
shown in FIG-31.20.
FIG-31.20
The XROTATE
XROTATE OBJECT objno , angle
OBJECT Statement
In the diagram:
objno is the integer value specifying the object.
angle is a real number giving the angle (in degrees) to

which the object is to be rotated.
For example, a cube (object 1) can be rotated about the x-axis to 45o using the line
XROTATE OBJECT 1, 45.0
FIG-31.21 shows the cube before and after rotation.
FIG-31.21
The Effect of the
XROTATE OBJECT
Statement
o o
Cube 0 rotation Cube at 45 rotation
Activity 31.7
Write a program (object3D04.dbpro) which implements the following logic:
Set screen resolution to 1280 by 1024
Create a cube (40 units in size)
Move the cube to (0,0,100)
FOR degree := 1 TO 360 DO

o
Rotate cube to degree
around the x-axis.
Wait 1 millisecond
ENDFOR
Run the program and check that it performs as expected.
Modify the program so that the cube revolves in the opposite direction about
the x-axis.

The YROTATE OBJECT Statement
To rotate an object about its local y-axis we use the YROTATE OBJECT statement
which has the format shown in FIG-31.22.
FIG-31.22
The YROTATE OBJECT YROTATE OBJECT objno , angle
Statement
In the diagram:
angle is a real number giving the angle (in degrees) to which the
object is to be rotated.
For example, a cube (ID 1) could be rotated about its y-axis to 60o using the line:
YROTATE OBJECT 1, 60.0
FIG-31.23 shows the cube after a rotation to 60o.
FIG-31.23
The Effect of the
YROTATE OBJECT
Statement
The ZROTATE OBJECT Statement

To rotate an object about its z-axis, we use the ZROTATE OBJECT statement which
has the format shown in FIG-31.24.
FIG-31.24
The ZROTATE OBJECT ZROTATE OBJECT objno , angle
Statement
In the diagram:
For example, a cube (ID 1) could be rotated about the z-axis to 110o using the line:
ZROTATE OBJECT 1, 110.0
FIG-31.25 shows the cube after a rotation to 110o.
FIG-31.25
The Effect of the
ZROTATE OBJECT
Statement

The program in LISTING-31.3 revolves a cube about all three axes at the same
time.
LISTING-31.3 REM ** Set display mode ***

Rotating an Object about
all Three Axes REM *** Create and position cube ***
POSITION OBJECT 1, 0, 0, 200
REM *** Rotate cube 1 degree at a time ***

REM *** around all three axes ***
FOR angle = 1 TO 360
XROTATE OBJECT 1, angle
YROTATE OBJECT 1, angle
ZROTATE OBJECT 1, angle
WAIT 10
NEXT angle

WAIT KEY
END
Activity 31.8
Type in and test the program given above (object3D05.dbpro).
The ROTATE OBJECT Statement

Rather than use three separate statements to rotate an object about all three axes,
the same effect can be achieved using the ROTATE OBJECT statement. This
statement takes three values specifying, for each axis, the degree of rotation. The
format of the statement is shown in FIG-31.26.
FIG-31.26
The ROTATE ROTATE OBJECT objno , xangle , yangle , zangle
OBJECT Statement
In the diagram:
xangle is a real number giving the angle (in degrees) to which the
object is to be rotated about its x-axis.
yangle is a real number giving the angle (in degrees) to which the
object is to be rotated about its y-axis.
zangle is a real number giving the angle (in degrees) to which the
object is to be rotated about its z-axis.
Activity 31.9
Rewrite the program you created in the previous Activity, replacing the
XROTATE, YROTATE and ZROTATE statements with a single ROTATE
OBJECT statement, producing the same effect as before.
All statements in the previous section rotate an object to a specific angle, irrespective

of that object’s current inclination. For example, if we use the YROTATE OBJECT
statement to turn a cube to 60o, the initial angle of the cube before the statement is
executed is irrelevant since the end result will be that the cube will end up at the
specified angle of 60o. This type of rotation is known as absolute rotation.
The SET OBJECT ROTATION Statement

When an object is rotated about all three axes at the same time, the action is normally
implemented by first rotating the object about the x-axis, then the y-axis and finally
the z-axis. Of course, it's all done so quickly that the operation will appear to be
instantaneous. However, should we want to reverse the order in which the rotations
take place (i.e. z-axis first, x-axis last) then we can use the SET OBJECT
ROTATION statement. Once set, the order in which the axes are handled will
remain on this new setting unless you revert to normal using a second option of the
SET OBJECT ROTATION statement, which has the format shown in FIG-31.27.
FIG-31.27
The SET OBJECT
ROTATION ZYX
Statement
SET OBJECT ROTATION
{ { XYZ
ZYX
objno
In the diagram:
XYZ Use this option to return the order of rotations to

the default x-axis, y-axis, z-axis order.
ZYX Use this option to set the order of rotations to

the z-axis, y-axis, x-axis order.
objno is an integer value specifying the object whose

order of rotation is to be modified.
Rotating Objects - Relative Rotation

It is also possible to make an object rotate by a specific angle from its current setting.
For example, if a cube has already been rotated 45o about the local y-axis, we can
command it to be rotated by a further 60o giving a final rotation position of 105o.
This type of rotation - where the angle specified is added to the initial tilt - is known
as relative rotation.
When using relative rotation, different terms are used for rotation about each axis.
Hence, we use the term PITCH for rotation about the x-axis, TURN for rotation
about the y-axis and ROLL for rotation about the z-axis (see FIG-31.28).
FIG-31.28
DOWN local z-axis
Relative Rotation Terms
RIGHT LEFT
RIGHT LEFT
local x-axis
PITCH TURN ROLL
UP
local y-axis

The PITCH OBJECT Statement
We can tilt an object upwards (i.e. rotate it in a positive direction about the x-axis)
using the PITCH OBJECT UP statement which has the format shown in FIG-31.29.
FIG-31.29
The PITCH OBJECT
Statement
PITCH OBJECT
{ { DOWN
UP
objno , angle
In the diagram:
DOWN, UP Choose DOWN to make the object rotate clockwise (as

viewed from the positive side of the y-axis); choose UP to
make the object rotate anticlockwise.
object is to be rotated relative to its current position.
The angle can be a positive or negative value.
The program in LISTING-31.4 performs the same function as the one you created
in Activity 31.7 where you used the XROTATE statement to rotate a cube through
360o. However, this time the XROTATE statement has been replaced by a PITCH
OBJECT UP command.
LISTING-31.4
REM *** Set screen resolution ***
Using Relative Rotation SET DISPLAY MODE 1280,1024,32
REM *** Create and position cube ***
REM *** Revolve the cube ***

FOR c = 1 TO 360
PITCH OBJECT UP 1, 1.0
WAIT 10
NEXT c
WAIT KEY
END
Activity 31.10
Type in the program given above (object3D06.dbpro) and make sure it is

equivalent to the earlier program in Activity 31.7.
Modify the program so that the cube rotates in the opposite direction.
The TURN OBJECT Statement

This statement allows relative rotation about the y-axis and has the format shown
in FIG-31.30.
{ {
FIG-31.30
LEFT
The TURN OBJECT TURN OBJECT objno , angle
Statement RIGHT

In the diagram:
LEFT, RIGHT Choose RIGHT to make the object rotate to the

right about the y-axis; choose LEFT to make the
object rotate to the left.
objno is an integer value giving the ID of the object to

be rotated.

which the object is to be rotated relative to its
current position. The angle can be a positive or
negative value.
Activity 31.11
Modify your previous program so that the cube rotates to the right about the
y-axis.
The ROLL OBJECT Statement

Relative rotation about the z-axis is achieved using the ROLL OBJECT statement
FIG-31.31
The ROLL OBJECT
Statement ROLL OBJECT
{ { LEFT
RIGHT
objno , angle
In the diagram:
LEFT, RIGHT Choose RIGHT to make the object rotate to the

right about the z-axis; choose LEFT to make the
object rotate to the left.

be rotated.

which the object is to be rotated relative to its
current position. The angle can be a positive or
negative value.
Activity 31.12
Modify your previous program so that the cube rotates to the left about the
z-axis.
The POINT OBJECT Statement

The main polygon of a 3D object is directed towards the player's viewpoint when
it is created. This polygon can be rotated to face any point in space using the POINT
OBJECT statement. This statement has the format shown in FIG-31.32.

FIG-31.32
The POINT OBJECT
POINT OBJECT objno , x , y , z
Statement
In the diagram:

be affected.
x,y,z are the coordinates of the point in space at which

the main polygon of the 3D object is to face.
In the program shown in LISTING-31.5 a cube is made to face the point (45,45,0)
using the statement:
POINT OBJECT 1,45,45,0
LISTING-31.5
REM *** Set display resolution and backdrop ***
Using the POINT
COLOR BACKDROP 0
OBJECT Statement
BACKDROP ON
REM *** Make the set of objects ***
MAKE OBJECT CUBE 1,40
REM *** Move cube to (0,0,100) after key press ***
WAIT KEY
REM *** point cube at (45,45,0)***
WAIT KEY
WAIT KEY
END
The result is shown in FIG-31.33.

FIG-31.33
Turning an Object to Face
a Specified Point
Initially, the cube’s main polygon faces the After the POINT OBJECT statement is
viewer. executed, the main polygon faces (45,45,0).
Activity 31.13
Type in and test the program given in LISTING-31.5 (object3D07.dbpro).
Modify the program to make the cube face the point (-20,17,-10).
The MOVE OBJECT distance Statement

We've already encountered a MOVE OBJECT statement which allows an object to
be moved up, down, left, or right, but a second version of MOVE OBJECT exists
which will move an object in the direction its main polygon is facing. This statement

FIG-31.34
The MOVE OBJECT
MOVE OBJECT objno , dist
distance Statement
In the diagram:
objno is an integer value specifying the object to be

moved.
dist is a real value specifying the distance to be

moved.
Activity 31.14
In your previous program, immediately after the POINT OBJECT statement,

add the following lines:
REM *** Move cube ***

WAIT KEY
MOVE OBJECT 1, 20
Run the updated program.
The FIX OBJECT PIVOT Statement

When an object rotates, its local axes rotate with it. In FIG-31.35 we see a cube and
its local axes before and after it has been rotated to -90o about its local z-axis.
FIG-31.35 local x-axis
local z-axis local z-axis
How the Local Axes are
Affected When an
Object is Rotated
local x-axis
local y-axis
local y-axis
o
Local Axes - Initial Position Local Axes - Cube Rotated to -90
about the z-axis
If we now rotate the cube about its own x-axis, it will turn left-to-right rather than
up-and-over, because its x-axis has shifted position. This is demonstrated in
LISTING-31.6 where the cube is rotated a full 360o about its x-axis, rotated by -90o
about its z-axis and then rotated a full 360o about its x-axis for a second time.
LISTING-31.6 REM *** Set display resolution ***
Local Axes Move with
REM *** Make and position cube ***
the Object
REM *** Rotate cube 360 about x-axis ***

FOR degree = 0 TO 360
XROTATE OBJECT 1, degree
WAIT 1
NEXT degree
continued on next page

LISTING-31.6 REM *** Rotate cube to -90 about z-axis
(continued) FOR degree = 0 TO -90 STEP -1
ZROTATE OBJECT 1, degree
Local Axes Move with WAIT 10
the Object NEXT degree
REM *** Rotate cube 360 about x-axis ***

FOR degree = 0 TO 360
XROTATE OBJECT 1, degree
WAIT 1
NEXT degree

WAIT KEY
END
Activity 31.15
When an object has been rotated, it is possible to reset the local axes so that they
are parallel to the main axes. This is done using the FIX OBJECT PIVOT statement
whose format is shown in FIG-31.36.
FIG-31.36
The FIX OBJECT PIVOT FIX OBJECT PIVOT objno
Statement
In the diagram:

local axes are to be reset.
When this statement is executed, the object in question has its local axes reset so
that the x-axis lies left-to-right, the y-axis top-to-bottom, and the z-axis in-to-out
(see FIG-31.37).
local x-axis
FIG-31.37 local z-axis
local z-axis
The Effect of Using FIX

OBJECT PIVOT
local x-axis
local y-axis
local y-axis
Cube rotated -90 degrees Cube’s local Axes are Reset

about the z-axis
local y-axis
local y-axis local x-axis local z-axis
local z-axis
local x-axis
Cube rotated -45 degrees

about the z-axis Cube’s local Axes are Reset

Activity 31.16
In your previous program (object3D08.dbpro), add the line
FIX OBJECT PIVOT 1
before the final FOR loop structure.
How does this affect the rotation of the cube?
Modify the program again so that the cube is only rotated to -45o in the second
FOR loop.
How is the cube's rotation affected this time?
Resizing Objects
It is possible to change the size of an object after it has been created. You have the
option to resize one, two, or all three of the object’s dimensions. This allows you
to make an object uniformly larger or smaller, or to distort the original shape by
changing each dimension by differing amounts.
The SCALE OBJECT Statement

Resizing an existing 3D object is achieved using the SCALE OBJECT statement
FIG-31.38
The SCALE OBJECT SCALE OBJECT objno , xperc , yperc , zperc
Statement
In the diagram:
objno is the integer value previously assigned to the

3D object.
xperc is a real number giving the new size of the

object’s x dimension as a percentage of its
original size in that dimension. For example, a
value of 100.0 will retain the current size, while
200.0 would double the object’s length in the x
dimension, and 50.0 would halve it.
yperc is a real number giving the new size of the

object’s y dimension as a percentage of its
original size in that dimension.
zperc is a real number giving the new size of the

object’s z dimension as a percentage of its
The program in LISTING-31.7 creates a sphere with a radius of 20 units. The sphere
is then resized so that the x dimension is doubled and the z dimension reduced to
10 units. The new shape is then rotated about the y-axis. The user must press ESC
to terminate the program.

Resizing an Object
REM *** Make and position the sphere **
MAKE OBJECT SPHERE 1, 20
REM *** Resize sphere ***

SCALE OBJECT 1, 200.0,100.0,50.0
REM *** Rotate shape ***

DO
TURN OBJECT RIGHT 1, 1.0
LOOP

WAIT KEY
END
Activity 31.17
Type in and test the program in LISTING-31.6 (object3D09.dbpro).
Modify the program so that a cone is used in place of the sphere.
Showing and Hiding Objects

Any 3D object is immediately visible from the moment it is created (assuming it’s
within view), but it is possible to hide an object using the HIDE OBJECT statement,
making it reappear later using the SHOW OBJECT command.
The HIDE OBJECT Statement

An object can be made invisible using the HIDE OBJECT statement which has the
format shown in FIG-31.39.
FIG-31.39
HIDE OBJECT objno
The HIDE OBJECT
Statement
In the diagram:

3D object which is to be hidden.
The SHOW OBJECT Statement

An object which has been previously hidden can be made to reappear using the
SHOW OBJECT statement which has the format shown in FIG-31.40.
FIG-31.40
The SHOW OBJECT SHOW OBJECT objno
Statement
In the diagram:

hidden 3D object which is to reappear.

The program in LISTING-31.8 rotates a cube continually, hiding the cube when ‘h’
is pressed and showing it again when ‘s’ is pressed.

Hiding and Showing
Objects REM *** Make and position cube ***
REM *** Rotate object ***

DO
PITCH OBJECT UP 1, 1.0
REM *** Read key ***
ch$ = INKEY$()
REM *** IF its s - show cube ***
IF ch$ = “s”
SHOW OBJECT 1
ENDIF
REM *** IF its h - hide cube ***
IF ch$ = “h”
HIDE OBJECT 1
ENDIF
LOOP
END
Activity 31.18
Type in and test the program given above (object3D10.dbpro).
The DELETE OBJECT Statement

When a 3D object is no longer required, its RAM space can be released using the
DELETE OBJECT statement which has the format shown in FIG-31.41.
FIG-31.41
DELETE OBJECT objno
The DELETE OBJECT
Statement
In the diagram:
objno is an integer value specifying the ID of the 3D

object to be deleted.
The DELETE OBJECTS Statement

If we need to delete several objects at one time, then the most efficient way to do
this is to use the DELETE OBJECTS statement which has the format shown in
FIG-31.42.
FIG-31.42
The DELETE
DELETE OBJECTS objno1 , objno2
OBJECTS Statement
In the diagram:
objno1 is an integer value specifying the lowest ID of the

3D objects to be deleted.
objno2 is an integer value specifying the highest ID of the
3D objects to be deleted.

For example, if we needed to delete 10 3D objects with ID values ranging from 8
to 17, then we would use the statement:
DELETE OBJECTS 8,17
Copying a 3D Object
We can create a copy of an existing 3D object in one of two ways, as described
below.
The CLONE OBJECT Statement

The CLONE OBJECT statement creates an independent copy of an existing 3D
object. The statement has the format shown in FIG-31.43.
FIG-31.43
The CLONE OBJECT CLONE OBJECT objno1 , objno2
Statement
In the diagram:
objno1 is an integer value specifying the ID to be

assigned to the object being created.
objno2 is an integer value specifying the ID of the

existing object to be copied.
The program in LISTING-31.9 uses the CLONE OBJECT statement to create a

duplicate cube.

Creating a Copy of a REM *** Make and position cube ***
3D Object MAKE OBJECT CUBE 1,40
POSITION OBJECT 1,-40,0,100
REM *** Rotate cube ***
WAIT KEY
XROTATE OBJECT 1 , -45
REM *** Copy cube ***
WAIT KEY
CLONE OBJECT 2,1
REM *** Position copy ***
REM *** Delete original cube ***
WAIT KEY
DELETE OBJECT 1
WAIT KEY
END
Activity 31.19
Is the copied cube rotated to the same angle as the original?
Is the copied cube deleted when the original cube is removed?

The INSTANCE OBJECT Statement
A second way to duplicate an existing object is to use the INSTANCE OBJECT
statement. This statement has the format shown in FIG-31.44.
FIG-31.44
The INSTANCE
INSTANCE OBJECT objno1 , objno2
OBJECT Statement
In the diagram:
objno1 is an integer value specifying the ID to be

objno2 is an integer value specifying the ID of the

existing object to be copied.
Although this may seem to have the same affect as the CLONE OBJECT statement,
in fact the two statements differ in how data about the copied object is held. When
CLONE OBJECT is used, the new object has its own independent data area; with
INSTANCE OBJECT the two objects share parts of the same data area. The
consequence of this is that objects created using INSTANCE OBJECT will
disappear if the original object from which they were created is deleted.
Activity 31.20
Modify your last program, replacing the CLONE OBJECT statement with a
INSTANCE OBJECT statement.
How does this change affect the operation of the program?
Change the DELETE OBJECT statement so that object 2, rather than object 1,
is deleted. How does this affect the program?
Retrieving Data on 3D Objects

The OBJECT EXIST Statement
We can check that an object of a specified ID actually exists using the OBJECT
EXIST statement which has the format shown in FIG-31.45.
FIG-31.45
The OBJECT EXIST OBJECT EXIST ( objno )
Statement
integer
In the diagram:
objno is an integer specifying the ID of the object to be

checked.
If the object exists, 1 is returned, otherwise zero is returned.

The OBJECT POSITION Statement
The exact position of a 3D object's centre can be determined using the OBJECT
POSITION statement. Three variations of the statement exist, with each variation
returning one of the object's ordinates. The statement has the format shown in
FIG-31.46.
FIG-31.46
{
The OBJECT POSITION X
Statement
OBJECT POSITION Y ( objno )
real
In the diagram:
X,Y,Z One of these options should be chosen. Choose X

if the x-ordinate of the specified object is required,
Y for the y-ordinate, and Z for the z-ordinate.

ordinate is to be returned.
For example, we could determine the position in space of object 1's centre using the
lines:
x = OBJECT POSITION X(1)

y = OBJECT POSITION Y(1)
z = OBJECT POSITION Z(1)
PRINT "Object 1 has its centre at (",x,",",y,",",z,")"
The OBJECT VISIBLE Statement

The OBJECT VISIBLE statement returns 1 if a specified 3D object is currently,
visible; if the object is hidden, the value zero is returned. The statement has the
FIG-31.47
OBJECT VISIBLE ( objno )
The OBJECT VISIBLE
Statement
integer
In the diagram:
objno is an integer value specifying the ID of the

object to be checked.
If the object is currently showing, 1 is returned, otherwise zero is returned.

The OBJECT SIZE Statement
The dimensions of a specified object can be determined using the OBJECT SIZE
statement. Like OBJECT POSITION, there are three variations available in this
statement, each returning one dimension of the object in question. The format for
this statement is given in FIG-31.48.
FIG-31.48
{
X
The OBJECT SIZE
Statement
OBJECT SIZE Y ( objno )
real
In the diagram:

if the width of the specified object is required,
Y for the height, and Z for the depth.
All three options can be omitted and the

statement will return an overall value for the size
of the 3D object.

dimension is to be returned.
The value returned by the statement is real and, because of rounding errors, this may
be slightly out. For example, if we create a cube (object 1) 40 units in all directions,
then the statement
PRINT "Width ", OBJECT SIZE X(1)
will display the value 39.9999961853.
Also, the OBJECT SIZE (1) statement - with no reference to any specific dimension
- gives an overall size based on all three dimensions.
Activity 31.21
Write a program (object3D12.dbpro) which creates a box of random size

(using limits 5 to 50) and then displays the box's width, height and depth.
The OBJECT ANGLE Statement

The angle to which an object has been rotated about any of its local axes can be
determined using the OBJECT ANGLE statement which has the format shown in
FIG-31.49.

FIG-31.49
X
The OBJECT ANGLE
Statement
OBJECT ANGLE Y ( objno )
real
In the diagram:

if the rotation about the local x axis is required,
Y for rotation about the y-axis, and Z for the
rotation about the z-axis.

rotation angle is to be returned.
Activity 31.22
Modify your previous program so that the box object is rotated by a random
number of degrees about all three axes. Display the amount of rotation in each
case.
Controlling an Object's Rotation Using the Mouse

In the example that follows, we're going to make a cube face towards the mouse
pointer. As the user moves the mouse pointer about the screen, so the cube will
continually re-orientate itself to face the pointer.
Before looking at the code, we have one main obstacle to overcome. The mouse
pointer commands (MOUSE X() and MOUSE Y() ) use a 2D coordinate system
with the origin at the top left corner of the screen; 3D objects use a coordinate system
in which the origin is (initially, at least) at the centre of the screen. To convert the
mouse's x ordinate readings to 3D space we need to use the line:
x3D = MOUSE X() - SCREEN WIDTH()/2
The y ordinate also needs to have it's sign changed, since for the mouse the positive
section of the y-axis is down, while in 3D space the positive section of the y-axis
is up! We can solve this with the line:
y3D = -(MOUSE Y() - SCREEN HEIGHT()/2)
Of course, there is no third dimension as far as the mouse pointer is concerned, so

we'll keep that set to zero.
We're now ready to describe the logic required by the program:
Create cube
Move cube backwards to reduce its apparent size
DO

Get mouse coordinates and convert to 3D space
Make cube point at these coordinates
LOOP
The code for the program is given in LISTING-31.10.

Making a 3D Object Face REM *** Make and position cube ***
the Mouse Pointer MAKE OBJECT CUBE 1, 40
DO
REM *** Convert mouse 2D coords to 3D ***
x3D = MOUSEX() - SCREEN WIDTH()/2
y3D = -(MOUSEY() - SCREEN HEIGHT()/2)
REM *** Re-orient cube ***
POINT OBJECT 1,x3D,y3D,0
LOOP
END
Activity 31.23
Type in and test the code given above (object3D13.dbpro).
Modify the program to use two spheres, set side-by-side, both of which should
face towards the mouse pointer.
Wireframe and Culling

The SET OBJECT WIREFRAME Statement
It is possible to show a 3D object in wireframe mode (which show only the edges
of the polygons that make up a shape) using the SET OBJECT WIREFRAME
statement which has the format shown in FIG-31.50.
FIG-31.50
The SET OBJECT
SET OBJECT WIREFRAME objno , mode
WIREFRAME Statement
In the diagram:
objno is an integer value identifying the object which

is to have its display mode altered.
mode is 0 or 1.
0 - solid mode
1 - wireframe mode
LISTING-31.11 demonstrates the use of this statement, switching between solid

and wireframe mode every time a key is pressed.
LISTING-31.11
REM *** Set screen mode ***
Switching Between SET DISPLAY MODE 1280,1024,32
Normal and Wireframe REM *** Make and position cube ***
Mode MAKE OBJECT CUBE 1, 40
REM *** Start in solid mode ***
wire = 0 continued on next page

LISTING-31.11 REM *** Rotate cube ***
(continued) DO
REM *** IF key pressed, switch mode ***
Switching Between IF INKEY$() <> ""
Normal and Wireframe wire = 1 - wire
Mode SET OBJECT WIREFRAME 1, wire
ENDIF
PITCH OBJECT DOWN 1, 1.0
TURN OBJECT LEFT 1, 1.0
LOOP
END
Activity 31.24
The SET OBJECT CULL Statement

Under normal circumstances, it is impossible to see every part of a 3D object at the
same time. The polygons that make up the hidden parts of an object are not drawn
by the computer. Obviously, this saves processing time and creates a more realistic
effect. If we take a second look at the previous program when running in wireframe
mode, we'll see that the polygons at the back of the cube are not drawn. This
elimination of hidden polygons is known as culling. Culling can be toggled on or
off using the SET OBJECT CULL statement which has the format shown in
FIG-31.51.
FIG-31.51
The SET OBJECT SET OBJECT CULL objno , mode
CULL Statement
In the diagram:
objno is an integer value identifying the object which

is to have its cull mode altered.
mode is 0 or 1.
0 - culling off
1 - culling on
Activity 31.25
Modify your last program by adding the line
SET OBJECT CULL 1,0
immediately before the DO..LOOP structure. Notice that the hidden polygons
are now being drawn.
Modify the program again so that pressing the w key toggles between
wireframe and solid mode and that pressing c toggles between culling on and
culling off.
There's a slight problem when it comes to displaying cylinders and cones, as we can
see from the output produced by LISTING-31.12.

LISTING-31.12 REM *** Set screen mode ***
A Problem with Cones REM *** Make and position cone and cylinder ***
and Cylinders MAKE OBJECT CONE 1, 5
POSITION OBJECT 1,-6,-5,0
XROTATE OBJECT 1, 45
MAKE OBJECT CYLINDER 2,5
POSITION OBJECT 2, 6,-5,0
WAIT KEY
END
Activity 31.26
What problem occurs with both shapes?
The inside surfaces of both shapes have not been drawn (see FIG-31.52).
FIG-31.52
Hidden Surfaces on
Cones and Cylinders
The inside surfaces

of both shapes are not
drawn
But we can solve this problem by switching off culling, so that the hidden polygons
are drawn.
Activity 31.27
Modify your previous program so that culling is switched off for both the cone
and the cylinder.
Storage Methods
When a 3D object is shown on screen, the coordinates of its vertices are stored in
memory in an area known as a vertex buffer. Normally, each object will have its
own vertex buffer. However, some video cards allow different objects to share
vertex buffers, other video cards don't. As a default, 3D objects in DARKBASIC
Pro do not share vertex buffers - this ensures compatibility with the maximum
number of video cards. However, it is possible to force vertex buffer sharing and,
if your video card can handle this, improve the performance of your program.
The SET GLOBAL OBJECT CREATION Statement

To force vertex buffer sharing, we use the SET GLOBAL OBJECT CREATION

FIG-31.53
SET GLOBAL OBJECT CREATION mode
The SET GLOBAL
OBJECT CREATION
Statement
In the diagram:
mode 0 or 1. 0 - no vertex buffer sharing (this is the

default setting); 1 - vertex buffer sharing allowed.
Even if your own video card does allow vertex buffer sharing, your customer's may
not, so it's probably best to ignore this option.
Summary
l A point in 3D space is specified using x, y, and z coordinates.
l There are three main planes in 3D space - XY, YZ, and XZ.
l The computer screen uses a positive-up, negative-down y-axis when operating

in 3D mode. This is the opposite from the 2D settings.
l The positive z-axis travels away from the viewer "into" the screen.
l A point in space is known as a vertex.
l A set of vertices, when joined, form a polygon.
l The join between two vertices is known as an edge.
l Basic 3D shapes are known as primitives.
l When first created a 3D object has it centre position at the origin.
l The x and y axes intersect at the centre of the screen at start-up.
l Use the MAKE OBJECT CUBE to create a cube.
l Use MAKE OBJECT BOX to create a cuboid.
l Use MAKE OBJECT SPHERE to create a sphere.
l The number of polygons used can be specified when creating a sphere.
l Use MAKE OBJECT CYLINDER to create a cylinder.
l Use MAKE OBJECT CONE to create a cone.
l Use MAKE OBJECT PLAIN to create a plane.
l A plane is initially oriented as an XY plane.
l Use MAKE OBJECT TRIANGLE to create a triangle.
l The initial position of a triangle is determined by the vertices given.
l Use POSITION OBJECT to place the centre of an object at a new location.

l Use MOVE OBJECT to move an object along the x or y axis.
l Every 3D object has its own local axes with the origin at the centre of the object.
l Use the XROTATE, YROTATE or ZROTATE OBJECT statements to rotate an

object to a specific angle about one of its local axes.
l Use ROTATE OBJECT to rotate an object to specific angles about all three local
axes at the same time.
l Use PITCH, TURN or ROLL OBJECT statements to rotate an object by a

number of degrees around a given axis.
l Use POINT OBJECT to make an object face towards a specified point.
l Use MOVE OBJECT distance to move the object a specified number of units in
the direction in which an object is pointing.
l Use FIX OBJECT PIVOT to reset an object's local axes to be in line with the
global axes.
l Use SCALE OBJECT to change the dimensions of an object.
l Use HIDE OBJECT to make an object invisible.
l Use SHOW OBJECT to make an invisible object reappear.
l Use DELETE OBJECT to erase an object from RAM.
l Use DELETE OBJECTS to erase a group og objects from RAM.
l Use CLONE OBJECT to make an independent copy of an existing object.
l Use INSTANCE OBJECT to create a dependent copy of an existing object.
l Use OBJECT EXIST to check if a specified object exists.
l Use OBJECT VISIBLE to check if a specified object is visible.
l Use OBJECT POSITION to determine the position in space of an object's centre.
l Use OBJECT SIZE to determine the dimensions of a specified object.
l Use OBJECT ANGLE to determine the current angle of rotation a specified

object has about its local axes.
l Use SET OBJECT WIREFRAME to display a 3D object in wireframe or normal

mode.
l Use SET OBJECT CULL to toggle culling for a specified 3D object.
l Use SET GLOBAL OBJECT CREATION to enable/disable vertex buffer

sharing.

Merging Primitives
Introduction
DarkBASIC Pro version 1.058 introduced statements which allow us to create new
shapes by merging two primitives. There are three basic options available:
Create the new shape from the combination of the two original shapes - known as
union (see FIG-31.54).
FIG-31.54
Shape Union
union
Shape 1 Shape 2 Resulting Shape
Create the new shape by removing the overlapping section of shape 2 from shape
1 - known as difference (see FIG-31.55).
FIG-31.55
Shape Difference
difference
Create the new shape from the overlapping area between shape 1 and shape 2 -
known as intersection (see FIG-56).
FIG-31.56
Shape Intersection
intersection
The Statements
When we join two shapes using the merge statements, the resulting shape is stored
in a format knows as Constructive Solid Geometry (CSG). We need not concern
ourselves with the details of this format, and it is only mentioned here so that you
know the meaning of the initials used in the statements.
The PERFORM CSG UNION Statement

We can create a new shape from the union of two existing shapes using the
PERFORM CSG UNION statement which has the format shown in FIG-31.57.

FIG-31.57
The PERFORM CSG
PERFORM CSG UNION objno1 , objno2
UNION Statement
In the diagram:
objno1 is an integer value specifying the ID of the first

3D object to be used in the union.
objno2 is an integer value specifying the ID of the second

3D object to be used in the union.
The statement modifies the shape of objno1 without affecting objno2. Normally,
the programmer would delete the second object once the union is completed.
The program in LISTING-31.13 demonstrates the union of a cube and a box. After
the box has been deleted, the resulting shape is then rotated about its local x and y
axes.
REM *** Set screen resolution and background ***

LISTING-31.13 SET DISPLAY MODE 1280,1024,32
COLOR BACKDROP 0
Creating a New 3D Shape
BACKDROP ON
using Union
REM *** Position camera ***
AUTOCAM OFF
This program positions POSITION CAMERA 0,0,-100
the camera which is
responsible for the picture REM *** Create two shapes used ***
we see on the screen. MAKE OBJECT CUBE 1,40
Full details of camera MAKE OBJECT BOX 2,10,30,10
usage are covered in POSITION OBJECT 2,0,15,0
Chapter 33.
REM *** Let viewer see position of shapes ***
WAIT KEY
REM *** Union shapes ***

PERFORM CSG UNION 1,2
REM *** Remove object 2 ***

DELETE OBJECT 2
REM *** Rotate new shape ***

DO
TURN OBJECT LEFT 1,1.0
PITCH OBJECT UP 1,1.0
LOOP

END
Activity 31.28
Try changing the second object to a sphere of diameter 10 and change the
POSITION OBJECT statement to read

As you've just discovered, the UNION statement only works predictably with cubes
and boxes. Other shapes give unpredictable results (although the cone is close).
The PERFORM CSG DIFFERENCE Statement

The PERFORM CSG DIFFERENCE statement removes from object 1 the volume
it shares in common with object 2. The statement has the format shown in
FIG-31.58.
FIG-31.58
The PERFORM CSG PERFORM CSG DIFFERENCE objno1 , objno2
DIFFERENCE Statement
In the diagram:

3D object to be used in the difference operation.

3D object to be used in the difference operation.
As before, the second object is unaffected by the operation and would normally be
deleted. Also, we are again restricted to cubes and boxes if we are to obtain
consistent results.
Activity 31.29
Restore your last project to its original code (as shown in LISTING-31.13).
Change the union operation to a difference operation and observe the new
shape created.
Modify the width and depth of the box to be 35.
The PERFORM CSG INTERSECTION Statement

The PERFORM CSG INTERSECTION statement does not perform intersection as
defined at the beginning of this chapter, but it does create a different shape from
that produced by the PERFORM CSG DIFFERENCE statement and so is worth
looking at. The statement has the format shown in FIG-31.59.
FIG-31.59
The PERFORM CSG
PERFORM CSG INTERSECTION objno1 , objno2
INTERSECTION
Statement
In the diagram:

3D object to be used in the intersection operation.

3D object to be used in the intersection operation.
Again, the second object would normally be deleted and we are restricted to cubes
and boxes if we are to obtain predictable results.

Activity 31.30
Modify your previous program and determine how the shape created by the
PERFORM CSG INTERSECTION differs from that produced by PERFORM
CSG DIFFERENCE.
Summary
l Cubes and boxes can be merged to create new shapes.
l Use PERFORM CSG UNION to modify a shape so that it becomes the

amalgamation of the original two shapes.
l Use PERFORM CSG DIFFERENCE to modify a shape so that the volume it

shares with a second shape is removed.
l Use PERFORM CSG INTERSECTION to modify a shape to remove polygons

which touch the second shape.

Solutions
Activity 31.1 Activity 31.5
No solution required. REM *** Set display & backdrop ***
COLOR BACKDROP 0
Activity 31.2 BACKDROP ON
REM *** Make the set of objects ***
1. MAKE OBJECT CUBE 1,4
REM *** Set display and backdrop *** MAKE OBJECT BOX 2,10,15,5
SET DISPLAY MODE 1280,1024,32 MAKE OBJECT SPHERE 3,7
COLOR BACKDROP 0 MAKE OBJECT CYLINDER 4,12
BACKDROP ON MAKE OBJECT CONE 5,12
REM *** Make the sphere *** REM *** Cube to (9,0,0) after key
MAKE OBJECT SPHERE 1,10 press ***
REM *** End program *** WAIT KEY
WAIT KEY POSITION OBJECT 1,9,0,0
END REM *** Box to (-60,0,100) ***
WAIT KEY
2. POSITION OBJECT 2,-60,0,100
REM *** Set display and backdrop *** REM *** Sphere to (-30,-40,50) ***
SET DISPLAY MODE 1280,1024,32 WAIT KEY
COLOR BACKDROP 0 POSITION OBJECT 3,-30,-40,50
BACKDROP ON REM *** Cylinder to (25,0,120) ***
REM *** Make the sphere *** WAIT KEY
MAKE OBJECT SPHERE 1,10,40,40 POSITION OBJECT 4,25,0,120
REM *** End program *** REM *** Cone to (0,25,100) ***
WAIT KEY WAIT KEY
END POSITION OBJECT 5,0,25,100
WAIT KEY
Activity 31.3 END
1.
REM *** Set display and backdrop *** Activity 31.6
COLOR BACKDROP 0 REM *** Set display & backdrop ***
BACKDROP ON SET DISPLAY MODE 1280,1024,32
REM *** Make the cylinder *** COLOR BACKDROP 0
MAKE OBJECT CYLINDER 1,5 BACKDROP ON
REM *** End program *** REM *** Make the set of objects ***
WAIT KEY MAKE OBJECT CUBE 1,10
END REM *** Cube to(9,0,100) after key
press ***
2. WAIT KEY
REM *** Set display and backdrop *** POSITION OBJECT 1,9,0,100
SET DISPLAY MODE 1280,1024,32 REM *** Cube 31.3 units to the right
COLOR BACKDROP 0 ***
BACKDROP ON WAIT KEY
REM *** Make the cone *** MOVE OBJECT RIGHT 1, 31.3
MAKE OBJECT CONE 1,5 REM *** End program ***
WAIT KEY END
END
Activity 31.7
Activity 31.4
Version 1
REM *** Set display & backdrop *** REM *** Set display & backdrop
SET DISPLAY MODE 1280,1024,32 SET DISPLAY MODE 1280,1024,32
COLOR BACKDROP 0 COLOR BACKDROP 0
BACKDROP ON BACKDROP ON
REM *** Make the cube *** REM *** Make the set of objects ***
MAKE OBJECT CUBE 1, 10 MAKE OBJECT CUBE 1,40
REM *** Cube to (50,0,0) after key REM ***Cube to(0,0,100) after key
press *** press ***
POSITION OBJECT 1,9,0,0 REM *** Rotate cube ***
REM *** Move the cube backwards *** FOR degree = 1 TO 360
WAIT KEY XROTATE OBJECT 1, degree
POSITION OBJECT 1, 9,0,30 WAIT 1
REM *** End program *** NEXT degree
WAIT KEY REM *** End program ***
END WAIT KEY
END

Version 2 REM *** Set screen resolution ***
REM *** Set display & backdrop *** SET DISPLAY MODE 1280,1024,32
SET DISPLAY MODE 1280,1024,32 REM *** Create and position cube ***
COLOR BACKDROP 0 MAKE OBJECT CUBE 1, 40
BACKDROP ON POSITION OBJECT 1, 0, 0, 200
REM *** Make the set of objects *** REM *** Revolve the cube ***
MAKE OBJECT CUBE 1,40 FOR c = 1 TO 360
REM ***Cube to(0,0,100) after key press *** ROLL OBJECT LEFT 1, 1.0
POSITION OBJECT 1,0,0,100 WAIT 10
REM *** Rotate cube *** NEXT c
FOR degree = 359 TO 0 STEP -1 REM *** End program ***
XROTATE OBJECT 1, degree WAIT KEY
WAIT 1 END
NEXT degree
WAIT KEY Activity 31.13
END
REM *** Set display & backdrop ***
Activity 31.8 COLOR BACKDROP 0
BACKDROP ON
No solution required. REM *** Make the set of objects ***
REM ***Cube to(0,0,100) after key press ***
WAIT KEY
Activity 31.9 POSITION OBJECT 1,0,0,100
REM ** Set display mode *** REM *** point cube at (-20,17,-10)***
SET DISPLAY MODE 1280,1024,32 WAIT KEY
REM *** Create and position cube *** POINT OBJECT 1,-20,17,-10
MAKE OBJECT CUBE 1, 40 REM *** End program ***
POSITION OBJECT 1, 0, 0, 200 WAIT KEY
REM *** Rotate cube 1 degree at a time *** END
REM *** around all three axes ***
FOR angle = 1 TO 360
ROTATE OBJECT 1, angle,angle,angle Activity 31.14
WAIT 10
NEXT angle The changes should cause the cube should move in the
REM *** End program *** direction it is pointing.
WAIT KEY
END
Activity 31.15
Activity 31.10 No solution required.
SET DISPLAY MODE 1280,1024,32 Activity 31.16
MAKE OBJECT CUBE 1, 40 The cube rotates in the same direction on both occasions.
FOR c = 1 TO 360
PITCH OBJECT UP 1, -1.0
Activity 31.17
WAIT 10 REM *** Set display resolution ***
NEXT c SET DISPLAY MODE 1280,1024,32
REM *** End program *** REM *** Make and position the cone ***
WAIT KEY MAKE OBJECT CONE 1, 20
END POSITION OBJECT 1, 0,0,200
REM *** Resize cone ***
SCALE OBJECT 1, 200.0,100.0,50.0
Activity 31.11 REM *** Rotate shape ***
DO
REM *** Set screen resolution *** TURN OBJECT RIGHT 1, 1.0
SET DISPLAY MODE 1280,1024,32 LOOP
REM *** Create and position cube *** REM *** End program ***
MAKE OBJECT CUBE 1, 40 END
FOR c = 1 TO 360
TURN OBJECT RIGHT 1, 1.0
Activity 31.18
WAIT 10
No solution required.
NEXT c
WAIT KEY
END
Activity 31.19
The second cube is created with the same rotation.
Activity 31.12 The copied cube does not disappear when the original is
deleted.

Activity 31.20 POSITION OBJECT 2, -25,0,100
DO
x3D = MOUSEX() - SCREEN WIDTH()/2
The duplicated cube is not rotated, but it is removed when
y3D = -(MOUSEY() - SCREEN HEIGHT()/2)
the original cube is deleted. POINT OBJECT 1, x3D,y3D,0
POINT OBJECT 2, x3D,y3D,0
When the second cube is deleted, the first cube is LOOP
unaffected. REM *** End program ***
END
Activity 31.21
Activity 31.24
REM *** Set display resolution ***
SET DISPLAY MODE 1280,1024,32 No solution required.
REM *** Seed random number generator ***
RANDOMIZE TIMER()
REM *** Make and position the box *** Activity 31.25
MAKE OBJECT BOX 1, RND(45)+5,RND(45)+5,
ÄRND(45)+5 REM *** Set screen mode ***
POSITION OBJECT 1, 0,0,0 SET DISPLAY MODE 1280,1024,32
REM *** Get dimensions of box *** REM *** Make and position cube ***
width# = OBJECT SIZE X(1) MAKE OBJECT CUBE 1, 40
height# = OBJECT SIZE Y(1) POSITION OBJECT 1,25,0,100
depth# = OBJECT SIZE Z(1) REM *** Start in solid and culling on ***
REM *** Display dimensions *** wire = 0
DO cull = 1
SET CURSOR 10,20 REM *** Rotate cube ***
PRINT "Width : ",width#, " Height :", DO
Ä height#, " Depth : ",depth# REM *** IF key pressed, switch mode ***
LOOP IF INKEY$() ="w"
REM *** End program *** wire = 1 - wire
WAIT KEY SET OBJECT WIREFRAME 1, wire
END ENDIF
IF INKEY$()="c"
cull = 1 - cull
Activity 31.22 SET OBJECT CULL 1,cull
ENDIF
REM *** Set display resolution *** PITCH OBJECT DOWN 1, 1.0
SET DISPLAY MODE 1280,1024,32 TURN OBJECT LEFT 1, 1.0
REM *** Seed random number generator *** LOOP
RANDOMIZE TIMER() REM *** End program ***
REM *** Make and position the box *** END
MAKE OBJECT BOX 1, RND(45)+5,RND(45)+5,
ÄRND(45)+5
POSITION OBJECT 1, 0,0,0 Activity 31.26
REM *** Rotate box at random ***
ROTATE OBJECT 1,RND(359),RND(359),RND(359) The inside surfaces of both shapes are not drawn.
REM *** Get dimensions of box ***
width# = OBJECT SIZE X(1)
height# = OBJECT SIZE Y(1) Activity 31.27
depth# = OBJECT SIZE Z(1)
REM *** Get rotations of box *** REM *** Set screen mode ***
x_axis_rotation = OBJECT ANGLE X(1) SET DISPLAY MODE 1280,1024,32
y_axis_rotation = OBJECT ANGLE Y(1)
z_axis_rotation = OBJECT ANGLE Z(1) REM *** Make & position cone and cylinder ***
REM *** Display details *** MAKE OBJECT CONE 1, 5
DO POSITION OBJECT 1,-6,-5,0
SET CURSOR 10,20 XROTATE OBJECT 1, 45
PRINT "Width : ",width#, " Height : " MAKE OBJECT CYLINDER 2,5
Ä, height#, " Depth : ",depth# POSITION OBJECT 2, 6,-5,0
SET CURSOR 10,40
PRINT "X-axis : ",x_axis_rotation, REM ** Culling off for both objects ***
Ä" y-axis : ",y_axis_rotation, SET OBJECT CULL 1,0
Ä" z-axis : ", z_axis_rotation SET OBJECT CULL 2,0
LOOP
REM *** End program *** REM *** End program ***
WAIT KEY WAIT KEY
END END

REM *** Set display resolution *** In fact, the UNION operation only works predictably with
SET DISPLAY MODE 1280,1024,32 cubes and boxes.
REM *** Make and position spheres ***

Activity 31.29
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POSITION CAMERA 0,0,-100
REM *** Create two shapes used ***
MAKE OBJECT BOX 2,10,30,10
WAIT KEY
REM *** Difference shapes ***
PERFORM CSG DIFFERENCE 1,2
DELETE OBJECT 2
DO
LOOP
END
Activity 31.30
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
REM *** Create two shapes used ***
WAIT KEY
REM *** Intersection shapes ***
PERFORM CSG INTERSETION 1,2
DELETE OBJECT 2
DO
LOOP
END

32
Applying a Texture Image to a 3D Object
Colouring a 3D Object
Loading a Texture Image
Mipmaps
Offsetting a Texture
Overlaying Textures
Seamless Tiling
Semi-Transparent 3D Object
Sky Spheres
Texture Mapping Options
Texture Transparency
Tiling
Video Texturing
DarkBASIC Pro: Texturing 797

Adding Texture
Introduction
The 3D shapes we have created so far look rather bland in white and shades of grey.
To make things more interesting we can wrap an image around a 3D shape and
thereby enhance its visual impact.
This is known as adding texture to the 3D object. An example of a textured cube

is shown in FIG-32.1.
FIG-32.1
A Textured Cube
Loading a Texture Image

We need to start by loading the picture we intend to use to texture the 3D object
into an image object with a statement such as:
LOAD IMAGE "texture01.bmp",1
After this has been done, we can transfer the image to the surface of one or more
3D objects.
Using the Image as a Texture

The TEXTURE OBJECT Statement
A loaded image can become the texture of a 3D object by executing the TEXTURE
OBJECT statement which has the format shown in FIG-32.2.
FIG-32.2
The TEXTURE TEXTURE OBJECT objno , imgno
OBJECT Statement
In the diagram:
objno is an integer value specifying the object to which

the texture is to be applied.
imgno is an integer value specifying the image to be

used as the texture.
798 DarkBASIC Pro: Texturing

For example, we could apply image 1 to object 2 using the line
TEXTURE OBJECT 2,1
The program in LISTING-32.1 demonstrates how a cube can be textured with a

wood image to create the impression of a wooden crate.
Adding Texture to an
Object REM *** Load texture image ***
LOAD IMAGE "textureWood.jpg",1
REM *** Create cube ***

REM *** Add texture to cube ***

TEXTURE OBJECT 1,1
REM *** Position cube ***

REM *** Rotate cube continuously ***

DO
LOOP

END
Activity 32.1
Type in and test the program in LISTING-32.1 (texture01.dbpro).
Change the texture image to eyecol.bmp.
We can see quite clearly from the results of the last Activity that the image is applied
separately to each face of the cube. For other shapes, the image may be applied
differently.
Activity 32.2
Modify your last program so that eyecol.bmp is applied as a texture to a box,

cylinder, cone and sphere (any dimensions will do). Create a separate program
for each shape.
How often is the image repeated on each of the shapes?
Mipmaps
Texturing a 3D object can be quite time consuming. It may be easy enough to map
a 300 by 300 pixel image onto a flat surface which occupies exactly 300 by 300
pixels on the screen, but if the 3D object moves off into the distance, the computer
has to work much harder to map the same 300 by 300 image onto an object which
now occupies just 23 by 23 pixels on the screen.
To help with this problem DarkBASIC Pro creates more than one copy of any image
that is loaded, with each copy being exactly half the size of the last (see FIG-32.3).

FIG-32.3
An Image with Added
Mipmaps
As a textured 3D object becomes smaller on the screen, the version of the image
used to texture that object changes from the the largest to the smallest.
The program in LISTING-32.2 demonstrates this effect.

Mipmaps in Action
REM *** Load texture image ***
LOAD IMAGE "eyecol.bmp",1
REM *** Make cube ***

TEXTURE OBJECT 1,1
REM *** Move cube away from viewer ***

DO
POSITION OBJECT 1,25,0,OBJECT POSITION Z(1)+10
LOOP

END
Activity 32.3
The effect is a fairly subtle one. Look closely at the image on the cube as it
moves away from your viewpoint. You should see it become less distinct as it
gets smaller.
The LOAD IMAGE Statement Again

We met the LOAD IMAGE statement back in Chapter 20 when we created image
objects which were then loaded into sprite objects. But the LOAD IMAGE
statement has an expanded form which allows us to dictate whether mipmaps are
to be created or not. This version of LOAD IMAGE has the format shown in
FIG-32.4.
FIG-32.4
The LOAD IMAGE
LOAD IMAGE filename , imgno , tflag
Statement
In the diagram:
filename is a string specifying the name of the file to be

loaded.

imgno is an integer specifying the ID to be allocated to
the image object being created.
tflag is an integer value specifying how the image is

to be stored.
0 - mipmaps are created
1 - no mipmaps are created
2 - loads the image in as a cubemap
texture (see the chapter on shaders)
When an image is loaded without mipmaps, any object using that image as a texture
must continue to use the original image even when the 3D object is greatly reduced
in size on the screen.
Activity 32.4
Modify your last program so that no mipmaps are used. To do this change the
LOAD IMAGE line to read:
LOAD IMAGE "eyecol.bmp",1,1
How does the texture on the cube differ in this program from the earlier
version?
Tiling
In the next example we'll create the floor of a dungeon by texturing a plane using
a cobblestone image.
The program uses the following logic:
Load cobblestone image

Create large plane
Rotate plane to be horizontal
Texture plane using the cobblestone image
The program itself is given in LISTING-32.3.
Creating a Floor Texture REM *** Load texture image ***
LOAD IMAGE "stonetile3.jpg",1
REM *** Create large plane ***
MAKE OBJECT PLAIN 1,400,400
REM *** Rotate plane to horizontal ***
XROTATE OBJECT 1,90
REM *** Texture plane ***
TEXTURE OBJECT 1,1

WAIT KEY
END
Activity 32.5

The screen dump in FIG-32.5 highlights the problem with the floor - the image has
stretched over the whole plane giving a floor that contains only a few unrealistically
large blocks rather than hundreds of smaller ones.
FIG-32.5
Floor Texturing
One way to solve the problem would be to use an image which actually shows the
hundreds of blocks that we need to create a realistic floor. However, this may not
be possible and the image would certainly have to be large if the visuals are to look
convincing as a character moves over the floor.
A second option is to make the texture image repeat itself several times over the
surface of the plane. This, for rather obvious reasons, is known as tiling.
The SCALE OBJECT TEXTURE Statement

Any image employed as a texture uses a UV coordinate system with the top left
being point (0,0) and the bottom right (1,1) no matter what the actual size of the
image is (see FIG-32.6).
FIG-32.6 (0,0) (1,0)
The UV Coordinate
System used by a Texture
Image
(0,1) (1,1)
When mapped to a flat plane, the image spreads itself over the object with point
(0,0) of the image mapping to the top-left corner of the plane and point (1,1) to the
bottom-right corner (see FIG-32.7).
FIG-32.7 Texture Image

(0,0)
The Default Mapping of
an Image to a Plane
(1,1)
Plane

Using the SCALE OBJECT TEXTURE statement, we can adjust this mapping
making only a part of the image stretch over the whole object, or have the image
duplicate itself several times creating a tiled effect as shown in FIG-32.8.
FIG-32.8
The Effects of Using
SCALE OBJECT
TEXTURE Part of the Image Used The Image Duplicated to Create
as Texture a Tiled Effect
The SCALE OBJECT TEXTURE statement has the format shown in FIG-32.9.
FIG-32.9
The SCALE OBJECT SCALE OBJECT TEXTURE objno , Uscale , Vscale
TEXTURE Statement
In the diagram:

the texture scaling is to be applied.
Uscale is a real number specifying the multiplication

factor along the U axis. Values less than 1 will
result in only part of the image being used.
Values greater than 1 will result in duplication
of the image over the 3D object.
Vscale is a real number specifying the multiplication

factor along the V axis. Values less than 1 will
of the image over the 3D object.
The first example shown in FIG-32.8 was created using the line:
SCALE OBJECT TEXTURE 1, 0.5, 0.5
while the second example was produced using:
SCALE OBJECT TEXTURE 1, 5.0, 5.0
The program in LISTING-32.4 demonstrates the effect of texture scaling by

applying cobblestones texture repeated 10 times in each direction to the plane.
Changing the Texture's REM *** Load texture image ***
Scaling LOAD IMAGE "stonetile3.jpg",1
REM *** Create and position plain ***
MAKE OBJECT PLAIN 1, 400,400
XROTATE OBJECT 1, -90
REM *** Texture plain ***
TEXTURE OBJECT 1,1
REM *** Scale texture ***
SCALE OBJECT TEXTURE 1,10.0,10.0
WAIT KEY
END

Activity 32.6
Type in and test the program given in LISTING-32.4 (texture04.dbpro).
Modify the scaling factors to each of the following settings and observe the
results:
Uscale Vscale
5.0 5.0
2.0 2.0
0.5 0.5
5.0 1.0
1.0 5.0
Change the 3D object used in your program from a plane to a sphere and retry
each of the settings given above.
Scaling a texture image in this way affects the image itself, so there is no way to
return to the original image settings within a program.
Seamless Tiling
For tiling to be convincing, the ends of the repeating image must butt together
without too obvious a join.
If we start with a simple picture and use it as a tiled texture (as shown in FIG-32.10)
we get a disappointing effect in which the edge of each image tile is very obvious.
FIG-32.10
A Visible Join Between
Tiles
To avoid this, we need to modify the image using a paint package such as Paint
Shop Pro or Photoshop.
The stages involved are shown in FIG-32.11.

FIG-32.11
Creating an Image for The main image has been
toned down to emphasise
Seamless Tiling the selected area
Load the original image into the Select a section of the image Copy selected area, mirror it,
paint program. along the whole of the left edge. and move it to the right edge.
Of course, it is equally
possible to copy the
right edge area to the left
side and the top to the
bottom.
You should choose

whatever combinations
suit the image in Erase parts of the copy to make Repeat the process, copying The image can now be tiled
question. it merge with the main image. a bottom section to the top. seamlessly.
It takes a bit of practice to achieve good results when creating a texture image, but
the results can be worth it.
Even when an image is not tiled, we can still have a problem with seams. For
example, when the image eyecol.bmp is applied as a texture to a sphere, the join
between the left and right edges of the image is quite apparent at the back of the
sphere, while the top and bottom edges are squeezed into single points at the two
"poles".
Activity 32.7
Attempt to modify eyecol.bmp (creating a new file named seamlesseye.bmp)

to give a seamless effect when textured onto a sphere.
Write a short program (act3207.dbpro) which applies the new file to a sphere
and then rotates the sphere continuously about its local y-axis.
Video Texture
It is even possible to use a video clip as a surface texture. To do this we need to start
by loading up a video with an instruction such as:
LOAD ANIMATION "mv1.mpg",1
The PLAY ANIMATION TO IMAGE Statement

Now we need to transfer the video to an image object and this is done using the
PLAY ANIMATION TO IMAGE statement which has the format shown in
FIG-32.12.

FIG-32.12
The PLAY ANIMATION TO IMAGE Statement
PLAY ANIMATION TO IMAGE varea , imgno , x1 , y1 , x2 , y2
In the diagram:
varea is an integer value specifying the video that is to

be copied to an image area.
imgno is an integer specifying the image area to which

the video is to be copied.
x1,y1,x2,y2 are the coordinates for the top-left and bottom

right space which the video is to occupy.
The size of the play area (as set by x1, y1, x2, y2) affects the quality of the image
when it appears on the 3D object; use too small a set of values and the video will
be heavily pixellated; use too large a set of values and displaying the video will put
too great a load on the processor/video card and slow the whole thing down.
Once the video has been transferred to the image object, we can then use the image
to texture a 3D object in the usual manner. LISTING-32.5 demonstrates the effect
by placing a video on a rotating cube.

Using a Video as a
Texture REM *** Load video ***
LOAD ANIMATION "lion.mpg",1
REM *** Transfer video to image ***

PLAY ANIMATION TO IMAGE 1,1,0,0,200,200
REM *** Create cube and texture with video ***

TEXTURE OBJECT 1,1
REM *** Move cube away from viewer ***


DO
PITCH OBJECT DOWN 1,1.0
LOOP

END
Activity 32.8
Modify the program to use very low values for the bottom right corner of the
video (i.e. 10,10) and very high values (i.e. 1000,1000). What affect do these
changes have on the final result?

Other Texture Effects
The SET OBJECT TEXTURE Statement
An alternative way to achieve a seamless tiled texture is to use the SET OBJECT
TEXTURE statement that adjusts the way in which each copy of the basic image is
tiled onto the surface of a 3D object. The statement has the format shown in
FIG-32.13.
FIG-32.13
The SET OBJECT SET OBJECT TEXTURE objno , tmode , mflag
TEXTURE Statement
In the diagram:

texture is to be modified.
tmode is an integer value (1,2,3,4) which directly affects

how the tiled texture is applied to the object.
1 - normal tiling.
2 - images are mirrored/flipped so that
identical edges meet.
3 - The last pixel along each edge is
extended over the remainder of the
surface.
4 - The image appears only once. The
remainder of the surface is black.
mflag Determines if mipmapping is to be used.

0 - mipmapping used
1 - no mipmapping
In FIG-32.14 we see the effects of each possible value for tmode when applying
eyecol.bmp as a tiled (2 by 2) texture on a cube.
FIG-32.14
The Effects of Using
Different tmode Settings
set object texture 1,1,0 set object texture 1,2,0 set object texture 1,3,0 set object texture 1,4,0
The program in LISTING-32.6 shows a tile textured cube with the tmode value of
the SET OBJECT TEXTURE statement set to 2.

Using the SET OBJECT
TEXTURE Statement REM *** Load image ***
REM *** Create a plain ***


LISTING-32.6 REM *** Texture object ***
(continued) TEXTURE OBJECT 1,1
SCALE OBJECT TEXTURE 1,2,2
TEXTURE Statement REM *** Modify tile mapping ***
SET OBJECT TEXTURE 1,4,0

WAIT KEY
END
Activity 32.9
Try other settings for tmode and check the effects produced.
The SCROLL OBJECT TEXTURE Statement

The texture image can be mapped onto an object with a varying degree of offset
along either the U or V axes. The overall effect is to modify which part of the texture
image is placed at the top left corner of the object.
The effect is created using the SCROLL OBJECT TEXTURE statement which has
the format shown in FIG-32.15.
FIG-32.15
The SCROLL OBJECT SCROLL OBJECT TEXTURE objno , Uoffset , Voffset
TEXTURE Statement
In the diagram:

texture is to be scrolled.
Uoffset,Voffset are a pair of real values representing the

coordinates of the image that are to be the
top-left corner of the texture when placed on a
3D object.
Examples of this statement in use are shown in FIG-32.16 where eyecol.bmp is

mapped to a plane object with various Uoffset, Voffset values.
FIG-32.16
The Effects of the
SCROLL OBJECT
TEXTURE Statement
scroll object texture 1,0.1,0.0 scroll object texture 1,0.0,0.1 scroll object texture 1,0.1,0.1
A complete program demonstrating the effect is given in LISTING-32.7.

LISTING 32.7 REM *** Set display resolution ***
Using the SCROLL
OBJECT TEXTURE REM *** Load image ***
Statement LOAD IMAGE "eyecol.bmp",1
REM *** Create and texture plane object ***

TEXTURE OBJECT 1,1
REM *** Offset the texture placed on the image***

SCROLL OBJECT TEXTURE 1,0.1,0.0

WAIT KEY
END
Activity 32.10
The effect makes a permanent change to the texture for that image, so repeating the
same statement creates a further offset.
Activity 32.11
In your last program, add another SCROLL OBJECT TEXTURE statement

immediately after the first using the same values.
How does this affect the texture?
By placing the SCROLL OBJECT TEXTURE statement in a loop, the texture can
scroll over the surface of the 3D object.
Activity 32.12
Remove the second SCROLL OBJECT TEXTURE statement from your last
program and insert the remaining SCROLL OBJECT TEXTURE statement in
a DO..LOOP structure.
What affect does this have on the 3D object's texture?
Modify the SCROLL OBJECT TEXTURE statement so that the texture

scrolls horizontally rather than vertically.
Activity 32.13
In Chapter 31 you created a program in which two spheres followed the

movement of the mouse pointer (object3D11.dbpro).
Modify that program so that the spheres are textured using seamlesseye.bmp.
Add the appropriate SCROLL OBJECT TEXTURE statements so that the

pupils of the eyes face the mouse pointer.

The SET OBJECT TRANSPARENCY Statement
A colour other than black When an image containing black is mapped to a sprite, any black areas in the image
can become the are automatically transparent when the sprite appears on the screen. However, this
transparent colour using is not the case with 3D objects.
the SET COLOR KEY
statement.
To demonstrate this, the next program (see LISTING-32.8) uses the image shown
in FIG-32.17 as the texture on a rotating cube.
FIG-32.17
Image used to Texture a
Cube

Black Textured Areas are
not Transparent on a 3D REM *** Load texture image ***
Object LOAD IMAGE "DoNot.bmp",2

REM *** Texture cube with image ***

TEXTURE OBJECT 1,2

DO
LOOP

END
Activity 32.14
However, we can force a 3D object to make black (or whatever other colour as been
set as the background colour using the SET COLOR KEY statement) transparent
using the SET OBJECT TRANSPARENCY statement which has the format shown
in FIG-32.18.
FIG-32.18
The SET OBJECT
SET OBJECT TRANSPARENCY objno , transflag
TRANSPARENCY
In the diagram:

background texture colour is to be made
transparent.

transflag is 0 or 1.
0 - background colour not transparent.
1 - background colour transparent.
Activity 32.15
Modify your last program by adding the line
SET OBJECT TRANSPARENCY 1,1
immediately before the DO..LOOP structure.
How does this affect the appearance of the cube.
The SET DETAIL MAPPING ON Statement

A second image can be combined with the basic texture image of an object to create
a new texture consisting of both images.
For example, if we take the images shown in FIG-32.19 with image 1 being the
basic texture and image 2 the overlaid texture, then we achieve the effect shown in
FIG-32.20 when these are applied to a cube.
FIG-32.19
The Images Used To
Texture a 3D Object
Image 1 Image 2
FIG-32.20
The Two Images Applied
to a Cube
Notice that any black areas in image 2 are automatically transparent.
A second image is applied to the texture of an object using the SET DETAIL
MAPPING ON statement which has the format shown in FIG-32.21.
FIG-32.21
The SET DETAIL SET DETAIL MAPPING ON objno , imgno
MAPPING ON Statement
In the diagram:

a second texture image is to be added.

imgno is an integer value specifying the image object
containing the picture to be used as a second
texture.
LISTING-32.9 creates the rotating cube shown above. The main lines of code are
LOAD IMAGE "DoNot.bmp",2
which loads the secondary image being used and

SET DETAIL MAPPING ON 1,2
which applies this image as an overlaid texture.

Using a Secondary REM *** Load texture images ***
Texture on a 3D Object LOAD IMAGE "textureWood.jpg",1
REM *** Create and texture cube ***
TEXTURE OBJECT 1,1
REM *** Add secondary texture ***

SET DETAIL MAPPING ON 1,2


DO
LOOP
END
Activity 32.16
Replace DoNot.bmp with DoNotMag.bmp which contains black text on a

magenta background. Set magenta as the transparent colour using the SET
IMAGE COLORKEY statement.
We are limited to a single image when overlaying an object's texture with detail.
So, attempting to add a second detail image will simply remove the first from the
object.
Activity 32.17
Add FlagMag.bmp as a second detail image to the cube object in your last
program.
What effect does this create?
If an object's texture has been tiled using the SCALE OBJECT TEXTURE
statement, any additional image added using SET DETAIL MAPPING ON will
also be tiled to the same extent as the original texture.

Activity 32.18
In your last program, remove all references to the flagmag.bmp file. Create a
tiled effect on the cube by using the SCALE OBJECT TEXTURE statement
with the Uscale and Vscale parameters both set to 2.
How is the DETAIL MAPPING image on the cube affected by the tiling?
The SET OBJECT FILTER Statement

Different methods of texturing can be specified using the SET OBJECT FILTER
statement. The differences achieved have little obvious effect on the visible
appearance of the textured object itself, but modify the algorithm used to create that
texture. The statement has the format shown in FIG-31-22.
FIG-32.22
The SET OBJECT SET OBJECT FILTER objno , filterflag
FILTER Statement
In the diagram:

texture is to be filtered.
filterflag is 0, 1, or 2
0 - always uses the original image
to texture (it never uses the smaller
images created by mipmapping).
1 - no smoothing is used.
2 - uses linear filtering.
The program in LISTING-32.10 creates 3 spheres, each textured using one of the
filter options. You may see a slight difference in the appearance of the spheres as
they move off into the background and reduce in size.
LISTING-32.10 REM *** Set screen resolution ***
REM *** Load image used as texture ***
FILTER Statement
LOAD IMAGE "grid8by8.bmp",1,1
REM *** Create three spheres ***
MAKE OBJECT SPHERE 1 ,40
REM *** Texture each sphere ***

TEXTURE OBJECT 1,1
TEXTURE OBJECT 2,1
TEXTURE OBJECT 3,1
REM *** Set different filter for each sphere ***

SET OBJECT FILTER 1, 0
REM *** Position spheres ***


LISTING-32.10 REM *** Moves spheres ***
(continued) FOR z = 1 TO 4000
MOVE OBJECT 1,1
Using the SET OBJECT MOVE OBJECT 2,1
FILTER Statement MOVE OBJECT 3,1
WAIT 10
NEXT z

WAIT KEY
END
Activity 32.19
Type in and test the program given above (texture10.dbpro).
Summary
l Texturing involves mapping an image onto the surface of a 3D object.
l Use LOAD IMAGE to load any image which is to be used to texture a 3D object.
l Use TEXTURE OBJECT to map an image to an object.
l Images stretch automatically to fit the surface of an object.
l On a cube or box the image is repeated on each face.
l On other 3D objects, the image appears only once.
l Mipmaps are smaller versions of the original image which are used to speed up
mapping when a textured object becomes much smaller than the original image.
l Tiling is the application of an image multiple times to the same surface.
l Use SCALE OBJECT TEXTURE to create a tiled texture.
l To create a seamless tile, make sure the opposite edges are complementary.
l Use PLAY ANIMATION TO IMAGE and TEXTURE OBJECT to display a

video on the surface of an object.
l Use SET OBJECT TEXTURE to specify how an image is mapped to a surface.
l Use SCROLL OBJECT TEXTURE to create an offset mapping of the image on

a 3D surface.
l Use SET OBJECT TRANSPARENCY to make black areas of a 3D object

disappear.
l Use SET DETAIL MAPPING ON to apply a second image to an already textured

object.
l Use SET OBJECT FILTER to modify how an image is filtered when being
mapped onto an object.

Other Visual Effects
Introduction
Although adding texture to a 3D object is the commonest way of changing an
object's appearance, it is by no means the only option available. In this section we'll
see some other options that are available to us in DarkBASIC Pro.
Changing Colour and Transparency

The COLOR OBJECT Statement
Rather than add a texture to a 3D object, we can give it a surface of a specific colour
using the COLOR OBJECT statement which has the format shown in FIG-32.23.
FIG-32.23
The COLOR OBJECT COLOR OBJECT objno , colour
Statement
In the diagram:
objno is an integer value specifying the object which is

to be coloured.
colour is an integer value specifying the colour to be

used on the object's surface.
For example, we could give object 1 a red surface using the line:
COLOR OBJECT 1, RGB(255,0,0)
An example of this statement in operation is given in LISTING-32.11 which colours

a rotating cube in red, changing to green when a random event occurs.
Using COLOR OBJECT
RANDOMIZE TIMER()
REM *** Colour cube red ***
REM *** Rotate Cube ***
DO
REM *** One chance in 1000 of changing to green ***
IF RND(1000) = 500
ENDIF
LOOP
END

Activity 32.20
A coloured surface cannot be used in conjunction with a main texture, but secondary
textures (created using SET DETAIL MAPPING ON) may still be used.
Activity 32.21
Modify your last program so that DoNot.bmp is used as a secondary texture on

the surface of the cube.
The GHOST OBJECT ON Statement

In FIG-32.24 we can see a cube which is semi-transparent with the grided plane in
the background showing through the cube.
FIG-32.24
A Transparent Cube
You'll have to look

closely to see the cube!
We can create this effect using the GHOST OBJECT ON statement which has the
FIG-32.25
The GHOST OBJECT
GHOST OBJECT ON objno , ghostflag
ON Statement
In the diagram:
objno is an integer value specifying the 3D object to

be made semi-transparent.
ghostflag is 0 to 5.
0 - object is semi-transparent
1 - object uses negative transparency
2 - object is semi-transparent but lighter
3 - uses the image's alpha channel
4 - similar to 1 but lighter
5 - object is opaque
The program in LISTING-32.12 demonstrates the effect of this instruction by

rotating the cube with a textured plane in the background.

Using the GHOST
OBJECT ON Statement REM *** Load texture images ***
LOAD IMAGE "grid8by8.bmp",1

REM *** Texture cube ***

TEXTURE OBJECT 1,1
REM *** Create background plane ***

TEXTURE OBJECT 2, 2
REM *** Create semi-transparent cube ***

GHOST OBJECT ON 1,0

DO
LOOP

END
Activity 32.22
Modify the transflag value in the GHOST OBJECT ON statement and

observe the effects of the various settings.
The GHOST OBJECT OFF Statement

The semi-transparency effect created by GHOST OBJECT ON can be disabled
using the GHOST OBJECT OFF statement which has the format shown in
FIG-32.26.
FIG-32.26
GHOST OBJECT OFF objno
The GHOST OBJECT
OFF Statement
In the diagram:
objno is an integer value specifying the 3D object in

which the semi-transparency mode is switched
off.
The FADE OBJECT Statement

The amount of light reflected from the surface of a 3D object can be modified using
the FADE OBJECT statement. This allows settings varying between no light
reflected and twice normal reflection. The statement has the format shown in
FIG-32.27.

FIG-32.27
The FADE OBJECT FADE OBJECT objno , value
Statement
In the diagram:
objno is an integer value specifying the 3D object

whose reflective index is to be changed.
value is an integer value between 0 and 200. A value of

zero means that the object will reflect no light;
a value of 100 creates normal amount of reflected
light; a value of 200 gives twice the normal
amount of reflected light.
The program in LISTING-32.13 demonstrates the effect of this statement by

gradually reducing the reflective value from 200 to zero.

Using FADE OBJECT to
REM *** Load texture images ***
make a Transparent
Object Disappear


TEXTURE OBJECT 1,1
REM *** Create background plain ***

TEXTURE OBJECT 2, 2
REM *** Start reflective value at 200 ***

reflectivity = 200
REM *** rotate cube ***

DO
FADE OBJECT 1,reflectivity
REM *** Decrement reflective value until it reaches zero ***
IF reflectivity > 0
DEC reflectivity
ENDIF
WAIT 10
LOOP

END
Activity 32.23
When used in combination with the GHOST OBJECT ON statement, the FADE
OBJECT statement can make a semi-transparent object disappear completely.

Activity 32.24
In your last program, add the line
GHOST OBJECT ON 1,0
How does this affect the cube?
There's still more to be said about texturing objects but we'll leave that to a later
chapter after we've covered other basic concepts such as cameras and lighting.
Summary
l Use COLOR OBJECT to tint the surface of a 3D object.
l Use GHOST OBJECT ON to make an object transparent.
l Use GHOST OBJECT OFF to make a transparent object opaque.
l Use FADE OBJECT to reduce the amount of light reflected by an object.
l When used on a transparent object, FADE OBJECT can make that object
invisible.
l When used on an opaque object, FADE OBJECT can make that object
completely black.

Images with an Alpha Channel
Introduction
As we saw back in Chapter 30, some picture files contain an alpha channel which
can affect the visibilty of the main image. This type of picture file, when used as a
texture, affects the overall result obtained.
Using Images with an Alpha Channel

The image windmillshaped.tga contains an image and an alpha channel as shown
in FIG-32.28.
FIG-32.28
An Image and its Alpha
Channel
Image Alpha Channel
We've already seen that the SET OBJECT TRANSPARENCY statement affects
the black area of an texture image, but the statement also controls how an alpha
channel within an image affects the final texture.
The program in LISTING 32-14 textures as cube using windmillshaped.tga. With

the default settings, the alpha channel within the image has no effect on the texturing
of the cube, but after a SET OBJECT TRANSPARENCY statement is used to
modify the texturing, the alpha channel changes the final look of the cube.
SET DISPLAY MODE 1280, 1024,32
Texturing with
Alpha-Channel Images REM *** Create cube ***

LOAD IMAGE "windmillshaped.tga",1
TEXTURE OBJECT 1,1

DO
REM *** IF key pressed, use alpha channel ***
IF INKEY$() <> ""
ENDIF
TURN OBJECT LEFT 1,1
LOOP
END
Activity 32.25

Notice that the other parts of the image do not disappear completely when the alpha
channel is activated. This is because the darkened areas of the alpha channel are
grey and not black. If black had been used, all other parts of the image would have
become invisible.
An image with an alpha channel also produces an effect when option 3 is used with
the GHOST OBJECT ON statement.
Activity 32.26
In your last program, change the line
to
GHOST OBJECT ON 1,3
Observe how this affect the program's display.
Summary
l Use SET OBJECT TRANSPARENCY with an alpha channel image to create
transparent or semitransparent texturing effects.
l Use GHOST OBJECT ON with option 3 to make use of the alpha channel
information in creating the final ghosting effect.

Creating a Complex 3D Shape
Introduction
By combining the basic 3D shapes available to us we can create almost any complex
shape. However, this may take a considerable amount of work and require a great
deal of code. In reality we would probably create objects such as these in a separate
3D drawing package and then import the resulting file into our DarkBASIC Pro
program (we'll see how to do this in Chapter 36). However, just to give us some
practice, we'll try creating a simple castle using only DarkBASIC Pro.
Designing the Castle

We are going to save ourselves a great deal of time later if we start by doing a grided
plan of the castle to give ourselves the basic layout and sizes. FIG-32.29 shows a
plan of the castle.
FIG-32.29
A Rough Sketch of the
Object Required
Next, we need to create a more traditional drawing showing the characteristics of

the castle (see FIG-32.30). We might also draw specific parts in more detail.

FIG-32.30
A More Detailed
Design of the Castle
Gathering the Components

The actual texture files being used need to be obtained or created. If you're not much
of an artist, then you'll find plenty of texture files on the Internet, but if you're
intending to create a commercial product, remember that almost everything you see
on the Internet will be owned by someone and they expect to be paid if you are
going to use their material. Even material that is advertised as free may still need
to be paid for when used in a commercial product.
The textures used on the castle are shown in FIG-32.31.

FIG-32.31
Textures Used
Filename : CobbleStones.jpg Filename : stone.jpg Filename : grass1.jpg Filename : ceil_U3_01.jpg

Object : path Object : castle walls and Object : lawn Object : castle ceiling
turrets
Filename : rock2.jpg Filename : tiles.jpg Filename : tree.jpg

Object : castle floor Object : castle and turret roofs Object : trees lining path
Creating the Code

The coding (see LISTING-32.15) is long but fairly straight forward. The complete
castle is created by a function, but this calls other functions which draw the various
parts of the castle.
Splitting the code in this way will help us keep the structure as understandable as
possible.

The Code
LISTING-32.15 REM *** Building Components ***
REM *** Texture Images ***
Drawing the Castle #CONSTANT grass 1
#CONSTANT road 2
#CONSTANT tree 3
#CONSTANT flooring 4
#CONSTANT wall 5
#CONSTANT roofing 6
#CONSTANT cover 8
#CONSTANT trellis 9
#CONSTANT transport 10
REM *** 3D Objects ***
#CONSTANT lawn 1
#CONSTANT approach 2
#CONSTANT floor1 3
#CONSTANT ceiling 5
#CONSTANT tree1 7
#CONSTANT frontwall 51
#CONSTANT backwall 52
#CONSTANT leftwall 53
#CONSTANT rightwall 54
#CONSTANT innerwall 55
#CONSTANT turret1 56
#CONSTANT turretroof1 57
#CONSTANT roof1 64
#CONSTANT column 70

DrawCastle()
WAIT KEY
END
FUNCTION DrawCastle()
LoadImages()
DrawGrounds()
DrawExternalWalls()
DrawRoofandCeiling()
DrawTurrets()
DrawInternalColumns()
ENDFUNCTION
FUNCTION LoadImages()
LOAD IMAGE "grass1.jpg",lawn
LOAD IMAGE "CobbleStones.jpg",road
LOAD IMAGE "tree.jpg",tree
LOAD IMAGE "rock2.jpg",flooring
LOAD IMAGE "stone.jpg",wall
LOAD IMAGE "tiles.jpg",roofing
LOAD IMAGE "ceil_U3_01.jpg",cover
ENDFUNCTION

LISTING-32.15 FUNCTION DrawGrounds()
(continued) REM *** Create lawn ***
MAKE OBJECT PLAIN lawn, 500,700
Drawing the Castle TEXTURE OBJECT lawn, grass
SCALE OBJECT TEXTURE lawn, 100,100
XROTATE OBJECT lawn,-90
POSITION OBJECT lawn,250,0,350
SET DETAIL MAPPING ON lawn,transport
REM *** Create approach road ***
MAKE OBJECT PLAIN approach, 50,700
TEXTURE OBJECT approach,road
SCALE OBJECT TEXTURE approach,8,50
XROTATE OBJECT approach,-90
POSITION OBJECT approach, -25,0,350
REM *** Create castle floor ***
MAKE OBJECT PLAIN floor1, 550,300
TEXTURE OBJECT floor1, flooring
SCALE OBJECT TEXTURE floor1,100,100
XROTATE OBJECT floor1,-90
POSITION OBJECT floor1,225,0,850
DrawTrees()
ENDFUNCTION
FUNCTION DrawExternalWalls()
REM *** Create front wall ***
MAKE OBJECT PLAIN frontwall,550,100
TEXTURE OBJECT frontwall,wall
SCALE OBJECT TEXTURE frontwall,30,6
POSITION OBJECT frontwall, 225,50,700
REM *** Create back wall ***

MAKE OBJECT PLAIN backwall,550,100
TEXTURE OBJECT backwall,wall
SCALE OBJECT TEXTURE backwall,50,10
POSITION OBJECT backwall, 225,50,1000
REM *** Create left wall ***

MAKE OBJECT PLAIN leftwall,300,100
TEXTURE OBJECT leftwall,wall
SCALE OBJECT TEXTURE leftwall,30,10
YROTATE OBJECT leftwall, -90
POSITION OBJECT leftwall,-50,50,850
REM *** Create right wall ***

MAKE OBJECT PLAIN rightwall,300,100
TEXTURE OBJECT rightwall,wall
SCALE OBJECT TEXTURE rightwall,30,10
YROTATE OBJECT rightwall,90
POSITION OBJECT rightwall,500,50,850
ENDFUNCTION
FUNCTION DrawTurrets()
REM *** Create first turret ***
MAKE OBJECT CYLINDER turret1,200
SCALE OBJECT turret1,40,100,40
MAKE OBJECT CONE turretroof1,81
REM *** Texture turret ***
TEXTURE OBJECT turret1,wall
SCALE OBJECT TEXTURE turret1,10,10
TEXTURE OBJECT turretroof1,roofing
SCALE OBJECT TEXTURE turretroof1,5,10
REM *** Position turret ***
POSITION OBJECT turret1, -25,100,970
POSITION OBJECT turretroof1,-25,240,970

LISTING-32.15 REM *** Second turret ***
(continued) MAKE OBJECT CYLINDER turret2,200
Drawing the Castle MAKE OBJECT CONE turretroof2,81
POSITION OBJECT turret2, 475,100,970
POSITION OBJECT turretroof2,475,240,970
REM *** Third turret ***
POSITION OBJECT turret3, -25,100,725
POSITION OBJECT turretroof3,-25,240,725
REM *** Fourth turret ***
POSITION OBJECT turret4, 475,100,725
POSITION OBJECT turretroof4,475,240,725
ENDFUNCTION
FUNCTION DrawRoofAndCeiling()
REM *** Create main roof ***
MAKE OBJECT PLAIN roof1,550,300
REM *** Texture main roof ***
TEXTURE OBJECT roof1,roofing
SCALE OBJECT TEXTURE roof1,50,10
REM *** Position roof ***
XROTATE OBJECT roof1, -90
POSITION OBJECT roof1,225,100,850
REM *** Create ceiling ***
MAKE OBJECT PLAIN ceiling,550,300
REM *** Texture ceiling ***
TEXTURE OBJECT ceiling,cover
SCALE OBJECT TEXTURE ceiling,5,2
REM *** Position ceiling ***
XROTATE OBJECT ceiling, -90
POSITION OBJECT ceiling,225,99,850
ENDFUNCTION
FUNCTION DrawInternalColumns()
RANDOMIZE TIMER()
FOR col = column TO column + 80
MAKE OBJECT BOX col,20,99.8,20
TEXTURE OBJECT col,wall
SCALE OBJECT TEXTURE col,5,30
POSITION OBJECT col,RND(515)-20,49.95, RND(270)+710
NEXT col
ENDFUNCTION

LISTING-32.15 FUNCTION DrawTrees()
(continued) REM *** Create trees ***
FOR c = tree1 TO tree1 + 30
Drawing the Castle MAKE OBJECT PLAIN c,25,35
TEXTURE OBJECT c, tree
SET OBJECT TRANSPARENCY c,1
POSITION OBJECT c,-45,17,(c-6) * 17.5
NEXT c
ENDFUNCTION
The function DrawInternalColumns() adds randomly placed columns within the

castle. This will allow our player to have obstacles to navigate without having to
go to a great deal of trouble designing an exact layout for the castle's interior.
The DrawTrees() function draws a set of trees by texturing a set of planes with a
tree image.
Activity 32.27
Type in and test the program given above (castle01.dbpro).
Activity 32.28
Remove the main section from the program, leaving only the constants and
functions. Save this as castle.dbpro.
Activity 32.29
Create a program (gallows.dbpro) containing a function, DrawGallows(), that

produces a 3D gallows similar to that in the sketch below. Use any appropriate
textures.
The gallows platform should be centred on (-125,7.5,0) and be 15 units high,

and 50 units in width and depth. Place the gallows on a 300 by 300 cobbled
plane.
In the main section of the program include the lines

POINT CAMERA -150,10,0
after the call to the DrawGallows() function. This will ensure that the camera
is pointing at the gallows.

Sky Spheres
Both the castle and the gallows look a bit out of place with the blue background.
One way to create a more natural environment is to place a large sphere round the
whole thing and texture that sphere with an image of the sky. This is known as a
sky sphere. FIG-32.32 shows the results obtained by adding a sky sphere to the
gallows program.
FIG-32.32
Using a Sky Sphere
To implement a sky sphere in our gallows program, we'll start by making the ground
plane set up in DrawGallows() a bit larger
MAKE OBJECT PLAIN GroundObj,2000,2000
and then increase the tiling so the cobbles don't get too large:
SCALE OBJECT TEXTURE GroundObj,150,150
These are the only changes required in the DrawGallows() function. Now we can
add a few lines to the main section. First we need the image to be used to texture
the sphere with a sky effect:
LOAD IMAGE "sky.jpg",1
Next we can create the sphere with the same diameter as the plane:
MAKE OBJECT SPHERE 1,-2000,50,50
Notice that the sphere has been created with extra polygons. This helps smooth out
the sky background.
Apparently all we need to do now is texture the sphere:
TEXTURE OBJECT 1,1

Activity 32.30
Modify your gallows.dbpro program using the lines given above. The main
section should be coded as:
REM *** Set up screen ***

DrawGallows()
REM *** Create the sky sphere ***

LOAD IMAGE "sky.jpg",1
MAKE OBJECT SPHERE 1,2000,50,50
TEXTURE OBJECT 1,1
REM *** Set up camera ***


WAIT KEY
END
Don't worry if you don't see any sky!
We can't see the texture on the sphere because we're on the inside of the sphere and
DarkBASIC Pro has culled the polygons that make up the sphere.
There are two ways to solve this problem. The first is to switch off culling on the
sphere. This can be done using the line:
SET OBJECT CULL 1,0
Activity 32.31
Add the above line to the main section of your code. Is the sky now visible?
An alternative way of displaying the sphere's texture is to turn the sphere inside out!
This is done by specifying a negative size for the sphere when it is being created.
Activity 32.32
Change the line
to have a negative size value:
MAKE OBJECT SPHERE 1,-2000,50,50
and remove the SET OBJECT CULL statement.
How does this affect the sphere's texture?

To cure the problem of an inverted mirror image (which isn't actually a problem in
this case) we need to save the original image as an inverted mirror image in the first
place and this will then be reversed when the image is used as a texture.
Activity 32.33
If you have an appropriate paint package, invert and mirror the image sky.jpg
(save the resulting image as skyIM.jpg) and use the new image as a texture for
the sky sphere.
If you don't have an appropriate package, the inverted mirror image is supplied
with the images for this chapter.
Summary
l A sky sphere allows us to create a sky affect around our 3D world.
l A sky sphere is a large sphere textured with an image of the sky.
l To make the sphere's texture visible from within the sphere, switch off the
sphere's culling or create and inverted mirror image of the sky and use it to texture
a sphere with a negative diameter value.

Solutions
REM *** Set display resolution *** By changing the coordinates of the bottom right corner of
SET DISPLAY MODE 1280,1024,32 the video to 10,10, the video itself is only 11 pixels by 11
pixels, and, when expanded to cover the surface of the
REM *** Create cube *** cube becomes indistinct and blocky.
REM *** Add texture to cube *** By changing the corner values from 10,10 to 1000,1000
TEXTURE OBJECT 1,1 we make the video 1001 pixels by 1001 pixels (actually
REM *** Position cube *** larger than the original recording). Since we cannot add
POSITION OBJECT 1,25,0,100 any detail which was not in the original recording, this
REM *** Rotate cube continuously ***
size does not achieve any better results than a lower
DO
resolution (say 640 by 640), but does increase the load on
LOOP the video hardware and slows down the whole process.
END
Activity 32.9
To change shape requires the line

Activity 32.10
to be replaced by each of the following in turn:
Activity 32.11
MAKE OBJECT BOX 1, 10,20,30
MAKE OBJECT CYLINDER 1, 15 REM *** Set display resolution ***
MAKE OBJECT CONE 1, 15 SET DISPLAY MODE 1280,1024,32
MAKE OBJECT SPHERE 1, 10 REM *** Load image ***
The cube and box repeat the texture image on each side; REM *** Create and texture plain object ***
other shapes show the image only once. MAKE OBJECT PLAIN 1,200,200
TEXTURE OBJECT 1,1
REM *** Offset texture on image***
Activity 32.3 WAIT KEY
No solution required. SCROLL OBJECT TEXTURE 1,0.1,0.0
WAIT KEY
END
Activity 32.4
The texture seems a little more blurred when using We can see from the results produced by the program that
mipmaps, but the texturing process seems to be carried out the scroll effect is cumulative, with the image moving
at a faster frame rate. another step to the left when the second SCROLL
OBJECT TEXTURE statement is applied.
Activity 32.5
Activity 32.12
Activity 32.6 REM *** Load image ***
No solution required. REM *** Create and texture plain object ***
TEXTURE OBJECT 1,1
REM *** Offset texture placed on image***
Activity 32.7 DO
REM *** Set display resolution *** SCROLL OBJECT TEXTURE 1,0.1,0.0
SET DISPLAY MODE 1280,1024,32 LOOP
REM *** Load texture image *** REM *** End program ***
LOAD IMAGE "seamlesseye.bmp",1 WAIT KEY
REM *** Create and texture sphere *** END
MAKE OBJECT SPHERE 1, 40,40,40
TEXTURE OBJECT 1,1 The image scrolls vertically.
DO
TURN OBJECT LEFT 1, 1.0 REM *** Set display resolution ***
LOOP SET DISPLAY MODE 1280,1024,32
END REM *** Load image ***

REM *** Create and texture plain object *** LOAD IMAGE "textureWood.jpg",1
MAKE OBJECT PLAIN 1,200,200 LOAD IMAGE "DoNotMag.bmp",2
TEXTURE OBJECT 1,1 REM *** Create and texture cube ***
REM *** Offset texture placed on image***
DO MAKE OBJECT CUBE 1, 40
SCROLL OBJECT TEXTURE 1,0.0,0.1 TEXTURE OBJECT 1,1
LOOP REM *** Add secondary texture ***
REM *** End program *** SET DETAIL MAPPING ON 1,2
WAIT KEY REM *** Position cube ***
DO PITCH OBJECT DOWN 1, 1.0
Activity 32.13 TURN OBJECT LEFT 1, 1.0
LOOP
REM *** Set display resolution *** REM *** End program ***
SET DISPLAY MODE 1280,1024,32 END
MAKE OBJECT SPHERE 1, 40 Activity 32.17
POSITION OBJECT 1,25,-20,100
MAKE OBJECT SPHERE 2, 40 REM *** Set display resolution ***
POSITION OBJECT 2, -25,-20,100 SET DISPLAY MODE 1280,1024,32
REM *** Texture spheres *** REM *** Set magenta as transparent ***
TEXTURE OBJECT 1,1 SET IMAGE COLORKEY 255,0,255
TEXTURE OBJECT 2,1 REM *** Load texture images ***
REM *** Offset so eyes are at front *** LOAD IMAGE "textureWood.jpg",1
SCROLL OBJECT TEXTURE 1, 0.51,0.0 LOAD IMAGE "DoNotMag.bmp",2
SCROLL OBJECT TEXTURE 2,0.51,0.0 LOAD IMAGE "FlagMag.bmp",3
REM *** Make eyes follow mouse *** REM *** Create and texture cube ***
DO MAKE OBJECT CUBE 1, 40
x3D = MOUSEX() - SCREEN WIDTH()/2 TEXTURE OBJECT 1,1
y3D = -(MOUSEY() - SCREEN HEIGHT()/2) REM *** Add text as secondary texture ***
POINT OBJECT 1, x3D,y3D,-300 SET DETAIL MAPPING ON 1,2
POINT OBJECT 2, x3D,y3D,-300 REM *** Try using another texture ***
LOOP SET DETAIL MAPPING ON 1,3
REM *** End program *** REM *** Position cube ***
END REM *** Rotate cube ***
DO PITCH OBJECT DOWN 1, 1.0
Activity 32.14 LOOP
No solution required. END
Only the wood and flag textures show; the text "DO NOT
Activity 32.15 OPEN" is missing.
REM *** Load texture image *** Activity 32.18
LOAD IMAGE "DoNot.bmp",2 REM *** Set display resolution ***
REM *** Make and position cube *** SET DISPLAY MODE 1280,1024,32
MAKE OBJECT CUBE 1, 40 REM *** Set magenta as transparent ***
POSITION OBJECT 1,25,0,100 SET IMAGE COLORKEY 255,0,255
REM *** Texture cube with image *** REM *** Load texture images ***
TEXTURE OBJECT 1,2 LOAD IMAGE "textureWood.jpg",1
REM *** Make black areas of texture *** LOAD IMAGE "DoNotMag.bmp",2
REM *** transparent *** REM *** Create and texture cube ***
SET OBJECT TRANSPARENCY 1,1 MAKE OBJECT CUBE 1, 40
REM *** Rotate cube *** TEXTURE OBJECT 1,1
DO REM *** Tile cube's texture ***
PITCH OBJECT DOWN 1, 1.0 SCALE OBJECT TEXTURE 1,2,2
TURN OBJECT LEFT 1, 1.0 REM *** Add secondary texture ***
LOOP SET DETAIL MAPPING ON 1,2
REM *** End program *** REM *** Position cube ***
Any part of the cube which is textured with black DO PITCH OBJECT DOWN 1, 1.0
disappears. TURN OBJECT LEFT 1, 1.0
LOOP
Activity 32.16 END
REM *** Set display resolution *** The DETAIL MAPPING image is also tiled.
REM *** Make magenta transparent ***
SET IMAGE COLORKEY 255,0,255

Activity 32.19 WAIT 10
LOOP
No solution required. REM *** End program ***
END
Activity 32.20 The cube fades until it is completely invisible.
Activity 32.25
RANDOMIZE TIMER() REM *** Set display resolution ***
REM *** Load image *** SET DISPLAY MODE 1280, 1024,32
LOAD IMAGE "DoNot.bmp",1 REM *** Create cube ***
REM *** Make and position cube *** MAKE OBJECT CUBE 1,10
MAKE OBJECT CUBE 1, 40 REM *** Texture cube ***
REM *** Create detail mapping *** LOAD IMAGE "windmillshaped.tga",1
SET DETAIL MAPPING ON 1,1 TEXTURE OBJECT 1,1
POSITION OBJECT 1,25,0,100 REM *** Rotate cube ***
REM *** Colour cube red *** DO
COLOR OBJECT 1, RGB(255,0,0) REM *** IF key pressed, ghost ***
REM *** Rotate Cube *** IF INKEY$() <> ""
DO GHOST OBJECT ON 1,3
PITCH OBJECT DOWN 1,1.0 ENDIF
TURN OBJECT LEFT 1, 1.0 TURN OBJECT LEFT 1,1
REM *** 1 in 1000 of going green *** LOOP
IF RND(1000) = 500 REM *** End program ***
COLOR OBJECT 1, RGB(0,255,0) END
ENDIF
LOOP
REM *** End program *** Activity 32.27
END
Activity 32.22
Activity 32.28
Activity 32.23
Activity 32.29
Activity 32.24 DrawGallows()
REM *** Set display resolution *** WAIT KEY
SET DISPLAY MODE 1280,1024,32 END
REM *** Make and position cube *** FUNCTION DrawGallows()
MAKE OBJECT CUBE 1, 40 REM *** Set up names ***
POSITION OBJECT 1,25,0,100 REM *** Object names ***
REM *** Texture cube *** #CONSTANT GroundObj 901
TEXTURE OBJECT 1,1 #CONSTANT PlatformObj 902
REM *** Create background plain *** #CONSTANT VerticalPostObj 903
MAKE OBJECT PLAIN 2,100,100 #CONSTANT HorizontalPostObj 904
TEXTURE OBJECT 2, 2 #CONSTANT DiagonalPostObj 905
POSITION OBJECT 2,0,0,200 #CONSTANT TopStepObj 906
REM *** Start reflective value at 200 *** #CONSTANT MiddleStepObj 907
reflectivity = 200 #CONSTANT BottomStepObj 908
REM *** Make cube transparent *** #CONSTANT StepEdgeRightObj 909
GHOST OBJECT ON 1,0 #CONSTANT StepEdgeLeftObj 910
REM *** rotate cube *** REM *** Image names ***
DO #CONSTANT CobbleImg 901
PITCH OBJECT DOWN 1, 1.0 #CONSTANT PlanksImg 902
TURN OBJECT LEFT 1, 1.0 #CONSTANT WoodImg 903
FADE OBJECT 1,reflectivity REM *** Load texture images ***
REM *** Reduce reflectivity to zero *** LOAD IMAGE "TextureWood.jpg",PlanksImg
IF reflectivity > 0 LOAD IMAGE "CobbleStones.jpg",CobbleImg
DEC reflectivity LOAD IMAGE "Wood.jpg",WoodImg
ENDIF REM *** Create cobbled square ***

TEXTURE OBJECT GroundObj,CobbleImg
XROTATE OBJECT GroundObj,-90
REM *** Create platform ***
MAKE OBJECT BOX PlatformObj,50,15,50
TEXTURE OBJECT PlatformObj,PlanksImg
SCALE OBJECT TEXTURE PlatformObj,2,2
POSITION OBJECT PlatformObj,-125,7.5,0
REM *** Create vertical post ***
MAKE OBJECT BOX VerticalPostObj,2,30,2
TEXTURE OBJECT VerticalPostObj,WoodImg
POSITION OBJECT VerticalPostObj,
Ä-147.5,30,0
REM *** Create horizontal post ***
MAKE OBJECT BOX HorizontalPostObj,
Ä2,15,2
TEXTURE OBJECT HorizontalPostObj,
ÄWoodImg
ZROTATE OBJECT HorizontalPostObj,90
POSITION OBJECT HorizontalPostObj,
Ä-139,44,0
REM *** Create diagonal post ***
MAKE OBJECT BOX DiagonalPostObj,1,10,1
TEXTURE OBJECT DiagonalPostObj,WoodImg
ZROTATE OBJECT DiagonalPostObj,-45
POSITION OBJECT DiagonalPostObj,
Ä-144,40,0
REM *** Make top step ***
MAKE OBJECT BOX TopStepObj,10,0.3,3
TEXTURE OBJECT TopStepObj,WoodImg
POSITION OBJECT TopStepObj,-130,12,-26
REM *** Make middle step ***
CLONE OBJECT MiddleStepObj,TopStepObj
POSITION OBJECT MiddleStepObj,-130,8,-29
REM *** Make bottom step ***
CLONE OBJECT BottomStepObj,TopStepObj
POSITION OBJECT BottomStepObj,-130,4,-32
REM *** Make right step edge***
MAKE OBJECT BOX StepEdgeRightObj,0.3,4,20
TEXTURE OBJECT StepEdgeRightObj,WoodImg
XROTATE OBJECT StepEdgeRightObj,-50
POSITION OBJECT StepEdgeRightObj,
Ä-125,6,-30
REM *** Make left step edge ***
CLONE OBJECT StepEdgeLeftObj,
ÄStepEdgeRightObj
POSITION OBJECT StepEdgeLeftObj,
Ä-135,6,-30
ENDFUNCTION
Activity 32.30
The changes to the cobbled square in the DrawGallows()
function are shown in bold below:
REM *** Create cobbled square ***

TEXTURE OBJECT GroundObj,CobbleImg
Activity 32.31
The sky should now be visible.
Activity 32.32
The sphere's texture is upside down and mirrored.
Activity 32.33

33
Changing the Camera's Field of View
Controlling Camera Movement with the Keyboard

Controlling Camera Movement with the Mouse
Creating and Deleting Extra Cameras
Linking the Camera to a 3D Object
Pointing the Camera
Positioning the Camera
Rotating the Camera
Setting the Camera's Output to the Screen
Specifying the Distance Limits of Camera Vision
DarkBASIC Pro : Cameras 835

Camera Basics
Introduction
Although we've seen how to move objects about in 3D space, our viewpoint has
remained fixed. That viewpoint is determined by the position of a camera. We can
think of what we see on the screen as a picture being broadcast by a TV camera; if
the camera moves then the picture we see on the screen shifts as well.
The camera in DarkBASIC Pro is a virtual one; totally invisible and with no chance
of accidentally appearing in the picture showing on the screen.
In this chapter we'll cover the DarkBASIC Pro statements that allow us to move the
main camera, create new cameras (so we can view the world from multiple
positions) and take full control of the view created on our screen.
We'll use the castle environment we created in the previous chapter to demonstrate
most of the camera statements.
Positioning the Camera

Any program containing a 3D object automatically contains a camera. It is this
camera (camera 0) whose "transmitted" image appears on the screen.
Just as we might position, move, or rotate a 3D object, so we can perform the same
type of operation on the camera. By moving the camera and making it point in some
other direction, so we generate a new viewpoint on the screen.
The POSITION CAMERA Statement

We can position the camera at any point in 3D space using the POSITION
CAMERA statement which has the format shown in FIG-33.1.
FIG-33.1
The POSITION
x , y , z
CAMERA Statement POSITION CAMERA camno ,
3DVector
In the diagram:
camno is an integer value specifying which camera is to

be moved. If no camera number is given then
the currently active camera is used.
x,y,z is a set of real values giving the position to which

the camera is to be moved.
3Dvector is an integer value giving the ID of the vector

containing the coordinates to which the camera
is to be moved.
Only camera zero is created automatically. If you don't intend to create any others,
836 DarkBASIC Pro : Cameras

then there is no need to specify a camera number in most camera-related
instructions.
The position to which the camera is to be moved can be given using three separate
values or a single 3-element vector object containing the same three values. We'll
ignore the vector option until we cover 3D vectors in a later chapter. So, using the
POSITION CAMERA statement we could place the main camera at position
(10,20,30) using the line:
POSITION CAMERA 0,10,20,30
Activity 33.1
Create a new program (camera01.dbpro) and include the code in castle.dba

(copy castle.dba to this project's folder and use #INCLUDE).
In the main section call DrawCastle(), followed by the lines:

WAIT KEY
WAIT KEY
END
How does this affect the view when the program is executed?
The MOVE CAMERA Statement

If we want the camera to travel in the direction in which it is currently facing, then
we can use the MOVE CAMERA statement. This statement has the format shown
in FIG-33.2.
FIG-33.2
The MOVE CAMERA MOVE CAMERA camno , distance
Statement
In the diagram:

distance is a real number specifying the number of units

the camera is to be moved. The camera moves in
the direction it is currently pointing. If a negative
value is given, the camera will move backwards.
For example, we could move the active camera forward 30 units using the line:
MOVE CAMERA 30
Activity 33.2
Replace the POSITION CAMERA statement in the last program with the line
MOVE CAMERA 50
How does this affect the view produced?

As a result of the last Activity the camera, which was facing slightly downwards,
has passed through the plane objects that make up the pathway and grass. We can
see this happen if we move the camera just one unit at a time.
Activity 33.3
Replace the MOVE CAMERA statement with the following lines:
FOR c = 1 TO 100
MOVE CAMERA 1
WAIT 50
NEXT c
Run the program and observe how the camera moves under the plane.
Changing the Viewpoint

The POINT CAMERA Statement
It's all very well to move a camera, but we also need to point the camera in the
desired direction. One way to do this is to use the POINT CAMERA statement
FIG-33.3
The POINT CAMERA POINT CAMERA camno , x , y , z
Statement
In the diagram:

x,y,z is a set of real values specifying the position in

space where the camera is to be aimed. The point
specified will appear in the middle of the screen.
For example, we could point the camera at the entrance to the castle using the line:
Activity 33.4
Add the above statement to your last program immediately before the FOR
loop used to move the camera.
Extend the FOR loop so that the camera moves right up to the castle's wall.
The ROTATE CAMERA Statement

The camera can be rotated to an exact angle about any of its own local axes (see
FIG-33.4) using the ROTATE CAMERA statement.

-
FIG-33.4
Absolute Camera +
y-axis
Rotation
Camera z-axis +
x-axis -
+
The ROTATE CAMERA statement has the format shown in FIG-33.5.

FIG-33.5
The ROTATE
CAMERA
Statement ROTATE CAMERA camno , xangle , yangle , zangle
In the diagram:
camno is the integer value specifying the camera to be rotated.
xangle is a real number giving the angle (in degrees) to which the
camera is to be rotated about its x-axis.
yangle is a real number giving the angle (in degrees) to which the
camera is to be rotated about its y-axis.
zangle is a real number giving the angle (in degrees) to which the
camera is to be rotated about its z-axis.
Rotation about the x-axis is the type of movement we would make filming a rocket
taking off or a diver jumping off a cliff. Rotation about the y-axis (also known as
panning) is the type of movement required to follow a racing car or a runner.
To make the camera point straight forward (that is, parallel to the main z-axis) we
would use the line:
ROTATE CAMERA 0,0,0
Activity 33.5
Remove the POINT CAMERA and FOR loop code from camera01.dbpro and
add the lines
FOR angle = -45 TO 45

ROTATE CAMERA angle,0,0
WAIT 10
NEXT angle
Run the program and observe the effect produced. Change the code to also
create rotation about the other two axes.

The SET CAMERA ROTATION Statement
When the camera is required to rotate about all three axes at the same time, it
normally performs rotation about the z-axis first, y-axis second and x-axis last. Of
course, this will be performed so quickly that you'll probably not be aware of the
order of events. However, if, for some reason we need to reverse the order of
rotations (x first, y second, and z last), we can use the SET CAMERA ROTATION
statement. Once set, the order in which the axes are handled will remain on this new
setting unless you revert to normal using a second option of the SET CAMERA
ROTATION statement. The statement has the format shown in FIG-33.6.
FIG-33.6
The SET CAMERA
ROTATION Statement
SET CAMERA ROTATION
{ { XYZ
ZYX
camno
In the diagram:
XYZ Use this option to set rotation order to x-axis,

y-axis, and, finally, z-axis.
ZYX Use this option to revert to the default rotation
order z-axis, y-axis, x-axis.
Rather than perform three rotations in a single statement, we can rotate about a
single axis only using one of the following:
The XROTATE CAMERA Statement

To rotate the camera about the x-axis we use the XROTATE OBJECT statement
FIG-33.7
The XROTATE
XROTATE CAMERA camno , angle
CAMERA Statement
In the diagram:
camno is an integer value giving the ID of the camera to be rotated.
FIG-33.8 shows the result of rotating the active camera to 35.0o and -35.0o about
the x-axis.
FIG-33.8
The Effects of the
XROTATE Statement
XROTATE CAMERA 35.0 XROTATE CAMERA -35.0
Initial camera orientation
Initial Camera View

The YROTATE CAMERA Statement
To rotate the camera about the y-axis we use the YROTATE OBJECT statement
FIG-33.9
The YROTATE
YROTATE CAMERA camno , angle
CAMERA Statement
In the diagram:
the y-axis.
FIG-33.10
The Effects of the
YROTATE
CAMERA Statement
YROTATE CAMERA 35.0 YROTATE CAMERA -35.0
Initial Camera View
The ZROTATE CAMERA Statement

To rotate the camera about the z-axis we use the ZROTATE OBJECT statement
FIG-33.11
The ZROTATE ZROTATE CAMERA camno , angle
CAMERA Statement
In the diagram:
the z-axis.
We might require such a rotation if we wanted to show the view from the inside as
a car rolls over onto its roof.

FIG-33.12
The Effects of the
ZROTATE CAMERA
Statement
ZROTATE CAMERA 35.0 ZROTATE CAMERA -35.0
Initial Camera View
The following statements allow relative rotation rather than absolute rotation.
The PITCH CAMERA Statement

We can tilt the camera upwards (i.e. rotate it in a positive direction about the x-axis)
or downwards (rotating in a negative direction about the x-axis) using the PITCH
FIG-33.13
The PITCH CAMERA
Statement
PITCH CAMERA
{ { DOWN
UP
camno , angle
In the diagram:
DOWN, UP Choose DOWN to make the camera tilt downwards;

choose UP to make the camera tilt upwards.
camno is the integer value representing the camera to be

rotated.
angle is a real number giving the number of degrees to

be added to the camera's current rotation about its
x-axis.
PITCH CAMERA UP moves the camera in the same direction as XROTATE

CAMERA using a positive angle.
PITCH CAMERA DOWN moves the camera in the same direction as XROTATE
CAMERA using a negative angle.
The TURN CAMERA Statement

We can turn the camera right (i.e. rotate it in a positive direction about the y-axis)
or left (rotating in a negative direction about the y-axis) using the TURN CAMERA
FIG-33.14
The TURN CAMERA
Statement TURN CAMERA
{ { LEFT
RIGHT
camno , angle

In the diagram:
LEFT, RIGHT Choose LEFT to make the camera rotate to the left;
choose RIGHT to make the camera rotate the right.

rotated.

y-axis.
TURN CAMERA RIGHT moves the camera in the same direction as YROTATE
CAMERA using a positive angle. TURN CAMERA LEFT moves the camera in
the same direction as YROTATE CAMERA using a negative angle.
The ROLL CAMERA Statement

We can roll the camera to the left (i.e. rotate it in a positive direction about the
z-axis) or right (rotating in a negative direction about the z-axis) using the ROLL
FIG-33.15
The ROLL CAMERA
Statement
ROLL CAMERA
{ { LEFT
RIGHT
camno , angle
In the diagram:
LEFT, RIGHT Choose LEFT to make the camera roll to the left;
choose RIGHT to make the camera roll to the
right.

rotated.

z-axis.
ROLL CAMERA RIGHT moves the camera in the same direction as ZROTATE
CAMERA using a positive angle. ROLL CAMERA LEFT moves the camera in the
same direction as ZROTATE CAMERA using a negative angle.
Activity 33.6
Create a new main section for camera01.dbpro, so that the camera can be
rotated using the keyboard. Each rotation should be by 1o in the required
direction. Start the camera with zero rotation. Control keys are as follows:
Key Direction Key Direction

z left y-axis / down x-axis
x right y-axis n left z-axis
' up x-axis m right z-axis

Retrieving Camera Data
Statements are available to retrieve the position and orientation of any camera.
These are explained below.
The CAMERA POSITION Statement

The CAMERA POSITION statement has three options, each allowing a single
ordinate of a specified camera's position to be discovered. The CAMERA
POSITION statement has the format shown in FIG-33.16.
FIG-33.16
The CAMERA
POSITION Statement
In the diagram:
X Use this option to return the x-ordinate of the

camera's current position.
Y Use this option to return the y-ordinate of the

Z Use this option to return the z-ordinate of the

camno is an integer value specifying the camera whose

position is required. If this option is omitted the
currently active camera is assumed.
The CAMERA ANGLE Statement

The angle to which a camera has been rotated about any of the local axes can be
determined using the CAMERA ANGLE statement which has the format shown in
FIG-33.17.
FIG-33.17
X
The CAMERA ANGLE
Statement
CAMERA ANGLE Y ( camno )
real

In the diagram:
X Use this option to return the camera's current

angle of rotation about its x-axis.
Y Use this option to return the camera's current

angle of rotation about its y-axis.
Z Use this option to return the camera's current

angle of rotation about its z-axis.

position is required. If this option is omitted the
currently active camera is assumed.
Activity 33.7
Modify your previous program so that the camera's angle of rotation about all
three local axes is always displayed on the screen.
Modifying Camera Characteristics

The SET CAMERA VIEW Statement
Normally, the picture created by our imaginary camera occupies the full screen, but
we can change this using the SET CAMERA VIEW statement which allows us to
specify a rectangular area of the screen where camera output should appear. The
SET CAMERA VIEW statement has the format shown in FIG-33.18.
FIG-33.18 The SET CAMERA VIEW Statement
SET CAMERA VIEW camno , x1 , y1 , x2 , y2
In the diagram:

output area on the screen is to be defined. If this
option is omitted the currently active camera is
assumed.
x1,y1,x2,y2 This is a set of integer values specifying the

coordinates of the top-left (x1,y1) and bottom-
right (x2,y2) corners of the rectangular area of
the screen to be used as camera output.
For example, we could change the output to occupy the area shown in FIG-33.19
using the statement:
SET CAMERA VIEW 0,0,1279,319

FIG-33.19
Changing the Area of the
Screen Occupied by the
Camera's Output
Activity 33.8
Modify your last program to use the area specified in the line above for
camera output.
Move the camera angle displays into the black area of the screen. Does this
cause any problems?
The SET CAMERA ASPECT Statement

If we use SET CAMERA VIEW to change the ratio of the width-to-height of the
picture, we'll end up with a distorted image, as we saw in FIG-33.19. However, we
can correct this using the SET CAMERA ASPECT statement which allows us to
define width-to-height aspect ratio. In effect, by making this ratio the same as that
created by the SET CAMERA VIEW we can remove any distortion. In the SET
CAMERA VIEW example on the previous page we set the area containing the
camera output to (0,0), (1279,319). This gives a width of 1280 and a height of 320
giving a width-to-height ratio of 4-to-1.
The SET CAMERA ASPECT statement has the format shown in FIG-33.20.
FIG-33.20
The SET CAMERA SET CAMERA ASPECT camno , ratio
ASPECT Statement
In the diagram:

aspect ratio is to be defined. If this option is
omitted, the currently active camera is assumed.
ratio is a real number giving the aspect ratio of width

to height.
To correct the distortion caused by the SET CAMERA VIEW statement we need
the line

SET CAMERA ASPECT 1280/320
giving the result shown in FIG-33.21.
FIG-33.21
Correcting the Camera
Output Distortion
Activity 33.9
In your last program, correct the distortion shown in the camera display.
The SET CAMERA FOV Statement

Real cameras often come with a zoom lens. This allows the user to zoom in on
distant objects or zoom out to get a wide-angle shot (useful when trying to record
in small spaces). Just exactly how much can be included in a shot is known as the
field of view (or FOV) and is measured as the angle between what appears on the
left edge of the screen and the right (see FIG-33.22). When zooming in on distant
objects, we have a very limited field of view; in wide-angle mode, we have a wide
field of view.
FIG-33.22
The Effect of Changing
the Field of View
Zoom out to
gain a large
field of view
When the camera is set to a wide field of ... we see more of the surroundings but
view... things looks smaller than normal.
Zoom in to
reduce the
field of view
When the camera is set to a narrow field ... we see less of the surroundings but
of view... things look larger than normal.
Our own virtual camera can perform the same trick using the SET CAMERA FOV
FIG-33.23
SET CAMERA FOV camno , angle
The SET CAMERA
FOV Statement
In the diagram:

field of view is to be defined. If this option is
omitted the currently active camera is assumed.
angle is a real number specifying the angle of field of

view.
The program in LISTING-33.1 demonstrates the effect of changes to the field of

view. It starts with an angle of 180o and reduces to 5o effectively zooming in on
one of the turrets on the castle. For comparison, a second loop in the program
actually moves the camera towards the turret without changing the field of view
from a starting value of 55o.
LISTING-33.1 #INCLUDE "Castle.dba"
Using Zoom SET DISPLAY MODE 1280,1024,32
REM *** Draw the castle ***
DrawCastle()
POSITION CAMERA 0,8,0
REM *** Set the camera screen area ***
REM *** Set the aspect ratio ***

REM *** Point camera at turret ***

REM *** Zoom in on turret ***

WAIT KEY
FOR angle = 180 to 5 STEP -1
SET CAMERA FOV angle
WAIT 50
NEXT angle
REM *** Move camera towards turret ***

WAIT KEY
SET CAMERA FOV 55
FOR move = 1 TO 580
MOVE CAMERA 1
WAIT 20
NEXT move

WAIT KEY
END
Activity 33.10
Type in and test the program given in LISTING-33.1 (camera02.dbpro).
The SET CAMERA RANGE Statement

It is possible to specify the range of the camera's vision. This means that objects
which are too close to the camera, or too far away, become invisible and hence do
not appear on the screen.
This effect is achieved using the SET CAMERA RANGE statement which has the

FIG-33.24
SET CAMERA RANGE camno , min , max
The SET CAMERA
RANGE Statement
In the diagram:
camno is an integer value specifying the camera affected.

If no camera number is given, the active camera
is assumed.
min is a real number specifying the closest distance

at which an object will be visible to the camera.
max is a real number specifying the furthest distance

at which an object will be visible to the camera.
For example, the statement
SET CAMERA RANGE 50,600
would create the effect shown in FIG-33.25.
FIG-33.25
Restricting the Visible
Area to Between 50 and
600 Units from the Only objects in this area
Camera are visible to the camera
Camera
50
600
Activity 33.11
immediately before the first FOR loop and observe how this affects the view
displayed on the screen.
Summary
l Whenever 3D objects are used in a DarkBASIC Pro program, a default camera
is automatically created.
l The image shown on the screen is the picture transmitted by the default camera.
l The camera is not a visible object.
l Use POSITION CAMERA to move the camera to a specific point.
l Use MOVE CAMERA to move the camera by a specified distance. Movement

can be forwards or backwards.

l Use POINT CAMERA to aim the camera at a specific position.
l Use ROTATE CAMERA to rotate the camera to absolute angles about all of its
three axes.
l Use SET CAMERA ROTATION to specify the order of axes rotations.
l Use XROTATE CAMERA to rotate the camera to an absolute angle about the
x-axis only.
l Use YROTATE CAMERA to rotate the camera to an absolute angle about the
y-axis only.
l Use ZROTATE CAMERA to rotate the camera to an absolute angle about the
z-axis only.
l Use PITCH CAMERA to rotate the camera by a number of degrees about its
x-axis.
l Use TURN CAMERA to rotate the camera by a number of degrees about its
y-axis.
l Use ROLL CAMERA to rotate the camera by a number of degrees about its
z-axis.
l Use CAMERA POSITION to retrieve the camera's current position.
l Use CAMERA ANGLE to retrieve the camera's angle of rotation about a specific
axis.
l Use SET CAMERA VIEW to set how much of the screen the camera's picture
occupies.
l Use SET CAMERA ASPECT to set the aspect ratio of the camera's picture.
l Use SET CAMERA FOV to set the camera's field of view. This is equivalent to
using a zoom lens on a standard camera.
l Use SET CAMERA RANGE to set the nearest and furthest point visible to the
camera.

Controlling Camera Movement
Introduction
Although we've already covered many commands that allow us to move, point and
rotate the camera, we still need to develop a few techniques to cover special
situations.
Automatic Camera Placement

The AUTOCAM Statement
The initial position of the camera is determined by the size and position of the 3D
objects that appear in the program. If we create a large cube, the camera will be
placed further back (its z-ordinate becoming more negative); with a smaller cube
the camera would be placed closer.
If, for some reason, we do not want automatic positioning to be implemented, we

can force the camera to start at position (0,0,0) using the line AUTCAM OFF. To
return to automatic positioning use AUTOCAM ON. The AUTOCAM statement
FIG-33.26
The AUTOCAM OFF ON
Statement AUTOCAM
OFF
The program in LISTING-33.2 demonstrates the effect of automatic camera

positioning by displaying the camera's position for a cube whose size can be set.

Automatic Camera REM *** Get size of cube ***
Positioning INPUT "Enter size of cube (2 to 100):",size
MAKE OBJECT CUBE 1, size
REM *** Display position of camera ***
DO
SET CURSOR 100,100
PRINT "(",CAMERA POSITION X(),",",CAMERA POSITION Y(),
Ä ",",CAMERA POSITION Z(),")"
LOOP
END
Activity 33.12
Type in and test the program in LISTING-33.2 (camera03.dbpro).
Modify the program by inserting the line

AUTOCAM OFF
immediately after the SET DISPLAY MODE statement.
How does this affect the camera's position?

Following the Action
In the next program (see LISTING-33.3) we see a strange object crash to the ground
outside the castle. In this first version of the program the camera is stationary and
we see only part of the object's journey.

#CONSTANT artifact 200
Creating a Falling Object #CONSTANT arttexture 200

COLOR BACKDROP RGB(64,0,128)
BACKDROP ON

AUTOCAM OFF
REM *** Draw castle ***

DrawCastle()
REM *** Show falling object ***

FallingObject()

WAIT KEY
END
FUNCTION FallingObject()
REM *** Create textured transparent sphere ***
MAKE OBJECT SPHERE artifact,10
LOAD IMAGE "lattice.bmp",arttexture
TEXTURE OBJECT artifact,arttexture
SET OBJECT TRANSPARENCY artifact, 1
REM *** Position artifact out in sky ***
POSITION OBJECT artifact, -200,300,900
REM *** Point the object towards the ground ***
POINT OBJECT artifact,300,0,30
REM *** Move object, stopping when it hits the ground ***
DO
MOVE OBJECT artifact,8
IF OBJECT POSITION Y(artifact) <= 2
EXIT
ENDIF
LOOP
ENDFUNCTION
You may want to put a

WAIT 20 statement at the Activity 33.13
end of the DO..LOOP if
your machine is a fast one. Type in and run the program given above (camera04.dbpro).
We can modify the program so that the camera is always pointing at the falling
object by adding the line
POINT CAMERA OBJECT POSITION X(artifact),

ÄOBJECT POSITION Y(artifact),OBJECT POSITION Z(artifact)
after the line MOVE OBJECT artifact,8 in the FallingObject() function.

Activity 33.14
Modify your previous program as described above and observe how this
affects the behaviour of the camera.
The SET CAMERA TO FOLLOW Statement

An even more dramatic effect can be produced by getting the camera to move along
FIG-33.27 with an object, keeping that object in the centre of the camera output. This can be
achieved using the SET CAMERA TO FOLLOW statement which has the format
The SET CAMERA TO shown in FIG-33.27.
FOLLOW Statement
SET CAMERA TO FOLLOW camno , x , y , z
, angle , distance , height , smooth , collision
In the diagram:
camno is an integer value specifying the camera affected.

If no camera number is given, the active camera
is assumed.
x,y,z is a set of real numbers giving the coordinates of

the position to which the camera is to point.
angle is a real number giving the angle (measured from

the z-axis) the camera is to maintain in relation to
the the point (x,y,z).
distance is a real number giving the distance the camera

is to be from the point (x,y,z).
height is a real number giving the camera's height above

(or below) the point (x,y,z).
smooth is a real number specifying how smoothly the

camera will move from its current position to the
new position it requires to be in to meet the
specifications given by the SET CAMERA TO
FOLLOW statement.
Collision boxes are collision is 0 or 1 and is used to specify how the camera
discussed in a later
chapter.
is to react to colliding with a static collision box.
This is one of the most complex statements in DarkBASIC Pro, so it's worth
spending a bit of time finding out just how each value in the statement affects the
results you achieve. To demonstrate how it works we'll track the falling object of
our previous program.
The x,y,z values give the position to be tracked, so in this case that means the

coordinates of the falling object which we can get using:
OBJECT POSITION X(artifact),OBJECT POSITION Y(artifact),

ÄOBJECT POSITION Z(artifact)
The distance and height values specify how far the camera is from the point being
tracked and the height above that point (see FIG-33.28).
FIG-33.28
The camera is not
rotated about the x-axis
The Camera Distance and by this statement...
Height ... which means it doesn’t
Camera point down at the object
distance automatically...
Angle of camera ... so if the camera is too high

is determined by height or the FOV too narrow, the object
distance and height will not appear in the camera view
Object
(x,y,z)
This is not enough information to specify a single position for the camera, since any
point on a circle which is distance units from the object and height units above the
object will meet the criteria specified (see FIG-33.29).
FIG-33.29 The camera could be

positioned at any point
on this circle
Possible Camera
Positions
distance distance
height height
y-axis
x-axis (x,y,z)
Object
z-axis
The final value required to create a unique position for the camera is the angle value
which specifies how far round the circle the camera is to be placed (see FIG-33.30).
FIG-33.30 The camera’s exact

position on the circle
The Exact Position is is determined by the
value of angle
Determined by angle
angle
distance
height
y-axis
x-axis (x,y,z)
Object
z-axis

Before we look at the purpose of the last two values used in the SET CAMERA TO
FOLLOW statement, we'll get a feel for how the statement works by modifying the
last program so that the FallingObject() function is now coded as:
REM *** Create textured transparent sphere ***
DO
MOVE OBJECT artifact, 8
SET CAMERA TO FOLLOW OBJECT POSITION X(artifact),
Ä
OBJECT POSITION Y(artifact),OBJECT POSITION Z(artifact),
Ä
0,20,0,1.0,1
EXIT
ENDIF
LOOP
ENDFUNCTION
Notice that the angle and height values are set to zero, while the distance setting is
20.
Activity 33.15
Modify program camera04.dbpro so that its FallingObject() function matches

the code given above.
Modify the program as follows, running the program after each change:
Set distance to 100

Set height to 20
Set angle to 90
The smooth parameter determines how smoothly the camera moves from its current
position to the position required by the latest call to SET CAMERA TO FOLLOW
statement. Using a value of 1.0 the camera will jump almost instantly to the required
position, no matter what its current location. On the other hand, by using a value of
100.0, the camera may take hundreds of frames (that is, several seconds) to move
into position.
Activity 33.16
Set the smooth value in your last program's SET CAMERA TO FOLLOW
statement to 100.0 and re-run the program.
What difference does this make to the camera's movements?
Change the smooth value to 20 and run the program again.
In our example, the smoothing factor affected the camera's movement from its
original position at (0,8,10) to its new position behind the sphere. The effect of
smoothing is also apparent if the object being tracked moves erratically in

unpredictable directions and varying speeds.
We can even use this instruction with a high smoothing factor to make the camera
move slowly along a path to a given point. For example, the lines

DO
SET CAMERA TO FOLLOW -20,200,900,1800,10,10,100.0,1
LOOP
will move the camera smoothly from the point (0,8,10) to within 10 units of the
point (-20,200,900).
Activity 33.17
Change the main section of your previous program to read
#INCLUDE "Castle.dba"
#CONSTANT arttexture 200

BACKDROP ON
AUTOCAM OFF
DrawCastle()
DO
SET CAMERA TO FOLLOW -20,200,900,180,10,10,100.0,1
LOOP
END
and check out how the camera moves and rotates between its starting and
finishing points.
You may have noticed that the camera moves straight through the first turret! Unless
it is playing the part of a ghostly apparition, we probably don't want this to happen.
To prevent this we use the final parameter of the SET CAMERA TO FOLLOW
statement which allows the camera to react to collisions. However, we need to know
how to handle collisions in general before examining this option, so we'll leave it
for a later chapter.
Giving the Player Control of the Camera

The CONTROL CAMERA USING ARROWKEYS Statement
The simplest way to give the player control of the camera position is to use the
CONTROL CAMERA USING ARROWKEYS statement. This allows the player
to rotate the camera about the y-axis using the left and right arrow keys and to move
forwards and backwards using the up and down arrow keys. The statement also
allows the speed of the camera's response to these keys to be specified. This
statement has the format shown in FIG-33.31

FIG-33.31
The CONTROL CAMERA USING
ARROWKEYS Statement
CONTROL CAMERA USING ARROWKEYS camno , move , turn
In the diagram:
camno is an integer value specifying the camera to be

used.
move is a real number specifying the speed at which the

camera moves in response to pressing the up and
down arrow keys.
turn is a real number specifying the speed at which the

camera rotates in response to pressing the left and
right arrow keys.
For example, we could control the movement of camera 0, allowing a move of 0.5
units each time an up or down arrow key is pressed and with a rotation of 0.1o every
time a left or right arrow key is pressed using the line:
CONTROL CAMERA USING ARROW KEYS 0,0.5,0.1
The statement must be placed in a loop structure (usually a DO..LOOP) so that the
statement is continually executed.
In LISTING-33.4 the user is allowed to move about the castle area by controlling
the camera.

Basic User Camera #CONSTANT arttexture 200
Control
BACKDROP ON

AUTOCAM OFF
REM *** Draw the castle and show falling object ***
DrawCastle()
FallingObject()
REM *** Re-position the camera ***

POINT CAMERA 0,8,200
REM *** Give user camera control ***

DO
CONTROL CAMERA USING ARROWKEYS 0,1,0.5
LOOP

END

LISTING-33.4 FUNCTION FallingObject()
(continued) REM *** Create textured transparent sphere ***
Basic User Camera LOAD IMAGE "lattice.bmp",arttexture
Control TEXTURE OBJECT artifact,arttexture


DO
MOVE OBJECT artifact, 8
SET CAMERA TO FOLLOW OBJECT POSITION X(artifact),
Ä
OBJECT POSITION Y(artifact),OBJECT POSITION Z(artifact),
Ä
90,100,20,20,1
EXIT
ENDIF
LOOP
ENDFUNCTION
Activity 33.18
Can you move inside the castle?
As you can see from the results of the Activity, the player takes on ghost-like
qualities; being able to walk through solid objects! Luckily, there's an easy way to
stop this.
The AUTOMATIC CAMERA COLLISION Statement

By using the AUTOMATIC CAMERA COLLISION statement we can stop a
camera moving through a 3D object. This is particularly useful when the user is
controlling camera movement. The statement has the format shown in FIG-33.32.
FIG-33.32
The AUTOMATIC CAMERA
COLLISION Statement
AUTOMATIC CAMERA COLLISION camno , radius , response
In the diagram:
camno is an integer value specifying the camera involved.
radius is a real number giving a collision radius. If the

camera comes within this distance of any 3D
object a collision between that object and the
camera will be assumed.
response is 0 or 1. Using the value zero, the camera reacts

to a collision with another object by "sliding off

to the side" of the object - this is known as a
sliding collision. When set to 1, the camera
halts when a collision is detected.
For example, we could make camera zero halt when it comes within 2 units of an
object using the line:
AUTOMATIC CAMERA COLLISION 0,2,1
Activity 33.19
By placing the line
before the DO loop in the main section of your last program, make the
user-control camera slide past any collisions.
Controlling the Camera with the Mouse

Arrow key camera control is of limited use since the camera can only be rotated
about the y-axis. Also, camera movement is only in the x-z plane (or a plane parallel
to this). To gain greater control, we'll have to write our own code.
We'll start by allowing the user to rotate the camera about both the x and y axes
using mouse movement. This is done by a function, PointCameraUsingMouse(),
which rotates a specified camera in response to mouse movement. The first version
of the routine is given below:
FUNCTION PointCameraUsingMouse(camno)
REM *** Get current camera angles ***
XcurrentAngle# = CAMERA ANGLE X(camno)
YcurrentAngle# = CAMERA ANGLE Y(camno)
REM *** Calculate the new angle based on mouse movement ***
XnewAngle# = XcurrentAngle#+ MOUSEMOVEY()
YnewAngle# = YcurrentAngle#+ MOUSEMOVEX()
REM *** Rotate camera ***
XROTATE CAMERA camno,XnewAngle#
YROTATE CAMERA camno,YnewAngle#
ENDFUNCTION
The routine stores the current angles of rotation and then calculates new angles
based on the movement of the mouse. Notice that mouse movement in the y
direction changes the angle of rotation about the x-axis, while mouse movement in
the x direction modifies the angle about the y-axis!
The routine is put to use in LISTING-33.5 which allows the user total freedom of
rotation in both the x and y axes.

REM *** set up screen ***
Camera Control Using SET DISPLAY MODE 1280,1024,32
the Mouse AUTOCAM OFF
DrawCastle()
REM *** Re-position camera ***

LISTING-33.5 REM *** Give user camera control ***
(continued) DO
PointCameraUsingMouse(0)
Camera Control Using LOOP
the Mouse
END
REM *** Rotates specified camera ***

XnewAngle# = XcurrentAngle#+ MOUSEMOVEY()
YnewAngle# = YcurrentAngle#+ MOUSEMOVEX()
ENDFUNCTION
Activity 33.20
Are there any problems with the rotation control?
Rotation is probably a bit too sensitive to mouse movement but we can overcome
this by reducing the effect of the mouse move values by changing the new angle
calculation statements to:
XnewAngle# = XcurrentAngle# + MOUSEMOVEY()*0.1

YnewAngle# = YcurrentAngle# + MOUSEMOVEX()*0.1
Rotation about the y-axis should be kept in the range 0 to 359.9 and we can do this
by applying the WRAPVALUE statement before assigning the value of the new
angle:
YnewAngle# = WRAPVALUE(YcurrentAngle# + MOUSEMOVEX()*0.1)
Given the restriction of a human's head movements, rotation about the x-axis should
probably remain within the range -90 to +90 so this angle should be calculated using
the lines:
XnewAngle# = XcurrentAngle# + MOUSEMOVEY()*0.1

IF XnewAngle# < -90
XnewAngle# = -90
ELSE
IF XnewAngle# > 90
XnewAngle# = 90
ENDIF
ENDIF
Activity 33.21
Make the changes described above to your previous program and check out
how this affects the performance of the camera.
Now that we can point the camera in any direction, the next thing we need to do is

move the camera. We already know how to do this using the CONTROL CAMERA
USING THE ARROW KEYS statement.
Activity 33.22
Add the line
CONTROL CAMERA USING ARROW KEYS 0,0.5,0
to the DO..LOOP structure in your last program.
Does this change create any problems?
As we can see from the last Activity, the CONTROL CAMERA USING
ARROWKEYS statement always returns the camera to zero degrees rotation about
the x-axis, effectively stopping us from pointing the camera up or down.
One way to deal with this is to write our own movement routine. The routine given
below takes the camera involved and the distance moved each time as parameters.
The camera moves in the direction it is pointing by pressing the left mouse button;
if the right mouse button is pressed, the camera moves backwards.
FUNCTION MoveCameraUsingMouse(camno, distance#)

IF MOUSECLICK() = 1
MOVE CAMERA camno,distance#
ELSE
IF MOUSECLICK() = 2
MOVE CAMERA camno,-distance#
ENDIF
ENDIF
ENDFUNCTION
Activity 33.23
Add the above function to your last program and replace the CONTROL
CAMERA USING ARROWKEYS line with the code
MoveCameraUsingMouse(0,1)
This gives us more freedom of movement than we might want for a player, but it is
useful when checking the layout of our virtual world.
Activity 33.24
By temporarily rotating the camera to zero degrees about the x-axis during
movement, modify the MoveCameraUsingMouse() function so that movement
is always parallel to the x-z plane irrespective of camera elevation.
To stop the camera moving through solid objects, add an AUTOMATIC

CAMERA COLLISION statement immediately before the DO..LOOP
structure in the main section of the program.
To stop the camera moving outside the modelled area, modify the
MoveCameraUsingMouse() function to limit the position of the camera to be
within the area bounded by the x-z points (-50,0) (500,700).

Activity 33.25
These mouse-camera control functions will prove useful elsewhere, so it will

be worthwhile saving these functions from the last program in a separate file
then using the #INCLUDE statement to add them to future programs.
Remove the code which limits movement to the area (-50,0),(500,700) in

MoveMouseUsing Camera().
Create a new project called Camera.dbpro.
Copy the code for the two functions, MoveCameraUsingMouse() and

PointCameraUsingMouse(), into the program file and then save and close the
project.
Summary
l Use AUTOCAM OFF to stop the camera being positioned automatically by the
program.
l Use AUTOCAM ON to reinstate automatic camera positioning.
l Use SET CAMERA TO FOLLOW to make the camera move smoothly so that
it is positioned to focus on a specific spot.
l Use CONTROL CAMERA USING ARROWKEYS to allow the user to control

x-z movement of the camera using the arrow keys.
l Use AUTOMATIC CAMERA COLLISION to stop the camera moving through

solid objects.
l For more flexible camera control, write your own functions.

Multiple Cameras
Introduction
So far we have used only the default camera (camera zero) which exists in every
DarkBASIC Pro 3D program, but it is possible to use multiple cameras. In the
statements that follow we'll see how this is done and how we can use several cameras
to enhance a game.
Using Additional Cameras

The MAKE CAMERA Statement
We can create a new camera using the MAKE CAMERA statement which has the
FIG-33.33
The MAKE CAMERA MAKE CAMERA camno
Statement
In the diagram:
camno is an integer value specifying the ID to be

assigned to the new camera. No two cameras may
be assigned the same ID.
For example, we could create a camera with ID 1 using the statement:
MAKE CAMERA 1
When a new camera is created, it is positioned at (0,0,0) and points in a positive

direction along the z-axis. The output from the camera is automatically displayed
on the screen, replacing the display from any previous camera.
The program in LISTING-33.6 demonstrates this effect.

LISTING-33.6
Creating a new Camera
BACKDROP ON
AUTOCAM OFF
DrawCastle()
REM *** Re-position the default camera ***

POSITION CAMERA -20,8,10
REM *** Make new camera ***

WAIT KEY
MAKE CAMERA 1

WAIT KEY
END

Activity 33.26
Why does the grass and path not show in the image from the second camera?
Notice that the background has turned green! Camera zero defaults to a blue
background, while camera 1's default background is always green.
Why did the background change colour when we saw the output from the second
camera? It turns out that the COLOR BACKDROP statement specifies the
background colour for the current camera only. In fact, the BACKDROP and
COLOR BACKDROP statements have a syntax slightly different from that given
back in volume 1, because both allow a camera to be specified. The updated version
of each statement is given below.
The COLOR BACKDROP Statement

This statement can be used to specify the background colour displayed by a specific
camera. The statement has the format shown in FIG-33.34.
FIG-33.34
The COLOR
BACKDROP Statement
COLOR BACKDROP [ camno , colour
In the diagram:
camno is an integer value giving the ID of the camera

whose background colour is to be changed. If
this value is omitted, it is the current camera that
is affected by the statement.
colour is an integer value representing the new

background colour.
To set the background colour for camera 1 to black we would use the statement
COLOR BACKGROUND 1,0
but for other colours we would probably use an RGB expression, as in the line:
COLOR BACKGROUND 1,RGB(255,255,0)
The BACKDROP Statement

The background colour specified using the COLOR BACKDROP statement is
activated (or deactivated) using the BACKDROP statement which has the format
shown in FIG-33.35.
FIG-33.35
The BACKDROP
Statement BACKDROP [ camno ,
{ ON
OFF

In the diagram:

whose background colour is to be (de)activated.
If this value is omitted, it is the current camera
that is affected by the statement.
ON Use this option to activate the background colour.
OFF Use this option to deactivate the background

colour and return to the default colour.
Activity 33.27
Modify your last program so that the background colour for camera 1 matches
that of camera zero.
The SET CURRENT CAMERA Statement

When we have more than one camera available, we can choose which one is the
active camera using the SET CURRENT CAMERA statement which has the format
shown in FIG-33.36.
FIG-33.36
The SET CURRENT SET CURRENT CAMERA camno
CAMERA Statement
In the diagram:

be the active camera.
When a program begins, camera zero is the active camera. Camera zero remains the
active camera even after other cameras have been created. Once a new active camera
has been specified using SET CURRENT CAMERA, any subsequent
camera-related statement which does not explicitly specify a camera number will
be assumed to be acting on this camera.
In LISTING-33.7 the previous program is modified so that camera 1 is moved above

and to the back of the castle. The view from camera 1 is activated by a key press.
Setting the Current REM *** set up screen ***

Camera SET DISPLAY MODE 1280,1024,32
BACKDROP ON
AUTOCAM OFF
DrawCastle()
REM *** Re-position camera zero ***

REM *** Make and position new camera ***

WAIT KEY

LISTING-33.7 MAKE CAMERA 1
(continued) SET CURRENT CAMERA 1
Setting the Current POINT CAMERA 200,0,0
Camera
WAIT KEY
END
Activity 33.28
Modify your previous program to match LISTING-33.7.
The DELETE CAMERA Statement

A camera's details can be deleted from RAM using the DELETE CAMERA
FIG-33.37
The DELETE CAMERA DELETE CAMERA camno
Statement
In the diagram:
camno is an integer value specifying the camera to be

deleted. Camera zero cannot be deleted.
For example, we could delete camera 1 using the statement:
DELETE CAMERA 1
Switching Between Cameras

The current version of DarkBASIC Pro (version 1.060) has no direct way of
switching between cameras. That is to say, if the computer screen is currently
showing the output from camera 2, there is no simple way to switch back to showing
the output from camera 1.
Since the output shown does change when a new camera is created, the only way
to switch between cameras is to delete and then recreate a camera. We also need to
save the position of the camera and the direction in which it is pointing so that the
recreated camera is correctly placed.
A function to perform this task is given below:
FUNCTION SwitchCamera(camno)
REM *** Save camera position ***
x# = CAMERA POSITION X(camno)
y# = CAMERA POSITION Y(camno)
z# = CAMERA POSITION Z(camno)
REM *** Save camera direction ***
angx# = CAMERA ANGLE X(camno)
angy# = CAMERA ANGLE Y(camno)
angz# = CAMERA ANGLE Z(camno)
REM *** Delete and recreate camera ***
DELETE CAMERA camno
MAKE CAMERA camno
REM *** Set camera as active camera ***
SET CURRENT CAMERA camno
REM *** Restore original position and direction ***

POSITION CAMERA camno,x#,y#,z#
ROTATE CAMERA camno,angx#,angy#,angz#
ENDFUNCTION
Obviously, the camera in question must already exist before the function is called.
Ideally, the function should start with the lines
IF NOT CAMERA EXIST(camno)

RETURN
ENDIF
but there is currently no CAMERA EXIST statement in DarkBASIC Pro.
After the SwitchCamera() function has been called, it will be necessary to reset the
background colour. For example, to switch to camera 2 with a red background we
would use the lines:
SwitchCamera(2)
BACKDROP ON
The program in LISTING-33.8 demonstrates the use of camera switching by

allowing the user to switch between cameras 1 and 2 using the "1" and "2" keys.
Switching Cameras
REM *** set up screen and use camera 1***
MAKE CAMERA 1
SET CURRENT CAMERA 1
BACKDROP ON
AUTOCAM OFF
DrawCastle()
REM *** Re-position camera 1 ***

REM *** Make and position camera 2 ***

WAIT KEY
MAKE CAMERA 2
SET CURRENT CAMERA 2
BACKDROP ON
REM *** Allow user to switch between cameras ***

DO
REPEAT
c$ = INKEY$()
UNTIL C$ = "1" OR C$ = "2"
SwitchCamera(VAL(c$))
BACKDROP ON
LOOP

END

LISTING-33.8 FUNCTION SwitchCamera(camno)
(continued) REM *** Save camera details ***
x# = CAMERA POSITION X(camno)
SwitchingCameras y# = CAMERA POSITION Y(camno)
z# = CAMERA POSITION Z(camno)
REM *** Save camera direction ***

angx# = CAMERA ANGLE X(camno)
angy# = CAMERA ANGLE Y(camno)
angz# = CAMERA ANGLE Z(camno)
REM *** Delete and recreate camera ***

DELETE CAMERA camno
MAKE CAMERA camno
REM *** Set camera as active camera ***

SET CURRENT CAMERA camno
REM *** Restore original position and direction ***

POSITION CAMERA camno,x#,y#,z#
ROTATE CAMERA camno,angx#,angy#,angz#
ENDFUNCTION
Activity 33.29
The SwitchCamera() function will be useful elsewhere. Add this function to

the Camera.dbpro project you created back in Activity 33.25.
Multiple Camera Output

It is possible to display the output from two or more cameras at the same time by
defining different output areas for each camera using the SET CAMERA VIEW
statement. The output shown in FIG-33.38 is produced by the program given in
LISTING-33.9.
FIG-33.38
Simultaneous Output
from Multiple Cameras

Picture-in-Picture SET DISPLAY MODE 1280,1024,32
AUTOCAM OFF
DrawCastle()

REM *** Create a second camera ...***

MAKE CAMERA 1
POINT CAMERA 1,200,0,0
REM *** ... with its own output area ***

SET CAMERA VIEW 1, 800,0,1279,300

WAIT KEY
END
Activity 33.30
Modify the program so that the user can control movement of camera 0 using
the mouse.
Modify the program again so that camera 1 slowly rotates to the left.
The CLEAR CAMERA VIEW Statement

We can clear the screen area to which a camera displays its output using the CLEAR
CAMERA VIEW statement which has the format shown in FIG-33.39.
FIG-33.39
The CLEAR CAMERA
VIEW Statement
CLEAR CAMERA VIEW camno , colour
In the diagram:

view port is to be cleared. If no camera ID is
given, the active camera is assumed.
colour is an integer value specifying the colour to be

used when clearing the view port.
We could clear the view area of camera 1 to red using the line:
CLEAR CAMERA VIEW 1, RGB(255,0,0)
Since the camera's output is recalculated for every frame, and the effect of the
CLEAR CAMERA VIEW statement only lasts for a single frame, there doesn't
seem to be much use for this statement.

Summary
l Use MAKE CAMERA to create a new camera.
l When a new camera is created, the output of that camera is displayed on the
screen.
l A new camera does not automatically become the active camera.
l Use COLOR BACKDROP and BACKDROP ON to select the background

colour for a camera's output.
l Use SET CURRENT CAMERA to set the active camera.
l Any statement in which the camera number is omitted is assumed to relate to the
active camera.
l Use DELETE CAMERA to delete a camera from RAM.
l Camera zero cannot be deleted.
l The output of multiple cameras can appear on the screen at the same time by
allocating different areas of the screen to each camera.
l Use CLEAR CAMERA VIEW to clear the camera's output area using a specified
colour.

Advanced Camera Techniques
Introduction
Some camera-related statements can be used in conjunction with other statements
to recreate effects such as mirrors or televisions or to show our character carrying
a weapon.
The Statements
The SET CAMERA TO IMAGE Statement
A rather neat trick is to display the output from a camera on the surface of an object.
This allows us to create things like mirrors or TV monitors. To achieve this affect
we need to first transfer the camera's output to an image object and then texture the
3D object with that image. The first stage in this process is achieved using the SET
CAMERA TO IMAGE statement whose format is given in FIG-33.40.
FIG-33.40 The SET CAMERA TO IMAGE Statement
SET CAMERA TO IMAGE camno , imgno , width , height
In the diagram:

output is to be mapped onto an image object.
imgno is an integer value specifying the ID of the image

to which camera output is to be sent. If an image
with this ID number does not already exist, it will
be created automatically.
width is an integer value giving the width of the image

object in pixels. This value must be a power of 2
(that is, 64, 128, 256, etc.)
height is an integer value giving the height of the image

object in pixels. This value must be a power of 2.
For example, we could output camera 1's display to image 10 which is set to a width
of 128 by 128 using the line:
SET CAMERA TO IMAGE 1,10,128,128
If image 10 does not exist when the statement is executed, it will be created.
However, if it does exist, its contents will be modified by this statement.
Next we need to use the image to texture a 3D object. For example, we could texture
object 5 with our new image using the line:
TEXTURE OBJECT 5,10

To make life easier we could create a function to do both stages, specifying the
camera number, image number, and object number to be used:
FUNCTION CameraOutputToObject(camno,objno,imgno)
SET CAMERA TO IMAGE camno,imgno,512,512
TEXTURE OBJECT objno,imgno
ENDFUNCTION
If you wanted to allow the dimensions of the image to be set, these could be added
to the parameter list for the function.
The program in LISTING-33.10 makes use of this routine to display the output of
camera 1 onto a cube placed in the grassy area in front of the castle.
#INCLUDE "Camera.dba"
Showing Camera Output
on an Object's Surface REM *** set up screen ***
AUTOCAM OFF
DrawCastle()
CreateReflectiveCube()
REM *** Create and use camera 2 ***

MAKE CAMERA 2
SwitchCamera(2)
REM *** Let user control camera ***

DO
LOOP

END
FUNCTION CreateReflectiveCube()
POSITION OBJECT 800,200,5.0001,50
REM *** Create camera inside cube ***
MAKE CAMERA 1
POSITION CAMERA 1,200,5.001,50
POINT CAMERA 1, -200,5.001,50
SET CAMERA VIEW 1, 0,0,10,10
COLOR BACKDROP 1, RGB(255,0,0)
BACKDROP ON 1
REM *** Texture cube with camera's output ***
CameraOutputToObject(1,800,222)
ENDFUNCTION
FUNCTION CameraOutputToObject(camno,objno, imgno)

ENDFUNCTION
Activity 33.31

Since all six sides of the cube show the same camera output, it is really only acting
as a mirror on the side which points in the same direction as the camera. To create
a more realistic mirror, we could show the camera's output on a plane.
Activity 33.32
In your previous program, replace the cube with a plane. The plane should be
perpendicular to the grass and facing in the same direction as camera 1.
Modify the main section so that the plane rotates slowly about its y-axis.
The SET CAMERA TO OBJECT ORIENTATION Statement

We can make a camera point in the same direction as a specific object using the
SET CAMERA TO OBJECT ORIENTATION statement. This statement has the
FIG-33.41 The SET CAMERA TO OBJECT ORIENTATION Statement
SET CAMERA TO OBJECT ORIENTATION camno , objno
In the diagram:
camno is an integer value specifying the camera

involved.
objno is an integer value specifying the ID of the object

whose orientation the camera is to match.
We can make use of this statement in the last program, making camera 1 always
point in the same direction as the plane. That way the illusion of a mirror is
maintained. We can achieve the effect using the statement:
SET CAMERA TO OBJECT ORIENTATION 1,800
The statement needs to be executed each time the object specified in the statement
moves.
Activity 33.33
In your last program, add the line given above as the last statement within the
DO..LOOP structure and observe how the image reflected by the plane
changes as the plane rotates.
The SET OBJECT TO CAMERA ORIENTATION Statement

You can't have failed to notice that the trees along the edge of the castle's pathway
look a little two dimensional and that shouldn't really be a surprise since they are
nothing more than textures on a plane. However, we can hide this 2D aspect of the
trees by making sure that the trees always face towards the camera. That way we
never get to see them edge on. We can achieve this in DarkBASIC Pro by using
the SET OBJECT TO CAMERA ORIENTATION statement which turns an object
to face in the same direction as the active camera. The statement has the format
shown in FIG-33.42.

FIG-33.42
SET OBJECT TO CAMERA ORIENTATION objno
The SET OBJECT TO
CAMERA
ORIENTATION Statement
In the diagram:

whose orientation is to be made to match that of
the active camera.
The following function turns all the trees in the castle scene to face away from the
player (hence they always have their back to the player).
FUNCTION OrientTrees()
SET OBJECT TO CAMERA ORIENTATION c
NEXT c
ENDFUNCTION
Activity 33. 34
Add the function given above to your last program.
Add a call to the function as your last statement in the DO..LOOP structure in
the main section of the program.
The LOCK OBJECT Statement

In First Person Shooter (FPS) games we almost always see the weapon our character
is carrying. No matter how the character turns, twists, or jumps, the weapon is
always there in the same position.
This effect can be toggled on or off in DarkBASIC Pro using the LOCK OBJECT
statement. LOCK OBJECT ON locks a specified object to a fixed position on the
screen no matter how the camera moves. To return an object to normal, use the
LOCK OBJECT OFF option. The statement has the format shown in FIG-33.43.
FIG-33.43
The LOCK OBJECT
Statement LOCK OBJECT
{ ON
OFF
objno
In the diagram:
ON Use this option to lock an object to the screen.
OFF Use this option to return a locked object to normal.
objno is an integer value giving the ID of the object

whose position is to be locked/unlocked.
Once this command has been executed, the objected specified will stay at exactly
the same spot on the screen irrespective of camera movement, but its position can
still be modified using statements such as POSITION OBJECT and ROTATE
OBJECT.

In LISTING-33.11 a wand (constructed from a cylinder) is locked to the screen -
we'll assume our character is a wizard rather than a warrior!
A background cube gives some depth to the program and the user can move the
camera using the arrow keys.
LISTING-33.11 REM *** Set up screen ***

Fixing An Object's HIDE MOUSE
Screen Position CreateWand()
REM *** Create a background cube ***
REM *** Control camera ***
DO
CONTROL CAMERA USING ARROWKEYS 0,0.1,0.1
LOOP
END
FUNCTION CreateWand()
REM *** Assign ID names ***
#CONSTANT wandobj 777
#CONSTANT wandimg 777
REM *** Make wand ***
MAKE OBJECT CYLINDER wandobj,1
LOAD IMAGE "wand.jpg",wandimg
TEXTURE OBJECT wandobj,wandimg
REM *** Fix wand to screen ***
LOCK OBJECT ON wandobj
REM *** Scale, rotate and position the wand ***
SCALE OBJECT wandobj, 10,100,10
XROTATE OBJECT wandobj,45
POSITION OBJECT wandobj,0.5,-1,2
ENDFUNCTION
Activity 33.35
Add the CreateWand() function to the last version of camera10.dbpro and

place a call to the function immediately before the DO..LOOP in the main
section of the program.
Test the result. Are there any problems with the mirrored plane?
There are a couple of camera statements that make use of 3D vectors. These
statements are just alternatives to other camera statements we've already covered,
but are included here for completeness. 3D vectors themselves are covered in a later
chapter.
The SET VECTOR3 TO CAMERA POSITION Statement

We can save the coordinates of a camera to a 3D vector using the SET VECTOR3
TO CAMERA POSITION statement. This statement has the format shown in
FIG-33.44.

FIG-33.44 The SET VECTOR3 TO CAMERA POSITION Statement
SET VECTOR3 TO CAMERA POSITION 3DVector , camno
In the diagram:
3DVector is an integer value specifying the 3D vector in

which the camera's position is to be stored.

whose position is to be stored.
For example, we could store the current position of camera zero in a 3D vector using
the lines:
result = MAKE VECTOR3(1) `Create 3D vector

SET VECTOR3 TO CAMERA POSITION 1,0
The SET VECTOR3 TO CAMERA ROTATION Statement

The current angles of rotation of a camera (about all three of its axes) can be stored
in a single 3D vector using the SET VECTOR3 TO CAMERA ROTATION
FIG-33.45 The SET VECTOR3 TO CAMERA ROTATION Statement
SET VECTOR3 TO CAMERA ROTATION 3DVector , camno
In the diagram:
3DVector is an integer value specifying the 3D vector in

which the camera's rotation is to be stored.

whose rotation details are to be stored.
For example, we could store the current rotation details of camera zero in a 3D
vector using the lines:
result = MAKE VECTOR3(2) `Create 3D vector

SET VECTOR3 TO CAMERA POSITION 2,0
Summary
l Use SET CAMERA TO IMAGE to send a camera's output to an image object.
l An image containing a camera's output can be used to texture a 3D object.
l Use SET CAMERA TO OBJECT ORIENTATION to make a camera face in

the same direction as a specified object.

l Use SET OBJECT TO CAMERA ORIENTATION to make an object face in
the same direction as the active camera.
l Use LOCK OBJECT ON to fix an object position on the screen irrespective of

camera movement.
l Use LOCK OBJECT OFF to return an object to normal positioning.
l Use SET VECTOR3 TO CAMERA POSITION to store a camera's coordinates

in a 3D vector.
l Use SET VECTOR3 TO CAMERA ROTATION to store a camera's angles of

rotation in a 3D vector.

Solutions
Activity 33.1 NEXT c
WAIT KEY
The main section should be coded as: END
REM *** Include DrawCastle() function *** To make the camera move right up to the castle wall,
#INCLUDE "castle.dba" change the FOR loop to read
SET DISPLAY MODE 1280,1024,32 FOR c = 1 TO 750
DrawCastle()
WAIT KEY
POSITION CAMERA 0,0,5,50 Activity 33.5
WAIT KEY
END REM *** Include DrawCastle() function ***
#INCLUDE "castle.dba"
Make sure you have copied Castle.dba into the same folder REM *** Set screen resolution ***
as this program. SET DISPLAY MODE 1280,1024,32
DrawCastle()
WAIT KEY
The camera has moved closer to the castle so that the 3D REM *** Turn camera by increments ***
objects now occupy the whole screen. FOR angle = -45 TO 45
ROTATE CAMERA angle,0,0
WAIT 10
Activity 33.2 NEXT angle
The program should now read: WAIT KEY
END
REM *** Include DrawCastle() function ***
#INCLUDE "castle.dba" To rotate the camera about the y-axis change the
ROTATE CAMERA line to
DrawCastle()
WAIT KEY ROTATE CAMERA 0,angle,0
MOVE CAMERA 50
WAIT KEY Rotation about the z-axis requires the line
END
ROTATE CAMERA 0,0,angle
The camera moves 50 units from its starting position in the
direction it is facing when a key is pressed. The jump from
the starting position to the final position is instant. Activity 33.6
Activity 33.3 REM *** Set screen resolution ***
The latest version of the program reads: DrawCastle()
REM *** Start with zero rotation ***
REM *** Include DrawCastle() function *** ROTATE CAMERA 0,0,0
#INCLUDE "castle.dba" DO
REM *** Set screen resolution *** REM *** Rotate camera using keys ***
SET DISPLAY MODE 1280,1024,32 SELECT INKEY$()
DrawCastle() CASE "z"
WAIT KEY TURN CAMERA LEFT 1
REM *** Move camera by increments *** ENDCASE
FOR c = 1 TO 50 CASE "x"
MOVE CAMERA 1 TURN CAMERA RIGHT 1
WAIT 50 ENDCASE
NEXT c CASE "'"
WAIT KEY PITCH CAMERA UP 1
END ENDCASE
CASE "/"
PITCH CAMERA DOWN 1
Activity 33.4 ENDCASE
CASE "n"
REM *** Include DrawCastle() function *** ROLL CAMERA LEFT 1
#INCLUDE "castle.dba" ENDCASE
REM *** Set screen resolution *** CASE "m"
SET DISPLAY MODE 1280,1024,32 ROLL CAMERA RIGHT 1
DrawCastle() ENDCASE
WAIT KEY ENDSELECT
REM *** Point camera in direction of move ** LOOP
POINT CAMERA -25,5,700 REM *** End program ***
REM *** Move camera by increments *** END
FOR c = 1 TO 100
MOVE CAMERA 1
WAIT 50

CASE "m"
Activity 33.7 ROLL CAMERA RIGHT 0.1
REM *** Include DrawCastle() function *** ENDCASE
#INCLUDE "castle.dba" ENDSELECT
REM *** Set screen resolution *** REM *** Get camera angles of rotation ***
SET DISPLAY MODE 1280,1024,32 xang# = CAMERA ANGLE X(0)
POSITION CAMERA 0,0,0 yang# = CAMERA ANGLE Y(0)
POINT CAMERA 0,0,0 zang# = CAMERA ANGLE Z(0)
DrawCastle() REM *** Display angles ***
ROTATE CAMERA 0,0,0 SET CURSOR 100,400
DO PRINT "Angle about x-axis: ",xang#
REM *** Rotate camera on key press *** SET CURSOR 100,420
SELECT INKEY$() PRINT "Angle about y-axis: ",yang#
CASE "z" SET CURSOR 100,440
TURN CAMERA LEFT 0.1 PRINT "Angle about z-axis: ",zang#
ENDCASE LOOP
CASE "x" REM *** End program ***
TURN CAMERA RIGHT 0.1 END
ENDCASE
CASE "'" Since the camera does not output to the black area of the
PITCH CAMERA UP 0.1 screen, that area is not continually redrawn and so we have
ENDCASE the same problem with overwriting screen areas that we
CASE "/" encountered way back in Chapter 2. We can solve this by
PITCH CAMERA DOWN 0.1
ENDCASE
changing the display lines to read:
CASE "n"
ROLL CAMERA LEFT 0.1 SET TEXT OPAQUE
ENDCASE SET CURSOR 100,400
CASE "m" PRINT " "
ROLL CAMERA RIGHT 0.1 SET CURSOR 100,400
ENDCASE PRINT "Angle about x-axis: ",xang#
ENDSELECT SET CURSOR 100,420
REM *** Get camera angles of rotation *** PRINT " "
xang# = CAMERA ANGLE X(0) SET CURSOR 100,420
yang# = CAMERA ANGLE Y(0) PRINT "Angle about y-axis: ",yang#
zang# = CAMERA ANGLE Z(0) SET CURSOR 100,440
REM *** Display angles *** PRINT " "
SET CURSOR 100,100 SET CURSOR 100,440
PRINT "Angle about x-axis: ",xang# PRINT "Angle about z-axis: ",zang#
SET CURSOR 100,120
PRINT "Angle about y-axis: ",yang#
SET CURSOR 100,140 Activity 33.9
PRINT "Angle about z-axis: ",zang#
LOOP
END
POINT CAMERA 0,0,0
Activity 33.8 DrawCastle()
REM *** Include DrawCastle() function *** ROTATE CAMERA 0,0,0
#INCLUDE "castle.dba" REM *** Set camera output area ***
REM *** Set screen resolution *** SET CAMERA VIEW 0,0,1279,319
SET DISPLAY MODE 1280,1024,32 REM *** Set aspect ratio ***
POSITION CAMERA 0,0,0 SET CAMERA ASPECT 1280/320
POINT CAMERA 0,0,0 DO
DrawCastle() REM *** Rotate camera on key press ***
ROTATE CAMERA 0,0,0 SELECT INKEY$()
REM *** Set camera output area *** CASE "z"
SET CAMERA VIEW 0,0,1279,319 TURN CAMERA LEFT 0.1
DO ENDCASE
REM *** Rotate camera on key press *** CASE "x"
SELECT INKEY$() TURN CAMERA RIGHT 0.1
CASE "z" ENDCASE
TURN CAMERA LEFT 0.1 CASE "'"
ENDCASE PITCH CAMERA UP 0.1
CASE "x" ENDCASE
TURN CAMERA RIGHT 0.1 CASE "/"
ENDCASE PITCH CAMERA DOWN 0.1
CASE "'" ENDCASE
PITCH CAMERA UP 0.1 CASE "n"
ENDCASE ROLL CAMERA LEFT 0.1
CASE "/" ENDCASE
PITCH CAMERA DOWN 0.1 CASE "m"
ENDCASE ROLL CAMERA RIGHT 0.1
CASE "n" ENDCASE
ROLL CAMERA LEFT 0.1 ENDSELECT
ENDCASE

REM *** Get camera angles of rotation *** Ä
CAMERA POSITION Y(),",",
xang# = CAMERA ANGLE X(0) Ä
CAMERA POSITION Z(),")"
yang# = CAMERA ANGLE Y(0) LOOP
zang# = CAMERA ANGLE Z(0) REM *** End program ***
REM *** Display angles *** END
SET TEXT OPAQUE
SET CURSOR 100,400
The cube becomes invisible because the camera is placed
PRINT " "
SET CURSOR 100,400 at its centre (0,0,0).
PRINT "Angle about x-axis: ",xang#
SET CURSOR 100,420
PRINT " " Activity 33.13
SET CURSOR 100,420
PRINT "Angle about y-axis: ",yang# No solution required.
SET CURSOR 100,440
PRINT " "
SET CURSOR 100,440 Activity 33.14
PRINT "Angle about z-axis: ",zang#
LOOP
#CONSTANT arttexture 200
END
No solution required. AUTOCAM OFF
REM *** Draw castle ***
Activity 33.11 DrawCastle()
REM *** Show falling object ***
FallingObject()
REM ***Set screen resolution ***
WAIT KEY
REM *** Draw the castle ***
END
DrawCastle()
REM *** Create textured trans. sphere ***
REM *** Set the camera screen area ***
REM *** Set the aspect ratio ***
SET OBJECT TRANSPARENCY artifact,1
REM *** Point camera at turret ***
REM *** Change camera range ***
REM *** Point object towards ground ***
REM *** Zoom in on turret ***
REM *** Move object, stopping when ***
WAIT KEY
REM *** it hits the ground ***
FOR angle = 180 to 5 STEP -1
DO
SET CAMERA FOV angle
MOVE OBJECT artifact,8
WAIT 50
REM *** Train camera on object ***
NEXT angle
POINT CAMERA OBJECT POSITION
REM *** Move camera towards turret ***
Ä
X(artifact),
WAIT KEY
Ä
OBJECT POSITION Y(artifact),
SET CAMERA FOV 55
POINT CAMERA -10,75,200 Ä
OBJECT POSITION Z(artifact),
FOR move = 1 TO 580 IF OBJECT POSITION Y(artifact) <= 2
MOVE CAMERA 1 EXIT
WAIT 20 ENDIF
NEXT move LOOP
REM *** End program *** ENDFUNCTION
WAIT KEY
END
Activity 33.15
Activity 33.12 The final version of the FallingObject() function should
read as follows:
REM *** Switch off auto camera placing ***
AUTOCAM OFF FUNCTION FallingObject()
REM *** Set screen resolution *** REM *** Create textured trans. sphere ***
SET DISPLAY MODE 1280,1024,32 MAKE OBJECT SPHERE artifact,10
REM *** Get size of cube *** LOAD IMAGE "lattice.bmp",arttexture
INPUT "Enter size of cube (2 to 100):",size TEXTURE OBJECT artifact,arttexture
REM *** Create cube *** SET OBJECT TRANSPARENCY artifact,1
MAKE OBJECT CUBE 1, size REM *** Position artifact out in sky ***
REM *** Display position of camera *** POSITION OBJECT artifact, -200,300,900
DO REM *** Point object towards ground ***
SET CURSOR 100,100 POINT OBJECT artifact,300,0,30
PRINT "(",CAMERA POSITION X(),",", REM *** Move object, stopping when ***

XnewAngle# = XcurrentAngle#
Ä
REM *** it hits the ground ***
DO + MOUSEMOVEY()*0.1
MOVE OBJECT artifact,8 IF XnewAngle# < -90
REM *** Make camera follow object *** XnewAngle# = -90
SET CAMERA TO FOLLOW OBJECT POSITION ELSE
Ä
X(artifact), IF XnewAngle# > 90
Ä
OBJECT POSITION Y(artifact), XnewAngle# = 90
Ä
OBJECT POSITION Z(artifact), ENDIF
Ä
90,100,20,1.0,1 ENDIF
YnewAngle# = WRAPVALUE(YcurrentAngle#
EXIT Ä+ MOUSEMOVEX()*0.1)
ENDIF REM *** Rotate camera ***
LOOP XROTATE CAMERA camno,XnewAngle#
ENDFUNCTION YROTATE CAMERA camno,YnewAngle#
ENDFUNCTION
Activity 33.16 The camera now moves more slowly in reaction to mouse
movements.
The SET CAMERA TO FOLLOW line in your program
should read
Activity 33.22
SET CAMERA TO FOLLOW
ÄOBJECT POSITION X(artifact), The updated program is now:
ÄOBJECT POSITION Y(artifact),
ÄOBJECT POSITION Z(artifact), #INCLUDE "Castle.dba"
Ä90,100,20,100.0,1 REM *** set up screen ***
AUTOCAM OFF
The camera now swings smoothly into place from its
DrawCastle()
original position, rather than jumping to the new position REM *** Re-position camera ***
abruptly. POSITION CAMERA 0,8,10
With a smooth value of 20 the jump between the original DO
position and the new position is faster than with a setting PointCameraUsingMouse(0)
of 100. CONTROL CAMERA USING ARROWKEYS 0,0.5,0
LOOP
END
Activity 33.17
No solution required. REM *** Rotates the active camera ***
Activity 33.18 YcurrentAngle# = CAMERA ANGLE Y(camno)
Yes. The camera can move straight through solid objects.
Ä+ MOUSEMOVEY()*0.1
IF XnewAngle# < -90
XnewAngle# = -90
Activity 33.19 ELSE
IF XnewAngle# > 90
No solution required. XnewAngle# = 90
ENDIF
ENDIF
Activity 33.20 YnewAngle# = WRAPVALUE(YcurrentAngle#
The camera turns too quickly to gain accurate control.
Ä+ MOUSEMOVEX()*0.1)
Activity 33.21 ENDFUNCTION
REM *** set up screen *** The camera can no longer point up or down.
AUTOCAM OFF
DrawCastle() Activity 33.23
POSITION CAMERA 0,8,10 #INCLUDE "Castle.dba"
POINT CAMERA 0,8,200 REM *** set up screen ***
DO SET DISPLAY MODE 1280,1024,32
PointCameraUsingMouse(0) AUTOCAM OFF
LOOP DrawCastle()
REM *** End program *** REM *** Re-position camera ***
END POSITION CAMERA 0,8,10
REM *** Rotates the active camera *** POINT CAMERA 0,8,200
FUNCTION PointCameraUsingMouse(camno) DO
REM *** Get current camera angles *** PointCameraUsingMouse(0)
XcurrentAngle# = CAMERA ANGLE X(camno) MoveCameraUsingMouse(0,1)
YcurrentAngle# = CAMERA ANGLE Y(camno) LOOP

REM *** End program *** Xangle# = CAMERA ANGLE X(camno)
END REM *** Point camera along ground ***
XROTATE CAMERA camno,0
REM *** Rotates the active camera *** IF MOUSECLICK() = 1
FUNCTION PointCameraUsingMouse(camno) MOVE CAMERA camno,distance
REM *** Get current camera angles *** ELSE
XcurrentAngle# = CAMERA ANGLE X(camno) IF MOUSECLICK() = 2
YcurrentAngle# = CAMERA ANGLE Y(camno) MOVE CAMERA camno,-distance
XnewAngle# = XcurrentAngle# ENDIF
Ä
+ MOUSEMOVEY()*0.1 ENDIF
IF XnewAngle# < -90 REM *** Restore angle about x-axis ***
XnewAngle# = -90 XROTATE CAMERA camno,Xangle#
ELSE REM *** Keep camera within area ***
IF XnewAngle# > 90 x# = CAMERA POSITION X(0)
XnewAngle# = 90 z# = CAMERA POSITION Z(0)
ENDIF IF x#<-50
ENDIF x# = -50
YnewAngle# = WRAPVALUE(YcurrentAngle# ELSE
Ä
+ MOUSEMOVEX()*0.1) IF x#>500
REM *** Rotate camera *** x#=500
XROTATE CAMERA camno,XnewAngle# ENDIF
YROTATE CAMERA camno,YnewAngle# ENDIF
ENDFUNCTION IF z# < 0
z# = 0
FUNCTION MoveCameraUsingMouse ELSE
Ä(camno,distance#) IF z# > 700
IF MOUSECLICK() = 1 z# = 700
MOVE CAMERA camno,distance# ENDIF
ELSE ENDIF
IF MOUSECLICK() = 2 POSITION CAMERA 0,x#,
MOVE CAMERA camno,-distance# Ä
CAMERA POSITION Y(0),z#
ENDIF ENDFUNCTION
ENDIF
ENDFUNCTION
Activity 33.25
REM *** set up screen *** Activity 33.26
AUTOCAM OFF The grass and path are not visible because camera 1 has
DrawCastle()
been created at position (0,0,0) and is therefore edge on to
POSITION CAMERA 0,8,10 those two planes.
DO Activity 33.27
MoveCameraUsingMouse(0,1) No solution required.
LOOP
END Activity 33.28
REM *** Rotates the active camera ***
BACKDROP ON
AUTOCAM OFF
Ä
+ MOUSEMOVEY()*0.1
DrawCastle()
REM *** Re-position the default camera ***
IF XnewAngle# < -90
XnewAngle# = -90
ELSE
REM *** Make new camera ***
IF XnewAngle# > 90
WAIT KEY
XnewAngle# = 90
MAKE CAMERA 1
ENDIF
REM *** Make background blue ***
ENDIF
YnewAngle# = WRAPVALUE(YcurrentAngle#
Ä
+ MOUSEMOVEX()*0.1)
WAIT KEY
REM *** Rotate camera *** END
ENDFUNCTION
Activity 33.29
FUNCTION MoveCameraUsingMouse
Ä(camno, distance)
REM *** Save angle about x axis ***

#INCLUDE "Camera.dba" Activity 33.33
SET DISPLAY MODE 1280,1024,32 The DO..LOOP in the main section should now read:
AUTOCAM OFF
DrawCastle() DO
REM *** Re-position camera *** PointCameraUsingMouse(2)
POSITION CAMERA 0,8,10 MoveCameraUsingMouse(2,1)
POINT CAMERA 0,8,200 TURN OBJECT LEFT 800,0.1
REM *** Create a second camera ...*** SET CAMERA TO OBJECT ORIENTATION 1,800
MAKE CAMERA 1 LOOP
REM *** ... with its own output area ***
SET CAMERA VIEW 1, 800,0,1279,300
Activity 33.34
REM *** Control camera movement *** #INCLUDE "Castle.dba"
DO #INCLUDE "Camera.dba"
PointCameraUsingMouse(0) REM *** set up screen ***
MoveCameraUsingMouse(0,0.5) SET DISPLAY MODE 1280,1024,32
TURN CAMERA LEFT 1,0.1 AUTOCAM OFF
LOOP DrawCastle()
REM *** End program *** CreateReflectivePlane()
WAIT KEY REM *** Create and use camera 2 ***
END MAKE CAMERA 2
Remember to copy Camera.dba to the current folder. POINT CAMERA 2,0,8,200
SwitchCamera(2)
Activity 33.31 DO
No solution required. MoveCameraUsingMouse(2,1)
Activity 33.32 OrientTrees()
LOOP
#INCLUDE "Castle.dba" REM *** End program ***
#INCLUDE "Camera.dba" END
SET DISPLAY MODE 1280,1024,32 FUNCTION CreateReflectivePlane()
AUTOCAM OFF REM *** Make cube ***
DrawCastle() MAKE OBJECT PLAIN 800,10,10
CreateReflectivePlane() POSITION OBJECT 800,200,5.001,50
REM *** Create and use camera 2 *** POINT OBJECT 800,-200,5,50
MAKE CAMERA 2 REM *** Camera in middle of plane***
POSITION CAMERA 2,0,8,10 MAKE CAMERA 1
POINT CAMERA 2,0,8,200 POSITION CAMERA 1,200,5,50
SwitchCamera(2) POINT CAMERA 1, -200,5,50
REM *** Let user control camera *** SET CAMERA VIEW 1, 0,0,10,10
DO COLOR BACKDROP 1, RGB(255,0,0)
PointCameraUsingMouse(2) BACKDROP ON 1
MoveCameraUsingMouse(2,1) REM *** Texture with camera's output ***
TURN OBJECT LEFT 800,0.1 CameraOutputToObject(1,800,222)
LOOP ENDFUNCTION
END FUNCTION CameraOutputToObject
Ä(camno,objno, imgno)
FUNCTION CreateReflectivePlane() SET CAMERA TO IMAGE camno,imgno,512,1024
REM *** Make plane *** TEXTURE OBJECT objno,imgno
MAKE OBJECT PLAIN 800,10,10 ENDFUNCTION
POINT OBJECT 800,-200,5,50 FUNCTION OrientTrees()
REM *** Create camera in mid-plane*** FOR c = tree1 TO tree1 + 30
MAKE CAMERA 1 SET OBJECT TO CAMERA ORIENTATION c
POSITION CAMERA 1,200,5,50 NEXT c
POINT CAMERA 1, -200,5,50 ENDFUNCTION
BACKDROP ON 1 Activity 33.35
REM *** Camera's output as texture ***
CameraOutputToObject(1,800,222) #INCLUDE "Castle.dba"
ENDFUNCTION #INCLUDE "Camera.dba"
FUNCTION CameraOutputToObject SET DISPLAY MODE 1280,1024,32
(camno,objno, imgno) AUTOCAM OFF
SET CAMERA TO IMAGE camno,imgno,512,1024 DrawCastle()
TEXTURE OBJECT objno,imgno CreateReflectivePlane()

REM *** Create and use camera 2 ***
MAKE CAMERA 2
SwitchCamera(2)
CreateWand()
DO
OrientTrees()
LOOP
END
FUNCTION CreateReflectivePlane()
POINT OBJECT 800,-200,5,50
REM *** Camera in middle of plane***
MAKE CAMERA
POINT CAMERA 1, -200,5,50
BACKDROP ON 1
REM *** Texture with camera's output ***
CameraOutputToObject(1,800,222)
ENDFUNCTION
FUNCTION CameraOutputToObject
Ä(camno,objno, imgno)
ENDFUNCTION
FUNCTION OrientTrees()
SET OBJECT TO CAMERA ORIENTATION c
NEXT c
ENDFUNCTION
FUNCTION CreateWand()
REM *** Assign ID names ***
#CONSTANT wandobj 777
#CONSTANT wandimg 777
REM *** Make wand ***
MAKE OBJECT CYLINDER wandobj,1
LOAD IMAGE "wand.jpg",wandimg
TEXTURE OBJECT wandobj,wandimg
REM *** Fix wand to screen ***
LOCK OBJECT ON wandobj
REM *** Scale, rotate & position wand ***
SCALE OBJECT wandobj, 10,100,10
XROTATE OBJECT wandobj,45
POSITION OBJECT wandobj,0.5,-1,2
ENDFUNCTION
The wand also appears on the output of any other cameras

being used.

34
Changing How Objects React to Light
Creating Coloured Lights
Fog
Modifying a Light's Brightness
Positioning Lights
Types of Lighting
DarkBASIC Pro: Lighting 885

Lighting
Introduction
By adding the correct lighting, we can transform a scene from something ordinary
into a place with atmosphere. Lighting can alter a location from appearing bright
and happy to somewhere dark and sinsister. Take your time thinking about lighting
and transform your game!
Types of Lighting
There are several types of lighting, each of which are described below.
Ambient Lighting
An ambient light is one with no obvious source and no obvious direction. Ambient
light is everywhere and appears to come from every direction. It's the sort of light
you might get on a heavily overcast day in the middle of winter.
Point Lighting
Point lighting is light that originates from a specific point in space and casts light
equally in all directions. The further away we move from the source, the duller the
light becomes. This is similar to a naked light bulb.
Spot Lighting
A spot light originates from a point in
FIG-34.1
space and the light itself shines in a
Lighting specific direction creating a cone-shaped
light. The angle over which the light
shines can be set and the light consists of
a primary and secondary area, with the
Lighting secondary area containing a duller light
than the primary area. Again the light
diminishes the further we travel from the
source. We can see examples of
spotlights in old war movies, the
spotlights shining into the skies
searching for enemy bombers.
Ambient: Light comes from every direction; Point: light comes from a single point;
brightness constant over distance. brightness reduces over distance. Directional Lighting
A directional light originates from an
infinitely far point, but its brightness
does not diminish over distance.
Sunshine on a cloudless day is the
obvious example for this type of lighting.
Spot: light comes from a single point; Directional: light comes from a distant point;
Fig-34.1 shows the basic characteristics
focuses on a specific area;
light reduces over distance.
rays are parallel; of each type of lighting.
brightness constant
886 DarkBASIC Pro: Lighting

Lighting in DarkBASIC Pro
When you run a 3D program in DarkBASIC Pro, two light sources are automatically
included: a directional light and an ambient light. If we examine the image shown
in FIG-34.2, we can see that the amount of shadow on each side of the cube gives
an indication of the position of the directional light.
FIG-34.2
Shading Caused by the
Default Directional Light
Every light source in DarkBASIC Pro must be allocated an ID number (only the
ambient light is exempt - it has no ID value). The default directional light, created
automatically, is assigned the ID number zero.
The HIDE LIGHT Statement

An existing light can be switched off using the HIDE LIGHT statement which has
FIG-34.3
The HIDE LIGHT HIDE LIGHT lightno
Statement
In the diagram:
lightno is an integer value giving the ID of the light to be

turned off.
For example, the statement
HIDE LIGHT 0
would switch off the default directional light, leaving only the weak ambient light.
The program in LISTING-34.1 shows a cube with default lighting. By pressing any
key, the default directional light is switched off, leaving only the ambient light.

Switching of the Default BACKDROP ON
Directional Light
AUTOCAM OFF

LISTING-34.1 POINT CAMERA 0,0,0
(continued)
Switching of the Default MAKE OBJECT CUBE 2,5
Directional Light POSITION OBJECT 2,0,2.5,0
REM *** Hide directional light ***

WAIT KEY
HIDE LIGHT 0

WAIT KEY
END
Activity 34.1
Type in and test the program given in LISTING-34.1 (lights01.dbpro).
The SHOW LIGHT Statement

A light which has been switched off using the HIDE LIGHT statement can be
switched back on using the SHOW LIGHT statement. This statement has the format
shown in FIG-34.4.
FIG-34.4
The SHOW LIGHT SHOW LIGHT lightno
Statement
In the diagram:

turned on.
The SET AMBIENT LIGHT Statement

The brightness of the ambient light can be set using the SET AMBIENT LIGHT
FIG-34.5
The SET AMBIENT SET AMBIENT LIGHT perc
LIGHT Statement
In the diagram:
perc is an integer value giving the percentage setting

for the default ambient light. A value of zero will
switch off the ambient light, while a value of 100
will give maximum brightness. The default
ambient setting is 25.
For example, we could switch off the ambient light using the statement:
SET AMBIENT LIGHT 0
Activity 34.2
Modify your previous program so that the ambient light level gradually
changes from an initial setting of 100 down to 0.

The COLOR AMBIENT LIGHT Statement
The color of the ambient light can also be set using the COLOR AMBIENT LIGHT
FIG-34.6
The COLOR AMBIENT COLOR AMBIENT LIGHT colour
LIGHT Statement
In the diagram:
colour is an integer value giving the colour setting for

the ambient light. Often this value will be
supplied by a call to the RGB statement.
For example, we could change the ambient light to red using the line:
COLOR AMBIENT LIGHT RGB(255,0,0)
Activity 34.3
Modify the last program to use a blue ambient light.
The MAKE LIGHT Statement

A new light source can be created using the MAKE LIGHT statement which has
FIG-34.7
The MAKE LIGHT MAKE LIGHT lightno
Statement
In the diagram:

created. No two lights can be allocated the same
ID value.
Like any other 3D object, lights are initially positioned at (0,0,0). When first
created, every light is a white, point light.
In LISTING-34.2 the default directional light is switched off and a new light
created. The effect of the new light can be seen on the side of the cube displayed
by the program.

Adding a new Light COLOR BACKDROP RGB(20,20,20)
BACKDROP ON
REM *** Use programmed screen updating ***

SYNC ON

AUTOCAM OFF
POSITION CAMERA -10,-3,20
POINT CAMERA 0,0,0

LISTING-34.2 REM *** Make and position cubes ***
(continued) MAKE OBJECT CUBE 2,5
POSITION OBJECT 2,0,2.5,-10
Adding a new Light MAKE OBJECT CUBE 3,5
SYNC:SYNC
REM *** Switch off directional light ***

WAIT KEY
SYNC is used so that HIDE LIGHT 0
the PRINT output SYNC
remains on the screen PRINT "Directional light off"
between key presses. SYNC
REM *** Create new light ***

WAIT KEY
MAKE LIGHT 1
SYNC
PRINT "New light created"
SYNC

WAIT KEY
END
Activity 34.4
Type in and test the program in LISTING-34.2 (lights02.dbpro).
The DELETE LIGHT Statement

Any existing light can be deleted from RAM using the DELETE LIGHT statement
FIG-34.8
The DELETE LIGHT DELETE LIGHT lightno
Statement
In the diagram:

deleted.
The COLOR LIGHT Statement

By default lights cast a white light, but this can be changed to another colour using
the COLOR LIGHT statement which has the format shown in FIG-34.9.
FIG-34.9
The COLOR LIGHT
Statement COLOR LIGHT lightno ,
{ colour
red
y ,z green , blue
In the diagram:
lightno is an integer value giving the ID of the light

whose colour is to be modified.
colour is an integer value specifying the colour of the

light to be produced.

red,green,blue is a set of integer values giving the red, green
and blue component values of the required colour.
Activity 34.5
Modify your previous program so that the added light is yellow.
The POSITION LIGHT Statement

A light can be moved to a new position using the POSITION LIGHT statement.
The format of this statement is given in FIG-34.10.
FIG-34.10
The POSITION LIGHT POSITION LIGHT lightno , x , y , z
Statement
In the diagram:

moved.
x,y,z is a set of real values specifying the new

position of the light.
For example, we could move light 1 to position (2,3,4) using the line:
POSITION LIGHT 1,2,3,4
Activity 34.6
In your previous program, make light 1 travel slowly away from position
(0,0,0) by decrementing its y-ordinate value (use a DO..LOOP structure so the
light continues to move indefinitely). Remember to use SYNC to update the
screen.
The SET LIGHT RANGE Statement

Spot and point lights only illuminate a limited volume, with the light levels reducing
as we move away from the source. Just how far the light's brightness extends can
be set using the SET LIGHT RANGE statement which has the format given in
FIG-34.11.
FIG-34.11
The SET LIGHT SET LIGHT RANGE lightno , distance
RANGE Statement
In the diagram:

for which the distance the light penetrates is to be
set.
distance is a real number giving the distance in units over

which the light is effective.
For example, we could set the light from light 1 to fade to complete darkness at a

distance of 10 units from the source using the line:
SET LIGHT RANGE 1,10
Activity 34.7
In your previous program, reduce the range of light 1 to 50 and observe how
this affects the time it takes for the light on the cubes to fade.
The SET SPOT LIGHT Statement

Although every new light starts off as a point light, it can be changed to other types.
An existing light can be changed to a spotlight using the SET SPOT LIGHT
statement whose format is shown in FIG-34.12.
FIG-34.12
The SET SPOT LIGHT SET SPOT LIGHT lightno , innerangle , outerangle
Statement
In the diagram:
lightno is an integer value giving the ID of the

light to be changed to a spotlight.
innerangle is a real number giving the angle defining the

cone of full light.
outerangle is a real number giving the angle defining the

cone of decreasing light.
A visual representation of a spot light is shown in FIG-34.13.
FIG-34.13 The inner angle

defines a cone inside
Spot Light Explained which the light is at
full brightness
inner outer
angle angle
light source
The outer angle

defines a cone inside
which the light fades from
full light to zero light
The SET DIRECTIONAL LIGHT Statement

We can change a light to a directional light using the SET DIRECTIONAL LIGHT
FIG-34.14
The SET DIRECTIONAL
LIGHT Statement
SET DIRECTIONAL LIGHT lightno , x , y , z

In the diagram:

light to be changed to a directional light.
x,y,z is a set of real numbers giving the position to

which the light is directed. This assumes the
light source is currently positioned at (0,0,0).
Unlike point and spot lights, a directional light has no specific location point. Light
comes from an undefined point with all rays parallel. However, in order to easily
specify a direction for these rays, when a directional light is created it is assumed
to be at position (0,0,0). The direction of the light emitted is then parallel to the line
drawn between (0,0,0) and (x,y,z) - this last point being given in the SET
DIRECTIONAL LIGHT statement. For example, we could transform an existing
light 1 to a directional light and have its rays shining from directly above using the
line:
SET DIRECTIONAL LIGHT 1,0,-1,0
Note that the length of the line between (0,0,0) and the point given when creating
the light has no effect on the brightness of the light. Hence, the line
SET DIRECTIONAL LIGHT 1, 0,-30,0
creates light coming from exactly the same direction and brightness as the previous
statement.
The SET POINT LIGHT Statement

A light can be changed back to a point light using the SET POINT LIGHT statement
which has the format shown in FIG-34.15
FIG-34.15
The SET POINT LIGHT SET POINT LIGHT lightno , x , y , z
Statement
In the diagram:

light to be changed to a point light.
x,y,z is a set of real numbers giving the new position

of the light.
The POINT LIGHT Statement

NOTE: This statement
has nothing to do with Spotlights and directional lights can both be made to point in a specific direction
point light types! using the POINT LIGHT statement which has the format shown in FIG-34.16.
FIG-34.16
The POINT LIGHT
POINT LIGHT lightno , x , y , z
Statement
In the diagram:

light to be pointed to a specific spot.

x,y,z is a set of real numbers giving the point in
space to which the light is to be directed.
The program in LISTING-34.3 shines a spot light onto a sphere and then moves the
light gradually away from the sphere.

Directing a Spot Light COLOR BACKDROP RGB(20,20,20)
BACKDROP ON
SYNC ON

AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Make and position sphere ***

SYNC:SYNC

WAIT KEY
HIDE LIGHT 0
SYNC
PRINT "Directional light off - press any key"
SYNC
REM *** Create spot light ***

WAIT KEY
MAKE LIGHT 1
SET SPOT LIGHT 1,40,60
POINT LIGHT 1,0,0,0
SYNC
PRINT "Spot light created - press any key"
SYNC
WAIT KEY
REM *** Move point light down the y-axis ***

PRINT "Press key to move light"
SYNC
WAIT KEY
zpos = 20
DO
POSITION LIGHT 1,0,0,zpos
SYNC
INC zpos
WAIT 50
LOOP

WAIT KEY
END
Activity 34.8
Type in and test the program given in LISTING-34.3 (lights03.dbpro).
Replace the sphere with a 20 by 20 plane. How does this affect the results?
Why has the spot light stopped working when we changed from a sphere to a plane?

When DarkBASIC Pro calculates lighting effects, it does so for each polygon that
makes up a 3D shape. A single polygon can only have a single light shading. Planes
are constructed from only two polygons - not nearly enough to create the subtle
effects of a spot light. It only works on the sphere because it was constructed from
a huge number of polygons.
The ROTATE LIGHT Statement

A spot or directional light can also be rotated to specific angles about all three axes
using the ROTATE LIGHT statement whose format is given in FIG-34.17
FIG-34.17
The ROTATE LIGHT ROTATE LIGHT lightno , xangle , yangle , zangle
Statement
In the diagram:

light to be rotated.
xangle is a real number giving the absolute angle to

which the light is to be rotated about the x-axis.
yangle is a real number giving the absolute angle to

which the light is to be rotated about the y-axis.
zangle is a real number giving the absolute angle to

which the light is to be rotated about the z-axis.
Activity 34.9
Modify your previous program as follows:
1. Remove the plane and return to a sphere as previously defined.

2. Set the spot to have an inner cone angle of 2 and outer cone angle of 8.
3. Position the spot light at (0,0,40).
4. In the DO..LOOP structure rotate the spot light about its y-axis rather than
move it away from the sphere.
The SET LIGHT TO OBJECT POSITION Statement

A light can be moved to the position of some other object using the SET LIGHT
TO OBJECT POSITION statement which has the format shown in FIG-34.18.
FIG-34.18
The SET LIGHT TO SET LIGHT TO OBJECT POSITION lightno , objno
OBJECT POSITION
Statement
In the diagram:

light to be moved.
objno is an integer value giving the ID of the

object to whose coordinates the light is to be
moved. The light will be placed at the exact
centre of the object specified.

For example, if we want light 1 to be moved to the position of object 10, then we
would use the line:
SET LIGHT TO OBJECT POSITION 1,10
We could achieve the same effect using the rather more complex line:
POSITION LIGHT 1,OBJECT POSITION X(10),OBJECT POSITION Y(10),

ÄOBJECT POSITION Z(10)
The program in LISTING-34.4 creates a row of spheres. A small cone with an

accompanying point light then moves past the spheres. As the cone and its light
move, each sphere, in turn, is affected by the light.
Moving a Light with an
Object
BACKDROP ON
REM *** Programmed screen updates ***

SYNC ON

AUTOCAM OFF
POINT CAMERA 0,0,0

xpost = -24
FOR objno = 3 TO 8
CLONE OBJECT objno,2
POSITION OBJECT objno,xpost,0,0
xpost = xpost + 12
NEXT objno
REM *** Make and position cone ***

MAKE OBJECT CONE 10, 2
POSITION OBJECT 10, -36,0,-10
REM *** Hide default directional light ***

HIDE LIGHT 0

MAKE LIGHT 1
SET POINT LIGHT 1,-12,0,-10
DO
POSITION LIGHT 1,OBJECT POSITION X(10),OBJECT POSITION Y(10),
Ä
OBJECT POSITION Z(10)
MOVE OBJECT RIGHT 10,1
SYNC
WAIT 1
LOOP

END
Activity 34.10
Type in and test the program in LISTING-34.4 (lights04.dbpro). Change the

first line in the DO loop to a SET LIGHT TO OBJECT POSITION statement.

The SET LIGHT TO OBJECT ORIENTATION Statement
A spot or directional light can be made to face in the same direction as some other
object using the SET LIGHT TO OBJECT ORIENTATION statement which has
FIG-34.19
The SET LIGHT
TO OBJECT
SET LIGHT TO OBJECT ORIENTATION lightno , objno
ORIENTATION
Statement
In the diagram:

light to be moved.
objno is an integer value giving the ID of an

object. The light specified will be made to point
in the same direction as this object.
To make spot light 1 point in the same direction as object 10 we would use the
statement:
SET LIGHT TO OBJECT ORIENTATION 1,10
The identical effect could be achieved using the line:
ROTATE LIGHT 1, OBJECT ANGLE X(10), OBJECT ANGLE Y(10),

ÄOBJECT ANGLE Z(10)
The program in LISTING-34.5 is a variation on the previous program. This time a

cube is positioned out from the centre sphere. The cube rotates about its own y-axis
and a spot light rotates in synchronisation with the cube. A second light - a point
light - has been added to make the cube itself more visible.

Pointing a Light in the COLOR BACKDROP RGB(20,20,20)
Same Direction as an BACKDROP ON
Object
SYNC ON

AUTOCAM OFF
POINT CAMERA 0,0,0

xpost = -24
FOR objno = 3 TO 8
xpost = xpost + 12
NEXT objno

POSITION OBJECT 10, 0,0,-10

LISTING-34.5 REM *** Dull directed light by changing it to dark grey ***
(continued) COLOR LIGHT 0, RGB(50,50,50)
Pointing a Light in the REM *** Create spot light at cube's position ***
Same Direction as an MAKE LIGHT 1
Object SET SPOT LIGHT 1,10,30
POSITION LIGHT 1, 0,0,-10
POINT LIGHT 1,0,0,0
REM *** Illuminate cube with second light ***

MAKE LIGHT 2
SET POINT LIGHT 2, 2,4,-11
SET LIGHT RANGE 2, 10
REM *** Rotate cube and light about y-axis***

yrotate = 0
DO
YROTATE OBJECT 10,yrotate
ROTATE LIGHT 1,0,OBJECT ANGLE Y(10),0
SET CURSOR 100,100
PRINT OBJECT ANGLE Y(10)
yrotate = WRAPVALUE(yrotate + 1)
SYNC
WAIT 1
LOOP

END
Activity 34.11
Try using a SET LIGHT TO OBJECT ORIENTATION statement in place of

the ROTATE LIGHT statement within the DO..LOOP structure.
Retrieving Light Data

Various statements exist which allow us to retrieve information about lights. These
are listed below.
The LIGHT EXIST Statement

The LIGHT EXIST statement returns 1 if a light of a given ID value exists; if the
light does not exist, the statement returns zero. This statement has the format shown
in FIG-34.20.
FIG-34.20
The LIGHT EXIST
LIGHT EXIST ( lightno )
Statement
integer
In the diagram:

checked.

The LIGHT VISIBLE Statement
The LIGHT VISIBLE statement returns 1 if a specified light is currently switched
on (that is, visible); otherwise zero is returned. This statement has the format shown
in FIG-34.21.
FIG-34.21
The LIGHT VISIBLE
LIGHT VISIBLE ( lightno )
Statement
integer
In the diagram:

checked.
The LIGHT RANGE Statement

The LIGHT RANGE statement returns a real number giving the range setting (as
specified using the SET LIGHT RANGE statement) for a specified light. This
FIG-34.22
The LIGHT RANGE
LIGHT RANGE ( lightno )
Statement
real
In the diagram:

checked.
The LIGHT TYPE Statement

The LIGHT TYPE statement returns an integer value giving a specified light's type.
This statement has the format shown in FIG-34.23.
FIG-34.23
The LIGHT TYPE LIGHT TYPE ( lightno )
Statement
integer
In the diagram:

checked.

Possible values for the returned integer are:
1 - directional light
2 - point light
3 - spot light
The LIGHT POSITION Statement

We can discover the position of a light (point or spot) using the LIGHT POSITION
statement, each form of the statement returning a single ordinate (x, y or z). The
FIG-34.24
The LIGHT POSITION X
Statement
LIGHT POSITION Y ( lightno )
real
In the diagram:
XYZ Use the appropriate option to return the ordinate

required.

involved.
A real value is returned by this statement. We could record the exact position of
light 1 using the following statements:
x# = LIGHT POSITION X(1)

y# = LIGHT POSITION Y(1)
z# = LIGHT POSITION Z(1)
The LIGHT DIRECTION Statement

The point at which a directional or spot light is aimed can be determined using the
LIGHT DIRECTION statement. This statement, which can be used in any of three
forms, returns the x, y, or z ordinate of the point at which the light is aimed. The
FIG-34.25
The LIGHT DIRECTION X
Statement
LIGHT DIRECTION Y ( lightno )
real

In the diagram:
XYZ Use the appropriate option to return the ordinate

required.

involved.
A real value is returned by this statement. The value returned always lies between
-1 and 1. FIG-34.26 shows how that value is calculated.
FIG-34.26
Light Direction
How the Value Returned
by LIGHT DIRECTION
is Calculated
A light (spot or directional) is created at a point Imagine a beam, exactly 1 unit in length, pointing
in space. in the direction in which the light is facing.
End point
The end point of this line matches the values As the beam changes direction, so the values
returned by appropriate calls to the LIGHT returned by LIGHT DIRECTION also change to
DIRECTION statement. match the new position of the end point.
Activity 34.12
Modify lights05.dbpro so that it displays the direction of the light in terms of

the values returned by LIGHT DIRECTION rather than the angle of rotation.
Fog
The FOG Statement
We can create a fog-like effect using FOG ON and disable the effect using FOG
OFF. The FOG statement has the format shown in FIG-34.27.
FIG-34.27
The FOG Statement
FOG
{ ON
OFF

The FOG COLOR Statement
The fog's colour can be set using the FOG COLOR statement which has the format
shown in FIG-34.28.
FIG-34.28
The FOG COLOR
Statement FOG COLOR
{ colour
red
y ,z green , blue
In the diagram:
colour is an integer value giving the colour of the fog.
red,green,blue is a set of integer values giving the red, green

and blue component values for the required fog
colour.
The FOG DISTANCE Statement

The distance from the active camera at which the fog becomes so dense as to obscure
any objects at that range can be set using the FOG DISTANCE statement which has
FIG-34.29
FOG DISTANCE distance
The FOG DISTANCE
Statement
In the diagram:
distance is an integer value giving the distance from the

active camera at which objects become totally
obscured by the fog.
The program in LISTING-34.6 demonstrates the use of the fog statements. The
program creates an enclosing cube, inside which a line of spheres recede into the
distance. The fog is switched on and the fog density varied.
Creating Fog COLOR BACKDROP RGB(20,20,20)
BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Switch off default directional light ***

HIDE LIGHT 0

FOR objno = 2 to 8
POSITION OBJECT objno,0,0,objno*5
NEXT objno

LISTING-34.6 REM *** Create an enclosing cube ***
(continued) MAKE OBJECT CUBE 10,-200
Creating Fog
REM *** Create a light ***
MAKE LIGHT 1
REM *** Set up Fog ***

WAIT 1000
FOG ON
REM *** Colour fog ***
FOG COLOR RGB(60,10,90)
REM *** Vary fog density ***

FOR dist = 100 to 20 STEP -1
FOG DISTANCE dist
WAIT 10
NEXT dist

WAIT KEY
END
Activity 34.13
The SET OBJECT FOG Statement

Individual objects can be made to ignore fog effects (and hence appear totally
unaffected by the fog) using the SET OBJECT FOG statement. This statement has
FIG-34.30
The SET OBJECT FOG
SET OBJECT FOG objno , fogflag
Statement
In the diagram:

whose reaction to fog is to be changed.
fogflag is 0 or 1. When set to zero, the object is unaffected

by the fog; when set to 1, the object is affected by
fog as normal.
Activity 34.14
In your previous program, create a moving cube using the following code:
WAIT 2000
FOR move = 1 TO 60
MOVE OBJECT 11,1
WAIT 10
NEXT move continued on next page

Activity 34.14 (continued)
The code should be inserted immediately before the final WAIT KEY
statement.
Remove the FOR loop which varies the fog distance and replace it with a
fixed fog distance of 50.
Run the modified program, observing the result.
Insert the line
SET OBJECT FOG 11,0
before the cube begins moving and check what difference this makes to the
cube's appearance.
Lights and 3D Vectors

Again, there are a couple of statements that make use of 3D vectors. These
statements are just alternatives to other light statements we've already covered, but
are included here for completeness. 3D vectors themselves are covered in a later
chapter.
The statements are summarised in TABLE-34.1.
TABLE-34.1
Statement Parameters Returned Description
Lights and 3D Vectors
Statements SET VECTOR3 TO LIGHT POSITION vno, lightno vno The3D vector with ID vno is set to
hold the coordinates of lightno.
SET VECTOR3 TO LIGHT ROTATION vno, lightno vno The 3D vector with ID vno is set to
hold lightno's rotation about each axis.
Summary
l There are four types of lighting in DarkBASIC Pro:
Ø Ambient - rays from every possible direction; each surface lit

equally.
Ø Point - rays come from a single point in space and radiate in all
directions from that point. The brightness grows weaker
over distance.
Ø Spot - rays come from a single point and radiate in one direction
only. Rays form a cone shape with a bright inner cone and
duller outer cone. Brightness grows weaker over distance.
Ø Directional - light comes from an undefined distance point; all rays are
parallel. Brightness does not diminish over distance.
l DarkBASIC Pro programs automatically create an ambient light and a

directional light.

l The default ambient light has no ID value.
l The default directional light has ID zero.
l Use HIDE LIGHT to switch off a specific light.
l Use SHOW LIGHT to switch a light back on.
l Use SET AMBIENT LIGHT to set the brightness of the default ambient light.
l Use COLOR AMBIENT to set the colour of the ambient light.
l Use MAKE LIGHT to create a new light. The light created will be a point light.
l Use DELETE LIGHT to remove a light from RAM.
l Use COLOR LIGHT to set the colour of a light.
l Use POSITION LIGHT to place a light at a specific point in space (point and
spot only).
l Use SET LIGHT RANGE to specifiy how far a light's brightness carries (point
and spot only).
l Use SET SPOT LIGHT to change a light to a spot light.
l Use SET DIRECTIONAL LIGHT to change a light to a directional light.
l Use SET POINT LIGHT to change a light to a point light.
l Use POINT LIGHT to specify the direction of a light (spot and directional only).
l Use ROTATE LIGHT to rotate the direction of a light (spot and directional only).
l Use SET LIGHT TO OBJECT POSITION to move a light to the centre of a

specified object.
l Use SET LIGHT TO OBJECT ORIENTATION to make a light point in the same
direction as a specified object.
l Use LIGHT EXIST to check if a specified light exists.
l Use LIGHT VISIBLE to check if a specified light is switched on.
l Use LIGHT RANGE to check the range of a specified light (spot and point only).
l Use LIGHT TYPE to determine the type of a specified light.
l Use LIGHT POSITION to determine the position of a specified light (spot and
point only).
l Use LIGHT DIRECTION to determine the end point of a 1 unit beam emitted
from a specified light assumed to be at position (0,0,0) (spot and directional
only).
l Spot and point lights only work effectively on surfaces constructed from many
polygons.

l Use the FOG statement to switch a fog effect on or off.
l Use FOG COLOR to set the colour of the fog.
l Use FOG DISTANCE to set the distance at which fog becomes opaque.
l Use SET OBJECT FOG to determine how an object reacts to fog.
l The type of light to which a statement may be applied is shown below:
Statement Light
A D P S
COLOR AMBIENT LIGHT
COLOR LIGHT
DELETE LIGHT
HIDE LIGHT
LIGHT DIRECTION
LIGHT EXIST
LIGHT POSITION
LIGHT RANGE
LIGHT TYPE
LIGHT VISIBLE
MAKE LIGHT
POINT LIGHT
POSITION LIGHT
ROTATE LIGHT
SET AMBIENT LIGHT
SET DIRECTIONAL LIGHT

SET LIGHT RANGE
SET LIGHT TO OBJECT POSITION
SET LIGHT TO OBJECT ORIENTATION

SET POINT LIGHT
SET SPOT LIGHT
SET VECTOR3 TO LIGHT POSITION

SET VECTOR3 TO LIGHT ROTATION
SHOW LIGHT

Solutions MAKE OBJECT CUBE 2,5
Activity 34.1 POSITION OBJECT 2,0,2.5,-10
SYNC:SYNC
REM *** Set up screen *** HIDE LIGHT 0
SET DISPLAY MODE 1280,1024,32 SYNC
COLOR BACKDROP RGB(0,200,0) PRINT "Directional light off"
BACKDROP ON SYNC
REM *** Position camera *** REM *** Create new light ***
AUTOCAM OFF WAIT KEY
POSITION CAMERA -10,10,20 MAKE LIGHT 1
POINT CAMERA 0,0,0 REM *** Make light yellow ***
REM *** Make and position cube *** COLOR LIGHT 1,255,255,0
MAKE OBJECT CUBE 2,5 SYNC
POSITION OBJECT 2,0,2.5,0 PRINT "New light created"
REM *** Hide directional light *** SYNC
HIDE LIGHT 0 WAIT KEY
FOR amblight = 100 TO 0 STEP -1 END
SET AMBIENT LIGHT amblight
WAIT 50
NEXT amblight Activity 34.6
WAIT KEY REM *** Set up screen ***
END SET DISPLAY MODE 1280,1024,32
BACKDROP ON
Activity 34.3 SYNC ON
REM *** Set up screen *** REM *** Position camera ***
SET DISPLAY MODE 1280,1024,32 AUTOCAM OFF
COLOR BACKDROP RGB(0,200,0) POSITION CAMERA -10,-3,20
BACKDROP ON POINT CAMERA 0,0,0
REM *** Position camera *** REM *** Make and position cubes ***
AUTOCAM OFF MAKE OBJECT CUBE 2,5
POSITION CAMERA -10,10,20 POSITION OBJECT 2,0,2.5,-10
POINT CAMERA 0,0,0 MAKE OBJECT CUBE 3,5
REM *** Make and position cube *** POSITION OBJECT 3,0,10,0
MAKE OBJECT CUBE 2,5 SYNC:SYNC
POSITION OBJECT 2,0,2.5,0 REM *** Switch off directional light ***
REM *** Hide directional light *** WAIT KEY
WAIT KEY HIDE LIGHT 0
HIDE LIGHT 0 SYNC
REM *** Make ambient blue *** PRINT "Directional light off"
COLOR AMBIENT LIGHT RGB(0,0,255) SYNC
FOR amblight = 100 TO 0 STEP -1 REM *** Create new light ***
SET AMBIENT LIGHT amblight WAIT KEY
WAIT 50 MAKE LIGHT 1
NEXT amblight COLOR LIGHT 1,255,255,0
REM *** End program *** SYNC
WAIT KEY PRINT "New light created"
END SYNC
REM *** Move light away ***
y# = 0
Activity 34.4 DO
POSITION LIGHT 1,0,y#,0
No solution required. y# = y# - 0.5
SYNC
LOOP
REM *** Set up screen *** END
BACKDROP ON Activity 34.7
SYNC ON REM *** Set up screen ***
REM *** Position camera *** SET DISPLAY MODE 1280,1024,32
AUTOCAM OFF COLOR BACKDROP RGB(20,20,20)
POSITION CAMERA -10,-3,20 BACKDROP ON
POINT CAMERA 0,0,0 REM *** Use programmed screen updating ***
REM *** Make and position cubes *** SYNC ON

REM *** Position camera *** DO
AUTOCAM OFF POSITION LIGHT 1,0,0,zpos
POSITION CAMERA -10,-3,20 SYNC
POINT CAMERA 0,0,0 INC zpos
REM *** Make and position cubes *** WAIT 50
MAKE OBJECT CUBE 2,5 LOOP
POSITION OBJECT 2,0,2.5,-10 REM *** End program ***
MAKE OBJECT CUBE 3,5 WAIT KEY
POSITION OBJECT 3,0,10,0 END
SYNC:SYNC
REM *** Switch off directional light *** The spotlight effect appears to have stopped working.
WAIT KEY
HIDE LIGHT 0
SYNC Activity 34.9
PRINT "Directional light off"
SYNC REM *** Set up screen ***
REM *** Create new light *** SET DISPLAY MODE 1280,1024,32
WAIT KEY COLOR BACKDROP RGB(20,20,20)
MAKE LIGHT 1 BACKDROP ON
COLOR LIGHT 1,255,255,0 REM *** Use programmed screen updating
SYNC ***
PRINT "New light created" SYNC ON
SYNC REM *** Position camera ***
REM *** Set light range *** AUTOCAM OFF
SET LIGHT RANGE 1, 50 POSITION CAMERA 0,0,20
REM *** Move light away *** POINT CAMERA 0,0,0
y# = 0 REM *** Make and position sphere ***
DO MAKE OBJECT SPHERE 2,20,150,150
POSITION LIGHT 1,0,y#,0 SYNC:SYNC
y# = y# - 0.5 REM *** Switch off directional light ***
SYNC WAIT KEY
LOOP HIDE LIGHT 0
REM *** End program *** SYNC
WAIT KEY PRINT "Directional light off - press any
END key"
SYNC
MAKE LIGHT 1
REM *** Set up screen *** REM *** Set inner and outer cones ***
SET DISPLAY MODE 1280,1024,32 SET SPOT LIGHT 1,2,8
COLOR BACKDROP RGB(20,20,20) POSITION LIGHT 1,0,0,40
BACKDROP ON POINT LIGHT 1,0,0,0
REM *** Use programmed screen updating SET LIGHT RANGE 1,100
*** SYNC
SYNC ON PRINT "Spot light created - press any
REM *** Position camera *** key"
AUTOCAM OFF SYNC
POSITION CAMERA 0,0,20 WAIT KEY
POINT CAMERA 0,0,0 REM *** Move point light down the y-axis
REM *** Make and position plane *** ***
MAKE OBJECT PLAIN 2,20,20 PRINT "Press key to move light"
SYNC:SYNC SYNC
REM *** Switch off directional light *** WAIT KEY
WAIT KEY yangle# = 0
HIDE LIGHT 0 DO
SYNC ROTATE LIGHT 1,0,yangle#,0
PRINT "Directional light off - press any SYNC
key" yangle# = WRAPVALUE(yangle#+1)
SYNC WAIT 10
REM *** Create spot light *** LOOP
MAKE LIGHT 1 WAIT KEY
SET SPOT LIGHT 1,40,60 END
POINT LIGHT 1,0,0,0
You may have to adjust the increment to yangle# and
SYNC the wait period depending on the power of the machine
PRINT "Spot light created - press any you are using.
key"
SYNC
REM *** Move point light down the y-axis
*** REM *** Set up screen ***
PRINT "Press key to move light" SET DISPLAY MODE 1280,1024,32
SYNC COLOR BACKDROP RGB(20,20,20)
WAIT KEY BACKDROP ON
zpos = 20 REM *** Programmed screen updates ***

SYNC ON SET POINT LIGHT 2, 2,4,-11
REM *** Position camera *** SET LIGHT RANGE 2, 10
AUTOCAM OFF REM *** Rotate cube/light about y-axis***
POSITION CAMERA 0,0,-60 yrotate = 0
POINT CAMERA 0,0,0 DO
REM *** Make and position sphere *** YROTATE OBJECT 10,yrotate
MAKE OBJECT SPHERE 2,10,100,100 SET LIGHT TO OBJECT ORIENTATION 1,10
POSITION OBJECT 2,-36,0,0 SET CURSOR 100,100
xpost = -24 PRINT OBJECT ANGLE Y(10)
FOR objno = 3 TO 8 yrotate = WRAPVALUE(yrotate + 1)
CLONE OBJECT objno,2 SYNC
POSITION OBJECT objno,xpost,0,0 WAIT 1
xpost = xpost + 12 LOOP
NEXT objno REM *** End program ***
REM *** Make and position cone *** END
POSITION OBJECT 10, -36,0,-10 In DarkBASIC Pro (v6.0) the SET LIGHT TO OBJECT
REM *** Hide default directional light
ORIENTATION does not achieve the required effect.
***
HIDE LIGHT 0
MAKE LIGHT 1
Activity 34.12
SET POINT LIGHT 1,-12,0,-10 REM *** Set up screen ***
SET LIGHT RANGE 1,10 SET DISPLAY MODE 1280,1024,32
DO COLOR BACKDROP RGB(20,20,20)
SET LIGHT TO OBJECT POSITION 1,10 BACKDROP ON
MOVE OBJECT RIGHT 10,1 REM *** Use programmed screen updating ***
SYNC SYNC ON
WAIT 50 REM *** Position camera ***
LOOP AUTOCAM OFF
REM *** End program *** POSITION CAMERA 0,0,-60
END POINT CAMERA 0,0,0
In the version of DarkBASIC Pro used (v6.0), the MAKE OBJECT SPHERE 2,10,100,100
SET LIGHT TO OBJECT POSITION did not POSITION OBJECT 2,-36,0,0
operate correctly, so you will have to revert to the xpost = -24
FOR objno = 3 TO 8
original POSITION LIGHT statement.
xpost = xpost + 12
Activity 34.11 NEXT objno
REM *** Set up screen *** REM *** Make and position cube ***
SET DISPLAY MODE 1280,1024,32 MAKE OBJECT CUBE 10, 2
COLOR BACKDROP RGB(20,20,20) POSITION OBJECT 10, 0,0,-10
BACKDROP ON REM *** Dull directed light by changing it
REM *** Use programmed screen updating to dark grey ***
*** COLOR LIGHT 0, RGB(50,50,50)
SYNC ON REM *** Create spot light at cube's
REM *** Position camera *** position ***
AUTOCAM OFF MAKE LIGHT 1
POSITION CAMERA 0,0,-60 SET SPOT LIGHT 1,10,30
POINT CAMERA 0,0,0 POSITION LIGHT 1, 0,0,-10
REM *** Make and position spheres *** POINT LIGHT 1,0,0,0
MAKE OBJECT SPHERE 2,10,100,100 SET LIGHT RANGE 1,100
POSITION OBJECT 2,-36,0,0 REM *** Illuminate cube with new light ***
xpost = -24 MAKE LIGHT 2
FOR objno = 3 TO 8 SET POINT LIGHT 2, 2,4,-11
CLONE OBJECT objno,2 SET LIGHT RANGE 2, 10
POSITION OBJECT objno,xpost,0,0 REM *** Rotate cube/light about y-axis***
xpost = xpost + 12 yrotate = 0
NEXT objno DO
REM *** Make and position cube *** YROTATE OBJECT 10,yrotate
MAKE OBJECT CUBE 10, 2 ROTATE LIGHT 1,0,OBJECT ANGLE Y(10),0
POSITION OBJECT 10, 0,0,-10 dirx# = LIGHT DIRECTION X(1)
REM *** Dull directed light by diry# = LIGHT DIRECTION Y(1)
changing it to dark grey *** dirz# = LIGHT DIRECTION Z(1)
COLOR LIGHT 0, RGB(50,50,50) SET CURSOR 100,100
REM *** Create spot light at cube's PRINT dirx#," ",diry#," ",dirz#
position *** yrotate = WRAPVALUE(yrotate + 1)
MAKE LIGHT 1 SYNC
SET SPOT LIGHT 1,10,30 WAIT 1
POSITION LIGHT 1, 0,0,-10 LOOP
POINT LIGHT 1,0,0,0 REM *** End program ***
SET LIGHT RANGE 1,100 END
REM *** Illuminate cube with new light
***
MAKE LIGHT 2

Activity 34.13
Activity 34.14
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
HIDE LIGHT 0
FOR objno = 2 to 8
POSITION OBJECT objno,0,0,objno*5
NEXT objno
REM *** Create an enclosing cube ***
MAKE OBJECT CUBE 10,-200
REM *** Create a light ***
MAKE LIGHT 1
REM *** Set up Fog ***
WAIT 1000
FOG ON
REM *** Colour fog ***
REM *** Set fog distance to 50 ***
FOG DISTANCE 50
REM *** fly in cube ***
WAIT 2000
FOR move = 1 TO 60
MOVE OBJECT 11,1
WAIT 50
NEXT move
WAIT KEY
END
The cube is seen emerging from the fog.
When the line
SET OBJECT FOG 11,0
is added before the final FOR loop, the cube is unaffected

by the fog.

35
Accessing Limb Data
Adding a Mesh to an Object
Changing a Limb's Mesh
Constructing Models from a Root Object and Limbs
Creating an Opening Door
Creating and Using Meshes
Creating an Object from a Mesh
Glueing an Object to a Limb
Loading and Saving Meshes
Manipulating Limbs
Offsetting and Rotating Limbs
Saving a Model in DBO Format

Scaling Limbs
Texturing and Colouring Limbs
DarkBASIC Pro: Meshes and Limbs 911

Meshes
Introduction
A 3D object's details (its vertices, texture, etc.) can be held by the computer in many
forms. In DarkBASIC Pro, the commonest of these is the object format, but other
options are available. One format holds an object in the form of a mesh.
To differentiate between the two formats we will refer to them as objects and
meshes. The most obvious difference between a mesh and an object is that meshes
are not visible on screen and do not retain texture details!
However, a mesh can be saved to, or loaded from, a file.
Meshes in themselves are of limited use. Their main purpose is as building blocks
for more complex shapes, as we'll see later in this chapter.
Handling Meshes
The MAKE MESH FROM OBJECT Statement
To create a mesh we need to convert an existing object to mesh format. This is done
using the MAKE MESH FROM OBJECT statement which has the format shown
in FIG-35.1.
FIG-35.1
The MAKE MESH
MAKE MESH FROM OBJECT meshno , objno
FROM OBJECT
Statement
In the diagram:
meshno is an integer value specifying the ID to be

assigned to the mesh being created. No other
mesh in the program may be assigned the same
value.
objno is an integer value giving the ID of the existing

3D object whose details are to be used to a create
the mesh.
The 3D object is unaffected by this operation.
The program in LISTING-35.1 creates a sphere and uses this to create a mesh; the
sphere is then deleted, leaving us with a mesh, but nothing visible on the screen.

Creating a Mesh COLOR BACKDROP 0
BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0
912 DarkBASIC Pro: Meshes and Limbs

LISTING-35.1 REM *** Make sphere ***
(continued) MAKE OBJECT SPHERE 1,10
WAIT KEY
Creating a Mesh
REM *** Make mesh from sphere ***
MAKE MESH FROM OBJECT 1,1
REM *** Delete the original object ***

DELETE OBJECT 1

WAIT KEY
END
Activity 35.1
Type in and test the program given in LISTING-35.1 (limbs01.dbpro).
The SAVE MESH Statement

Once created, a mesh can be saved to a file using the SAVE MESH statement. The
format for this statement is given in FIG-35.2.
FIG-35.2
The SAVE MESH SAVE MESH filename , meshno
Statement
In the diagram:
filename is a string specifying the name to be given to the

file in which the mesh data is to be saved.
The string may include path information.
meshno is an integer value giving the ID of the mesh to

be saved.
Files are saved in DirectX format and, therefore, should be given a .x extension
when being named.
For example, to save mesh 1 to a file named sphere01.x in a folder named meshdata
on the C: drive, we would use the line:
SAVE MESH "C:\meshdata\sphere01.x",1
If the file already exists, its existing data will be overwritten; if the file does not
exist, it will be created. Any folders named in the path details must already exist or
the save operation will fail.
If path details are omitted, then the file will be saved in the current folder - normally,
this will be the folder containing the program code.
Activity 35.2
Modify your last program so that the sphere created is saved in a file called
sphere01.x in the current project's folder.
Modify the program to texture the sphere using marble.bmp before converting
it to a mesh. Save the resulting mesh in file sphere02.x.

The LOAD MESH Statement
A previously saved mesh (or most other 3D objects saved in .x format) can be loaded
into a program using the LOAD MESH statement which has the format given in
FIG-35.3.
FIG-35.3
The LOAD MESH LOAD MESH filename , meshno
Statement
In the diagram:
filename is a string specifying the name of the file

containing the mesh.
The string may include path information.
meshno is an integer value specifying the ID to be

assigned to the mesh when loaded.
For example, we could load the file sphere01.x (as created by the last program) into
a mesh with ID 1 using the line:
LOAD MESH "C:\limbs01\sphere01.x",1
The MAKE OBJECT Statement

If we are to make a loaded mesh visible, then we need to turn it back into a regular
object. This can be achieved using the MAKE OBJECT statement which has the
FIG-35.4
imgno
The MAKE OBJECT
Statement MAKE OBJECT objno , meshno ,
0
In the diagram:
objno is an integer value specifying the ID to be

meshno is an integer value specifying the ID of the mesh

to be used to create the object.
0 Use this option (zero) if no texture is to be applied

to the new object.
imageno is an integer value specifying the ID of the image

to be used to texture the object. This image must
have been loaded previously.
In the program in LISTING-35.2, the previously saved mesh, sphere01.x, is loaded

and converted to an object. A marble texture is applied to the new object.

Creating an Object from a COLOR BACKDROP 0
Mesh BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Load the mesh ***
LOAD MESH "sphere01.x",1
REM *** Convert mesh to object ***
This code assumes that
MAKE OBJECT 1,1,1
the file sphere01.x is in
the same folder as the
WAIT KEY
program code.
END
Activity 35.3
Make sure that the files sphere01.x, sphere02.x and marble.bmp have been
copied into the current project's folder.
Modify the program to load sphere02.x instead of sphere01.x. Is the object

created from loading sphere02.x textured?
The DELETE MESH Statement

A mesh can be erased from RAM using the DELETE MESH statement which has
FIG-35.5
DELETE MESH meshno
The LOAD MESH
Statement
In the diagram:

to be deleted.
After a mesh has been used to create an object, we could use this command to delete
the original mesh.
The MESH EXIST Statement

We can check to see if a mesh of a specific ID currently exists using the MESH
FIG-35.6
MESH EXIST ( meshno )
The MESH EXIST
Statement
integer
In the diagram:
meshno is an integer value specifying the mesh ID that is

to be checked.

If a mesh of the specified number exists, then 1 is returned by the statement,
otherwise zero is returned.
Summary
l A mesh is a 3D shape held in a different format from the more usual 3D object.
l Mesh shapes are not normally visible on the screen.
l Use MAKE MESH FROM OBJECT to create a mesh.
l Use SAVE MESH to save a mesh to a file.
l Use LOAD MESH to load a mesh file.
l Use MAKE OBJECT to create an object from an existing mesh.
l When creating an object from a mesh, a texture can be applied to the object once
it is created.
l Use DELETE MESH to delete an existing mesh from RAM.
l Use MESH EXIST to check if a mesh of a specified number currently exists.

Limbs
Introduction
A regular object and one or more meshes can be linked together to create a group
object or model. This can be useful, allowing us to do something as simple as
applying different textures to different parts of the group object, or to make the
moving of one object result in the automatic movement of a second object. When
a mesh is linked to a standard 3D object, the mesh is known as a limb and the original
object is known as the root.
Getting Started
At the core of any model is the basic root object. This can be created in the normal
way. We'll start by creating a sphere object:
MAKE OBJECT SPHERE 1,10
Now we can create a second object and convert it to a mesh, deleting the object
when we're done:
REM *** Make a box ...***

MAKE OBJECT BOX 2,10,10,0.1
REM *** ...and turn it into a mesh ***
REM *** Now delete the original box object ***
DELETE OBJECT 2
The ADD LIMB Statement

Using the ADD LIMB statement we can attach our mesh to the sphere object. Since
many meshes may be attached to a model, it is necessary to assign an attachment
(limb) number to each mesh. The format for the ADD LIMB statement is given in
FIG-35.7.
FIG-35.7
The ADD LIMB ADD LIMB objno , limbno , meshno
Statement
In the diagram:
objno is an integer value identifying the object to

which the mesh is being joined.
limbno is an integer value specifying the limb number to

be assigned to the mesh when it is joined to the
object. Each limb joined to the object
must have an unique limb number.
meshno is an integer value identifying the mesh from

which the limb is to be created.
For example, we could join our box mesh to the sphere object using the line

ADD LIMB 1,1,1
The program in LISTING-35.3 joins a box mesh to a sphere object and then allows
the user to move the camera to view the resulting model.

Adding a Limb to an COLOR BACKDROP RGB(0,200,0)
Object BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0

REM *** Make mesh from object ***

REM *** Delete original object ***

DELETE OBJECT 2
REM *** Attach mesh to object

ADD LIMB 1,1,1
REM *** Allow user to move camera ***

DO
CONTROL CAMERA USING ARROWKEYS 0,1,1
LOOP

END
Activity 35.4
The output of this program is shown in FIG-35.8.

FIG-35.8
Adding a Limb to a
Sphere
Notice that the limb's centre is positioned at the centre of the object to which it is
joined.
Once we have created a group object by adding a limb to a standard object, all the
items within that group become linked. As a result, if the root object is moved, its
limbs will also move.

Activity 35.5
In your last program add the lines
REM *** Move object ***

WAIT KEY
immediately after the mesh is attached to the sphere.
Run the program and check that the sphere's limb moves along with the sphere.
In fact, any other operation which is applied to the root object, affects all the items
in the group object. For example, using the line HIDE OBJECT 1 in the last
program, would hide not only the sphere, but also the limb attached to it. If we
colour or texture the object, its limbs will also take on the same colour or texture.
Activity 35.6
Using the WAIT KEY statement to initiate each action, modify your last
program so that the root object changes colour to red and then is textured
using the image marble.bmp.
The MAKE OBJECT FROM LIMB statement

It's possible to turn a limb into an object using the MAKE OBJECT FROM LIMB
FIG-35.9
The MAKE OBJECT FROM LIMB
Statement
MAKE OBJECT FROM LIMB objno , rootobjno , limbno
In the diagram:
objno is an integer value specifying the ID to be

assigned to the object that is to be created.
rootobjno is an integer value identifying the limb's root

object.
limbno is an integer value giving the number of

the limb being used to create an object.
For example, in the last program we could create a new object, 5, from limb number
1 of group object 1 using the line:
MAKE OBJECT FROM LIMB 5,1,1
The new object will not retain any colour or texture details from the root object.

Activity 35.7
In your last program, create object 2 from limb 1 of the group object. Move
the newly created object to position (20,0,0).
Once you've tested the program, return the program to its original coding.
The OFFSET LIMB Statement

When a limb is first added, its centre will be placed at the root object's centre.
However, the limb can be repositioned using the OFFSET LIMB statement. This
statement has the format given in FIG-35.10.
FIG-35.10
The OFFSET LIMB
OFFSET LIMB objno , limbno , x , y , z
Statement
In the diagram:
objno is an integer value giving the limb's root object.
limbno is an integer value specifying the limb to be

moved. A limb of this value must previously
have been attached to the root object.
x,y,z are real values giving the offset distances of the

limb along each axis, as measured from its original
position.
For example, we could move the box attached to the sphere to the back of that sphere
(a distance of 5 units along the z-axis) using the line:
OFFSET LIMB 1,1,0,0,5
Activity 35.8
Add the above line to your last program immediately after the limb is attached
to the sphere. How does this change the position of the limb?
Now we'll add a second box limb at the front of the sphere using the lines:
ADD LIMB 1,2,1

OFFSET LIMB 1,2,0,0,-5
Notice that the mesh used to create the first limb is also used to create the second
limb.
Activity 35.9
Add the above lines to your program and check out the position of the two
limbs.
The ROTATE LIMB Statement

In positioning a limb, it can be rotated about any of the three axes using the ROTATE

LIMB statement which has the format shown in FIG-35.11.
FIG-35.11 The ROTATE LIMB Statement
ROTATE LIMB objno , limbno , xang , yang , zang
In the diagram:
objno is an integer value identifying the limb's root

object.

rotated.
xang is a real number giving the angle of rotation

about the x-axis.
yang is a real number giving the angle of rotation

about the y-axis.
zang is a real number giving the angle of rotation

about the z-axis.
All angles are given in degrees.
The next lines create a third box limb on the sphere and place that limb on the base
of the sphere:
ADD LIMB 1,3,1

ROTATE LIMB 1,3,90,0,0
OFFSET LIMB 1,3,0,-5,0
As a result of adding this code, we get the model shown in FIG-35.12.
FIG-35.12
A Rotated Limb is Added
to the Sphere
Activity 35.10
Modify your last program so that a total of 6 limbs are added to the sphere
creating a cube around the sphere.
Remove the code which moves the model and allows the camera to be
controlled by the user.
Replace it with code which makes the model rotate about all three of its axes.

The SCALE LIMB Statement
An individual limb within a model can be resized using the SCALE LIMB statement
FIG-35.13 The SCALE LIMB Statement
SCALE LIMB objno , limbno , xperc , yperc , zperc
In the diagram:
objno is the integer value specifying the limb's root

object.
limbno is an integer value identifying the limb to be

scaled.
xperc is a real number giving the new size of the

object’s x dimension as a percentage of its
original size in that dimension. For example, a
value of 100.0 will retain the original size, while
200.0 would double the object’s length in the x
dimension, and 50.0 would halve it.
yperc is a real number giving the new size of the

object’s y dimension as a percentage of its
zperc is a real number giving the new size of the

object’s z dimension as a percentage of its
Activity 35.11
In your previous program, modify the width (x) and height (y) of limb 4 to be
200% of their original size; do not modify the depth of the limb.
Modify the program again to make limb 4 grow slowly in width and height
from 100% to 200% and then back again. This action should happen
continuously.
Remove all changes made in this Activity before continuing.
We have ended up with a cube! But surely it would have been much easier to use
the MAKE OBJECT CUBE statement. However, the advantage of using limbs is
that we can now colour or texture the limbs individually.
The COLOR LIMB Statement

To colour an individual limb, we use the COLOR LIMB statement which has the

FIG-35.14
COLOR LIMB objno , limbno , colour
The COLOR LIMB
Statement
In the diagram:

object.

coloured.
colour is an integer value giving the colour to be

assigned to the limb. Often this will be the value
returned by an RGB statement.
For example, limb 1 of object 1 could be assigned the colour yellow using the line
COLOR LIMB 1,1,RGB(255,255,0)
Activity 35.12
Modify your last program so that the cube's sides are coloured as follows:
Side Colour
1 yellow
2 red
3 blue
4 magenta
5 cyan
6 green
The TEXTURE LIMB Statement

Rather than just colour a limb we can texture it with an image by using the
TEXTURE LIMB statement whose format is shown in FIG-35.15.
FIG-35.15
The TEXTURE LIMB TEXTURE LIMB objno , limbno , imgno
Statement
In the diagram:

object.

textured.
imgno is an integer value giving the ID of the image to

be used as the texture.
For example, limb 1 of object 1 could be textured using the image onespot.bmp
using the lines:

LOAD IMAGE "onespot.bmp",1
TEXTURE LIMB 1,1,1
By using a set of six images, we can create the dice shown in FIG-35.16.
FIG-35.16
Creating a Dice
Images Used
Final Group Object
The program which produces the dice is given in LISTING-35.4.

LISTING-35.4 REM *** Main program ***
SetUpScreen()
Texturing a Dice CreateDiceShape()
TextureDice()
RotateDice()
END
FUNCTION SetUpScreen()
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION CreateDiceShape()
DELETE OBJECT 2
REM *** Attach back ***
ADD LIMB 1,1,1
REM *** Add front ***
ADD LIMB 1,2,1
OFFSET LIMB 1,2,0,0,-5
REM *** Add bottom ***
ADD LIMB 1,3,1
REM *** Add top ***
ADD LIMB 1,4,1
REM *** Add left side ***
ADD LIMB 1,5,1
OFFSET LIMB 1,5,-5,0,0
REM *** Add right side ***
ADD LIMB 1,6,1

LISTING-35.4 ROTATE LIMB 1,6,0,90,0
(continued) OFFSET LIMB 1,6,5,0,0
ENDFUNCTION
Texturing a Dice
FUNCTION TextureDice()
REM *** Load Images ***
LOAD IMAGE "Spot1.bmp",1
REM *** Colour top red ***
TEXTURE LIMB 1,1,1
TEXTURE LIMB 1,2,6
TEXTURE LIMB 1,3,2
TEXTURE LIMB 1,4,5
TEXTURE LIMB 1,5,3
TEXTURE LIMB 1,6,4
ENDFUNCTION
FUNCTION RotateDice()
DO
PITCH OBJECT UP 1, 1
ROLL OBJECT LEFT 1,1
WAIT 10
LOOP
ENDFUNCTION
Activity 35.13
Activity 35.14
Create a new project (limbsvideo.dbpro). Create a cube using a sphere and six
box limbs. Colour all limbs blue. Load the video vd.mpg on limb 2 of the cube
and make the cube rotate about all three of its axes.
The SCALE LIMB TEXTURE Statement

Just as we can scale the texture applied to an object, so we can also scale the texture
of a limb. This is done using the SCALE LIMB TEXTURE which has the format
FIG-35.17 shown in FIG-35.17.
The SCALE LIMB Statement
SCALE LIMB TEXTURE objno , limbno , Uscale , Vscale
In the diagram:
objno is an integer value specifying the limb's root

object.
limbno is an integer value identifying the limb whose

texture is to be scaled.
Uscale is a real number specifying the multiplication

factor along the U axis. Values less than 1 will
of the image over the limb's surface.
Vscale is a real number specifying the multiplication

factor along the V axis. Values less than 1 will
of the image over the limb's surface.
For example, we could make the single spot on limb 1 of our dice look like four
spots by duplicating the texture twice along both axes. This would require the
statement:
SCALE LIMB TEXTURE 1,1,2.0,2.0
Activity 35.15
Modify your limbs04.dpro program's RotateDice() function so that the dice is

stationary with the single spot facing the camera.
Add a SCALE LIMB TEXTURE statement to the main section so that the
single spot looks like six spots.
When the program is working as required, remove the SCALE LIMB

TEXTURE statement and resave your program.
Activity 35.16
Write a program (Act3516.dbpro) to create a corridor as shown in the image

below. Each surface of the corridor should be a limb of a zero-sized root
Files Used: object. The functions within the program should be:
Floor - floor_m_01.bmp SetUpScreen Sets the display mode and positions the camera
Ceiling - ceil_m_03.bmp
Wall - wall_m_44.bmp at (0,4,-100)
CreateModel Constructs the corridor
The main section should allow the user to move the camera using the arrow
keys.

The SCROLL LIMB TEXTURE Statement
The texture image can be mapped onto a limb with a varying degree of offset along
either the U or V axes. The overall effect is to modify which part of the texture
image is placed at the top left corner of the limb. This is achieved using the SCROLL
LIMB TEXTURE statement which has the format shown in FIG-35.18.
FIG-35.18 The SCROLL LIMB TEXTURE Statement
SCROLL LIMB TEXTURE objno , limbno , Uoffset , Voffset
In the diagram:

object.

texture is to be offset.
Uoffset is a real number specifying the offset along the

x-axis. This value should lie between 0 and 1.
Voffset is a real number specifying the offset along the

y-axis. This value should lie between 0 and 1.
The result of using the statement
SCROLL LIMB 1,1,0.5,0.5
on the one spot side of the dice is shown in FIG-35.19.

FIG-35.19
Scrolling a Limb's Texture
The HIDE LIMB Statement

Individual limbs within a group object can be hidden using the HIDE LIMB
FIG-35.20
The HIDE LIMB HIDE LIMB objno , limbno
Statement
In the diagram:

object.

hidden.

Activity 35.17
Modify limbs04.dbpro to hide limb 1 of the dice.
Notice that the root object (the sphere) becomes visible when we remove the side
panel. However, we can avoid this by giving the root object a diameter of zero. This
won't affect the size or visibility of the limbs attached to the sphere.
Activity 35.18
Modify your last program so that the sphere is created with a diameter of zero.
The SHOW LIMB Statement

A hidden limb can be made visible again using the SHOW LIMB statement which
FIG-35.21
SHOW LIMB objno , limbno
The SHOW LIMB
Statement
In the diagram:

object.

redisplayed.
Activity 35.19
In limbs04.dbpro, modify the main section so that limb 1 of the dice reappears
after a key press.
The REMOVE LIMB Statement

A limb can be deleted using the REMOVE LIMB statement which has the format
shown in FIG-35.22.
FIG-35.22
The REMOVE LIMB REMOVE LIMB objno , limbno
Statement
In the diagram:

object.

deleted.
The LINK LIMB Statement

Although every limb in a group model must be added to the root object, a limb need
not be attached directly to that object. Limbs can also be attached (or linked) to

existing limbs within the model. For example, in FIG-35.23 we see a model with a
sphere as its root object to which a cylinder has been attached. Linked to the cylinder
is a cone.
FIG-35.23 The cylinder is a

limb of the sphere
Linking a New Limb to
an Existing Limb
The cone has been

attached to the cylinder
The cone limb is known as the child limb, while the limb to which it is joined (the
cylinder) is known as the parent limb. This effect is achieved using the LINK LIMB
FIG-35.24
The LINK LIMB
LINK LIMB objno , parentlimb , childlimb
Statement
In the diagram:

object.
parentlimb is an integer value identifying to which limb the

child limb is to be attached.
childlimb is an integer value giving the ID of the limb being

attached to the parent limb.
The steps involved in adding a child limb are shown in FIG-35.25.

FIG-35.25
Creating a
Child Limb
Start by creating the root object: Create and scale the objects to be Convert objects to meshes:
MAKE OBJECT SPHERE 1,8 used as limbs: MAKE MESH FROM OBJECT 1,2
MAKE OBJECT CYLINDER 2,10 MAKE MESH FROM OBJECT 2,3
SCALE OBJECT 2, 20,300,20 and delete the original objects:
MAKE OBJECT CONE 3,5 DELETE OBJECT 2
DELETE OBJECT 3
The cone is
The cone is attached to the
hidden within centre of the
the sphere cylinder
Add the meshes as limbs of the Link the cone as a child of the cylinder: Move the cone to the end of the
sphere: LINK LIMB 1,1,2 cylinder:
ADD LIMB 1,1,1 OFFSET LIMB 1,2,0,15,0
ADD LIMB 1,2,2

It is best to attach and position child limbs before moving the parent limb from its
original position.
Once attached to a parent limb, a child limb will move, rotate or resize automatically
when the parent limb is altered. The program in LISTING-35.5 demonstrates the
use of a child limb, and shows how that child limb is affected when the parent limb
is moved.
LISTING-35.5 REM *** Main program ***
SetUpScreen()
Using Child Limbs CreateModel()
ManipulateModel()
WAIT KEY
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION CreateModel()
REM *** Make model root ***
REM *** Create objects used as limbs ***
SCALE OBJECT 2, 20,300,20
MAKE OBJECT CONE 3,5
REM *** Convert limb objects to meshes ***
REM *** Delete limb objects ***
DELETE OBJECT 2
DELETE OBJECT 3
REM *** Add limbs to root ***
ADD LIMB 1,1,1
ADD LIMB 1,2,2
REM *** Link cone as child of cylinder ***
LINK LIMB 1,1,2
REM *** Move cone to end of cylinder ***
ENDFUNCTION
FUNCTION ManipulateModel()
REM *** Move cylinder up ***
WAIT KEY
FOR y = 1 TO 15
OFFSET LIMB 1,1,0,y,0
WAIT 100
NEXT y
REM *** Rotate whole model ***
WAIT KEY
FOR angle = 0 TO 45
WAIT 100
NEXT angle
ENDFUNCTION

Activity 35.20
Type in and test the program in LISTING-35.5 (limbs05.dbpro).
Find out what happens to the cone when the cylinder is deleted, then return the
program to its previous state.
The CHANGE MESH Statement

The mesh used on a specific limb of a model can be changed using the CHANGE
MESH statement. This allows us to change the shape of a limb. The CHANGE
MESH statement has the format shown in FIG-35.26.
FIG-35.26
The CHANGE MESH CHANGE MESH objno , limbno , meshno
Statement
In the diagram:

object.

mesh is to be changed.
meshno is an integer value giving the ID of the new mesh

to be assigned to the limb.
To change the cone in the model created in the last program to a sphere, we would
use the lines:
MAKE OBJECT SPHERE 2,5 `Create sphere

MAKE MESH FROM OBJECT 3,2 `Change it into a mesh
CHANGE MESH 1,2,3 `Replace cone with sphere
Activity 35.21
Modify the main section of your last program to change the cone to a sphere.
Activate the change with a key press from the user.
The GLUE OBJECT TO LIMB Statement

An object can be temporarily linked to a limb. Such an object would then move in
unison with that limb (even though it may not be visually attached to the limb). This
can be useful to create the effect of a player picking up an item or to show items
within a moving container. To attach an object to a limb, we use the statement GLUE
OBJECT TO LIMB which has the format shown in FIG-35.27.
FIG-35.27 The GLUE OBJECT TO LIMB Statement
GLUE OBJECT TO LIMB objno1 , objno2 , limbno [, mode
In the diagram:
objno1 is an integer value specifying the object to be

glued to the limb.

objno2 is an integer value specifying the limb's root
object.
limbno is an integer value identifying the limb to which

objno1 is to be glued.
mode 0, 1, or 2. This determines how the glued object

attaches to the limb.
0 - any displacement that has been applied
to the limb is also applied to the object.
This is the default behaviour if the mode
value is omitted from the statement.
1 - appears to result in a similar effect to
the zero option.
2 - the object will be placed at the centre of
the limb to which it has been glued.
FIG-35.28 shows the effect of the mode parameter when a cube is glued to the cone
of the model we created earlier.
FIG-35.28
The Effects of the mode
Parameter of the GLUE
OBJECT TO LIMB Object 4 is displaced Object 4 is placed at
The cone has been 15 units along the y-axis
Statement the centre of the limb
displaced 15 units - just as the limb to which to which it is attached
along the y-axis it is attached
Initial set-up. GLUE OBJECT TO LIMB 4,1,2,0 GLUE OBJECT TO LIMB 4,1,2,2
We'll see the GLUE OBJECT TO LIMB statement in action by adding a new
routine, DemoGlue(), to the last program. This new routine creates a cube object,
moves the cylinder limb (and hence the cone limb) over to the cube before attaching
the cube to the cone. After a key press, the whole model is rotated back to its starting
position, with the cube following the movement. The new function is coded as:
FUNCTION DemoGlue()
REM *** Rotate model so that cone is beside cube ***

FOR angle = 0 TO 85
ROLL OBJECT RIGHT 1,1
WAIT 50
NEXT angle
REM *** Glue cube to cone
GLUE OBJECT TO LIMB 4,1,2,0
WAIT KEY
REM *** Rotate model back to start position ***

FOR angle = 0 TO 85
WAIT 50
NEXT angle
ENDFUNCTION

Activity 35.22
Make sure limbs05.dbpro matches the original listing.
Add the following lines - which move the cylinder - to the end of the
CreateModel() function:
REM *** Position cylinder ***

Add the code for DemoGlue() to your program.
Delete the function ManipulateModel().
Change the line in the main section of the program from
ManipulateModel()
to
DemoGlue()
Run the new version of the program. What happens when the cube is linked to
the cone?
Modify the GLUE OBJECT TO LIMB statement to use mode 1 and then
mode 2. How does each change affect the result?
We can use POSITION OBJECT to reposition the glued object relative to the limb
to which it is joined - but we need to take into account the rotation of the cube's
axes. For example, rather than have the cube placed within the centre of the cone
(as you observed in the last Activity) we could make the cube appear to be "glued"
to the cone's surface by moving the cube slightly (see FIG-35.29).
FIG-35.29
Moving the Object after
Glueing
The Cube Attached to the Centre of the Cone The Cube Repositioned Slightly after Attachment
From the moment an object is glued to a limb, the object's new position becomes
its new origin (0,0,0) and the axes are orientated with those of the limb to which it
is attached. For our cube, the consequences are shown in FIG-35.30.
So to move the cube down slightly (as shown in FIG-35.29) we need to use the line:
Notice that the movement is along the x-axis.
Activity 35.23
Try adding the POSITION OBJECT statement to your previous program

immediately after the GLUE statement.

FIG-35.30 +y
How an Object's Axes and

Spatial Origin move when
Glued to a Limb
-x +x
The cone, and its -x

axes have been rotated
through 85 o
-y
+y
-y
(0,0,0)
+x
The UNGLUE OBJECT Statement
To detach an object from a group model, we need to use the UNGLUE OBJECT
FIG-35.31
UNGLUE OBJECT objno
The UNGLUE OBJECT
Statement
In the diagram:
objno is an integer value specifying the object to be

unglued.
In our previous program we could unglue the cube from the cone using the line:
UNGLUE OBJECT 4
Activity 35.24
Modify function DemoGlue() in your last program so that the cylinder swings
back down to 85o, unglues the cube, and then swings back up to the vertical
position. Use a key press to activate each operation. How is the cube affected
when it is unglued from the cone?
The SET LIMB SMOOTHING Statement

When two polygons drawn in different planes meet, often a sharp edge is created.
By using the SET LIMB SMOOTHING statement, this affect can be smoothed out
where it occurs within a specific limb. The SET LIMB SMOOTHING statement
FIG-35.32
The SET LIMB
SET LIMB SMOOTHING objno , limbno , smperc
SMOOTHING Statement
In the diagram:
objno is an integer value specifying limb's root

object.

edges are to be smoothed.
smperc is a integer percentage giving the degree of

smoothing required. A value of 0 gives no
smoothing effect while 100 gives maximum
smoothing.
Activity 35.25
In your last program:

Comment out the call to DemoGlue().
In SetUpScreen(), change the camera's position to (0,0,-55).
In the main section add code to smooth the cone limb after a key press
(set the smperc value to 100).
Observe the subtle effect of the smoothing operation on the cone.
Creating Doors
Most doors are hinged on one side and open by rotating about the hinged edge.
When we need to create a door in a game, it is simple enough to create a thin box
with the appropriate texture (see FIG-35.33), but it takes a bit more effort to get the
door to swing open. Using ROTATE OBJECT won't work because the door will
rotate about its centre rather than its edge.
FIG-35.33
Creating a Door
To create a door that will swing open we need to resort to making the box a limb.
We can then open the door by rotating the root object. The necessary code is shown
in LISTING-35.6.
LISTING-35.6 SetupScreen()
CreateDoor()
Opening a Hinged Door WAIT KEY
SwingDoor(80)
WAIT KEY
END
REM *** Set screen resolution and position camera ***
COLOR BACKDROP 0

LISTING-35.6 BACKDROP ON
(continued) AUTOCAM OFF
Opening a Hinged Door POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION CreateDoor()
#CONSTANT doorimg 1
#CONSTANT doorhingeobj 1
#CONSTANT doorobj 2
#CONSTANT doormesh 1
#CONSTANT doorlimb 1

LOAD IMAGE "door_05.jpg",doorimg
REM *** Create root object ***
MAKE OBJECT SPHERE doorhingeobj,0
REM *** Add door ***
MAKE OBJECT BOX doorobj,3,5,0.1
MAKE MESH FROM OBJECT doormesh,doorobj
DELETE OBJECT doorobj
ADD LIMB doorhingeobj,doorlimb,doormesh
TEXTURE LIMB doorhingeobj,doorlimb,doorimg
OFFSET LIMB doorhingeobj,doorlimb,1.5,0,0
ENDFUNCTION
FUNCTION SwingDoor(angle)
REM *** Swing door 1 degree at a time ***
FOR c = 1 TO angle
ROTATE OBJECT doorhingeobj,0,c,0
WAIT 50
NEXT c
ENDFUNCTION
Activity 35.26
What happens if the door swings past the 90o mark?
Modify the SwingDoor() routine so that the door swings from its current
starting place and allow the delay between each 1o move to be specified as a
parameter.
Retrieving Limb Data

There are several statements available to determine various limb settings. These
statements are described below.
The LIMB EXIST Statement

We can check to see if a limb of a specific number exists within a model using the
LIMB EXIST statement which has the format shown in FIG-35.34.
FIG-35.34
The LIMB EXIST
LIMB EXIST ( objno , limbno )
Statement
integer

In the diagram:

object.
limbno is an integer value giving the limb number to be

checked.
If the specified limb exists, 1 is returned, otherwise zero is returned.
The LIMB VISIBLE Statement

To check if a limb is visible, we can use the LIMB VISIBLE statement which has
FIG-35.35
The LIMB VISIBLE LIMB VISIBLE ( objno , limbno )
Statement
integer
In the diagram:

object.
limbno is an integer value giving the limb number to be

checked.
If the specified limb is visible, 1 is returned, otherwise zero is returned.
The LIMB OFFSET Statement

When a limb is first added to the root object (or another limb), the new limb is
positioned at the centre of its parent component. In this position the new limb has
an offset of zero in all three directions (x, y, and z). However, if we now move the
new limb (as we moved the cone to the end of the cylinder in LISTING-35.5) an
offset has been applied to that limb.
The current offset values of a limb can be determined using the LIMB OFFSET
FIG-35.36 X
The LIMB OFFSET
Statement LIMB OFFSET Y ( objno , limbno )
real

In the diagram:
X,Y,Z Use one of these options to retrieve the offset value

along the required axis.
objno is an integer value identifying the root object to

which the limb in question is attached.
limbno is an integer value giving the number of the limb

whose offset value is to be discovered.
If a limb is attached to the root object, then the values returned by this statement
relate to the limb centre's displacement from the root object's centre. If a limb is a
child limb then the values returned are measured from the parent limb's centre to
the child limb's centre.
Activity 35.27
Modify your last version of limbs05.dpro to display the offset values in each
direction of both the cylinder and the cone.
The LIMB SCALE Statement

If a limb is resized after being attached to a group object, the degree to which a limb
has been scaled in each dimension can be determined using the LIMB SCALE
statement, which has the format shown in FIG-35.37.
FIG-35.37
X
The LIMB SCALE
Statement
LIMB SCALE Y ( objno , limbno )
real
In the diagram:
X,Y,Z Use one of these options to retrieve the scale value

along the required axis.


whose scale value is to be returned.
For example, we could determine the scaling that has been applied to the cylinder
(limb 1) in our previous program using the lines:
xscale# = LIMB SCALE X(1,1)

yscale# = LIMB SCALE Y(1,1)
zscale# = LIMB SCALE Z(1,1)

The LIMB ANGLE Statement
If a limb is rotated independently of its root or parent component (using ROTATE
LIMB), the degree to which a limb has been rotated about each of its local axes can
be determined using the LIMB ANGLE statement which has the format shown in
FIG-35.38.
FIG-35.38
X
The LIMB ANGLE
Statement
LIMB ANGLE Y ( objno , limbno )
real
In the diagram:
X,Y,Z Use one of these options to retrieve the angle of

rotation about the required axis.


whose angle of rotation is to be returned.
The LIMB POSITION Statement

The position of the centre of a limb within the 3D world can be discovered using
the LIMB POSITION statement which has the format shown in FIG-35.39.
FIG-35.39
The LIMB POSITION X
Statement
LIMB POSITION Y ( objno , limbno )
real
In the diagram:
X,Y,Z Use one of these options to retrieve the position of

the limb in the required direction.


whose position is to be returned.

The LIMB DIRECTION Statement
Whereas LIMB ANGLE returns the angle of a limb relative to its own local axes,
the LIMB DIRECTION gives the angle of the limb relative to the main world axes.
The statement has the format shown in FIG-35.40.
FIG-35.40
X
The LIMB DIRECTION
Statement
LIMB DIRECTION Y ( objno , limbno )
real
In the diagram:
X,Y,Z Use one of these options to retrieve the direction of

the limb along the required axis.


whose world angle of rotation is to be returned.
FIG-35.41 shows a screen dump from the program in LISTING-35.7 which displays
all the limb details for the cone in our previous model as it carries out various actions.
REM *** Main program ***
FIG-35.41
Displaying Limb Details

LISTING-35.7 SetUpScreen()
CreateModel()
Displaying Limb Data REPEAT
HandleUser()
UNTIL key$="5"
END
FUNCTION HandleUser()
REM *** Display options ***
SET CURSOR 0,100
PRINT "1 - Rotate arm"
PRINT "2 - Rotate cone"
PRINT "3 - Move model"
PRINT "4 - Scale cone"
PRINT "5 - QUIT"
SYNC
REM *** Get user's choice ***
REPEAT
key$ = INKEY$()
UNTIL key$ >= "1" AND key$ <= "5"
REM *** Execute option chosen ***
SELECT key$
CASE "1"
RotateArm()
WAIT KEY
ENDCASE
CASE "2"
RotateCone()
WAIT KEY
ENDCASE
CASE "3"
POSITION OBJECT 1,OBJECT POSITION X(1)-5, 0,0
WAIT KEY
ENDCASE
CASE "4"
ScaleCone()
WAIT KEY
ENDCASE
ENDSELECT
ENDFUNCTION
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
SYNC ON
ENDFUNCTION
DELETE OBJECT 2
DELETE OBJECT 3

LISTING-35.7 ADD LIMB 1,1,1
(continued) ADD LIMB 1,2,2
Displaying Limb Data LINK LIMB 1,1,2
SYNC
ENDFUNCTION
FUNCTION RotateCone()
REM *** Rotate cone 1 degree at a time ***
FOR angle = 1 TO 90
ROTATE LIMB 1,2,0,0,angle
REM *** Show limb data ***
DisplayLimbData(1,2)
SYNC
WAIT 50
NEXT angle
ENDFUNCTION
FUNCTION ScaleCone()
REM *** Scale the cone 1% at a time ***
FOR scale = 100 TO 50 STEP -1
SCALE LIMB 1,2,100,scale,100
REM *** Display limb data ***
SYNC
WAIT 50
NEXT scale
ENDFUNCTION
FUNCTION RotateArm()
REM *** Rotate model showing limb data ***
FOR angle = 0 TO 90
SYNC
WAIT 50
NEXT angle
WAIT KEY
REM *** Rotate model back to start position ***
FOR angle = 0 TO 90
SYNC
WAIT 50
NEXT angle
ENDFUNCTION
FUNCTION DisplayLimbData(objno,limbno)
REM *** Get limb data ***
xoffset# = LIMB OFFSET X(objno,limbno)
yoffset# = LIMB OFFSET Y(objno,limbno)
zoffset# = LIMB OFFSET Z(objno,limbno)
xangle# = LIMB ANGLE X(objno,limbno)
yangle# = LIMB ANGLE Y(objno,limbno)
zangle# = LIMB ANGLE Z(objno,limbno)
xpost# = LIMB POSITION X(objno,limbno)
ypost# = LIMB POSITION Y(objno,limbno)
zpost# = LIMB POSITION Z(objno,limbno)
xdir# = LIMB DIRECTION X(objno,limbno)
ydir# = LIMB DIRECTION Y(objno,limbno)
zdir# = LIMB DIRECTION Z(objno,limbno)
xscale# = LIMB SCALE X(objno,limbno)
yscale# = LIMB SCALE Y(objno,limbno)
zscale# = LIMB SCALE Z(objno,limbno)

LISTING-35.7 REM *** Display data on screen ***
(continued) SET CURSOR 100,70
PRINT "Position x-axis : ", Format$(xpost#,6,2),
Displaying Limb Data " Position y-axis : ",Format$(ypost#,6,2),
" Position z-axis : ",Format$(zpost#,6,2)
SET CURSOR 100,100
PRINT "Direction x-axis: ", Format$(xdir#,6,2),
" Direction y-axis: ",Format$(ydir#,6,2),
" Direction z-axis: ",Format$(zdir#,6,2)
SET CURSOR 100,130
PRINT "Angle x-axis : ", Format$(xangle#,6,2),
" Angle y-axis : ",Format$(yangle#,6,2),
" Angle z-axis : ",Format$(zangle#,6,2)
SET CURSOR 100,160
PRINT "Scaling x-axis : ", Format$(xscale#,6,2),
" Scaling y-axis : ",Format$(yscale#,6,2),
" Scaling z-axis : ",Format$(zscale#,6,2)
SET CURSOR 100,190
PRINT "Offset x-axis : ", Format$(xoffset#,6,2),
" Offset y-axis : ",Format$(yoffset#,6,2),
" Offset z-axis : " ,Format$(zoffset#,6,2)
ENDFUNCTION
FUNCTION Format$(v#,totaldigits,decplaces)
REM *** Check number's sign ***
The Format$() routine is IF v# < 0
described in the May 2006 sign = -1
Support page of our ELSE
website. sign = 1
ENDIF
REM *** Round number ***
v# = sign*ABS(v#+0.5/10^decplaces)
REM *** Convert number to string ***
v$ = STR$(v#)
REM *** Find position of decimal point ***
FOR size = 1 TO LEN(v$)
IF MID$(v$,size) = "."
EXIT
ENDIF
NEXT size
REM *** Extract required decimal places from string ***
result$ = LEFT$(v$,size+decplaces)
REM *** Pad string to size required ***
WHILE LEN(result$) < totaldigits
result$ = " "+result$
ENDWHILE
ENDFUNCTION result$
The purpose of each routine is described briefly below:
SetUpScreen Initialises the screen resolution and camera

position.
HandleUser Displays user options, accepts keyboard selection,

and executes the option chosen.
CreateModel Creates the 3D model used in the program.
RotateCone Rotates the cone relative to the parent cylinder

limb.
ScaleCone Scales the cone after it is attached to the

cylinder.

RotateArm Rotates the complete model 90o about the z-axis.
DisplayLimbData Displays all details about the specified limb.
Format$ Displays a real number to a specified format.
Activity 35.28
The PERFORM CHECKLIST FOR OBJECT LIMBS Statement

It is possible to create a checklist giving details of the structure of a group object.
This is done using the PERFORM CHECKLIST FOR OBJECT LIMBS statement
which takes the format shown in FIG-35.42.
FIG-35.42
The PERFORM PERFORM CHECKLIST FOR OBJECT LIMBS objno
CHECKLIST FOR
OBJECT LIMBS
Statement In the diagram:
objno is an integer value giving the ID of the root object

whose limbs are to be listed.
For example, we could create a checklist of the limbs for a model whose root object
had the ID 5 using the line:
PERFORM CHECKLIST FOR OBJECT LIMBS 5
The checklist produced contains details of the root object and all limbs added to
that object.
To access the checklist created by this command we need to use the following
statements:
These statements were CHECKLIST QUANTITY
covered in pages 364 - CHECKLIST STRING$
366.
The program in LISTING-35.8 demonstrates the use of the PERFORM

CHECKLIST FOR OBJECT LIMBS statement, displaying the details of the model
containing the sphere, cylinder and cone that we created earlier.
LISTING-35.8 REM *** Main section ***
SetupScreen()
Listing Limb Names CreateModel()
ModelDetails()
WAIT KEY
END
AUTOCAM OFF
POINT CAMERA 0,0,0
SYNC ON
ENDFUNCTION

LISTING-35.8 FUNCTION CreateModel()
(continued) REM *** Make model root ***
Listing Limb Names REM *** Create objects used as limbs ***
SCALE OBJECT 2,20,300,20
DELETE OBJECT 2
DELETE OBJECT 3
ADD LIMB 1,1,1
ADD LIMB 1,2,2
LINK LIMB 1,1,2
REM *** Move cone to end of cylinder ***
SYNC
ENDFUNCTION
FUNCTION ModelDetails()
SET CURSOR 0,100
PRINT "This object contains ", CHECKLIST QUANTITY(),
Ä" components"
FOR c = 1 TO CHECKLIST QUANTITY()
PRINT CHECKLIST STRING$(c)
NEXT c
SYNC
ENDFUNCTION
Activity 35.29
The LIMB NAME$ Statement

When a model is constructed using a 3D package, any limb within that model can
be assigned a name. The name of that limb can be found using the LIMB NAME$
statement. This has the format shown in FIG-35.43.
FIG-35.43
The LIMB NAME$
Statement
LIMB NAME$ ( objno , limbno )
string
In the diagram:

the limb in question is attached.

name is to be found.

For example, we could discover the name of limb 2 in model 1 using the line:
PRINT LIMB NAME$(1,2)
Activity 35.30
Modify the previous program so that the function ModelDetails() contains the
lines:
FOR c = 1 TO 2
PRINT LIMB NAME$(1,c)
NEXT c
SYNC
The LIMB TEXTURE Statement

If a limb has been textured, the ID of the image used to texture the limb can be
discovered using the LIMB TEXTURE statement which has the format shown in
FIG-35.44.
FIG-35.44
The LIMB TEXTURE LIMB TEXTURE ( objno , limbno )
Statement
integer
In the diagram:


texture ID is to be found.
For example, the texture ID for limb 2 of model 1 can be found using a line such
as:
limbtextureID = LIMB TEXTURE(1,2)
The LIMB TEXTURE NAME Statement

Rather than just get the ID number of a texture, we can retrieve a texture's name
using the LIMB TEXTURE NAME statement. Note that the name is not the name
of the image file used to texture the limb, but the name assigned to the texture during
the building of the model. This means that you'll only get meaningful names when
using this statement with models created using a 3D modelling package. The LIMB
TEXTURE NAME statement has the format shown in FIG-35.45.
FIG-35.45
The LIMB TEXTURE
LIMB TEXTURE NAME ( objno , limbno )
NAME Statement
string

In the diagram:


texture name is to be found.
If no name is assigned to the texture, or no texture has been applied to the limb, then
an empty string is returned.
For example, the texture name for limb 2 of model 1 can be found using a line such
as:
texturename$ = LIMB TEXTURE NAME(1,2)
The CHECK LIMB LINK Statement

We can check that a new limb can be linked to a specified limb number using the
CHECK LIMB LINK statement which has the format shown in FIG-35.46.
FIG-35.46
The CHECK LIMB
CHECK LIMB LINK ( objno , limbno )
LINK Statement
integer
In the diagram:

limbno is an integer value identifying the limb to which

the new limb is to be attached.
The statement returns a value of 1 if a link to the limb specified by limbno is legal;
For example, we could check that a link to limb number 2 in model 1 was possible
with code such as:
IF CHECK LIMB LINK(1,2) = 1
PRINT "Link to limb 2 is valid"
ENDIF
Saving a Model in DBO Format

Introduction
Once a model has been created, it is possible to save it to a file. The model in that
file can then be loaded by another DarkBASIC Pro program. For example, we could
create a ladder from a set of cylinders and then save that ladder to a file. The ladder
would then be available for use in any other DarkBASIC Pro program.

The DBO File Format
When we save a DarkBASIC Pro model it is stored in DBO (DarkBasic Object)
format. This is a text-based file which can be viewed in a normal text editor such
as Notepad. A discussion of the structure of this file format is beyond the scope of
this text.
Creating an Elevator Model

FIG-35.47 shows the simple platform elevator model we are going to create.
FIG-35.47
The Elevator Model
The model is made from a zero-sized sphere as the root object, with a cylinder and
a box as limbs.
The necessary code is given in LISTING-35.9.

LISTING-35.9 ScreenSetUp()
MakeElevator()
Creating a Model for RaiseElevator()
Saving to a File WAIT KEY
END
FUNCTION ScreenSetUp()
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION MakeElevator()
REM *** Create Objects required ***
SCALE OBJECT 2,0.5,100,0.5
MAKE OBJECT BOX 3,5,0.2,5
REM *** Create limb meshes ***

LISTING-35.9 REM *** Delete objects ***
(continued) DELETE OBJECT 2
DELETE OBJECT 3
Creating a Model for REM *** Add limbs to root ***
Saving to a File ADD LIMB 1,1,1
ADD LIMB 1,2,2
REM *** Texture limbs ***
LOAD IMAGE "steel.bmp",1
LOAD IMAGE "metal.bmp",2
TEXTURE LIMB 1,1,1
TEXTURE LIMB 1,2,2
The RaiseElevator()
SCALE LIMB TEXTURE 1,2,5,5
function is included only
REM *** Move platfrom to base ***
to show the elevator in
OFFSET LIMB 1,2,0,-50,-2
action and is not required
ENDFUNCTION
in order to save the
model to a file.
FUNCTION RaiseElevator()
FOR c = -50 TO 50
OFFSET LIMB 1,2,0,c,-2
WAIT 60
NEXT c
ENDFUNCTION
Activity 35.31
The SAVE OBJECT Statement

To save an object created in a program to a DBO file, we use the SAVE OBJECT
FIG-35.48
The SAVE OBJECT SAVE OBJECT filename , objno
Statement
In the diagram:
filename is a string giving the name of the file to be used

when saving the model. The filename should be
given the extension ".DBO" to correctly identify
the type of file being created.
The string may include path details.

be saved. This can be a single primitive object or
a complex model containing limbs. For a limbed
model, the root object's ID is used.
For example, to save the elevator model created in the previous program, we would
use the line:
SAVE OBJECT "elevator.dbo",1
Activity 35.32
Replace the call to RaiseElevator() in the main section of your last program
with the line given above and check that the file elevator.dbo is created.

The LOAD OBJECT Statement
The LOAD OBJECT statement is used to load an existing 3D object from a file.
This statement can be used to load DBO files, but models stored in other 3D file
formats can also be loaded (see Chapter 36).
The LOAD OBJECT statement has the format shown in FIG-35.49.

FIG-35.49
The LOAD OBJECT LOAD OBJECT filename , objno
Statement
In the diagram:
filename is a string giving the name of the file containing

the 3D object. The string may contain drive and
folder details.
objno is the integer value to be assigned to the object

loaded from the file. No other object may be
assigned the same value.
For example, we could load the elevator model using the line:
LOAD OBJECT "elevator.dbo",1
The above line assumes that the file elevator.dbo is in the current project's folder.
In fact, this is the simplest form of the LOAD OBJECT statement. We'll leave a full
explanation of the statement until the next chapter.
The program in LISTING-35.10 loads the elevator model.

Loading a DBO Model
WAIT KEY
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 35.33
Type in and save the program in LISTING-35.10 (limbs10.dbpro).
Copy the elevator.dbo file to this project's folder.
Run the program.
What is wrong with the elevator model?
If a model uses textures, the images employed must also be copied to the same folder

as the .DBO file.
Activity 35.34
Copy the steel.bmp and metal.bmp files used by the elevator model into the
limbs10 folder.
Run the program again.
Is the model correctly textured this time?
Add the RaiseElevator() routine to your program and check that the platform
moves in the same way as before.
Summary
l A 3D object which is constructed from several component parts is known as a
group object or model.
l A model contains a root object and a set of limbs.
l The root object is a standard 3D object.
l The root object may have zero width, height and depth.
l Colouring or texturing the root object will result in the limbs of the model taking
on the same colour or texture.
l Meshes are linked to the root object and become limbs within the model.
l Use ADD LIMB to attach a mesh to the root object of a model.
l Limbs are initially positioned with their centre at the centre of the root object.
l Use MAKE OBJECT FROM LIMB to create a normal 3D object from a limb
within a model. The limb is unaffected by this operation.
l Use OFFSET LIMB to reposition a limb relative to its root.
l Use ROTATE LIMB to rotate a limb relative to its root.
l Use SCALE LIMB to resize a limb.
l Use COLOR LIMB to colour a specific limb.
l Use TEXTURE LIMB to texture a specific limb.
l Use SCALE LIMB TEXTURE to tile the texture on a specific limb.
l Use SCROLL LIMB TEXTURE to reposition the texture on a specific limb.
l Use HIDE LIMB to make a specific limb invisible.
l Use SHOW LIMB to make a previously invisible limb reappear.
l Use REMOVE LIMB to remove a specific limb from a model.

l Use LINK LIMB to link a limb (child limb) to another limb (parent limb) rather
than to the root. The child limb's centre is initially positioned at the parent limb's
centre.
l Use CHANGE MESH to change which mesh is used for an existing limb.
l Use GLUE OBJECT TO LIMB to attach an object to a limb.
l Use UNGLUE OBJECT to detach a previously glued object from a limb.
l Use SET LIMB SMOOTHING to smooth the surface of a specific limb.
l Use LIMB EXIST to check if a limb of a specified ID exists.
l Use LIMB VISIBLE to check if a specific limb is currently visible.
l Use LIMB OFFSET to discover the current offset settings for a specific limb.
l Use LIMB SCALE to discover the current scaling factors for a specific limb.
l Use LIMB ANGLE to discover a limb's rotation relative to its parent limb or the
root.
l Use LIMB POSITION to discover a limb's position relative to the world axes.
l Use LIMB DIRECTION to discover a limb's rotation relative to the world axes.
l Use PERFORM CHECKLIST FOR OBJECT LIMBS to create a list of limbs in

a given model.
l Individual limbs can be assigned names if the model was created by a 3D

modelling package.
l The root object is included in the list created by the PERFORM CHECKLIST
FOR OBJECT LIMBS.
l Use CHECKLIST QUANTITY and CHECKLIST STRING$ to access the

contents of the model's checklist.
l Use LIST NAME$ to access the name of a specific limb.
l Use LIMB TEXTURE to access the image ID of the texture used on a specific
limb.
l Use LIMB TEXTURE NAME to access the name assigned to the texture used
on a specific limb.
l Textures may be named if the 3D model was created by a modelling package.
l Use CHECK LIMB LINK to check if a specific limb can be the parent of a new
limb.
l Use SAVE OBJECT to save primitive shapes or complex models to a file.
l Models saved to a file are stored in DBO format.
l Use LOAD OBJECT to load an existing model from a DBO file.

Solutions
Activity 35.1
No solution required. REM *** Delete original object ***
DELETE OBJECT 2
REM *** Attach mesh to object ***
Activity 35.2 ADD LIMB 1,1,1
REM *** Move object ***
WAIT KEY
COLOR BACKDROP 0
BACKDROP ON
DO
AUTOCAM OFF
LOOP
POINT CAMERA 0,0,0
END
REM *** Make sphere ***
WAIT KEY When the sphere is moved, any limbs attached to it also
REM *** Make mesh from sphere *** move.
REM *** Delete the original object ***
DELETE OBJECT 1 Activity 35.6
REM *** Save mesh to file ***
SAVE MESH "sphere01.x",1 REM *** Set up screen ***
REM *** End program *** SET DISPLAY MODE 1280,1024,32
WAIT KEY COLOR BACKDROP 0
END BACKDROP ON
REM *** Set up screen *** AUTOCAM OFF
SET DISPLAY MODE 1280,1024,32 POSITION CAMERA 0,10,-100
COLOR BACKDROP 0 POINT CAMERA 0,0,0
BACKDROP ON REM *** Make and position sphere ***
REM *** Position camera *** MAKE OBJECT SPHERE 1,10
AUTOCAM OFF MAKE OBJECT BOX 2,10,10,0.1
POSITION CAMERA 0,10,-100 REM *** Make mesh from object ***
POINT CAMERA 0,0,0 MAKE MESH FROM OBJECT 1,2
REM *** Make sphere *** REM *** Delete original object ***
MAKE OBJECT SPHERE 1,10 DELETE OBJECT 2
REM *** Texture sphere *** REM *** Attach mesh to object ***
LOAD IMAGE "marble.bmp",1 ADD LIMB 1,1,1
TEXTURE OBJECT 1,1 REM *** Move object ***
WAIT KEY WAIT KEY
REM *** Make mesh from sphere *** POSITION OBJECT 1,-10,0,0
MAKE MESH FROM OBJECT 1,1 REM *** Colour root ***
REM *** Delete the original object *** WAIT KEY
DELETE OBJECT 1 COLOR OBJECT 1, RGB(255,0,0)
REM *** Save mesh to file *** REM *** Texture root ***
SAVE MESH "sphere02.x",1 WAIT KEY
REM *** End program *** LOAD IMAGE "marble.bmp",1
WAIT KEY TEXTURE OBJECT 1,1
END REM *** Allow user to move camera ***
DO
Activity 35.3 LOOP
No texture is displayed on the sphere with either .x file. END
Any colour or texture applied to the root object is also

Activity 35.4 applied to the limbs.
Activity 35.7
Activity 35.5 REM *** Set up screen ***
REM *** Set up screen *** COLOR BACKDROP 0
SET DISPLAY MODE 1280,1024,32 BACKDROP ON
COLOR BACKDROP 0 REM *** Position camera ***
BACKDROP ON AUTOCAM OFF
REM *** Position camera *** POSITION CAMERA 0,10,-100
AUTOCAM OFF POINT CAMERA 0,0,0
POSITION CAMERA 0,10,-100 REM *** Make and position sphere ***
POINT CAMERA 0,0,0 MAKE OBJECT SPHERE 1,10
REM *** Make and position sphere *** MAKE OBJECT BOX 2,10,10,0.1
MAKE OBJECT SPHERE 1,10 REM *** Make mesh from object ***

MAKE MESH FROM OBJECT 1,2 BACKDROP ON
REM *** Delete original object *** REM *** Position camera ***
DELETE OBJECT 2 AUTOCAM OFF
REM *** Attach mesh to object *** POSITION CAMERA 0,10,-100
ADD LIMB 1,1,1 POINT CAMERA 0,0,0
REM *** Move object *** REM *** Make and position sphere ***
WAIT KEY MAKE OBJECT SPHERE 1,10
POSITION OBJECT 1,-10,0,0 MAKE OBJECT BOX 2,10,10,0.1
REM *** Colour root *** REM *** Make mesh from object ***
WAIT KEY MAKE MESH FROM OBJECT 1,2
COLOR OBJECT 1, RGB(255,0,0) REM *** Delete original object ***
REM *** Texture root *** DELETE OBJECT 2
WAIT KEY REM *** Attach mesh to object ***
LOAD IMAGE "marble.bmp",1 ADD LIMB 1,1,1
TEXTURE OBJECT 1,1 OFFSET LIMB 1,1,0,0,5
REM *** Create an object from the limb *** ADD LIMB 1,2,1
WAIT KEY OFFSET LIMB 1,2,0,0,-5
MAKE OBJECT FROM LIMB 2,1,1 ADD LIMB 1,3,1
REM *** Reposition new object *** ROTATE LIMB 1,3,90,0,0
POSITION OBJECT 2,20,0,0 OFFSET LIMB 1,3,0,-5,0
REM *** Allow user to move camera *** ADD LIMB 1,4,1
DO ROTATE LIMB 1,4,90,0,0
CONTROL CAMERA USING ARROWKEYS 0,0.1,0.1 OFFSET LIMB 1,4,0,5,0
LOOP ADD LIMB 1,5,1
REM *** End program *** ROTATE LIMB 1,5,0,90,0
END OFFSET LIMB 1,5,-5,0,0
ADD LIMB 1,6,1
Activity 35.8 OFFSET LIMB 1,6,5,0,0
REM *** Rotate model ***
The limb moves to the far edge of the sphere. DO
PITCH OBJECT UP 1,1
Activity 35.9 ROLL OBJECT LEFT 1,1
WAIT 20
REM *** Set up screen *** LOOP
SET DISPLAY MODE 1280,1024,32 REM *** End program ***
COLOR BACKDROP 0 END
BACKDROP ON
AUTOCAM OFF Activity 35.11
POINT CAMERA 0,0,0 The first version requires the line
MAKE OBJECT SPHERE 1,10 SCALE LIMB 1,4, 200,200,100
The second version of the program is shown below:
DELETE OBJECT 2 REM *** Set up screen ***
REM *** Attach mesh to object *** SET DISPLAY MODE 1280,1024,32
ADD LIMB 1,1,1 COLOR BACKDROP 0
REM *** Move object *** BACKDROP ON
WAIT KEY REM *** Position camera ***
POSITION OBJECT 1,-10,0,0 AUTOCAM OFF
REM *** Colour root *** POSITION CAMERA 0,10,-100
WAIT KEY POINT CAMERA 0,0,0
COLOR OBJECT 1, RGB(255,0,0) REM *** Make and position sphere ***
REM *** Texture root *** MAKE OBJECT SPHERE 1,10
WAIT KEY MAKE OBJECT BOX 2,10,10,0.1
REM *** Add second limb *** REM *** Make mesh from object ***
ADD LIMB 1,2,1 MAKE MESH FROM OBJECT 1,2
OFFSET LIMB 1,2,0,0,-5 REM *** Delete original object ***
LOAD IMAGE "marble.bmp",1 DELETE OBJECT 2
TEXTURE OBJECT 1,1 REM *** Attach mesh to object ***
REM *** Allow user to move camera *** ADD LIMB 1,1,1
DO OFFSET LIMB 1,1,0,0,5
CONTROL CAMERA USING ARROWKEYS 0,0.1,0.1 ADD LIMB 1,2,1
LOOP OFFSET LIMB 1,2,0,0,-5
REM *** End program *** ADD LIMB 1,3,1
END ROTATE LIMB 1,3,90,0,0
ADD LIMB 1,4,1
REM *** Set up screen *** ADD LIMB 1,5,1
SET DISPLAY MODE 1280,1024,32 ROTATE LIMB 1,5,0,90,0
COLOR BACKDROP 0 OFFSET LIMB 1,5,-5,0,0

ADD LIMB 1,6,1 SET DISPLAY MODE 1280,1024,32
ROTATE LIMB 1,6,0,90,0 COLOR BACKDROP 0
OFFSET LIMB 1,6,5,0,0 BACKDROP ON
REM *** Scale limb 4 *** REM *** Position camera ***
DO AUTOCAM OFF
FOR scale = 100 TO 200 POSITION CAMERA 0,10,-100
SCALE LIMB 1,4,scale,scale,100 POINT CAMERA 0,0,0
WAIT 10 REM *** Make and position sphere ***
NEXT scale MAKE OBJECT SPHERE 1,10
FOR scale = 200 TO 100 STEP -1 MAKE OBJECT BOX 2,10,10,0.1
SCALE LIMB 1,4,scale,scale,100 REM *** Make mesh from object ***
WAIT 10 MAKE MESH FROM OBJECT 1,2
NEXT scale REM *** Delete original object ***
LOOP DELETE OBJECT 2
REM *** End program *** REM *** Attach 6 limbs ***
END ADD LIMB 1,1,1
ADD LIMB 1,2,1
Activity 35.12 OFFSET LIMB 1,2,0,0,-5
ADD LIMB 1,3,1
REM *** Set up screen *** ROTATE LIMB 1,3,90,0,0
SET DISPLAY MODE 1280,1024,32 OFFSET LIMB 1,3,0,-5,0
COLOR BACKDROP 0 ADD LIMB 1,4,1
BACKDROP ON ROTATE LIMB 1,4,90,0,0
REM *** Position camera *** OFFSET LIMB 1,4,0,5,0
AUTOCAM OFF ADD LIMB 1,5,1
POSITION CAMERA 0,10,-100 ROTATE LIMB 1,5,0,90,0
POINT CAMERA 0,0,0 OFFSET LIMB 1,5,-5,0,0
REM *** Make and position sphere *** ADD LIMB 1,6,1
MAKE OBJECT SPHERE 1,10 ROTATE LIMB 1,6,0,90,0
MAKE OBJECT BOX 2,10,10,0.1 OFFSET LIMB 1,6,5,0,0
REM *** Make mesh from object *** REM *** Colour all limbs blue ***
MAKE MESH FROM OBJECT 1,2 FOR limbno = 1 TO 6
REM *** Delete original object *** COLOR LIMB 1,limbno,RGB(0,0,255)
DELETE OBJECT 2 NEXT limbno
REM *** Attach 6 limbs *** REM *** Display video on limb 2 ***
ADD LIMB 1,1,1 LOAD ANIMATION "vd.mpg",1
OFFSET LIMB 1,1,0,0,5 PLAY ANIMATION TO IMAGE 1,1,0,0,120,80
ADD LIMB 1,2,1 TEXTURE LIMB 1,2,1
OFFSET LIMB 1,2,0,0,-5 REM *** rotate model ***
ADD LIMB 1,3,1 DO
ROTATE LIMB 1,3,90,0,0 PITCH OBJECT UP 1,1
OFFSET LIMB 1,3,0,-5,0 TURN OBJECT LEFT 1,1
ADD LIMB 1,4,1 ROLL OBJECT LEFT 1,1
ROTATE LIMB 1,4,90,0,0 WAIT 20
OFFSET LIMB 1,4,0,5,0 LOOP
ADD LIMB 1,5,1 REM *** End program ***
ROTATE LIMB 1,5,0,90,0 END
OFFSET LIMB 1,5,-5,0,0
ADD LIMB 1,6,1
Activity 35.15
REM *** Colour limbs *** REM *** Main program ***
COLOR LIMB 1,1,RGB(255,255,0) SetUpScreen()
COLOR LIMB 1,2,RGB(255,0,0) CreateDiceShape()
COLOR LIMB 1,3,RGB(0,0,255) TextureDice()
COLOR LIMB 1,4,RGB(255,0,255) RotateDice()
COLOR LIMB 1,5,RGB(0,255,255) SCALE LIMB TEXTURE 1,1,2,3
COLOR LIMB 1,6,RGB(0,255,0) REM *** End program ***
REM *** rotate model *** WAIT KEY
DO END
PITCH OBJECT UP 1,1
TURN OBJECT LEFT 1,1 FUNCTION SetUpScreen()
ROLL OBJECT LEFT 1,1 REM *** Set up screen ***
WAIT 20 SET DISPLAY MODE 1280,1024,32
LOOP COLOR BACKDROP RGB(0,200,0)
REM *** End program *** BACKDROP ON
END REM *** Position camera ***
AUTOCAM OFF
Activity 35.13 POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION CreateDiceShape()
Activity 35.14 MAKE OBJECT SPHERE 1,10
REM *** Set up screen *** REM *** Make mesh from object ***

MAKE MESH FROM OBJECT 1,2 #CONSTANT rootobj 1
REM *** Delete original object *** #CONSTANT planeobj 2
DELETE OBJECT 2 REM *** Meshes ***
REM *** Attach back *** #CONSTANT planemesh 1
ADD LIMB 1,1,1 REM *** Limbs ***
OFFSET LIMB 1,1,0,0,5 #CONSTANT floorlimb 1
REM *** Add front *** #CONSTANT ceilinglimb 2
ADD LIMB 1,2,1 #CONSTANT wall1limb 3
OFFSET LIMB 1,2,0,0,-5 #CONSTANT wall2limb 4
REM *** Add bottom *** REM *** Load texture images ***
ADD LIMB 1,3,1 LOAD IMAGE "floor_m_01.bmp",floorimg
ROTATE LIMB 1,3,90,0,0 LOAD IMAGE "ceil_m_03.bmp",ceilingimg
OFFSET LIMB 1,3,0,-5,0 LOAD IMAGE "wall_m_44.bmp",wallimg
REM *** Add top *** REM *** Create plane mesh ***
ADD LIMB 1,4,1 MAKE OBJECT BOX planeobj,10,200,0.1
ROTATE LIMB 1,4,90,0,0 MAKE MESH FROM OBJECT planemesh,planeobj
OFFSET LIMB 1,4,0,5,0 DELETE OBJECT planeobj
REM *** Add left side *** REM *** Create corridor model ***
ADD LIMB 1,5,1 REM *** Make invisible root ***
ROTATE LIMB 1,5,0,90,0 MAKE OBJECT SPHERE rootobj,0
OFFSET LIMB 1,5,-5,0,0 REM *** Add floor ***
REM *** Add right side *** ADD LIMB rootobj,floorlimb,planemesh
ADD LIMB 1,6,1 TEXTURE LIMB rootobj,floorlimb,ceilingimg
ROTATE LIMB 1,6,0,90,0 SCALE LIMB TEXTURE rootobj,floorlimb,2,40
OFFSET LIMB 1,6,5,0,0 ROTATE LIMB rootobj,floorlimb,90,0,0
ENDFUNCTION ADD LIMB rootobj,ceilinglimb,planemesh
TEXTURE LIMB rootobj,ceilinglimb,
FUNCTION TextureDice() ceilingimg
REM *** Load Images *** SCALE LIMB TEXTURE rootobj,ceilinglimb,
LOAD IMAGE "Spot1.bmp",1 2,40
LOAD IMAGE "Spot2.bmp",2 ROTATE LIMB rootobj,ceilinglimb,90,0,0
LOAD IMAGE "Spot3.bmp",3 OFFSET LIMB rootobj,ceilinglimb,0,10,0
LOAD IMAGE "Spot4.bmp",4 ADD LIMB rootobj,wall1limb,planemesh
LOAD IMAGE "Spot5.bmp",5 TEXTURE LIMB rootobj,wall1limb,wallimg
LOAD IMAGE "Spot6.bmp",6 SCALE LIMB TEXTURE rootobj,wall1limb,2,40
REM *** Colour top red *** ROTATE LIMB rootobj,wall1limb,90,0,90
TEXTURE LIMB 1,1,1 OFFSET LIMB rootobj,wall1limb,5,5,0
TEXTURE LIMB 1,2,6 ADD LIMB rootobj,wall2limb,planemesh
TEXTURE LIMB 1,3,2 TEXTURE LIMB rootobj,wall2limb,wallimg
TEXTURE LIMB 1,4,5 SCALE LIMB TEXTURE rootobj,wall2limb,2,40
TEXTURE LIMB 1,5,3 ROTATE LIMB rootobj,wall2limb,90,0,90
TEXTURE LIMB 1,6,4 OFFSET LIMB rootobj,wall2limb,-5,5,0
ENDFUNCTION HIDE LIGHT 0
ENDFUNCTION
FUNCTION RotateDice()
XROTATE OBJECT 1,180
ENDFUNCTION Activity 35.17
The main section of the program should be changed to
Activity 35.16
REM *** Main program ***
REM *** Main section ***
SetUpScreen()
SetupScreen()
CreateDiceShape()
CreateModel()
TextureDice()
REM *** Give user control of camera ***
RotateDice()
DO
HIDE LIMB 1,1
LOOP
WAIT KEY
END
END
FUNCTION SetUpScreen() The resulting model is shown below:


BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
REM *** Declare named constants ***
REM *** Images ***
#CONSTANT floorimg 1
#CONSTANT ceilingimg 2
#CONSTANT wallimg 3
REM *** Objects ***

In the function CreateDiceShape(), the line REM *** Main program ***
SetUpScreen()
CreateModel()
DemoGlue()
should be changed to WAIT KEY
END
The effect of this change on the model displayed on screen REM *** Set up screen ***
is shown below SET DISPLAY MODE 1280,1024,32
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 35.19 SCALE OBJECT 2, 20,300,20
REM *** Main program *** REM *** Convert objects to meshes ***
SetUpScreen() MAKE MESH FROM OBJECT 1,2
CreateDiceShape() MAKE MESH FROM OBJECT 2,3
TextureDice() REM *** Delete limb objects ***
RotateDice() DELETE OBJECT 2
HIDE LIMB 1,1 DELETE OBJECT 3
REM *** Make limb reappear *** REM *** Add limbs to root ***
WAIT KEY ADD LIMB 1,1,1
SHOW LIMB 1,1 ADD LIMB 1,2,2
REM *** End program *** REM *** Make cone child of cylinder ***
WAIT KEY LINK LIMB 1,1,2
END REM *** Move cone to end of cylinder ***
REM *** Position cylinder ***
ENDFUNCTION
Modify the main section to
FUNCTION DemoGlue()
REM *** Main program *** REM *** Create and position cube ***
SetUpScreen() MAKE OBJECT CUBE 4,3
CreateModel() POSITION OBJECT 4,30,0,0
ManipulateModel() REM *** Rotate model to cube ***
REM *** Remove cylinder *** FOR angle = 0 TO 85
WAIT KEY ROLL OBJECT RIGHT 1,1
REMOVE LIMB 1,1 WAIT 50
REM *** End program *** NEXT angle
WAIT KEY REM *** Glue cube to cone ***
END GLUE OBJECT TO LIMB 4,1,2,0
WAIT KEY
Since the cone is attached to the cylinder, it is also REM *** Rotate model upright ***
FOR angle = 0 TO 85
removed from the model. ROLL OBJECT LEFT 1,1
WAIT 50
NEXT angle
Modify the main section to:
When the program runs, the cube takes on the same offset
REM *** Main program *** and rotational values as the cone.
SetUpScreen()
CreateModel() A mode value of 1 gives the same effect, but with mode
ManipulateModel() set to 2, the centre of the cube is linked to the centre of the
REM *** Change cone to sphere *** cone.
WAIT KEY
CHANGE MESH 1,2,3
WAIT KEY
END

Activity 35.23 REM *** Smooth cone ***
WAIT KEY
SET LIMB SMOOTHING 1,2,100
The code for DemoGlue() changes to:
`DemoGlue()
FUNCTION DemoGlue()
WAIT KEY
END
REM *** Rotate model to cube ***
FOR angle = 0 TO 85
Activity 35.26
ROLL OBJECT RIGHT 1,1 The new version of SwingDoor():
WAIT 50
NEXT angle FUNCTION SwingDoor(angle,delay)
REM *** Glue cube to cone *** REM *** Get the current door angle ***
GLUE OBJECT TO LIMB 4,1,2,2 currentangle =
POSITION OBJECT 4,2,0,0 OBJECT ANGLE Y(doorhingeobj)
WAIT KEY REM *** Check direction of swing ***
REM *** Rotate model upright *** IF angle > currentangle
FOR angle = 0 TO 85 incr = 1
ROLL OBJECT LEFT 1,1 ELSE
WAIT 50 incr = -1
NEXT angle ENDIF
ENDFUNCTION REM *** Swing from starting angle ***
REM *** in 1 degree steps ***
FOR c = currentangle TO angle STEP incr
Activity 35.24 ROTATE OBJECT doorhingeobj,0,c,0
WAIT delay
DemoGlue() should be coded as: NEXT c
ENDFUNCTION
FUNCTION DemoGlue()
REM *** Create and position cube *** We could also modify the routine to add checks on the
parameters to make sure angle and delay lie wihin
REM *** Rotate model to cube *** acceptable limits.
FOR angle = 0 TO 85
ROLL OBJECT RIGHT 1,1 To test the function, we could change the main section to:
WAIT 50
NEXT angle SetupScreen()
REM *** Glue cube to cone *** CreateDoor()
GLUE OBJECT TO LIMB 4,1,2,2 WAIT KEY
POSITION OBJECT 4,2,0,0 SwingDoor(20,50)
WAIT KEY WAIT KEY
REM *** Rotate model upright *** SwingDoor(80,50)
FOR angle = 0 TO 85 WAIT KEY
ROLL OBJECT LEFT 1,1 SwingDoor(45,50)
WAIT 50 WAIT KEY
NEXT angle END
REM *** Model back to "ground" level ***
WAIT KEY If we swing the door too far, we'll see that the hinges and
FOR angle = 0 TO 85
handle are on the wrong sides on the reverse of the door!
WAIT 50 We could cure this by making the door itself a group
NEXT angle object.
REM *** Release cube ***
WAIT KEY
UNGLUE OBJECT 4 Activity 35.27
REM *** Model back to start position ***
WAIT KEY To see the offsets for each limb the main section could be
FOR angle = 0 TO 85 coded as:
WAIT 50 REM *** Main program ***
NEXT angle SetUpScreen()
ENDFUNCTION CreateModel()
DO
When the cube is released it repositions itself at the centre SET CURSOR 100,100
PRINT "Cylinder offset : ",
of the cone. We could use POSITION OBJECT to return it
ÄLIMB OFFSET X(1,1),":",
to its original position.
ÄLIMB OFFSET Y(1,1),":",
ÄLIMB OFFSET Z(1,1)
SET CURSOR 100,130
Activity 35.25 PRINT "Cone offset : ",
ÄLIMB OFFSET X(1,2),":",
The main section should be coded as:
ÄLIMB OFFSET Y(1,2),":",
REM *** Main program *** ÄLIMB OFFSET Z(1,2)
SetUpScreen() LOOP
CreateModel() REM *** End program ***
END

Activity 35.28 FUNCTION RaiseElevator()
FOR c = -50 TO 50
No solution required. OFFSET LIMB 1,2,0,c,-2
WAIT 60
NEXT c
ENDFUNCTION
Activity 35.29
Activity 35.30
The updated version of Modeldetails() should be:
FUNCTION ModelDetails()
PRINT "This object contains ",
CHECKLIST QUANTITY()," components"
NEXT c
SYNC
FOR c = 1 To 2
PRINT LIMB NAME$(1,c)
NEXT c
SYNC
ENDFUNCTION
The limbs are titled new limb in each case. We would

have to create the model in an external package to
allocate names to the limbs.
Activity 35.31
Activity 35.32
The main section of your program should now be as
follows:
ScreenSetUp()
MakeElevator()
SAVE OBJECT "elevator.dbo",1
WAIT KEY
END
Activity 35.33
The model is not textured.
Activity 35.34
The model is correctly textured when the required images
are placed in the project folder.
The final version of the program should be:
ScreenSetUp()
RaiseElevator()
WAIT KEY
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION

36
3D File Formats
How to Load Existing 3D Objects
Frame Interpolation
Joining Animated Models
Playing Animated Models
Setting the Animation Speed
DarkBASIC Pro: Importing 3D Objects 961

Importing 3D Objects
Introduction
Creating 3D objects using DarkBASIC Pro is all very well, but ultimately is
probably too limiting to allow us to achieve the sophistication we need in most
games. For example, it would be rather difficult to create a 3D gun, never mind a
humanoid, using only spheres, boxes, and cylinders.
Realistically, we need to create these complex shapes in some other package - and
there are many packages out there which are designed specifically to do just that.
If you've already made a few million from selling your DarkBASIC Pro game, then
you can afford professional-level 3D creation software such as Maya, 3D Studio
Max, or XSI SoftImage. For those with a more limited budget, have a look at some
of the packages available from The Game Creators such as 3D Canvas Pro or
gameSpace. There are even free or shareware packages such as Anim8or and
Milkshape.
Don't worry if you have absolutely no artistic ability (like me!), you'll still be able
to create models with 3D modelling software that will truly amaze you; if you do
have an artistic side, then you will be able to produce just about anything you can
imagine.
One word of warning, when creating 3D models that are to be used in a game, you
need to keep the number of polygons used as low as possible. The more polygons
an object contains, the more load you place on the processor and graphics card. With
more processing to do, the game's frame rate will drop, and when it gets too low,
your game will become jerky and unconvincing.
As usual, there is software out there to help overcome even this problem.
Polygon-reducing software attempts to reduce the number of polygons a 3D object
uses, while still retaining as much detail as possible.
Of course, learning how to use a 3D modelling package takes time - a long time if
you want to be really proficient - so, as an alternative, you can buy ready-to-use 3D
models such as the Dark Matter series offered by The Game Creators (see FIG-36.1).
FIG-36.1
A 3D Model from the
Dark Matter CD
962 DarkBASIC Pro: Importing 3D Objects

It is not within the scope of this book to show you how to use a 3D modelling
package, but if you would like to have a go at creating your own objects - and you
should - you'll find plenty of tutorial material on the Internet for most well-known
packages.
File Formats
Almost every 3D modelling package has its own format for saving objects to a file.
DarkBASIC Pro accepts several formats, BSP, MD2, MD3, MDL, PAK , but the
commonest are Microsoft's DirectX format (these files end with a .x extension) and
3D Studio format (.3ds extension) as well as DarkBASIC Pro's own .dbo format.
The examples given in this chapter make use of files available from Dark Matter
Volume 1.
Statements for Loading and Using 3D Objects

The LOAD OBJECT Statement Again
In the last chapter we saw that the LOAD OBJECT statement can be used to load
an existing 3D model stored in DBO format. However, the same instruction can
be used to load .x and .3ds files created by separate modelling packages. The
extended syntax for the LOAD OBJECT statement is shown in FIG-36.2.
FIG-36.2 The LOAD OBJECT Statement
LOAD OBJECT filename , objno

[[ , loadmode
[ , reducemode
In the diagram:

the model to be loaded.
objno is an integer value giving the ID to be assigned to

the model once it is loaded.
loadmode is an integer value specifying how texture and

other surface effects are to be handled.
Possible values are:
0 - DarkBASIC default method
1 - DarkBASIC Pro default method
2 - Retain material/diffuse effects
3 - Retain material/texture effects
4 - Blend texture/diffuse effects
5 - Retain multi-material effects
reducemode is an integer value used to scale loaded texture

x represents any positive images.
integer value. 0 - No scaling
x - Original size divided by x
The program in LISTING-36.1 demonstrates how to load the high resolution

version of an AK47 rifle, available in Dark Matter 1. The object is then rotated.

Loading a Static Model COLOR BACKDROP RGB(155,155,0)
BACKDROP ON
REM *** Load the gun object ***

LOAD OBJECT "H-AK47-Static.X",1
REM *** rotate gun ***
DO
PITCH OBJECT UP 1,1
WAIT 10
LOOP
END
Activity 36.1
Type in and test the program in LISTING-36.1 (loaded01.dbpro). Make sure

you've copied the .x file used into the program's directory.
Although the gun is the correct shape, it doesn't have any texturing. The weapon
has been designed with a texture, but this is held in a separate file called AK47.DDS.
If we want the texture to appear on the gun, we must copy the .dds file to the current
program folder.
Activity 36.2
By clicking on DarkBASIC Pro's Media button and selecting Add, copy the
AK47.DDS file into the loaded01 folder created by the last program.
Re-run your program. (The gun should now be displaying its texture.)
Activity 36.3
Modify your last program to load and rotate the jet fighter held in file H-Jet
Fighter-Move.x. The associated textures are held in JetFighter.dds.
Once loaded, a 3D object can be manipulated using all the standard OBJECT
statements, thus allowing us to do such things as position, rotate and resize objects.
Not all 3D objects are static, unmoving things. Animated 3D objects will actually
be held as a series of frames; a bit like an animated sprite. For example, Dark Matter's
Hive-Brain alien comes in several files, each of these contain a few frames showing
the alien performing some basic movement.
The program in LISTING-36.2 loads the walking alien.
LISTING-36.2 REM *** Set screen resolution and backdrop ***

SET DISPLAY MODE 1280, 1024, 32
Loading an Animated COLOR BACKDROP 0
Image BACKDROP ON
AUTOCAM OFF

LISTING-36.2 REM *** Create a point light ***
(continued) SET POINT LIGHT 0,-10,10,0
Loading an Animated
Image REM *** Load object and position camera ***
LOAD OBJECT "H-Alien Hivebrain-Move.X",1

WAIT KEY
END
Activity 36.4
Type in and test the program given in LISTING-36.2 (loaded02.dbpro).
Despite being an animated model, the alien shows no sign of any movement.
The PLAY OBJECT Statement

To get the object to move, we need to play each frame in the animation and this is
done using the PLAY OBJECT statement. The statement offers to play every frame
in the animation or just a specified set of frames. The PLAY OBJECT statement
FIG-36.3
The PLAY OBJECT PLAY OBJECT objno , start , finish
Statement
In the diagram:
objno is an integer value specifying the 3D object to be

played. Only objects which consist of two or more
frames can be specified.
start is an integer value specifying the first frame to

be played. The first frame of an animation is
frame zero.
finish is an integer value specifying the last frame to be

played.
If start and finish are omitted, the whole animation is played. If finish is omitted
every frame from start to the end of the animation is played.
For example,
PLAY OBJECT 1
will play every frame in object 1 while
PLAY OBJECT 1,0,10
will play frames 0 to 10 only and

PLAY OBJECT 1,5
will play from frame 5 to the end of the animation.

Activity 36.5
In your last program, add the lines

WAIT KEY
PLAY OBJECT 1
immediately after the LOAD OBJECT statement.
The LOOP OBJECT Statement

If we want the frames of an animated 3D object to play repeatedly, then we need to
use the LOOP OBJECT statement which has the format shown in FIG-36.4.
FIG-36.4
The LOOP OBJECT
LOOP OBJECT objno , start , finish
Statement
In the diagram:
objno is an integer value specifying the 3D object to be

looped. Only objects which consist of two or more
frames can be specified.
start is an integer value specifying the first frame to

be included in the loop.
finish is an integer value specifying the last frame to be

included in the loop.
If start and finish are omitted, the whole animation is played. If finish is omitted
every frame from start to the end of the animation is played.
Activity 36.6
In your program, change the PLAY OBJECT statement to LOOP OBJECT

and re-run the program.
To create a smooth animation, the first and last frames need to be almost identical.
The walking alien judders slightly as it cycles back to the first frame. We'll improve
this in a moment.
The TOTAL OBJECT FRAMES Statement

If we didn't create the animated 3D object ourselves, then we need to find out just
how many frames are in the animation. We can do this using the TOTAL OBJECT
FRAMES statement which returns the number of frames in a specified object. The
FIG-36.5
The TOTAL OBJECT
TOTAL OBJECT FRAMES ( objno )
FRAMES Statement
integer

In the diagram:
objno is an integer value specifying the 3D object whose

frame count is to be returned.
Activity 36.7
Add the appropriate code to your program so that the number of frames in the
alien animation is displayed on the screen.
Now that we know just how many frames are in the animated object, we can limit
exactly which frames are played.
Activity 36.8
Modify the LOOP OBJECT statement in your program so that only frames 0
to 20 are used in the animation shown on the screen.
Moving the Alien

Our alien is walking on the spot, but it would be more impressive if it actually went
somewhere. In LISTING-36.3 the alien walks towards camera and eventually
disappears as it passes the camera.
Moving an Animated
COLOR BACKDROP 0
Object
BACKDROP ON
AUTOCAM OFF
REM *** Create a point light ***
SET POINT LIGHT 0,-10,10,0
REM *** Load object; position camera ***
REM *** Play object continuously ***
LOOP OBJECT 1,0,20
REM *** Move alien ***
DO
MOVE OBJECT 1,-0.03
LOOP
END
Activity 36.9
Type in and test the program in LISTING-36.3 (loaded03.dbpro).
The SET OBJECT SPEED Statement

The speed at which the frames of an animated 3D object are played can be set using
the SET OBJECT SPEED statement which has the format shown in FIG-36.6.
FIG-36.6
The SET OBJECT
SET OBJECT SPEED objno , perc
SPEED Statement

In the diagram:
objno is an integer value giving the ID of the 3D

object whose speed is to be set.
perc is an integer value giving the new speed as a

percentage of normal speed. For example, a
value of 50 will play the frames at half normal
speed, while 200 gives twice normal speed.
Activity 36.10
In loaded03.dbpro, change the speed of the alien's walk to 50% of normal.
The STOP OBJECT Statement

A playing or looping animated object can be stopped using the STOP OBJECT
FIG-36.7
STOP OBJECT objno
The STOP OBJECT
Statement
In the diagram:

object whose animation is to be stopped.
When stopped, the model will continue to show the last frame played. This could
be any frame within the animation.
Activity 36.11
Modify your program so that the hive alien's animation (and repositioning) is
halted when it moves to a z-ordinate value of less than -6.
The SET OBJECT FRAME Statement

When an animated model is first loaded, it displays frame zero. We can set the model
to show any specific frame using the SET OBJECT FRAME statement which has
FIG-36.8
The SET OBJECT
FRAME Statement
SET OBJECT FRAME objno , frame [ , boundsflg
In the diagram:

object to which this operation is to be applied.
frame is an integer value specifying the frame of the 3D

animated object to be displayed.
boundsflg 0 or 1.

0 - the collision volume around the object is not
recalculated.
1 - the collision volume is recalculated.
As we'll see in a later chapter, every 3D object is surrounded by an invisible box or

sphere which is used to help determine when a collision between two objects has
occurred. This is known as the bounding volume.
If, when we move to a different frame in an animated 3D object, that object has
changed orientation, then we have the option to reposition the bounding volume to
match the new position of the object (see FIG-36.9).
FIG-36.9 Recalculating the Bounding Volume for a New Frame
Every 3D object is surrounded by If the object is animated and moved to a However, if boundsflg = 0, then the
an invisible bounding volume used to new frame, using boundsflg = 1 will move bounding volume will not be
help detect collisions. the bounding volume with the object. repositioned.
Failing to recalculate the bounding volume will give us inaccurate collision results.
Activity 36.12
Modify your program so that when the hive alien stops, frame 13 is shown.
The SET OBJECT INTERPOLATION Statement

When an animation has to jump to a specific frame, this can create a rather jerky
effect as the 3D object jumps suddenly from one position to another. By using the
SET OBJECT INTERPOLATION statement, we can create a much smoother
transition with the program actually creating temporary frames to move gradually
from the current position to the new one. The format for this statement is shown in
FIG-36.10.
FIG-36.10
The SET OBJECT SET OBJECT INTERPOLATION objno , value
INTERPOLATION
Statement
In the diagram:

object to which this operation is to be applied.
value is 1 to 100 and specifies the degree of

interpolation to be used. A value of 100 gives
instant transition to the final frame; a value of 1
will create a slow transition.

The program in LISTING-36.4 demonstrates the effect of the SET OBJECT
INTERPOLATION statement by creating a jump from frame 0 to frame 13, first
without interpolation, and, the second time, with interpolation.

Smoother Frame COLOR BACKDROP 0
Movement using BACKDROP ON
Interpolation AUTOCAM OFF
REM *** Jump to frame 13; no interpolation ***
REPEAT
SET CURSOR 100,100
PRINT "Press key for jump without interpolation"
UNTIL INKEY$() <> ""
SET OBJECT FRAME 1,13
WAIT 500
REM *** Jump to frame 13; no interpolation ***
REPEAT
SET CURSOR 100,100
PRINT "Press key to return to frame 0"
WAIT 500
REPEAT
SET CURSOR 100,100
PRINT "Press key for jump with interpolation"
SET OBJECT INTERPOLATION 1,1
WAIT KEY
END
Activity 36.13
Type in and test the program in LISTING-36.4 (loading04.dbpro).
The APPEND OBJECT Statement

Often several animation files will exist which show the same object performing
different actions. For example, the hive alien is shown moving into attack position
in a file called H-Alien Hivebrain-Attack1.x. Rather than load this animation as a
separate object, the frames the new file contains can be copied into the original
object using the APPEND OBJECT statement which has the format shown in
FIG-36.11.
FIG-36.11
The APPEND OBJECT
APPEND OBJECT filename , objno , frame
Statement
In the diagram:
filename is a string giving the name of the file

containing the 3D animation. This string may
contain drive and path details.

objno is an integer value giving the ID of the existing 3D
object to which the new animation is to be added.
frame is an integer value specifying the frame number at

which the new animation is to be inserted into
the existing object.
To add the alien attack animation to the end of the object containing the moving
alien (object 1), we'd use the line:
APPEND OBJECT "H-Alien Hivebrain-Attack1.x", 1,

ÄTOTAL OBJECT FRAMES(1)+1
Activity 36.14
Modify program loading03.dbpro so that the alien attacks when it is at its

closest point to the camera.
Retrieving Animation Object Information

The OBJECT PLAYING Statement
We can determine if an animated 3D object is currently playing using the OBJECT
PLAYING statement which has the format shown in FIG-36.12.
FIG-36.12 OBJECT PLAYING ( objno )

The OBJECT PLAYING
Statement
integer
In the diagram:
objno is an integer value giving the ID of the animated

3D object being checked.
The statement returns 1 if the specified object is playing; otherwise zero is returned.
The OBJECT LOOPING Statement

To check if an animated 3D object is looping (rather than just playing) we can use
the OBJECT LOOPING statement which has the format shown in FIG-36.13.
FIG-36.13
OBJECT LOOPING ( objno )
The OBJECT
LOOPING Statement
integer
In the diagram:


The statement returns 1 if the specified object is looping; otherwise zero is returned.
The OBJECT FRAME Statement

To discover which frame of an animated 3D object is currently being displayed, use
the OBJECT FRAME statement which has the format shown in FIG-36.14.
FIG-36.14
OBJECT FRAME ( objno )
The OBJECT FRAME
Statement
integer
In the diagram:

The function returns the number of the frame currently being displayed by the
specified object.
The OBJECT SPEED Statement

The speed at which an animated 3D object is playing can be determined using the
OBJECT SPEED statement whose format is shown in FIG-36.15.
FIG-36.15
OBJECT SPEED ( objno )
The OBJECT SPEED
Statement
integer
In the diagram:

The statement returns the current speed setting. The default speed of a 3D animated
object is 100, and that will be the value returned by this statement, unless the object
in question has had its speed changed using the SET OBJECT SPEED.
The OBJECT INTERPOLATION Statement

The current interpolation value (given using SET OBJECT INTERPOLATION)
can be determined using the OBJECT INTERPOLATION statement whose format
FIG-36.16
The OBJECT
OBJECT INTERPOLATION ( objno )
INTERPOLATION
Statement
integer

In the diagram:

The OBJECT SIZE Statement

The size of an object in each dimension can be discovered using the OBJECT SIZE
statement. In fact, there are two separate versions of this statement. The first gives
a single value representing the overall size of a 3D object, taking into account all
three dimensions. The format for this statement is shown in FIG-36.17.
FIG-36.17
The OBJECT SIZE OBJECT SIZE ( objno )
Statement (Version 1)
real
In the diagram:

whose size is to be determined.
For example, we could display the overall size of object 1 using the line:
PRINT OBJECT SIZE(1)
The size is given in world units.
The second version of OBJECT SIZE is used to discover the actual size of the 3D
object in each dimension. This version of the statement has the format shown in
FIG-36.18.
FIG-36.18
X
The OBJECT SIZE
Statement (Version 2)
OBJECT SIZE Y ( objno )
real
In the diagram:
X,Y,Z Choose the appropriate option for the dimension

required (X = width; Y = height; Z = depth).

whose size is to be determined.
For example, we could discover the height of object 1 using the line:
PRINT OBJECT SIZE Y(1)
The program in LISTING-36.5 loads the H-Alien Hivebrain-Move.x and displays

the model's overall size as well as its width, height and depth.

Finding the Size of a 3D COLOR BACKDROP 0
Object BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Load object ***

size# = OBJECT SIZE(1)
sizex# = OBJECT SIZE X(1)
sizey# = OBJECT SIZE Y(1)
sizez# = OBJECT SIZE Z(1)
REPEAT
SET CURSOR 100,100
PRINT "Overall size: ",size#," width:",sizex#,"
height:",sizey#," depth:",sizez#
WAIT 500
END
Activity 36.15
Limbs
We looked at limbs in the previous chapter, but here we have a chance to examine
the limbs of a model created outwith DarkBASIC Pro.
The program in LISTING-36.6 displays the names of all the limbs in the model
H-Alien Hivebrain-Move.x and then hides one of the limbs.
Examining the Limbs of COLOR BACKDROP 0
an Imported Model BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
SYNC ON
REM *** Load object ***

SYNC
REM *** Get and display model limb details ***

nooflimbs = CHECKLIST QUANTITY()
PRINT "Model contains ",nooflimbs," limbs"
FOR c = 1 TO nooflimbs
NEXT c
SYNC
REM *** Return to automatic screen updating ***

WAIT KEY
SYNC OFF
WAIT KEY
END

Activity 36.16
The limbs of an imported model can be manipulated using the statements covered
in the previous chapter.
Activity 36.17
Modify your last program to hide limb 2 from the Hivebrain model.
Summary
l Complex 3D models are usually created using a 3D drawing package.
l DarkBASIC Pro can import .DBO, .3DS, .X and other format models.
l Models may be static or animated.
l Animated models consist of several frames - a bit like frames from a movie.
l Each frame of an animated model shows the model in a slightly different pose
from the previous frame.
l Use LOAD OBJECT to load a model stored in a file.
l Make sure the model and the texturing it uses are stored in the current project
folder.
l Use PLAY OBJECT to play a selected number of frames in an animated model.
l PLAY OBJECT plays through the specified frames only once.
l Use LOOP OBJECT to continuously play a set of frames from an animated

model.
l Use TOTAL OBJECT FRAMES to discover exactly how many frames are used
in an animated model.
l Use SET OBJECT SPEED to set the playing speed of a playing model animation.
l Use STOP OBJECT to halt the playing of an animated model.
l When an animation is stopped it continues to show the frame at which it halted.
l Use SET OBJECT FRAME to set an animated model to a specific frame.
l When setting the frame, the bounding volume may be recalculated.
l Use SET OBJECT INTERPOLATION to smooth the transition when jumping

from one frame to another in an animated model.
l Use APPEND OBJECT to add the frames of one model to those of another.
l Use OBJECT PLAYING to determine if a specific animated model is playing.

l Use OBJECT LOOPING to determine if a specific animated model is currently
looping.
l Use OBJECT FRAME to discover which frame is currently showing on an

animated model.
l Use OBJECT SPEED to discover the current speed setting for an animated
model.
l Use OBJECT INTERPOLATION to discover the interpolation setting for an

animated model.
l The limbs of an imported model can be identified using the PERFORM

CHECKLIST FOR OBJECT LIMBS statement covered in the previous chapter.
l The limbs of an imported model can be manipulated in the same way as limbs
created using DarkBASIC Pro code.

Solutions Activity 36.7
Activity 36.1
No solution required. REM *** Set screen resolution ***
BACKDROP ON
Activity 36.2 REM *** Position camera ***
No solution required. AUTOCAM OFF
Activity 36.3 SET LIGHT RANGE 0,15
REM *** Set screen resolution *** REM *** Load object ***
SET DISPLAY MODE 1280,1024,32 LOAD OBJECT "H-Alien Hivebrain-Move.X",1
COLOR BACKDROP RGB(155,155,0) WAIT KEY
BACKDROP ON LOOP OBJECT 1
REM *** Load the jet object *** REM *** Display number of frames ***
LOAD OBJECT "H-Jet Fighter-Move.X",1 totalframes = TOTAL OBJECT FRAMES(1)
REM *** Position camera *** DO
POSITION CAMERA 0,0,-20 SET CURSOR 100,100
REM *** rotate jet *** PRINT "Total frames = ",totalframes
DO LOOP
TURN OBJECT LEFT 1,1 REM *** End program ***
PITCH OBJECT UP 1,1 WAIT KEY
WAIT 10 END
LOOP
END
Activity 36.8
Activity 36.4 SET DISPLAY MODE 1280, 1024, 32
BACKDROP ON
AUTOCAM OFF
Activity 36.5 POSITION CAMERA 0,1,-10
REM *** Set screen resolution *** REM *** Create a point light ***
SET DISPLAY MODE 1280, 1024, 32 SET POINT LIGHT 0,-10,10,0
COLOR BACKDROP RGB(155,155,0) SET LIGHT RANGE 0,15
BACKDROP ON REM *** Load object ***
REM *** Position camera *** LOAD OBJECT "H-Alien Hivebrain-Move.X",1
POSITION CAMERA 0,1,-10 LOOP OBJECT 1,0,20
REM *** Create a point light *** totalframes = TOTAL OBJECT FRAMES(1)
SET POINT LIGHT 0,-10,10,0 DO
SET LIGHT RANGE 0,15 SET CURSOR 100,100
REM *** Load object *** PRINT "Total frames = ",totalframes
LOAD OBJECT "H-Alien Hivebrain-Move.X",1 LOOP
PLAY OBJECT 1 WAIT KEY
REM *** End program *** END
WAIT KEY
END
Activity 36.9

SET DISPLAY MODE 1280, 1024, 32 Activity 36.10
BACKDROP ON REM *** Set screen resolution ***
REM *** Position camera *** SET DISPLAY MODE 1280, 1024, 32
AUTOCAM OFF COLOR BACKDROP 0
POSITION CAMERA 0,1,-10 BACKDROP ON
REM *** Create a point light *** AUTOCAM OFF
SET POINT LIGHT 0,-10,10,0 REM *** Create a point light ***
SET LIGHT RANGE 0,15 SET POINT LIGHT 0,-10,10,0
REM *** Load object *** SET LIGHT RANGE 0,15
LOAD OBJECT "H-Alien Hivebrain-Move.X",1 REM *** Load object; position camera ***
WAIT KEY LOAD OBJECT "H-Alien Hivebrain-Move.X",1
LOOP OBJECT 1 POSITION CAMERA 0,1,-10
REM *** End program *** REM *** Play object continuously ***
WAIT KEY LOOP OBJECT 1,0,20
END REM *** Change play speed ***
SET OBJECT SPEED 1,50

REM *** Move alien *** REM *** Set screen resolution ***
DO SET DISPLAY MODE 1280, 1024, 32
MOVE OBJECT 1,-0.03 COLOR BACKDROP 0
LOOP BACKDROP ON
REM *** End program *** AUTOCAM OFF
END REM *** Create a point light ***
Activity 36.11 REM *** Load object ***
REM *** Set screen resolution *** REM *** Add attack frames ***
SET DISPLAY MODE 1280, 1024, 32 APPEND OBJECT "H-Alien
COLOR BACKDROP 0 Hivebrain-Attack1.x",1,26
BACKDROP ON REM *** Position camera ***
AUTOCAM OFF POSITION CAMERA 0,1,-10
REM *** Create a point light *** REM *** Play object continuously ***
SET POINT LIGHT 0,-10,10,0 LOOP OBJECT 1,0,20
SET LIGHT RANGE 0,15 REM *** Change play speed ***
REM *** Load object; position camera *** SET OBJECT SPEED 1,50
LOAD OBJECT "H-Alien Hivebrain-Move.X",1 REM *** Move alien ***
POSITION CAMERA 0,1,-10 DO
REM *** Play object continuously *** MOVE OBJECT 1,-0.03
LOOP OBJECT 1,0,20 REM *** Stop alien when its close ***
REM *** Change play speed *** IF OBJECT POSITION Z(1) < -6
SET OBJECT SPEED 1,50 STOP OBJECT 1
REM *** Move alien *** PLAY OBJECT 1,26,
DO Ä
TOTAL OBJECT FRAMES(1)
MOVE OBJECT 1,-0.03 EXIT
REM *** Stop alien when its close *** ENDIF
IF OBJECT POSITION Z(1) < -6 LOOP
STOP OBJECT 1 REM *** End program ***
EXIT WAIT KEY
ENDIF END
LOOP
END
Activity 36.12
Activity 36.16
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
REM *** Create a point light *** Activity 36.17
SET LIGHT RANGE 0,15 To hide limb 2 add the lines
LOAD OBJECT "H-Alien Hivebrain-Move.X",1 REM *** Hide a limb ***
POSITION CAMERA 0,1,-10 HIDE LIMB 1,2
REM *** Play object continuously ***
LOOP OBJECT 1,0,20 just before the line
REM *** Change play speed ***
SET OBJECT SPEED 1,50 REM *** End program ***
REM *** Move alien ***
DO
MOVE OBJECT 1,-0.03
REM *** Stop alien when its close ***
IF OBJECT POSITION Z(1) < -6
STOP OBJECT 1
EXIT
ENDIF
LOOP
WAIT KEY
END
Activity 36.13
Activity 36.14

37
How to Check if an Object is on Screen
How to Select an Object on the Screen
Calculating Distance from the Camera
Calculating Position Relative to the Camera
Screen Coordinates and World Coordinates
Using the Mouse for Selection
DarkBASIC Pro: Screen Control 979

User Control
Introduction
There must be very few games in which the user does not directly control the
movement of at least one 3D object. For example, if we created a 3D chess game,
we'd need to allow the player to select which piece is to be moved next.
In this chapter we're going to look at some techniques for helping the player to select
from a group of 3D objects, as well as finding out the commands available to help
with this task.
Selecting an Object
FIG-37.1 shows a line of three balls. The aim of this section is to examine ways of
selecting one of the three balls for deletion.
FIG-37.1
A Choice of Objects
The code required to create this set up is given in LISTING-37.1.
LISTING-37.1 FUNCTION ScreenSetUp()

Creating the Objects SET DISPLAY MODE 1280,1024,32
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION CreateBalls()
#CONSTANT firstballobj 11
#CONSTANT lastballobj 13
#CONSTANT ballimg 2
REM *** Load ball texture ***

LOAD IMAGE "lava1.bmp",ballimg
REM *** Create balls ***

x = -9
FOR objno = firstballobj TO lastballobj
MAKE OBJECT SPHERE objno,2,20,20
TEXTURE OBJECT objno,ballimg
POSITION OBJECT objno,x,0,0
x = x + 3
NEXT objno
ENDFUNCTION
980 DarkBASIC Pro: Screen Control

Activity 37.1
Type in the code given above (control01.dbpro) and add a main section to call
these functions and display the balls on the screen.
Next, we will add a selector. This would give the player a visual clue as to which
object is currently selected (see FIG-37.2).
FIG-37.2
Adding a Selector
The new object is simply a cube textured with a blue square and a black centre (see
FIG-37.3). The black area of the object has been made transparent.
FIG-37.3
The Texture Used on
the Selector Cube
A function to add the selector is given below:
FUNCTION CreateSelector()
#CONSTANT selectorobj 2
#CONSTANT selectorimg 3
REM *** Load selector texture ***

LOAD IMAGE "selector.bmp",selectorimg
REM *** Create selector cube ***

MAKE OBJECT CUBE selectorobj,2
TEXTURE OBJECT selectorobj, selectorimg
SET OBJECT TRANSPARENCY selectorobj,1
POSITION OBJECT selectorobj,0,0,0
SET OBJECT CULL selectorobj,0
ENDFUNCTION
Activity 37.2
Add the new function to your program (calling it from the main section) and
check that the selector is created successfully.
The next function allows the user to move the selector left and right using the arrow
keys and to make a final selection by pressing the Enter key.
FUNCTION MoveSelector()
REPEAT
REM *** Get current selector position ***
x = OBJECT POSITION X(selectorobj)
y = OBJECT POSITION Y(selectorobj)

z = OBJECT POSITION Z(selectorobj)
REM *** Move selector left or right ***
IF LEFTKEY()
POSITION OBJECT selectorobj,x-3,y,z
ELSE IF RIGHTKEY()
POSITION OBJECT selectorobj,x+3,0,0
ENDIF
ENDIF
CHR$(13) is the Enter WAIT 100
(or Return) key. UNTIL INKEY$() = CHR$(13)
ENDFUNCTION
The WAIT 100 statement at the end of the function is there to slow down the whole
process. If not included the selector is too responsive to key presses.
Activity 37.3
Add the new function to your program and test that it works correctly.
When the selector has moved over the required ball, we can see on the screen which
item is selected, but we also need to get the program itself to realise which ball has
been picked. To do that we need to learn a few more DarkBASIC Pro statements.
The OBJECT SCREEN Statement

3D objects exist in a 3 dimensional space and yet are displayed on a two-dimensional
screen. Although an object may remain at a fixed point in 3D space, still it can move
position on the screen if the camera were to move (see FIG-37.4).
Screen 1 Screen 2
FIG-37.4
Screen Position Changes
though 3D Position is
Unchanged
The objects shift to a different position on the screen as the camera is moved
but remain fixed at the same location in world space.
We can use the OBJECT SCREEN statement to discover the x,y coordinates of a
3D object on the screen. The statement has the format shown in FIG-37.5
{
FIG-37.5
X
The OBJECT SCREEN OBJECT SCREEN ( objno )
Statement
Y
real
In the diagram:
X,Y Use one of these to select the required ordinate.

screen position is to be determined.
For example, we could discover the screen coordinates of the selector object in the
program we are developing using the lines:
selectorx# = OBJECT SCREEN X(selectorobj)

selectory# = OBJECT SCREEN Y(selectorobj)
The PICK OBJECT Statement

The PICK OBJECT statement complements the OBJECT SCREEN statement by
returning the ID of the object which lies at a particular spot on the screen. The
statement searches through all objects within a given set of object IDs and returns
the ID of the first object it finds at the screen coordinates given. The PICK OBJECT
FIG-37.6
The PICK OBJECT PICK OBJECT ( x , y , objno1 , objno2 )
Statement
integer
In the diagram:
x,y are a pair of integer values giving the screen

coordinates to be checked.
objno1 is an integer value giving the lowest object ID to

be checked.
objno2 is an integer value giving the highest object ID to

be checked.
If any part of an object with an ID between objno1 and objno2 is positioned at screen
coordinates (x,y), then the ID of that object is returned by the statement. If point
(x,y) does not lie within any of the objects, zero is returned.
For example, we could check which ball covers the screen position (selectorx#,
selectory#) with the line:
objno =
PICK OBJECT(selectorx#,selectory#,firstballobj,lastballobj)
Now we have enough information to create a new function, GetObjectSelected(),

which returns the ID of the ball selected.
FUNCTION GetObjectSelected()
objno =
PICK OBJECT(selectorx#,selectory#,firstballobj,lastballobj)
ENDFUNCTION objno
To finish off, we'll create a routine to handle the ball selected. For the moment we'll
just make the ball disappear.

FUNCTION HandleObject(objno)
REM *** IF no object selected, exit ***
IF objno = 0
EXITFUNCTION
ENDIF
REM *** Hide object selected ***
HIDE OBJECT objno
ENDFUNCTION
Activity 37.4
Add the functions GetObjectSelected() and HandleObject(), to your program.
Code the main section of the program as follows:
ScreenSetUp()
CreateBalls()
CreateSelector()
DO
MoveSelector()
objno = GetObjectSelected()
HandleObject(objno)
LOOP
WAIT KEY
END
Test the program by hiding all three balls.
When a PICK OBJECT statement is executed, it generates several items of data.

These are the distance from the camera to the object selected and the position of
that object in 3D space relative to the camera. The following two statements allow
us to access this data.
The GET PICK DISTANCE Statement

Executing the PICK OBJECT statement causes DarkBASIC Pro to calculate the
distance between the camera and the object selected (see FIG-37.7).
FIG-37.7
Camera distanc
The Camera-Object e
Object
Distance
We can access this information using the GET PICK DISTANCE statement which
FIG-37.8
The GET PICK GET PICK DISTANCE ( )
DISTANCE Statement
real
The statement returns the distance in world units between the camera and the surface
of the object returned by the previous PICK OBJECT statement.

This information would be useful if we needed to calculate if an enemy is within
hitting distance in a First Person Shooter game.
We could rewrite the HandleObject() routine to display the distance of the selected
sphere:
IF objno = 0
EXITFUNCTION
ENDIF
REM *** Find distance from camera to ball ***
distance# = GET PICK DISTANCE()
INK 0,0
REM *** Display distance for 2 seconds ***
t = TIMER()
WHILE TIMER() - t < 2000
SET CURSOR 100,100
PRINT "Distance = ",distance#
ENDWHILE
ENDFUNCTION
Activity 37.5
Modify your HandleObject() function to match the code given above.
Test the routine by checking the distance of all three balls from the camera.
The PICK VECTOR Statement

The second data item created by executing the PICK OBJECT statement is a 3D
vector containing the coordinates of the selected object. The coordinates are not the
world coordinates of the object (which we could obtain using OBJECT POSITION)
but the position of the object as measured from the camera (see FIG-37.9).
FIG-37.9 xoffset
PICK VECTOR gives Camer

a yoffset
the Position of the Object
Selected Object relative
to the Camera
zoffset
We can examine the contents of this vector using the PICK VECTOR statement
which is shown in FIG-37.10.
FIG-37.10
{
X
The PICK VECTOR
Statement
PICK VECTOR Y ( )
real
In the diagram:
X,Y,Z Use the appropriate option for the ordinate

required.

For example, if we assume the statement
objno = PICK OBJECT (640,512,0,2)
returns a non-zero value, we could get the returned object's position (relative to the
camera) using the lines:
x# = GET PICK VECTOR X()

y# = GET PICK VECTOR Y()
z# = GET PICK VECTOR Z()
The program in LISTING-37.2 allows the user to move the camera using the arrow
keys and displays a continual readout of the relative position of a sphere in relation
to the camera.

An Object's AUTOCAM OFF
Coordinates Relative to POSITION CAMERA 0,0,-10
the Camera POINT CAMERA 0,0,0
REM *** Create sphere ***

REM *** main loop ***

DO
objno = PICK OBJECT (640,512,0,2)
DisplayCoords(objno)
LOOP
END
FUNCTION DisplayCoords(objno)
REM *** Get object's relative coords ***
REM *** Display info on screen ***
SET CURSOR 100,100
PRINT "Object:",objno,"Relative coords x:",x#," Y:",y#,
" z:",z#
ENDFUNCTION
Activity 37.6
Type in and test the program given in LISTING-37.2 (control02.dbpro).
The PICK SCREEN Statement

It is possible to pick an arbitrary point on the screen, an arbitrary distance "into" the
screen and have DarkBASIC Pro calculate the coordinates of that point in 3D space
relative to the camera. This information is stored in the PICK vector.
The operation is performed using the PICK SCREEN statement which has the
FIG-37.11
The PICK SCREEN PICK SCREEN x , y , distance
Statement

In the diagram:
x,y are a pair of real values giving the screen

coordinates chosen.
distance is a real value giving the depth factor to be taken

into account.
For example, if we were to execute the line

PICK SCREEN 200,300,15
the computer would work out the equivalent point in the 3D world and return the
coordinates of that point relative to the active camera's position.
The program in LISTING-37.3 allows the user to mouse click anywhere on the
screen and returns the coordinates of the equivalent point at a screen depth of 15
units.
Screen Coordinates AUTOCAM OFF
Converted to World POSITION CAMERA 0,0,-10
Coordinates POINT CAMERA 0,0,0
REM *** Create object to create correct screen mode ***
DO
DisplayCoords()
LOOP
END
FUNCTION DisplayCoords()
REM *** Get mouse coords ***
PICK SCREEN MOUSEX(), MOUSEY(), 15
SET CURSOR 100,100
PRINT "Relative coords x:",x#," Y:",y#," z:",z#
ENDFUNCTION
Activity 37.7
Type in and test the program in LISTING-37.3 (control03.dbpro). How could

we convert the relative coordinates to absolute coordinates?
The OBJECT IN SCREEN Statement

We can check to see if a specific object actually appears within the screen area using
the OBJECT IN SCREEN statement which has the format shown in FIG-37.12.
FIG-37.12
The OBJECT IN OBJECT IN SCREEN ( objno )
SCREEN Statement
integer

In the diagram:

to be checked.
The routine returns 1 if the object specified is wholly or partially on the screen;
The program in LISTING-37.4 displays the on-screen/off-screen status of a sphere

as it moves slowly towards the edge of the screen.
Checking if an Object is AUTOCAM OFF
On Screen POSITION CAMERA 0,0,-20
POINT CAMERA 0,0,0

DO
MOVE OBJECT 1,0.01
DisplayStatus(1)
LOOP
END
FUNCTION DisplayStatus(objno)
SET CURSOR 100,100
IF OBJECT IN SCREEN(objno)
PRINT "On screen"
ELSE
PRINT "Off screen"
ENDIF
ENDFUNCTION
Activity 37.8
Type in and test the program in LISTING-37.4 (control04.dbpro).
Are the same results obtained when the object is invisible?
What happens when an object passes behind the camera?
Selecting Objects using the Mouse

We can use the mouse to select an object on the screen by moving the mouse pointer
over the required object and then feeding the mouse coordinates as parameters to
the PICK OBJECT statement.
The program in LISTING-37.5 is a variation on the three-ball example we created

earlier in this chapter. By clicking on a ball with the mouse, the ball will disappear.
LISTING-37.5
ScreenSetUp()
Selecting an Object CreateBalls()
Using the Mouse DO
objno = GetObjectSelected()
HandleObject(objno)
LOOP

LISTING-37.5 REM *** End program ***
(continued) END
Selecting an Object FUNCTION ScreenSetUp()

Using the Mouse REM *** Set up screen ***
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
#CONSTANT ballimg 2
x = -9
x = x + 3
NEXT objno
ENDFUNCTION
FUNCTION GetObjectSelected()
REM *** Wait for mouse click ***
WHILE MOUSECLICK()=0
ENDWHILE
REM *** Select object under mouse pointer ***
objno = PICK OBJECT(MOUSEX(),MOUSEY(),firstballobj,lastballobj)
ENDFUNCTION objno
IF objno = 0
EXITFUNCTION
ENDIF
HIDE OBJECT objno
ENDFUNCTION
Activity 37.9
Type in and test the program in LISTING-37.5 (control05.dbpro).
Modify the program to have a 3 by 3 grid of balls. Each new row of balls
should be displaced by 3 units along the z-axis (see below).
Check that each of the nine balls can be deleted.

Summary
l DarkBASIC Pro has several statements that allow us to relate 2D screen
coordinates to the world coordinates used by 3D objects.
l Use the OBJECT SCREEN statement to discover the screen coordinates of a

specific 3D object.
l Use the PICK OBJECT statement to discover the ID of an object at a given point
on the screen.
l Use GET PICK DISTANCE to discover how far the object detected by PICK
OBJECT is from the camera.
l Use PICK VECTOR to discover the coordinates of the object detected by PICK
OBJECT relative to the camera.
l Use PICK SCREEN to convert a 2D screen position and a depth value to a point
in 3D space. The coordinates generated are relative to the camera position.
l Use OBJECT IN SCREEN to determine if a specific object is within the view

of the camera (and hence, appearing on the screen).

Solutions x = -9
Activity 37.1
The main section should be coded as: FOR objno = firstballobj TO lastballobj
ScreenSetUp()
x = x + 3
CreateBalls()
NEXT objno
ENDFUNCTION
WAIT KEY
END
Activity 37.2 REM *** Load selector texture ***
The main section is now: REM *** Create selector cube ***
MAKE OBJECT CUBE selectorobj,2
REM *** Main section *** TEXTURE OBJECT selectorobj, selectorimg
ScreenSetUp() SET OBJECT TRANSPARENCY selectorobj,1
CreateBalls() POSITION OBJECT selectorobj,0,0,0
CreateSelector() SET OBJECT CULL selectorobj,0
REM *** End program *** ENDFUNCTION
WAIT KEY
END FUNCTION GetObjectSelected()
Activity 37.3 objno = PICK OBJECT
(selectorx#,selectory#,firstballobj,
The main section changes to: lastballobj)
ENDFUNCTION objno
ScreenSetUp() FUNCTION HandleObject(objno)
CreateBalls() REM *** IF no object selected, exit ***
CreateSelector() IF objno = 0
MoveSelector() EXITFUNCTION
REM *** End program *** ENDIF
WAIT KEY REM *** Hide object selected ***
END HIDE OBJECT objno
ENDFUNCTION
Activity 37.4 FUNCTION MoveSelector()

REPEAT
The complete version of the program is: REM *** Get selector position ***
x = OBJECT POSITION X(selectorobj)
REM *** Main section *** y = OBJECT POSITION Y(selectorobj)
ScreenSetUp() z = OBJECT POSITION Z(selectorobj)
CreateBalls() REM *** Move selector left/right ***
CreateSelector() IF LEFTKEY()
DO POSITION OBJECT selectorobj,x-3,y,z
MoveSelector() ELSE IF RIGHTKEY()
objno = GetObjectSelected() POSITION OBJECT selectorobj,x+3,0,0
HandleObject(objno) ENDIF
LOOP ENDIF
REM *** End program *** WAIT 100
END UNTIL INKEY$() = CHR$(13)
ENDFUNCTION
REM *** Set up screen *** Activity 37.5
COLOR BACKDROP RGB(255,255,100) No solution required.
BACKDROP ON
POINT CAMERA 0,0,0 No solution required.
ENDFUNCTION
FUNCTION CreateBalls() Activity 37.7

#CONSTANT lastballobj 13 Adding the camera's coordinates to the relative
#CONSTANT ballimg 2 coordinates gives us the absolute coordinates.

Activity 37.8
Being hidden makes no difference to an object's on-screen
status.
We could make the object move behind the camera by

moving it along the z-axis. This only requires the direction
in which the sphere is pointing to be changed:
POINT OBJECT 1,0,0,-1
The object is considered to be off-screen when it passes

behind the camera.
Activity 37.9
To create nine balls we need to modify the CreateBalls()
function to create 3 rows of 3. The new version of the
routine given below keeps changes to a minimum.
#CONSTANT ballimg 2
x = -9
z = 0
POSITION OBJECT objno,x,0,z
x = x + 3
IF x = 0
x = -9
z = z + 3
ENDIF
NEXT objno
ENDFUNCTION

38
Adding Help to a Game
Constructing a 3D Game of Skill
Recording Game State
DarkBASIC Pro: Solitaire 993

Solitaire - The Board Game
Introduction
In this chapter we are going to develop a 3D version of the board game Solitaire
(see FIG-38.1) - not to be confused with the card game of the same name.
FIG-38.1
The Game of Solitaire
The Equipment
The game consists of 32 marbles and a board with 33 indentations, or pits. The pits
form the shape of a plus sign (+). A marble is placed in every pit, except the central
one.
The Aim
The aim of the game is to remove all but one marble from the board. This last marble
should be in the central pit at the end of the game.
The Rules
Any marble can be moved by jumping over one other marble into an empty pit.
Jumps can be horizontal or vertical but not diagonal. Only a single marble can be
jumped over on each move. The marble which is jumped over is removed from the
board.
Creating a Computer Version of the Game

User Controls
Each turn involves the player first selecting a marble and then an empty pit into
which the marble is to be moved.
To achieve this, the player will move a selector cube about the board using the arrow
keys. Once over a required pit, the Enter key is pressed to indicate a selection.
994 DarkBASIC Pro: Solitaire

Pressing U will allow the player to deselect a marble.
Pressing N will initiate a new game.
Pressing F will terminate the current game.
Pressing F1 will display the help screen. Within the help screen, pressing R will
display the rules, and K the keys which may be used in the game.
Game Responses
The game will respond to player actions in the following ways:
Ø When a marble is selected, it will change colour.

Ø When an empty pit is selected as the second part of a move, the selected
marble will jump to the new position and return to its original colour.
The jumped marble will be removed.
Ø When U is pressed immediately after a marble has been selected, the
selected marble will return to its original colour and the player will have
to select a new marble.
Ø If an empty pit is selected when a marble selection is required, an error
message will be displayed and the player must select a pit containing a
marble.
Ø If an occupied pit is selected when a empty pit is required, an error
message will be displayed and a new destination pit must be selected.
Ø If the marble and destination pit selected do not generate a valid move, an
error message will be displayed and a new destination pit must be
selected.
Ø If N is pressed, a new game is started. No prompt is given.
Ø If F is pressed, the program will terminate after displaying a shut down
message.
Ø If F1 is pressed, the introductory page of the Help section will be
displayed. A second press of the same key will cause the Help section to
disappear.
Ø If R is pressed while a help page is visible, the help page containing the
rules will be displayed.
Ø IF K is pressed while a help page is visible, the help page showing the
keys which can be used in the game will be displayed.
Screen Layout
The screen layout is shown in FIG-38.1.
Media Used
The images used to texture the board, marbles and selector cube are shown in
FIG-38.2. Notice that even the help text is held as a single image with three frames.

FIG-38.2 The Images Used in the Program
Marble texture
Board texture Selector texture Help image (3 frames)
Data Structures
Although we can use 3D objects to represent the board visually, we need a second
way of representing the board as data. This will make the coding easier when trying
to decide on valid moves.
If we had a square playing area, the obvious approach would be to use a

two-dimensional array with one cell representing one pit on the board. However,
despite using a cross-shaped playing area, with a little thought we can still make
use of a two-dimensional array. At the centre the board is 7 pits wide and 7 pits
deep, so a 7 by 7 integer array could be used to represent the board:
DIM board (7,7)
Now we'll fill the array with a zero in any cell that corresponds to a pit and with -1
where a cell corresponds to a flat part of the board. FIG-38.3 shows the board and
the corresponding array contents.
FIG-38.3
The Board and the Array -1 -1 0 0 0 -1 -1
used to Represent it -1 -1 0 0 0 -1 -1
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0
-1 -1 0 0 0 -1 -1
-1 -1 0 0 0 -1 -1
board Array
Actual Board Layout
Lower ID values are Rather than just store zeros, it will suit us better to store the IDs of each marble in
reserved for other objects. the cells of the array. So, if we assume the marbles have object IDs starting at 11,
the board array would start off as shown in FIG-38.4.
FIG-38.4 -1 -1 11 12 13 -1 -1
-1 -1 14 15 16 -1 -1
The board array with 17 18 19 20 21 22 23
Marble IDs
24 25 26 27 28 29 30
31 32 33 34 35 36 37
-1 -1 38 39 40 -1 -1
-1 -1 41 42 43 -1 -1
board Array

It would be useful to employ a second 7 by 7 array in which to store the world
coordinates of each pit. Although we could work the positions out mathematically
(since the pits are evenly spaced) it will probably create less complex code if we
just store the coordinates of each position in another array. Since the board is flat,
it won't be necessary to store the y value of any position (since these will all be
identical), just the x and z values.
When we create a 3D object, its centre is at (0,0,0), so, assuming we don't move the
board, the centre pit will have x, z values of 0,0.
Since we need to store two values in every cell of the array (the x and z ordinates),
we'll need to create a record structure for this
TYPE Coords
x AS INTEGER
z AS INTEGER
ENDTYPE
before creating our array:
DIM boardCoords(7,7) AS Coords
Now, assuming the centre of each pit is 3 world units in both directions from any
neighbour, the new array would contain the values shown in FIG-38.5.
FIG-38.5
-1 -1 (-3,9) (0,9) (3,9) -1 -1
The boardCoords Array
-1 -1 (-3,6) (0,6) (3,6) -1 -1
(-9,3) (-6,3) (-3,3) (0,3) (3,3) (6,3) (9,3)
(-9,0) (-6,0) (-3,0) (0,0) (3,0) (6,0) (9,0)
(-9,-3) (-6,-3) (-3,-3) (0,-3) (3,-3) (6,-3) (9,-3)
-1 -1 (-3,-6) (0,-6) (3,-6) -1 -1

-1 -1 (-3,-9) (0,-9) (3,-9) -1 -1
boardCoords Array
In practice, we'll probably store coordinates in the invalid positions of the array (the
four cells in each corner), since it is easier to write the code that way.
We'll also need to record the position of the selector cube, but this time we'll use its
position on the board, not its screen position. For example, if the selector is at the
centre of the board, then it is in row 4, column 4. To store this information, we'll
need another record structure
TYPE BoardPosition
row AS INTEGER
col AS INTEGER
ENDTYPE
then we can declare a variable for the selector's position:
selectorposition AS BoardPosition
On each turn the player chooses a marble and a position to which the marble is to

be moved. We'll also need to record these two positions:
GLOBAL DIM moves(2) AS BoardPosition
The final important variable that we will require is a game-state variable. There are
three main stages during game play. These are:
Ø About to choose a marble

Ø About to choose a move-to space
Ø Looking at the help text
We need to be aware of which state the game is in, because this dictates what options
are currently available to the player. When the player is about to choose a marble,
it would be invalid to select an empty pit; however, choosing a pit containing a
marble would be invalid when the player is about to choose the move-to position.
Trying to play while viewing the Help is also disallowed.
So our new variable will be assigned only the values 1, 2 or 3:
gamestate '1-choose marble; 2-choose move-to; 3-showing help
Lastly, we need to record how many marbles are currently on the board; when this
reduces to one, the game is over. We will do this using a variable named
marbleremaining.
All the variables above will be made global so that they can be accessed throughout
the program:
REM *** Declare data structure ***

TYPE Coords
x AS INTEGER
z AS INTEGER
ENDTYPE
TYPE BoardPosition
row AS INTEGER
col AS INTEGER
ENDTYPE
REM *** Declare main variables ***

GLOBAL DIM board(7,7) `Marble positions
GLOBAL DIM boardCoords(7,7) AS Coords `Marble coordinates
GLOBAL selectorposition AS BoardPosition `Position of selector
GLOBAL DIM move(2) AS BoardPosition `marble and move-to
`positions
GLOBAL gamestate `Game state
`1-choose marble;
`2-choose space;
`3-showing help
GLOBAL marblesremaining `marbles on board
The images and 3D objects also need to be assigned ID values and it will make the
code clearer if these are given names in #CONSTANT statements.
REM **********************************
REM *** Constants ***
REM **********************************
REM *** Object constants ***
#CONSTANT boardobj 1
#CONSTANT firstmarbleobj 3

#CONSTANT lastmarbleobj 34
REM *** Image constants ***
#CONSTANT boardimg 1
#CONSTANT marbleimg 2
#CONSTANT helpimage 4
REM *** Sprite constants ***
#CONSTANT helpsprite 1
Now it's time to have a look at the logic of the game. As usual, for a larger project,
the main section should consist mainly of function calls so that we can get an
overview of what's going on in just a few lines. In English, the whole logic could
be described as follows:
Initialise screen
Set up game
REPEAT
Get player's move
UNTIL game over
In the main, this translates into straight function calls but we'll have to know if the
player has chosen to quit the game, so the function which gets the player's move
will have to return a flag to indicate if the game should be ended. This means that
the main section should be coded as:
SetUpScreen()
SetUpGame()
REPEAT
quit = GetPlayersMove()
UNTIL quit OR marblesremaining = 1
END
Activity 38.1
Create a new project (Solitaire.dbpro) and copy the code for the constants and
data into the source file.
Code the main section as given above and add test stubs for each of the
functions called. The test stub for GetPlayersMove() should be:
FUNCTION GetPlayersMove()
PRINT "GetPlayersMove() executed"
ENDFUNCTION 1
Run the program and make sure it executes.
To stop the program finishing before you get time to see all the messages
displayed, add a WAIT KEY statement before the END command in the
program's main section.
Make sure all the media files have been copied to the project's folder.
Now we'll begin to build up the project by working our way through each routine
needed.
Adding SetUpScreen()
This routine sets the screen resolution and backdrop as well as positioning the
camera, so its code is quite straight forward:

BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 38.2
Replace the test stub for SetUpScreen() with the actual code as given above.
Once a 3D object (or sprite) appears on the screen, or we start using a camera, the
screen refresh method employed means that the output from any PRINT statement
vanishes almost as soon as it appears. Because of this, we can no longer see the
messages confirming that routines have been executed.
You might feel that we no longer need to see these messages now that we're up and
running but, unfortunately, you're probably wrong! Remember every program is
just waiting its opportunity to catch us out - so don't take any chances!
There are various ways we might overcome the problem of the disappearing
messages, but in this case we're going to create a replacement for the PRINT
statement that displays a string at any point on the screen for a given number of
seconds. The new function, DisplayMessage() is coded below:
FUNCTION DisplayMessage(x,y,message$,seconds#)
t = TIMER()
INK 0,0
millisecs = seconds# * 1000
WHILE TIMER() - t < millisecs
SET CURSOR x,y
PRINT message$
ENDWHILE
ENDFUNCTION
Activity 38.3
Add DisplayMessage() to the end of your program and replace all the PRINT
statements in the test stubs with calls to this function.
Adding SetUpGame()
This routine is a bit more complex since we need to set up the board, add the marbles,
position the selector and set up the Help feature. Since each of these will take up
several lines of code, it is best if we create new routines to handle each stage. This
means that the SetUpGame() function itself mainly consists of calls to other
functions, but it also initialises the gamestate variable:
FUNCTION SetUpGame()
CreateBoard()
CreateMarbles()
CreateSelector()
SetUpHelp()
gamestate = 1
ENDFUNCTION

Activity 38.4
Add the code for SetUpGame() to your program and add test stubs for the
routines that it calls.
Adding CreateBoard()
This routine is responsible for setting up the board. This means, not only does it
create a thin 3D box to represent the board, texturing it with the board image shown
in FIG-38.2, but it also fills the board and boardCoords arrays with relevant data.
While the visual aspects of creating the board require only a few lines, initialising
the arrays is much more complex, so again we'll make another routine to do that
part of the job (CreateInternalBoard()). The code for CreateBoard() is:
FUNCTION CreateBoard()
REM *** Set up board arrays ***
CreateInternalBoard()
REM *** Create and texture board ***
MAKE OBJECT BOX boardobj,27,27,0.5
LOAD IMAGE "board.bmp",boardimg
TEXTURE OBJECT boardobj,boardimg
REM *** Turn and position board ***
ROTATE OBJECT boardobj,90,0,0
POSITION OBJECT boardobj,-0.2,-1.2,0
ENDFUNCTION
Activity 38.5
Update your program with the code above.
Create a test stub for CreateInternalBoard().
Adding CreateInternalBoard()
Although this is a rather long routine, it doesn't do anything too complicated; just
fills the board and boardCoords arrays with the values shown earlier. The code for
the routine is:
FUNCTION CreateInternalBoard()
REM *** Initialise the board array ***
#CONSTANT EMPTYPIT 0
#CONSTANT NOTALLOWED -1
REM *** Assume every position empty ***
FOR row = 1 TO 7
FOR col = 1 TO 7
board(row,col) = EMPTYPIT
NEXT col
NEXT row
REM *** then mark the invalid positions ***
board(1,1) = NOTALLOWED

REM *** Initialise the boardContents array ***

REM *** with the pit coordinates ***
x = -9
z = 9
FOR row
= 1 TO 7
FOR
col = 1 TO 7
BoardCoords(row,col).x = x
BoardCoords(row,col).z = z
x = x + 3
NEXT col
z = z - 3
x = -9
NEXT row
ENDFUNCTION
Activity 38.6
Add the CreateInternalBoard() function to your program.
Although it won't make any visible difference to your program, execute the
code just to make sure there are no syntax errors or other problems.
Adding CreateMarbles()
Now we need to create the 3D marble objects. These will be placed just above the
pits on the board, except for the central pit which remains empty.
The logic for this routine can be described in structured English as:
Load marble texture

Set objno to firstmarbleobj
FOR each row on the board DO
FOR each column on the board DO
IF the pit is empty AND NOT the central pit THEN
Create sphere (2 units diameter)
Texture sphere
Position sphere at the coordinates given in boardCoords(row,col)
Store the sphere's ID value, objno, in board(row,col)
Increment objno
ENDIF
ENDFOR
ENDFOR
Activity 38.7
Use the logic given above to produce the code for the CreateMarbles()
function and test your updated program.
Adding CreateSelector()
A blue outline cube is used to select which marble is to be moved and its destination.
The CreateSelector() function creates the cube placing it initially in the central pit.
The logic for the routine is:

Load selector texture
Create selector cube object (1.5 units)
Texture cube
Make black areas of cube invisible
Switch off culling so back lines are drawn
Position the cube at central pit
Set selectorposition to row 4, column 4
Activity 38.8
Use the logic given to create the final version of the CreateSelector() routine
and add it to your program.
Check that the selector appears at the correct position on the board.
Adding SetUpHelp()
The player will be able to get help on the game rules and on which keys can be used
by pressing the F1 key (the traditional help key for Microsoft Windows programs).
The Help itself will be created as an animated sprite containing three frames; one
for each page of the Help (see FIG-38.6).
FIG-38.6 The Help Image
This routine will create the required sprite, but then hide the sprite. The sprite will
appear whenever the F1 key is pressed. If the user presses the F1 key a second time,
the Help sprite will disappear.
If you want to create your own Help image using a paint package, remember not to
use any black since it will automatically become transparent when the sprite is
displayed.
The code for this routine is given below:
FUNCTION SetUpHelp()
REM *** Create the sprite required by help ***
CREATE ANIMATED SPRITE helpsprite,"rules.bmp",3,1,helpimg
REM *** Position sprite ***

SPRITE helpsprite,10,10,helpimg
REM *** Set sprite to show first page ***
SET SPRITE FRAME helpsprite,1
REM *** Hide sprite ***
HIDE SPRITE helpsprite
ENDFUNCTION
Of course, we'll need another routine later to make the Help sprite appear when
required.
Activity 38.9
Add the SetUpHelp() function to your program. Check that the program still
compiles correctly.
Adding GetPlayerMove()
The GetPlayerMove() is the core routine of the game, so it will probably come as
no surprise that it is the most complex. However, it really only involves reading the
keyboard, recognising which key has been pressed and carrying out the required
action if it is valid. Invalid actions will produce error messages.
Unlike all the other functions, this one returns a value. If the user has chosen the
quit option, 1 is returned and zero for all other options.
The logic for this routine is:
Set quit to zero

Read a key
IF
arrow key pressed:
Move selector
Enter key pressed:
IF
gamestate = 1:
Select a marble
IF okay THEN
Set gamestate to 2
ENDIF
gamestate = 2:
Select a pit
IF okay THEN
Move marble
Set gamestate to 1
ENDIF
ENDIF
U pressed:
IF gamestate = 2 THEN
Undo marble selection
Set gamestate to 1
ENDIF
N pressed:
Restart game
F1 pressed:
IF gamestate < 10 THEN
Show first page of help
Add 10 to gamestate
ELSE
Hide help
Subtract 10 from gamestate
ENDIF
K pressed:
IF gamestate > 10 THEN
Show keys help page

ENDIF
R pressed:
IF gamestate > 10 THEN
Show rules help page
ENDIF
F pressed:
Set quit to 1
ENDIF
Wait for all keys to be released
Return value of quit
The code for this routine is:

REM *** Assume player will not quit this turn ***
quit = 0
REM *** Read a key ***
REPEAT
code = SCANCODE()
UNTIL code <> 0
SELECT code
CASE 200,208,205,203 Àrrow keys
MoveSelector(code)
ENDCASE
CASE 28 Ènter key
REM *** IF at select marble stage, select marble ***
IF gamestate = 1
REM *** IF marble selected, gamestate = 2, ***
IF SelectMarble()
gamestate = 2
ENDIF
ELSE
REM *** IF in select empty pit stage,select pit
IF gamestate = 2
REM *** IF pit okay... ***
IF SelectPit()
REM *** Move marble and return to state 1
MoveMarble()
gamestate = 1
ENDIF
ENDIF
ENDIF
ENDCASE
CASE 22 Ù - undo selection
REM *** IF marble selected THEN ***
IF gamestate = 2
REM *** Return marble to normal colour ***
objno = board(move(1).row,move(1).col)
TEXTURE OBJECT objno,marbleimg
REM *** return to gamestate 1 ***
gamestate = 1
ENDIF
ENDCASE
CASE 49 `N - new game
SetUpScreen()
SetUpGame()
ENDCASE
CASE 59 `F1 - Help
IF gamestate < 10
SelectHelpPage(1)
gamestate = gamestate + 10
ELSE
SelectHelpPage(0)
gamestate = gamestate - 10
ENDIF
ENDCASE
CASE 37 `K - Help page 2
IF gamestate > 10

SelectHelpPage(2)
ENDIF
ENDCASE
CASE 19 `R - Help page 3
IF gamestate > 10
SelectHelpPage(3)
ENDIF
ENDCASE
CASE 33 `F - Quit game
quit = 1
ENDCASE
ENDSELECT
REM *** Wait till all keys released ***
WHILE SCANCODE() <> 0
ENDWHILE
ENDFUNCTION quit
Activity 38.10
Add the GetPlayerMove() function to your program.
Add test stubs for any routines called by GetPlayerMove() that are yet to be
written.
Test the program.
Adding MoveSelector()
The MoveSelector() function takes a parameter. This is the arrow key code and is
used to decide in which direction the selector is to be moved.
The function's logic is:
Set row to copy of selectorposition.row

Set col to copy of selectorposition.col
IF
left arrow AND not at left hand column :
Decrement col
right arrow AND not at right-hand column:
Increment col
down arrow AND not at bottom row:
Decrement row
up arrow AND not at top row:
Increment row
ENDIF
IF board(row,col) is a valid position THEN
Set selectorposition.row to copy of row
Set selectorposition.col to copy of col
Move selector to new position
ENDIF
and this is coded as:
FUNCTION MoveSelector(code)
REM *** Move selector in response to arrow key ***
row = selectorposition.row
col = selectorposition.col
IF code = 203 AND col>1
DEC col
ELSE IF code = 205 AND col < 7
INC col
ELSE IF code = 200 AND row > 1
DEC row
ELSE IF code = 208 AND row < 7

INC row
ENDIF
ENDIF
ENDIF
ENDIF
REM *** IF a valid position, move selector ***
IF board(row,col) <> -1
selectorposition.row = row
selectorposition.col = col
POSITION OBJECT selectorobj,
BoardCoords(row,col).x,0,BoardCoords(row,col).z
ENDIF
ENDFUNCTION
Activity 38.11
Add the new routine to your program and check that the selector moves
correctly.
Adding SelectMarble()
In the first phase of a move (when gamestate = 1), the player has to select a marble.
This involves the player moving the selector over the required pit position and then
pressing Enter. At this point, control jumps to the SelectMarble() routine which
checks that the selected pit does, in fact, contain a marble (if not, an error message
is displayed) and changes the colour of the marble to indicate its selection. The
routine returns 1 when a marble has been selected and zero when no marble was
selected.
The logic for this routine is
IF selector not over marble THEN

Display an error message
Return 0 from function
ENDIF
Change colour of selected marble
Record position of marble in move(1)
and is coded as:
FUNCTION SelectMarble()
REM *** IF selector not over marble THEN ***
IF board(selectorposition.row,selectorposition.col) < 1
REM *** Display message; return 0 ***
DisplayMessage(100,100,"Choose a marble",2)
EXITFUNCTION 0
ENDIF
REM *** Change marble colour ***
COLOR OBJECT
board(selectorposition.row,selectorposition.col),
RGB(255,255,0)
REM *** record position of marble ***
move(1).row = selectorposition.row
move(1).col = selectorposition.col
ENDFUNCTION 1
Activity 38.12
Add SelectMarble() to your program and check that it functions correctly.

Adding SelectPit()
The second part of a move involves choosing an empty pit into which the selected
marble is to be moved.
As well as the obvious check that the selected pit must be empty, we also need to
check that it is exactly two pits from the selected marble in either a horizontal or
vertical direction. But even that's not all the checking required: moving the marble
to the selected pit must involve "jumping over" one other marble (the one that is to
be removed).
The logic for the routine is:
IF selected pit not empty THEN

Display "Choose empty pit"
ENDIF
Record selector position in move(2)
IF not a valid move THEN
Display "Invalid move"
ENDIF
Checking for a valid move in the final IF mentioned above will be quite complex,
so we'll use another function, isValidMove(), to carry out the task. This function
will return 1 when the move is valid and zero when it's invalid.
Activity 38.13
Using the logic given, add the SelectPit() function to your program.
Add a test stub for isValidMove() which returns the value 1.
Adding IsValidMove()
There are three checks needed in this function:
Ø That the selected marble is being moved 2 places horizontally

Ø That the selected marble is being moved 2 places vertically
Ø That the move involves "jumping" another marble
One of the first two tests must be true and the third test must also be true in order
for the move to be a valid one.
Activity 38.14
Write the complete version of the IsValidMove() function and test it within
your program.
Adding MoveMarble()
Once a valid move has been entered, the selected marble must be moved to its new
position. On the screen, this not only involves moving the selected 3D sphere and
changing it back to its original texture, but also involves removing the "jumped"

marble. The equivalent changes must be made to the board array to reflect the new
situation. Also, since a marble is removed from the board, the marblesremaining
variable must be decremented. The logic required by the routine is:
Move marble in board array

Move marble on screen
Return marble to original texture
Calculate position of jumped marble
Remove jumped marble from screen
Remove jumped marble from board array
Decrement marblesremaining
Activity 38.15
Code the MoveMarble() function and add it to your program.
Adding SelectHelpPage()
The final routine to be created is the SelectHelpPage() function. This function takes
a parameter (pageno) specifying the page number to be displayed. Since the Help
pages are stored as frames in an animated sprite, the pageno parameter represents
the frame to be displayed. If pageno is outside the range 0 to 3, then the function
exits; if the pageno is zero, the Help sprite is removed from the screen. The function
requires the following logic:
IF invalid value for pageno THEN

EXIT
ENDIF
IF pageno > 0 THEN
Set sprite frame to pageno
ELSE
Hide Help sprite
ENDIF
Activity 38.16
Code the SelectHelpPage() from the logic given and test your completed
program.
Using the Mouse

Introduction
An alternative way of selecting a marble and the pit it is to be moved to would be
to use the mouse. As we saw in the last chapter, the mouse can be used in conjunction
with the PICK OBJECT statement to select 3D objects on the screen. However,
although this would work fine for selecting the marble, picking the destination pit
for the marble would prove more difficult, since there is no object within an empty
pit that can be selected with a mouse click.
One way to overcome this problem is to place an invisible object at every valid pit
position, since the PICK OBJECT statement can be used to select invisible objects.
This means that pits which show a marble will actually contain two objects: a marble
and an invisible object; pits without a marble will contain only an invisible object.
To make a move the player will first click on a pit containing a marble and the mouse
position will be compared with the range of marble IDs to discover which one has

been selected. When the destination is being selected as the second part of a move,
the mouse position will be compared with the set of invisible objects.
Although this will give us the ID values of the marble and invisible objects, we still
have to translate this into row and column values so that the board array can be
updated and the move array can be filled. We can find the selected marble's row and
column position by searching the board array for a match with the ID returned by
the PICK OBJECT statement (see FIG-38.7).
FIG-38.7
-1 -1 11 12 13 -1 -1
Finding the ID of the -1 -1 14 15 16 -1 -1
Selected Marble
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31 32 33 34 35 36 37
-1 -1 38 39 40 -1 -1
-1 -1 41 42 43 -1 -1
PICK OBJECT will return the ID of The ID can be looked up in the
the selected object board array finding the row and column
If we set up an array containing the IDs of the invisible objects, we can use the same
technique to find the row/column position of the empty pit selected as the second
part of a move.
Updating the Program
Activity 38.17
Create a new project (Solitaire2.dbpro) and copy the source code for the first
version of the game into the new source file.
Copy all of the media files used into the new folder.
Since we are using a copy of the original program, it won't matter if things go wrong
- we can always go back.
We need to start by removing all references to the selector object since it won't be
needed in the new version of the game.
Activity 38.18
Remove all lines containing references to selectorimg, selectorobj, and

selectorposition from the program.
The next step is to add our invisible objects to each pit. We can't hide the objects
In the text we continue to
since PICK OBJECT doesn't detect hidden objects, but we can texture the objects
refer to the invisible with a black image and then use SET OBJECT TRANSPARENCY to make the
objects as being hidden. objects disappear.
To do this we'll create some new constants:

#CONSTANT hiddenimg 3
#CONSTANT firsthiddenobj 35
#CONSTANT lasthiddenobj 67

and an array in which to store the IDs of the invisible objects:
GLOBAL DIM PitObjects(7,7) ÌDs of the invisible objects
The routine to create the invisible objects is then coded as:
FUNCTION CreateHiddenPitObjects()
REM *** Set first hidden object ID ***
pitobjno = firsthiddenobj
REM *** Load image used to texture objects ***
LOAD IMAGE "black.bmp",hiddenimg
REM *** FOR each position on board ***
FOR row = 1 TO 7
FOR col = 1 TO 7
REM *** IF its a valid position ***
IF board(row,col) > -1
REM *** Record hidden object ID for that pit
PitObjects(row,col) = pitobjno
REM *** Make, texture and hide object ***
MAKE OBJECT SPHERE pitobjno,2
TEXTURE OBJECT pitobjno,hiddenimg
POSITION OBJECT pitobjno,boardCoords(row,col).x,
Ä
0,boardCoords(row,col).z
SET OBJECT TRANSPARENCY pitobjno,1
REM *** Increment object ID
INC pitobjno
ENDIF
NEXT col
NEXT row
ENDFUNCTION
This routine should be called from SetUpGame().
Activity 38.19
Add the constants and global variables mentioned above to your program.
Add the routine CreateHiddenPitObjects() and place a call to it in

SetUpGame().
Copy the file black.bmp to your folder.
Now we have to make some changes to the GetPlayersMove() function. The original
version of this function reacts to key presses, but in the update we need to allow for
both mouse and key presses to be detected. The routine starts by waiting for a key
press using the lines
REPEAT
code = SCANCODE()
UNTIL code <> 0
but in the new version, we need to wait for a key press or a mouse click and this is
achieved by the code:
REPEAT
mouse = MOUSECLICK()
code = SCANCODE()
UNTIL code <> 0 OR mouse <>0
The original SELECT statement that follows was based on the scan code of the key
pressed. Since we now have no use for a selector, detecting the arrow keys is no

longer required, so the lines
CASE 200,208,205,203 Àrrow keys

MoveSelector(code)
ENDCASE
can be removed from this new version of the program.
The next CASE statement within the SELECT structure is CASE 28 which was
used to detect the Enter key. But now the mouse button needs to activate this option.
To keep our changes to a minimum, we can set code to 28 when the mouse is clicked:
IF mouse <> 0
code = 28
ENDIF
These lines need to be placed just before the SELECT structure.
At the end of the SELECT structure, we waited for the button pressed by the player
to be released using the code
REM *** Wait till all keys released ***

WHILE SCANCODE() <> 0
ENDWHILE
but now we also need to wait for the mouse button to be released, so our code should
be changed to:
REM *** Wait till all keys and mouse buttons released ***
WHILE SCANCODE() <> 0 OR MOUSECLICK() <> 0
ENDWHILE
Activity 38.20
Make the changes described to the GetPlayersMove() routine.
The SelectMarble() routine needs to be changed. This version needs to detect the
mouse pointer position and select the marble over which it has been placed. The
logic for the updated routine is:
Find mouse coordinates

Get ID of marble at this screen position
IF no ID detected THEN
Display "Choose a marble"
Return 0 from the routine
ENDIF
Find position of ID in board array and record row and column
Change colour of marble selected
Record row and column in moves(1)
Return 1 from routine
This is coded as:
REM *** Check if position chosen contains marble ***
x = MOUSEX()
y = MOUSEY()
objno = PICK OBJECT(x,y,firstmarbleobj,lastmarbleobj)
IF objno = 0
DisplayMessage(100,100,"Choose a marble",2)
EXITFUNCTION 0
ENDIF

REM *** Find marble's ID in board array ***
FOR row = 1 TO 7
FOR col = 1 TO 7
IF board(row,col) = objno
prow = row
pcol = col
ENDIF
NEXT col
NEXT row
REM *** Change marble colour ***
COLOR OBJECT objno,RGB(255,255,0)
REM *** record position of marble
move(1).row = prow
move(1).col = pcol
ENDFUNCTION 1 `Valid marble selected
The search for the marble ID within the board array is very inefficient coding, but
for such a small search area will have no real effect on the speed of the game - and
it has the advantage of being quite simple.
Activity 38.21
Update your SelectMarble() function to match the code given.
The last routine to be changed is the SelectPit() function, which again needs to use
the mouse's position to determine which empty pit has been selected. The routine's
new logic is:
Find mouse coordinates

Get ID of hidden object at this screen position
Get ID of marble at this screen position
IF no hidden ID detected OR a marble ID was detected THEN
Display "Choose an empty space"
Return 0 from the routine
ENDIF
Find position of hidden object ID in pitObjects array and record the row and column
Record row and column in moves(2)
Check to see if move is valid
IF not valid THEN
Display "Invalid move"
ENDIF
Activity 38.22
Use the above logic to code the SelectPit() function and test your program.
Suggested Enhancements
What we've developed in this chapter is the core of a game, but it still needs all those
nice finishing touches that gives a game that professional look.
Some of these would be simple to implement: use a larger font for messages; change
the textures used; an introductory splash screen; a finish screen; displaying the
number of marbles remaining; displaying the time elapsed; allowing the camera to
be moved.
Others would be a bit more complicated: keeping a top 5 best scores (least marbles
remaining); allowing the player to backtrack - undoing an unlimited number of

moves (you would need a stack structure for this one).
If you have time, see what enhancements you can come up with - you'll learn a lot
from the exercise.

Solutions
REM ********************************** REM **********************************
REM *** Constants *** REM *** Constants ***
REM ********************************** REM **********************************
REM *** Object constants *** REM *** Object constants ***
#CONSTANT boardobj 1 #CONSTANT boardobj 1
#CONSTANT selectorobj 2 #CONSTANT selectorobj 2
#CONSTANT firstmarbleobj 3 #CONSTANT firstmarbleobj 3
#CONSTANT lastmarbleobj 34 #CONSTANT lastmarbleobj 34
REM *** Image constants *** REM *** Image constants ***
#CONSTANT boardimg 1 #CONSTANT boardimg 1
#CONSTANT marbleimg 2 #CONSTANT marbleimg 2
#CONSTANT selectorimg 3 #CONSTANT selectorimg 3
#CONSTANT helpimg 4 #CONSTANT helpimg 4
REM *** Sprite constants *** REM *** Sprite constants ***
#CONSTANT helpsprite 1 #CONSTANT helpsprite 1
REM *** Declare data structure *** REM *** Declare data structure ***
TYPE Coords TYPE Coords
x AS INTEGER x AS INTEGER
z AS INTEGER z AS INTEGER
ENDTYPE ENDTYPE
TYPE BoardPosition TYPE BoardPosition

row AS INTEGER row AS INTEGER
col AS INTEGER col AS INTEGER
ENDTYPE ENDTYPE
REM *** Declare main variables *** REM *** Declare main variables ***
GLOBAL DIM board(7,7)`Marble positions GLOBAL DIM board(7,7) `Marble positions
GLOBAL DIM boardCoords(7,7) AS Coords GLOBAL DIM boardCoords(7,7) AS Coords
`Marble coordinates `Marble coordinates
GLOBAL selectorposition AS BoardPosition GLOBAL selectorposition AS BoardPosition
`Position of selector `Position of selector
GLOBAL DIM move(2) AS BoardPosition GLOBAL DIM move(2) AS BoardPosition
`marble and move-to position `marble and move-to position
GLOBAL gamestate GLOBAL gamestate
`Game state `Game state
`1-choose marble; `1-choose marble;
`2-choose space; `2-choose space;
`3-showing help `3-showing help
GLOBAL marblesRemaining GLOBAL marblesRemaining
`marbles on board `marbles on board
REM *** Main section *** REM *** Main section ***
SetUpScreen() SetUpScreen()
SetUpGame() SetUpGame()
REPEAT REPEAT
quit = GetPlayersMove() quit = GetPlayersMove()
UNTIL quit OR marblesremaining = 1 UNTIL quit OR marblesremaining = 1
WAIT KEY `***** Remove later **** WAIT KEY `***** Remove later ****
END END
`**************************************** FUNCTION SetUpScreen()

`*************** Test stubs************** REM *** Set up screen ***
`**************************************** SET DISPLAY MODE 1280, 1024, 32
FUNCTION SetUpScreen() COLOR BACKDROP RGB(255,255,150)
PRINT "SetUpScreen() executed" BACKDROP ON
ENDFUNCTION REM *** Position camera ***
AUTOCAM OFF
FUNCTION SetUpGame() POSITION CAMERA 10,10,-20
PRINT "SetUpGame() executed" POINT CAMERA 0,0,0
ENDFUNCTION ENDFUNCTION
FUNCTION GetPlayersMove() `****************************************

PRINT "GetPlayerMove() executed" `*************** Test stubs**************
ENDFUNCTION 1 `****************************************
PRINT "SetUpGame() executed"
ENDFUNCTION

FUNCTION GetPlayersMove() WHILE TIMER() - t < millisecs
PRINT "GetPlayerMove() executed" SET CURSOR x,y
ENDFUNCTION 1 PRINT message$
ENDWHILE
ENDFUNCTION
Activity 38.3 `***************************************
`*************** Test stubs**************
REM ********************************** `****************************************
REM *** Constants *** FUNCTION SetUpGame()
REM ********************************** DisplayMessage
(100,100,"SetUpGame() executed",2)
REM *** Object constants *** ENDFUNCTION
#CONSTANT selectorobj 2 FUNCTION GetPlayersMove()
#CONSTANT firstmarbleobj 3 DisplayMessage
#CONSTANT lastmarbleobj 34 (100,100,"GetPlayersMove() executed",2)
ENDFUNCTION 1
The final WAIT KEY statement may also be removed at
#CONSTANT selectorimg 3 this stage.
#CONSTANT helpimg 4
REM *** Sprite constants *** Activity 38.4

REM **********************************
REM **********************************
TYPE Coords
x AS INTEGER
z AS INTEGER
ENDTYPE
TYPE BoardPosition
row AS INTEGER
col AS INTEGER
ENDTYPE
#CONSTANT helpimg 4
GLOBAL DIM boardCoords(7,7) AS Coords
`Marble coordinates
REM *** Sprite constants ***
GLOBAL selectorposition AS BoardPosition
`Position of selector
GLOBAL DIM move(2) AS BoardPosition
`marble and move-to position
TYPE Coords
GLOBAL gamestate
x AS INTEGER
`Game state
z AS INTEGER
`1-choose marble;
ENDTYPE
`2-choose space;
`3-showing help
TYPE BoardPosition
GLOBAL marblesRemaining
row AS INTEGER
`marbles on board
col AS INTEGER
ENDTYPE
SetUpScreen()
SetUpGame()
REPEAT
GLOBAL DIM boardCoords(7,7) AS Coords
`Marble coordinates
`Position of selector
END
GLOBAL DIM move(2) AS BoardPosition
`marble and move-to position
GLOBAL gamestate
`Game state
`1-choose marble;
`2-choose space;
BACKDROP ON
`3-showing help
AUTOCAM OFF
`marbles on board
POINT CAMERA 0,0,0
ENDFUNCTION
SetUpScreen()
SetUpGame()
FUNCTION DisplayMessage
REPEAT
(x,y,message$,seconds#)
t = TIMER()
INK 0,0
END

FUNCTION SetUpScreen() REM *** Sprite constants ***
REM *** Set up screen *** #CONSTANT helpsprite 1
COLOR BACKDROP RGB(255,255,150) REM *** Declare data structure ***
BACKDROP ON TYPE Coords
REM *** Position camera *** x AS INTEGER
AUTOCAM OFF z AS INTEGER
POSITION CAMERA 10,10,-20 ENDTYPE
POINT CAMERA 0,0,0
ENDFUNCTION TYPE BoardPosition
row AS INTEGER
FUNCTION SetUpGame() col AS INTEGER
CreateBoard() ENDTYPE
CreateMarbles()
CreateSelector() REM *** Declare main variables ***
SetUpHelp() GLOBAL DIM board(7,7) `Marble positions
gamestate = 1 GLOBAL DIM boardCoords(7,7) AS Coords
ENDFUNCTION `Marble coordinates
FUNCTION DisplayMessage `Position of selector
Ä(x,y,message$,seconds#) GLOBAL DIM move(2) AS BoardPosition
t = TIMER() `marble and move-to position
INK 0,0 GLOBAL gamestate
millisecs = seconds# * 1000 `Game state
WHILE TIMER() - t < millisecs `1-choose marble;
SET CURSOR x,y `2-choose space;
PRINT message$ `3-showing help
ENDWHILE GLOBAL marblesRemaining
ENDFUNCTION `marbles on board
`**************************************** REM *** Main section ***

`*************** Test stubs************** SetUpScreen()
`**************************************** SetUpGame()
REPEAT
FUNCTION GetPlayersMove() quit = GetPlayersMove()
DisplayMessage UNTIL quit OR marblesremaining = 1
Ä
(100,100,"GetPlayerMove() executed",2) REM *** End program ***
ENDFUNCTION 1 END
FUNCTION CreateBoard() FUNCTION SetUpScreen()

DisplayMessage(100,100, REM *** Set up screen ***
Ä
"CreateBoard() executed",2) SET DISPLAY MODE 1280, 1024, 32
ENDFUNCTION COLOR BACKDROP RGB(255,255,150)
BACKDROP ON
FUNCTION CreateMarbles() REM *** Position camera ***
DisplayMessage(100,100, AUTOCAM OFF
Ä
"CreateMarbles() executed",2) POSITION CAMERA 10,10,-20
POINT CAMERA 0,0,0
ENDFUNCTION
ENDFUNCTION
DisplayMessage(100,100, FUNCTION SetUpGame()
Ä
"CreateSelector() executed",2) CreateBoard()
CreateMarbles()
ENDFUNCTION
CreateSelector()
FUNCTION SetUpHelp() SetUpHelp()
DisplayMessage(100,100, gamestate = 1
Ä
"SetUpHelp() executed",2) ENDFUNCTION
ENDFUNCTION
Activity 38.5 REM *** Create and texture board ***
REM ********************************** MAKE OBJECT BOX boardobj,27,27,0.5
REM *** Constants *** LOAD IMAGE "board.bmp",boardimg
REM ********************************** TEXTURE OBJECT boardobj,boardimg
REM *** Turn and position board ***
REM *** Object constants *** ROTATE OBJECT boardobj,90,0,0
#CONSTANT boardobj 1 POSITION OBJECT boardobj,-0.2,-1.2,0
#CONSTANT selectorobj 2 ENDFUNCTION
Ä
(x,y,message$,seconds#)
REM *** Image constants *** t = TIMER()
#CONSTANT boardimg 1 INK 0,0
#CONSTANT marbleimg 2 millisecs = seconds# * 1000
#CONSTANT selectorimg 3 WHILE TIMER() - t < millisecs
#CONSTANT helpimg 4 SET CURSOR x,y

PRINT message$ GLOBAL marblesRemaining
ENDWHILE `marbles on board
ENDFUNCTION
`**************************************** SetUpScreen()
`*************** Test stubs************** SetUpGame()
`**************************************** REPEAT
FUNCTION GetPlayersMove() UNTIL quit OR marblesremaining = 1
DisplayMessage REM *** End program ***
Ä (100,100,"GetPlayerMove() executed",2) END
ENDFUNCTION 1
FUNCTION CreateMarbles() REM *** Set up screen ***
DisplayMessage(100,100, SET DISPLAY MODE 1280, 1024, 32
Ä "CreateMarbles() executed",2) COLOR BACKDROP RGB(255,255,150)
ENDFUNCTION BACKDROP ON
FUNCTION CreateSelector() AUTOCAM OFF
DisplayMessage(100,100, POSITION CAMERA 10,10,-20
Ä "CreateSelector() executed",2) POINT CAMERA 0,0,0
FUNCTION SetUpHelp() FUNCTION SetUpGame()

DisplayMessage(100,100, CreateBoard()
Ä "SetUpHelp() executed",2) CreateMarbles()
ENDFUNCTION CreateSelector()
SetUpHelp()
FUNCTION CreateInternalBoard() gamestate = 1
DisplayMessage(100,100, ENDFUNCTION
Ä "CreateInternalBoard() executed",2)
ENDFUNCTION
Activity 38.6 REM *** Create and texture board ***
MAKE OBJECT BOX boardobj,27,27,0.5
REM ********************************** LOAD IMAGE "board.bmp",boardimg
REM *** Constants *** TEXTURE OBJECT boardobj,boardimg
REM ********************************** REM *** Turn and position board ***
ROTATE OBJECT boardobj,90,0,0
REM *** Object constants *** POSITION OBJECT boardobj,-0.2,-1.2,0
#CONSTANT boardobj 1 ENDFUNCTION
#CONSTANT firstmarbleobj 3 FUNCTION CreateInternalBoard()
#CONSTANT lastmarbleobj 34 REM *** Initialise the board array ***
REM *** Image constants *** #CONSTANT NOTALLOWED -1
#CONSTANT boardimg 1 REM *** Assume every position empty ***
#CONSTANT marbleimg 2 FOR row = 1 TO 7
#CONSTANT selectorimg 3 FOR col = 1 TO 7
#CONSTANT helpimg 4 board(row,col) = EMPTYPIT
REM *** Sprite constants *** NEXT col
#CONSTANT helpsprite 1 NEXT row
REM *** then mark invalid positions ***
REM *** Declare data structure *** board(1,1) = NOTALLOWED
TYPE Coords board(1,2) = NOTALLOWED
x AS INTEGER board(2,1) = NOTALLOWED
z AS INTEGER board(2,2) = NOTALLOWED
ENDTYPE board(1,6) = NOTALLOWED
TYPE BoardPosition board(2,6) = NOTALLOWED
row AS INTEGER board(2,7) = NOTALLOWED
col AS INTEGER board(6,1) = NOTALLOWED
ENDTYPE board(6,2) = NOTALLOWED
REM *** Declare main variables *** board(7,2) = NOTALLOWED
GLOBAL DIM board(7,7) `Marble positions board(6,6) = NOTALLOWED
GLOBAL DIM boardCoords(7,7) AS Coords board(6,7) = NOTALLOWED
`Marble coordinates board(7,6) = NOTALLOWED
GLOBAL selectorposition AS BoardPosition board(7,7) = NOTALLOWED
`Position of selector REM *** Initialise boardContents ***
GLOBAL DIM move(2) AS BoardPosition REM *** with the pit coordinates ***
`marble and move-to position x = -9
GLOBAL gamestate z = 9
`Game state FOR row = 1 TO 7
`1-choose marble; FOR col = 1 TO 7
`2-choose space; BoardCoords(row,col).x = x
`3-showing help BoardCoords(row,col).z = z

x = x + 3
NEXT col
z = z - 3
x = -9
NEXT row
Activity 38.8
ENDFUNCTION
The code for CreateSelector() is
Ä(x,y,message$,seconds#)
REM *** Load selector texture ***
t = TIMER()
INK 0,0
REM *** Create selector cube ***
MAKE OBJECT CUBE selectorobj,1.5
WHILE TIMER() - t < millisecs
TEXTURE OBJECT selectorobj, selectorimg
SET CURSOR x,y
SET OBJECT TRANSPARENCY selectorobj,1
PRINT message$
SET OBJECT CULL selectorobj,0
ENDWHILE
ENDFUNCTION
POSITION OBJECT selectorobj,0,0,0
REM *** Record position of selector ***
`****************************************
selectorposition.row = 4
`*************** Test stubs**************
selectorposition.col = 4
`****************************************
ENDFUNCTION
DisplayMessage The routine should be placed immediately after
Ä
(100,100,"GetPlayerMove() executed",2) CreateMarbles().
ENDFUNCTION 1
FUNCTION CreateMarbles() Activity 38.9

DisplayMessage(100,100,
Ä
"CreateMarbles() executed",2) Place SetUpHelp() immediately after CreateSelector().
ENDFUNCTION
FUNCTION CreateSelector() Activity 38.10

Ä
"CreateSelector() executed",2) Place GetPlayersMove() after SetUpHelp().
ENDFUNCTION
The following test stubs should also be added:
Ä
"SetUpHelp() executed",2)
FUNCTION MoveSelector(move)
ENDFUNCTION Ä"MoveSelector() executed",2)
ENDFUNCTION
Activity 38.7 FUNCTION SelectMarble()

The code for CreateMarbles() is Ä"SelectMarble() executed",2)
ENDFUNCTION 1
FUNCTION CreateMarbles()
REM *** Load marble texture *** FUNCTION SelectPit()
LOAD IMAGE "lava1.bmp",marbleimg DisplayMessage(100,100,
REM *** Create marbles *** Ä
"SelectPit() executed",2)
objno = firstmarbleobj ENDFUNCTION 1
FOR row = 1 TO 7
FOR col = 1 TO 7 FUNCTION MoveMarble()
IF board(row,col) = 0 AND
Ä
((row <> 4) OR (col <> 4))
Ä
"MoveMarble() executed",2)
ENDFUNCTION
POSITION OBJECT
Ä
FUNCTION SelectHelpPage(page)
objno,BoardCoords(row,col).x,0,
Ä
boardCoords(row,col).z Ä
"SelectHelpPage() executed",2)
board(row,col) = objno ENDFUNCTION
INC objno
ENDIF
NEXT col Activity 38.11
NEXT row
ENDFUNCTION No solution required.
Add the code to your program immediately after the

CreateInternalBoard() function. Activity 38.12
If you've entered this separately (rather than modified the No solution required.
test stub), don't forget to remove the test stub version of
the routine.
Activity 38.13
The code for SelectPit() is:

FUNCTION SelectPit() Activity 38.16
REM *** IF selected pit not empty, exit
IF board(selectorposition.row, The completed program should be coded as:
selectorposition.col) <> 0
Ä "Choose an empty space",2)
REM **********************************
EXITFUNCTION 0
REM **********************************
ENDIF
IF NOT isValidMove()
DisplayMessage(100,100,"Invalid move",2)
EXITFUNCTION 0
ENDIF
ENDFUNCTION 1
The new test stub is: #CONSTANT marbleimg 2
FUNCTION isValidMove() #CONSTANT helpimg 4
Ä "isValidMove() executed",2) REM *** Sprite constants ***
ENDFUNCTION 1 #CONSTANT helpsprite 1

Activity 38.14 TYPE Coords
x AS INTEGER
The code for isValidMove() is: z AS INTEGER
ENDTYPE
FUNCTION IsValidMove()
REM *** Invalid if start & finish pits TYPE BoardPosition
REM *** not exactly 2 pits apart *** row AS INTEGER
REM *** or no marble in-between *** col AS INTEGER
test1 = (ABS(move(1).row-move(2).row)=2) ENDTYPE
ÄAND (move(1).col = move(2).col)
test2 = (ABS(move(1).col-move(2).col)=2)
ÄAND (move(1).row = move(2).row) GLOBAL DIM board(7,7)
`Marble positions
midrow = ABS(move(1).row+move(2).row)/2
midcol = ABS(move(1).col+move(2).col)/2 GLOBAL DIM boardCoords(7,7) AS Coords
test3 = board(midrow,midcol) > 1 `Marble coordinates
REM *** Return 1 if tests checks okay *** GLOBAL selectorposition AS BoardPosition
IF (test1 OR test2) AND test3 `Position of selector
result = 1 GLOBAL DIM move(2) AS BoardPosition
ELSE òtherwise return zero `marble and move-to positions
result = 0 GLOBAL gamestate `Game state
ENDIF `1-choose marble;
ENDFUNCTION result `2-choose space;
`3-showing help
`marbles on board
Activity 38.15
The code for MoveMarble() is:
SetUpScreen()
SetUpGame()
FUNCTION MoveMarble()
REPEAT
REM *** Move marble in board array data ***
board(moves(2).row,moves(2).col) =
board(moves(1).row,moves(1).col)
board(moves(1).row,moves(1).col) = 0
END
REM *** Move marble on screen ***
POSITION OBJECT
Äboard(moves(2).row,moves(2).col),
ÄboardCoords(moves(2).row,moves(2).col).x,
Ä0,boardCoords(moves(2).row,moves(2).col).z
TEXTURE OBJECT
Ä
BACKDROP ON
board(moves(2).row,moves(2).col),marbleimg REM *** Position camera ***
REM *** Calc position of jumped marble *** AUTOCAM OFF
jumpedrow = (moves(1).row+moves(2).row)/2 POSITION CAMERA 10,10,-20
jumpedcol = (moves(1).col+moves(2).col)/2 POINT CAMERA 0,0,0
REM *** Remove jumped marble from screen *** ENDFUNCTION
jumpedmarble = board(jumpedrow,jumpedcol)
HIDE OBJECT jumpedmarble FUNCTION SetUpGame()
DELETE OBJECT jumpedmarble CreateBoard()
REM *** Remove marble from board array *** CreateMarbles()
board(jumpedrow,jumpedcol)=0 CreateSelector()
REM *** Decrement marblesremaining *** SetUpHelp()
DEC marblesremaining gamestate = 1

FUNCTION CreateBoard() FUNCTION CreateSelector()
REM *** Set up board arrays *** REM *** Load selector texture ***
CreateInternalBoard() LOAD IMAGE "selector.bmp",selectorimg
REM *** Create and texture board *** REM *** Create selector cube ***
MAKE OBJECT BOX boardobj,27,27,0.5 MAKE OBJECT CUBE selectorobj,1.5
LOAD IMAGE "board.bmp",boardimg TEXTURE OBJECT selectorobj, selectorimg
TEXTURE OBJECT boardobj,boardimg SET OBJECT TRANSPARENCY selectorobj,1
REM *** Turn and position board *** SET OBJECT CULL selectorobj,0
ROTATE OBJECT boardobj,90,0,0 REM *** Position cube ***
POSITION OBJECT boardobj,-0.2,-1.2,0 POSITION OBJECT selectorobj,0,0,0
ENDFUNCTION REM *** Record position of selector ***
selectorposition.row = 4
FUNCTION CreateInternalBoard() selectorposition.col = 4
REM *** Initialise the board array *** ENDFUNCTION
#CONSTANT NOTALLOWED -1 FUNCTION SetUpHelp()
REM *** Assume every position empty *** REM *** Create sprite requ'd by help ***
FOR row = 1 TO 7 CREATE ANIMATED SPRITE helpsprite,
FOR col = 1 TO 7 Ä"rules.bmp",3,1,helpimg
board(row,col) = EMPTYPIT REM *** Position sprite ***
NEXT col SPRITE helpsprite,10,10,helpimg
NEXT row REM *** Set sprite to show first page ***
REM *** then mark invalid positions *** SET SPRITE FRAME helpsprite,1
board(1,1) = NOTALLOWED REM *** Hide sprite ***
board(1,2) = NOTALLOWED HIDE SPRITE helpsprite
board(2,1) = NOTALLOWED ENDFUNCTION
board(1,6) = NOTALLOWED FUNCTION GetPlayersMove()
board(1,7) = NOTALLOWED REM *** Read a key ***
board(2,6) = NOTALLOWED REPEAT
board(2,7) = NOTALLOWED code = SCANCODE()
board(6,1) = NOTALLOWED UNTIL code <> 0
board(6,2) = NOTALLOWED quit = 0
board(7,1) = NOTALLOWED SELECT code
board(7,2) = NOTALLOWED CASE 200,208,205,203 Àrrow keys
board(6,6) = NOTALLOWED MoveSelector(code)
board(6,7) = NOTALLOWED ENDCASE
board(7,6) = NOTALLOWED CASE 28 Ènter key
board(7,7) = NOTALLOWED REM *** IF state 1, get marble ***
REM *** Initialise boardContents *** IF gamestate = 1
REM *** with the pit coordinates *** REM *** Select marble ***
x = -9 REM *** IF okay, move to state 2
z = 9 IF SelectMarble()
FOR row = 1 TO 7 gamestate = 2
FOR col = 1 TO 7 ENDIF
BoardCoords(row,col).x = x REM *** If state 2, select pit
BoardCoords(row,col).z = z ELSE IF gamestate = 2
x = x + 3 REM *** IF okay THEN ***
NEXT col IF SelectPit()
z = z - 3 REM *** Move marble ***
x = -9 MoveMarble()
NEXT row REM *** Return to state 1
ENDFUNCTION gamestate = 1
ENDIF
FUNCTION CreateMarbles() ENDIF
REM *** Load marble texture *** ENDIF
LOAD IMAGE "lava1.bmp",marbleimg ENDCASE
REM *** Create marbles *** CASE 22 Ù - undo selection
objno = firstmarbleobj REM *** IF marble selected THEN ***
FOR row = 1 TO 7 IF gamestate = 2
FOR col = 1 TO 7 REM *** return original texture
IF board(row,col) = 0 AND objno =
((row <> 4) OR (col <> 4)) board(move(1).row,move(1).col)
MAKE OBJECT SPHERE objno,2,20,20 TEXTURE OBJECT objno,marbleimg
TEXTURE OBJECT objno,marbleimg REM *** Return to state 1 ***
POSITION OBJECT gamestate = 1
Ä objno,BoardCoords(row,col).x,0, ENDIF
Ä boardCoords(row,col).z ENDCASE
board(row,col) = objno CASE 49 `N - new game
INC objno SetUpScreen()
ENDIF SetUpGame()
NEXT col ENDCASE
NEXT row CASE 59 `F1 - Help
ENDFUNCTION REM *** IF help screen not showing
IF gamestate < 10
REM *** Show help page 1 ***
SelectHelpPage(1)
REM *** In Help state ***

gamestate = gamestate + 10 IF NOT isValidMove()
ELSE DisplayMessage(100,100,
SelectHelpPage(0) Ä
"Invalid move",2)
gamestate = gamestate - 10 EXITFUNCTION 0
ENDIF ENDIF
ENDCASE ENDFUNCTION 1
CASE 37 `K - Help page 2
IF gamestate > 10 FUNCTION IsValidMove()
SelectHelpPage(2) REM *** Invalid if start and finish pits
ENDIF REM *** not exactly 2 distant ***
ENDCASE REM *** or no marble between ***
CASE 19 `R - Help page 3 test1 = (ABS(move(1).row-move(2).row)=2)
IF gamestate > 10 Ä
AND (move(1).col = move(2).col)
SelectHelpPage(3) test2 = (ABS(move(1).col-move(2).col)=2)
ENDIF Ä
AND (move(1).row = move(2).row)
ENDCASE midrow = ABS(move(1).row+move(2).row)/2
CASE 33 `F - Quit game midcol = ABS(move(1).col+move(2).col)/2
quit = 1 test3 = board(midrow,midcol) > 1
ENDCASE REM *** Return 1 if tests okay ***
ENDSELECT IF (test1 OR test2) AND test3
REM *** Wait till all keys released *** result = 1
WHILE SCANCODE() <> 0 ELSE òtherwise return zero
ENDWHILE result = 0
ENDFUNCTION quit ENDIF
ENDFUNCTION result
FUNCTION MoveSelector(code)
REM *** Move selector using arrow key *** FUNCTION MoveMarble()
row = selectorposition.row REM *** Move marble in board array ***
col = selectorposition.col board(move(2).row,move(2).col) =
IF code = 203 AND col>1 board(move(1).row,move(1).col)
DEC col board(move(1).row,move(1).col) = 0
ELSE IF code = 205 AND col < 7 REM *** Move marble on screen ***
INC col POSITION OBJECT
ELSE IF code = 200 AND row>1 Ä
board(move(2).row,move(2).col),
DEC row Ä
boardCoords(move(2).row,move(2).col).x,
ELSE IF code = 208 AND row<7
INC row
Ä
0,boardCoords(move(2).row,move(2).col).z
TEXTURE OBJECT
ENDIF
ENDIF
Ä
board(move(2).row,move(2).col),marbleimg
REM *** Calc jumped marble position ***
ENDIF jumpedrow = (move(1).row+move(2).row)/2
ENDIF jumpedcol = (move(1).col+move(2).col)/2
REM *** IF a valid, move selector *** REM *** Remove marble from screen ***
IF board(row,col) <> -1 jumpedmarble = board(jumpedrow,jumpedcol)
selectorposition.row = row HIDE OBJECT jumpedmarble
selectorposition.col = col DELETE OBJECT jumpedmarble
POSITION OBJECT
Ä
selectorobj,BoardCoords(row,col).x,
REM *** Remove marble from board ***
Ä
0,BoardCoords(row,col).z
board(jumpedrow,jumpedcol)=0
REM *** Decrement marblesremaining ***
ENDIF DEC marblesremaining
FUNCTION SelectMarble() FUNCTION SelectHelpPage(pageno)

REM *** IF selector not on marble THEN REM *** IF invalid page selected, exit
IF board(selectorposition.row,
Ä
IF pageno < 0 OR pageno > 3
selectorposition.col) < 1 EXITFUNCTION
REM *** Display message; return 0 *** ENDIF
Ä
REM *** IF no zero page no. display page
"Choose a marble",2) IF pageno > 0
EXITFUNCTION 0 SET SPRITE FRAME helpsprite,pageno
ENDIF SHOW SPRITE helpsprite
REM *** Change marble colour *** ELSE
COLOR OBJECT board(selectorposition.row, REM *** ELSE, hide help ***
Ä
selectorposition.col),RGB(255,255,0) HIDE SPRITE helpsprite
REM *** record position of marble ENDIF
move(1).row = selectorposition.row ENDFUNCTION
ENDFUNCTION 1 FUNCTION DisplayMessage
Ä(x,y,message$,seconds#)
FUNCTION SelectPit() t = TIMER()
REM *** IF selected pit not empty, exit INK 0,0
IF board(selectorposition.row, millisecs = seconds# * 1000
Ä
selectorposition.col) <> 0 WHILE TIMER()-t < millisecs
DisplayMessage(100,100, SET CURSOR x,y
Ä
"Choose an empty space",2) PRINT message$
EXITFUNCTION 0 ENDWHILE
ENDIF ENDFUNCTION

Activity 38.17
No solution required. REM *** Declare data structure ***
TYPE Coords
x AS INTEGER
Activity 38.18 z AS INTEGER
ENDTYPE
The following lines should be removed from the main
section of the program:: TYPE BoardPosition
row AS INTEGER
#CONSTANT selectorobj 2 col AS INTEGER
ENDTYPE
GLOBAL selectorposition AS BoardPosition GLOBAL DIM board(7,7)
`Position of selector `Marble positions on board
GLOBAL DIM BoardCoords(7,7) AS Coords
In SetUpGame() remove the line `Marble world coordinates
GLOBAL DIM moves(2) AS BoardPosition
CreateSelector() `marble and move-to positions
GLOBAL gamestate
`Game state 1-choose marble; 2-choose
The function CreateSelector() should be deleted. space; 3-showing help
GLOBAL marblesremaining
In GetPlayersMove(), remove the line `Marbles left on board
GLOBAL DIM PitObjects(7,7)
MoveSelector(code) ÌD of the hidden objects
from the first CASE structure.

Activity 38.20/22
The routine MoveSelector() should be removed.
The complete program code is:
In SelectMarble() remove the lines: REM **********************************
IF board(selectorposition.row, REM **********************************
selectorposition.col) < 1 REM *** Object constants ***
COLOR OBJECT board(selectorposition.row,
Äselectorposition.col),RGB(255,255,0)
#CONSTANT firstmarbleobj
#CONSTANT lastmarbleobj
3
34
#CONSTANT firsthiddenobj 35
move(1).row = selectorposition.row #CONSTANT lasthiddenobj 67
move(1).col = selectorposition.col REM *** Image constants ***

In SelectPit() remove the lines: #CONSTANT marbleimg 2
#CONSTANT hiddenimg 3
IF board(selectorposition.row, #CONSTANT helpimg 4
Äselectorposition.col) <> 0 REM *** Sprite constants ***
move(2).col = selectorposition.col TYPE Coords
x AS INTEGER
z AS INTEGER
Activity 38.19 ENDTYPE
The constants and globals should now be coded as: TYPE BoardPosition
row AS INTEGER
REM *** Constants *** col AS INTEGER
ENDTYPE
REM *** Object Constants ***
#CONSTANT boardobj 1 REM *** Declare main variables ***
#CONSTANT firstmarbleobj 3 GLOBAL DIM board(7,7)
#CONSTANT lastmarbleobj 34 `Marble positions
#CONSTANT firsthiddenobj 35 GLOBAL DIM boardCoords(7,7) AS Coords
#CONSTANT lasthiddenobj 67 `Marble coordinates
REM *** Image Constants *** `Position of selector
#CONSTANT boardimg 1 GLOBAL DIM move(2) AS BoardPosition
#CONSTANT marbleimg 2 `marble and move-to positions
#CONSTANT hiddenimg 3 GLOBAL gamestate
#CONSTANT helpimage 4 GLOBAL gamestate `Game state
`1-choose marble;
REM *** Sprite Constants *** `2-choose space;
#CONSTANT helpsprite 1 `3-showing help

FOR row = 1 TO 7
GLOBAL marblesRemaining FOR col = 1 TO 7
`marbles on board BoardCoords(row,col).x = x
GLOBAL DIM pitObjects(7,7) BoardCoords(row,col).z = z
ÌDs of hidden objects x = x + 3
NEXT col
REM *** Main section *** z = z - 3
SetUpScreen() x = -9
SetUpGame() NEXT row
REPEAT ENDFUNCTION
UNTIL quit OR marblesremaining = 1 FUNCTION CreateMarbles()
REM *** End program *** REM *** Load marble texture ***
END LOAD IMAGE "lava1.bmp",marbleimg
REM *** Create marbles ***
FUNCTION SetUpScreen() objno = firstmarbleobj
REM *** Set up screen *** FOR row = 1 TO 7
SET DISPLAY MODE 1280,1024,32 FOR col = 1 TO 7
COLOR BACKDROP RGB(255,255,150) IF board(row,col) = 0
BACKDROP ON AND ((row <> 4) OR (col <> 4))
REM *** Position camera *** MAKE OBJECT SPHERE objno,2,20,20
AUTOCAM OFF TEXTURE OBJECT objno,marbleimg
POSITION CAMERA 10,10,-20 POSITION OBJECT objno,
POINT CAMERA 0,0,0 boardCoords(row,col).x,0,
ENDFUNCTION boardCoords(row,col).z
board(row,col) = objno
FUNCTION SetUpGame() INC objno
CreateBoard() ENDIF
CreateMarbles() NEXT col
CreateHiddenPitObjects() NEXT row
SetUpHelp() ENDFUNCTION
gamestate = 1
ENDFUNCTION FUNCTION CreateHiddenPitObjects()
REM *** Set first hidden object ID ***
FUNCTION CreateBoard() pitobjno = firsthiddenobj
REM *** Set up board arrays *** REM *** Load texture image ***
CreateInternalBoard() LOAD IMAGE "black.bmp",5
REM *** Create and texture board *** REM *** FOR each position on board ***
MAKE OBJECT BOX boardobj,27,27,0.5 FOR row = 1 TO 7
LOAD IMAGE "board.bmp",boardimg FOR col = 1 TO 7
TEXTURE OBJECT boardobj,boardimg REM *** IF its a valid position ***
REM *** Turn and position board *** IF board(row,col)>-1
ROTATE OBJECT boardobj,90,0,0 REM *** Record hidden object ID
POSITION OBJECT boardobj,-0.2,-1.2,0 Ä for that pit
ENDFUNCTION pitObjects(row,col) = pitobjno
REM *** Make, texture and hide
FUNCTION CreateInternalBoard()
REM *** Initialise the board array ***
Ä object ***
MAKE OBJECT SPHERE pitobjno,2
#CONSTANT EMPTYPIT 0 TEXTURE OBJECT pitobjno,5
#CONSTANT NOTALLOWED -1 POSITION OBJECT pitobjno,
REM *** Assume every position empty *** Ä boardCoords(row,col).x,0,
FOR row = 1 TO 7
FOR col = 1 TO 7
Ä boardCoords(row,col).z
SET OBJECT TRANSPARENCY
board(row,col) = EMPTYPIT
NEXT col
Ä pitobjno,1
REM *** Increment object ID
NEXT row
INC pitobjno
REM *** then mark invalid positions ***
ENDIF
NEXT col
NEXT row
ENDFUNCTION
REM *** Create sprite requ'd by help ***
CREATE ANIMATED SPRITE helpsprite,
Ä"rules.bmp",3,1,helpimg
REM *** Position sprite ***
SPRITE helpsprite,10,10,helpimg
REM *** Set sprite to show first page ***
SET SPRITE FRAME helpsprite,1
REM *** Hide sprite ***
ENDFUNCTION
REM *** Initialise boardContents ***
REM *** with the pit coordinates ***
x = -9
z = 9

FUNCTION GetPlayersMove() y = MOUSEY()
REM *** Read keyboard/mouse *** objno = PICK OBJECT
REPEAT Ä(x,y,firstmarbleobj,lastmarbleobj)
mouse = MOUSECLICK() REM *** IF no marble ***
code = SCANCODE() IF objno = 0
UNTIL code <> 0 OR mouse <>0 REM *** Show message; return 0
REM *** Assume not quitting *** DisplayMessage(100,100,
quit = 0 Ä
"Choose a marble",2)
REM *** Mouse click same as Enter *** EXITFUNCTION 0
IF mouse <> 0 ENDIF
code = 28 REM *** Find marble's ID in board ***
ENDIF FOR row = 1 TO 7
SELECT code FOR col = 1 TO 7
CASE 28 Ènter key IF board(row,col) = objno
REM *** IF start of turn THEN *** prow = row
IF gamestate = 1 pcol = col
REM *** Select marble and ENDIF
REM *** IF OK, state 2, *** NEXT col
IF SelectMarble() NEXT row
gamestate = 2 REM *** Change marble colour ***
ENDIF COLOR OBJECT objno,RGB(255,255,0)
REM *** IF second stage of turn *** REM *** record position of marble
ELSE IF gamestate = 2 move(1).row = prow
REM *** Select pit and move(1).col = pcol
REM *** IF OK, ENDFUNCTION 1 `Valid marble selected
IF SelectPit()
REM *** Move marble *** FUNCTION SelectPit()
MoveMarble() REM *** Check for hidden obj or marble
REM *** Start new turn *** x = MOUSEX()
gamestate = 1 y = MOUSEY()
ENDIF objno = PICK OBJECT
ENDIF Ä(x,y,firsthiddenobj,lasthiddenobj)
ENDIF marbleno = PICK OBJECT(
ENDCASE
CASE 22 Ù - undo selection
Äx,y,firstmarbleobj,lastmarbleobj)
REM *** IF no hidden obj OR marble THEN
IF gamestate = 2 IF objno = 0 OR marbleno <> 0
objno =
Ä
REM *** Display message, return 0 ***
board(move(1).row,move(1).col) DisplayMessage(100,100,
gamestate = 1
Ä
"Choose an empty space",2)
EXITFUNCTION 0
ENDIF ENDIF
ENDCASE REM *** Find pit's ID in pitObjects ***
CASE 49 `N - new game FOR row = 1 TO 7
SetUpScreen() FOR col = 1 TO 7
SetUpGame() IF pitObjects(row,col) = objno
ENDCASE prow = row
CASE 59 `F1 - Help pcol = col
IF gamestate < 10 ENDIF
SelectHelpPage(1) NEXT col
gamestate = gamestate + 10 NEXT row
ELSE REM *** Record move-to position ***
SelectHelpPage(0) move(2).row = prow
gamestate = gamestate - 10 move(2).col = pcol
ENDIF REM *** IF not a valid move THEN ***
ENDCASE IF NOT isValidMove()
CASE 37 `K - Help page 2 REM *** Display message;return 0 ***
IF gamestate > 10 DisplayMessage(100,100,
SelectHelpPage(2)
ENDIF
Ä
"Invalid move",2)
EXITFUNCTION 0
ENDCASE ENDIF
CASE 19 `R - Help page 3 ENDFUNCTION 1 `Valid move-to pit selected
IF gamestate > 10
SelectHelpPage(3) FUNCTION IsValidMove()
ENDIF REM *** Invalid if start and finish pits
ENDCASE REM *** not exactly 2 apart distant ***
CASE 33 `F - Quit game REM *** or no marble in-between ***
quit = 1 test1 = (ABS(move(1).row-move(2).row)=2)
ENDCASE
ENDSELECT
ÄAND (move(1).col = move(2).col)
test2 = (ABS(move(1).col-move(2).col)=2)
REM *** Wait keys and mouse released ***
WHILE SCANCODE() <> 0 OR
ÄAND (move(1).row = move(2).row)
midrow = ABS(move(1).row+move(2).row)/2
MOUSECLICK() <> 0
midcol = ABS(move(1).col+move(2).col)/2
ENDWHILE
test3 = board(midrow,midcol) > 1
ENDFUNCTION quit
REM *** Return 1 if tests check okay ***
IF (test1 OR test2) AND test3
result = 1
REM *** Check for marble ***
ELSE òtherwise return zero
x = MOUSEX()

result = 0
ENDIF
ENDFUNCTION result
FUNCTION MoveMarble()
REM *** Move marble in board ***
board(move(2).row,move(2).col) =
Ä
board(move(1).row,move(1).col)
board(move(1).row,move(1).col) = 0
REM *** Move marble on screen ***
POSITION OBJECT
Ä
board(move(2).row,move(2).col),
Ä
boardCoords(move(2).row,move(2).col).x,0,
Ä
boardCoords(move(2).row,move(2).col).z
TEXTURE OBJECT board(move(2).row,move(2).col),
Ä
marbleimg
REM *** Calculate position of jumped marble ***
jumpedrow = (move(1).row+move(2).row)/2
jumpedcol = (move(1).col+move(2).col)/2
REM *** Remove jumped marble from screen ***
jumpedmarble = board(jumpedrow,jumpedcol)
HIDE OBJECT jumpedmarble
DELETE OBJECT jumpedmarble
REM *** Remove jumped marble from board ***
board(jumpedrow,jumpedcol)=0
REM *** Decrement marblesremaining ***
DEC marblesremaining
ENDFUNCTION
FUNCTION SelectHelpPage(pageno)
IF pageno < 0 OR pageno > 3
EXITFUNCTION
ENDIF
IF pageno > 0
SET SPRITE FRAME helpsprite,pageno
SHOW SPRITE helpsprite
ELSE
ENDIF
ENDFUNCTION
FUNCTION DisplayMessage(x,y,message$,seconds#)
t = TIMER()
INK 0,0
WHILE TIMER()-t < millisecs
SET CURSOR x,y
PRINT message$
ENDWHILE
ENDFUNCTION

39
Ambient Reflection
Bump Mapping
Diffuse Reflection
Emissive Light Effect
Light Mapping
Shadows
Specular Reflection
DarkBASIC Pro: Advanced Lighting and Texturing 1027

Advanced Lighting and Texturing
Introduction
We've seen how basic texturing and lighting can create impressive results, but even
more spectacular effects can be created with the various additional statements
available in DarkBASIC Pro.
Surface Reflection
In the real world, different materials reflect light in different ways. For example, a
red woollen coat, a red shiny metal sphere, and a red plastic box each reflect light
differently and hence give us a visual clue as to the material from which the object
is constructed. By modifying how a 3D object reflects light we can change its look
and so give the impression of different materials.
In DarkBASIC Pro there are three types of reflection:
Ø ambient reflection - the reflection of ambient light.

Ø diffuse reflection - the scattered reflection of a directed light.
Rough surfaces (even at a microscopic scale)
scatter reflected light equally in all directions.
Ø specular reflection - the mirrored reflection of directed light.
Light is reflected at the same angle as it
arrives at the surface.
FIG-39.1 shows the conceptual ideas behind diffuse and specular reflection.
FIG-39.1
Parallel Parallel
rays ... ...are reflected in rays ... ...are reflected at the same
Diffuse and Specular all directions... angle as they arrive...
Reflection ? ?
...by the ...by the

rough surface smooth surface
Diffuse Reflection Specular Reflection
Diffuse reflection creates surfaces that have a matt finish; specular reflection creates
a glossy surface.
FIG-39.2 shows the visual effect of both diffuse and specular reflection.
FIG-39.2
Diffuse and Specular
Reflection
Diffuse Specular
1028 DarkBASIC Pro: Advanced Lighting and Texturing

A surface's reflective properties can be modified by several statements in
DarkBASIC Pro.
The SET OBJECT AMBIENT Statement

The simplest of these statements is the SET OBJECT AMBIENT statement which
affects how an object reflects ambient light. This statement has the format given in
FIG-39.3.
FIG-39.3
SET OBJECT AMBIENT objno , ambflag
The SET OBJECT
AMBIENT Statement
In the diagram:

whose ambient light reflectivity is to be changed.
ambflag is 0 or 1. If ambflag is set to zero, the object

reflects no ambient light; if set to 1, the object
reflects ambient light normally.
For example, we could switch off object 1's reflection of ambient light using the
statement:
SET OBJECT AMBIENT 1,0
The program in LISTING-39.1 creates a sphere, switches off the directional light,
and, after a key press, switches off the sphere's reflection of ambient light.

Manipulating Ambient
BACKDROP ON
Light Reflection
REM *** Programmed screen updates ***
SYNC ON

AUTOCAM OFF
POINT CAMERA 0,0,0


HIDE LIGHT 0
SYNC
REM *** Switch off ambient reflection ***

PRINT "Change sphere's reflection of ambient light"
PRINT "Press any key"
SYNC
WAIT KEY
SYNC
SYNC

WAIT KEY
END

Activity 39.1
Type in and test the program in LISTING-39.1 (advanced01.dbpro).
Modify the program so that the sphere's ambient light reflection is reactivated
after another key press.
The SET OBJECT DIFFUSE Statement

How an object produces a reflected diffuse light can be modified using the SET
OBJECT DIFFUSE statement. This statement allows the colour of the reflected
diffuse light to be specified. The statement has the format shown in FIG-39.4.
FIG-39.4
SET OBJECT DIFFUSE objno , colour
The SET OBJECT
DIFFUSE Statement
In the diagram:

whose diffuse setting is to be changed.
If the light falling on an colour is an integer value specifying the diffuse colour
object is not white, the to be reflected. Where the colour of an object
diffuse colour should be
a combination of the has already been specified, it makes sense to use
object and light colours. that same colour value here (this assumes the
object is being lit by a white light).
For example, we could create a red diffuse reflection from object 1 using the line:
SET OBJECT DIFFUSE 1, RGB(255,0,0)
Activity 39.2
Create a program showing a blue sphere. Switch off the ambient reflection of
the sphere and set its diffuse reflection using cyan.
The SET OBJECT SPECULAR Statement

To recreate the mirror-like reflection of a shiny object, we can use the SET OBJECT
SPECULAR statement. The colour of the light reflected from the specular area of
an object can also be specified in the statement. The SET OBJECT SPECULAR
FIG-39.5
The SET OBJECT SET OBJECT SPECULAR objno , colour
SPECULAR Statement
In the diagram:

whose specular reflection setting is to be changed.
colour is an integer value specifying the specular colour

to be used. The colour used should match that of
the light shining on the object.

For example, we could create a white specular reflection on object 1 using the line:
SET OBJECT SPECULAR 1, RGB(255,255,255)
Activity 39.3
Add a yellow spot light - position (-1,3,-5) - to your previous program and
create a key-activated yellow specular area on the sphere.
The SET OBJECT SPECULAR POWER Statement

If the SET OBJECT SPECULAR statement has been applied to an object, the
amount of specular reflection (shininess) produced can be adjusted using the SET
OBJECT SPECULAR POWER statement. This statement has the format shown in
FIG-39.6.
FIG-39.6
The SET OBJECT
SPECULAR POWER SET OBJECT SPECULAR POWER objno , value
Statement
In the diagram:

whose specular reflection power is to be changed.
value is an integer value ranging from 0 to 100.

A setting of zero creates specular reflection over
a large area of the object; a setting of 100 gives
a small reflection area.
The result of using value settings of 100 and 10 are shown in FIG-39.7.
FIG-39.7
Varying the Specular
Power Setting
Low Specular High Specular

(value = 100) (value = 10)
The following code can be used to show the full range of specular reflection:
FOR power = 100 TO 0 STEP -1
SET OBJECT SPECULAR POWER 1,power
WAIT 25
NEXT power
Activity 39.4
Add the code given above to your last program and observe the effect.
The SET OBJECT EMISSIVE Statement

Some objects emit their own light (a bulb, the sun, etc.). Where we want an object
to simulate this situation, we can use the SET OBJECT EMISSIVE statement. The

effect of this statement is to colour the object independently from the colour of any
light hitting the object. The format for this statement is given in FIG-39.8.
FIG-39.8
The SET OBJECT SET OBJECT EMISSIVE objno , colour
EMISSIVE Statement
In the diagram:

whose emissive light setting is to be changed.
colour is an integer value specifying the emissive light

colour to be used.
For example, we could create a blue emissive light from object 1 using the line:
SET OBJECT EMISSIVE 1,RGB(0,0,255)
FIG-39.9 shows the difference between diffuse and emissive lighting.

FIG-39.9
Emissive Lighting Effect
Diffuse Emissive
The program in LISTING-39.2 demonstrates this effect by creating a sphere and

switching off both the default ambient and directional lights. With no light to reflect,
the sphere appears black. After a key press, blue emissive light is assigned to the
sphere which then becomes uniformly blue.

Creating Emissive Light COLOR BACKDROP RGB(155,155,0)
BACKDROP ON
REM *** position camera ***

AUTOCAM OFF
REM *** Create the sphere ***

REM *** Switch off lights ***

WAIT KEY
SET AMBIENT LIGHT 0
HIDE LIGHT 0
REM *** Produce fake emissive light ***

WAIT KEY
SET OBJECT EMISSIVE 1, RGB(0,0,255)
WAIT KEY
END

Activity 39.5
It should be noted that the SET OBJECT EMISSIVE statement does not create true
light since it will not illuminate other, nearby objects.
Activity 39.6
In your last program, create a 10 unit cube centred on (15,15,20).
Is the cube affected by the blue sphere?
Of course, all we have to do to make the sphere appear to emit light is to place a
point light at the centre of the sphere. This light should be the same colour as the
sphere's emissive light.
Activity 39.7
In your last program, add the following lines immediately before the final
WAIT KEY statement:
WAIT KEY
MAKE LIGHT 1
SET POINT LIGHT 1,0,0,0
COLOR LIGHT 1, RGB(0,0,255)
Test the program, observing how the new light affects the cube.
When another light shines on an object with emissive properties, the surface of that
object is shown in a colour created by a combination of the emissive colour and the
external light's colour. So a red directed light shining on a blue emissive sphere will
create magenta highlights on the sphere. This is demonstrated in LISTING-39.3.

Combining Emissive and BACKDROP ON
External Lighting REM *** Position camera ***
AUTOCAM OFF
REM *** Switch off ambient light ***

SET AMBIENT LIGHT 0
REM *** Make directional light red ***


REM *** Create blue emissive effect ***

WAIT KEY
SET OBJECT EMISSIVE 1, RGB(0,0,255)

WAIT KEY
END

Activity 39.8
Type in and test the program given in LISTING-39.3 (advanced03.dbpro).
The SET OBJECT LIGHT Statement

How an object reacts to lights other than the ambient light can be modified using
the SET OBJECT LIGHT statement which has the format given in FIG-39.10.
FIG-39.10
The SET OBJECT SET OBJECT LIGHT objno , lightflag
LIGHT Statement
In the diagram:

whose light reflectivity is to be changed.
lightflag is 0 or 1. If lightflag is set to zero, the object

absorbs no light and all surfaces reflect 100% of
the light ; if set to 1, the object reflects light
normally.
For example, maximum reflection from object 1 would be achieved using the line:
SET OBJECT LIGHT 1,0
The effects of the two possible settings are shown in FIG-39.11.
FIG-39.11
The Effects of using SET
OBJECT LIGHT
SET OBJECT LIGHT 1,1 SET OBJECT LIGHT 1, 0
Activity 39.9
Reload program effects02.dbpro.
Use the SET OBJECT LIGHT statement to switch off normal light reflection
on the cube.
How does this affect the cube?
Change the colour of the ambient light to red. Does this affect the colour
reflected by the cube?
Bear in mind that the final reflection created by an object is determined by many
factors including the light sources, the object's reflective characteristics, and the
texture of the object.

Mappings
Images can be applied to an object, not to create a basic texture on that object, but
to further affect the final appearance of that object.
The SET LIGHT MAPPING ON Statement

Not all parts of an object's surface reflect equal amounts of light, and although
DarkBASIC Pro will take care of this for us automatically, in many cases there are
occasions when we need to give it a helping hand. We can modify how parts of an
object reflect light using a light map. A light map is a grey-scale image which is
applied to an object in much the same way as a texture image. Once wrapped onto
the object, the light map affects how light is reflected from that object. Areas of the
surface which correspond to white parts in the light map image reflect light as
normal, while black areas reflect little or no light. The screen dumps in FIG-39.12
show the results of a street at night with and without light mapping.
FIG-39.12
Using Light Mapping
No light mapping applied
Light mapping applied

The default directional light is switched off and ambient light remains on in both
scenes. In the second scene, light maps have been applied to the wall, street and
lamppost.
The function, CreateScene(), which creates the first version of the scene as shown
in FIG-39.12 is given below:
FUNCTION CreateScene()
REM *** Create cobbled street ***
XROTATE OBJECT 1,90
LOAD IMAGE "cobbleslarge.jpg",1
TEXTURE OBJECT 1,1
REM *** Create wall ***
LOAD IMAGE "brickslarge.jpg",2
TEXTURE OBJECT 2,2
REM *** Create lamppost ***
LOAD IMAGE "lamppost.jpg",3
TEXTURE OBJECT 3,3
REM *** Make sphere appear to glow ***
REM *** Put light in sphere ***
MAKE LIGHT 1
ENDFUNCTION
Activity 39.10
Write a program to execute the function given above (lightmap.dbpro).
The main section should include the following logic:
Set the backdrop colour to RGB(20,20,20) - a dull grey.

Position the camera at (20,-4,-5) pointing at (0,0,3)
Test the program and observe the results.
The light map images used to create the second version of the scene are shown in
FIG-39.13.
FIG-39.13
Examples of Light Maps
Light map for wall and street Light map for lamppost
(lightmap.jpg) (lampmap.jpg)

We can apply light mapping to a specific object using the SET LIGHT MAPPING
ON statement which has the format shown in FIG-39.14.
FIG-39.14
The SET LIGHT SET LIGHT MAPPING ON objno , imgno
In the diagram:

object to which the light map is to be applied.
imgno is an integer value giving the ID of the image

object containing the light map picture.
The program in LISTING-39.4 uses the scene produced by CreateScene() and the
light maps shown in FIG-39.13 to create the effect shown in the second image of
FIG-39.12.
Using a Light Map COLOR BACKDROP RGB(20,20,20)
BACKDROP ON

AUTOCAM OFF
POSITION CAMERA 20,-4,-5
POINT CAMERA 0,0,0
REM *** Make scene ***

CreateScene()
REM *** Load and apply light map to street and wall ***
WAIT KEY
LOAD IMAGE "lightmap.jpg",4
SET LIGHT MAPPING ON 1,4
REM *** Load light map for lamppost ***

WAIT KEY
LOAD IMAGE "lampmap.jpg",5

WAIT KEY
END
REM ************ Code for CreateScene goes here ***
Activity 39.11
Type in and test the program given above (advanced04.dbpro). Remember to

include the actual code for function CreateScene().
Load and examine the file lightgrey.jpg.
Modify your program to use this file as the light map for the wall and ground.
How does this affect the result?
Unfortunately, if a 3D object is textured using tiling, the light map will also be tiled.

The result of a tiled texture on the wall as shown in FIG-39.15 demonstrates the
problem.
FIG-39.15
Problems with Tiling and
Light Maps
It's unlikely that we'll want this effect, so we must make sure that any object to
which we are going to apply a light map is textured by a single, untiled, image.
The SET BUMP MAPPING ON Statement

Every 3D object we create is perfectly smooth, whereas many real-life objects are
not. Although we can give some impression of roughness by applying a texture to
an object, that doesn't actually affect how the object reflects light. A better way to
simulate some roughness on a surface is to use a bump map. Again, this is a
greyscale image which is applied to an object and is used to give the impression of
the lumps and bumps that real surfaces contain. Normally, the bump map will be a
greyscale version of the original image used to texture an object. For example, the
texture and corresponding bump map for the cobbled street in our last program is
shown in FIG-39.16.
FIG-39.16
Bump Maps are Often
Based on the Texture
Image
Texture image Bump map image

To add a bump map to a 3D object, we use the SET BUMP MAPPING ON statement
FIG-39.17
The SET BUMP SET BUMP MAPPING ON objno , imgno

In the diagram:

object to which the bump map is to be applied.

object containing the bump map picture.
For example, we could use the file named cobblesbump.jpg as a bump map for the
cobbled street in our previous program with the lines:
LOAD IMAGE "cobblesbump.jpg",6

SET BUMP MAPPING ON 1,6
Activity 39.12
Add the lines given above, preceded by a WAIT KEY statement, to your last
program.
How does this affect the appearance of the street?
The SET SPHERE MAPPING ON Statement

A shiny spherical object can mirror the items that are positioned nearby. A quick
way to achieve this affect without the processing power that would be needed if it
was done for real is to use the SET SPHERE MAPPING ON statement. This allows
a fisheye-style image (see FIG-39.18) to be mapped onto a spherical object.
FIG-39.18
A Fisheye Image
The Fisheye Image The Fisheye Image applied to a Sphere

FIG-39.19
The SET SPHERE SET SPHERE MAPPING ON objno , imgno
In the diagram:

object to which the sphere map is to be applied.

object containing the sphere map picture.
Since a sphere map is meant to represent the reflection of objects on the surface of
a sphere, it only makes sense to apply this mapping to a spherical object. At first
glance, we may be tempted to think that the overall effect of this statement is really

no different from that of TEXTURE OBJECT, but the reflection shown on the
surface of a sphere will not change when the sphere is rotated. A texture, on the
other hand, being part of the object, will turn with the sphere. This is demonstrated
in LISTING-39.5.

Reflection in a Sphere AUTOCAM OFF
POINT CAMERA 0,0,0

WAIT KEY
REM *** Load and apply sphere map ***

LOAD IMAGE "fish.jpg",1
SET SPHERE MAPPING ON 1,1
WAIT KEY
REM *** Rotate sphere ***

DO
WAIT 10
LOOP
END
Activity 39.13
Load the image stripes.jpg and use it to texture the sphere. The sphere should
now have a texture and sphere map image.
How does this affect the overall result when the sphere is rotated?
However, although the image displayed by sphere mapping does not change when
the object rotates, if the object begins to change position, then to maintain the
charade of being a reflection, the sphere mapped image needs to change its
perspective. This effect is demonstrated by the program in LISTING-39.6.
The Reflection Changes
AUTOCAM OFF
as the Object Moves

LOAD IMAGE "fish.jpg",1
SET SPHERE MAPPING ON 1,1
REM *** Rotate sphere and move ***

POINT OBJECT 1, 0,1,0
DO
MOVE OBJECT 1, 1
WAIT 20
LOOP
END

Activity 39.14
The SET BLEND MAPPING ON Statement

Blend mapping allows the original texture image and a second image to be merged
in various ways to create a texture based on both images. An example of this is
shown in FIG-39.20 where the original texture, the blend image and the final result
are displayed.
FIG-39.20
One Result Obtained by
Blending Two Images
Texture Image Blend Image Result (Mode 5)
We can achieve this effect in DarkBASIC Pro by using the SET BLEND MAPPING
ON statement which takes the format shown in FIG-39.21.
FIG-39.21
The SET BLEND SET BLEND MAPPING ON objno , imgno , mode
MAPPING ON
Statement
In the diagram:

object to which the blend map is to be applied.

object containing the blend map picture.
mode is an integer value giving the blend option to

be used. The actual range for this value is not
given but probably lies in the range 1 to 11.
The program in LISTING-39.7 blends the two images shown above, switching
mode (FROM 1 TO 11) every 5 seconds.

Blend Mapping AUTOCAM OFF
POINT CAMERA 0,0,0

REM *** Load texture and blend images ***

LOAD IMAGE "vstripes.jpg",1
LOAD IMAGE "hstripes.jpg",2

LISTING-39.7 REM *** Texture sphere ***
(continued) TEXTURE OBJECT 1,1
REM *** Try each mode ***
Blend Mapping FOR mode = 1 TO 11
REM *** Change blend mode ***
SET BLEND MAPPING ON 1,2,mode
REM *** Rotate sphere for 5 seconds ***
time = TIMER()
REM *** Rotate for 5 seconds ***
WHILE timer() - time < 5000
REM *** Display current mode ***
SET CURSOR 100,100
PRINT "Mode ",mode
ENDWHILE
NEXT mode
END
Activity 39.15
Type in and test the program in LISTING-39.7 (advanced07.jpg).
The SET CUBE MAPPING ON Statement

Imagine a cube constructed from mirrors floating in the exact centre of a cubic room.
Each wall of the room has a different image painted on it. The floating cube begins
to rotate and, as it does so, the reflection from its mirrored sides change. This is the
setup we create when using the SET CUBE MAPPING ON statement. Six images,
invisible to us, are placed around the specified 3D object and reflected off that object
(see FIG-39.22).
The 6 images are

FIG-39.22 placed around the axes,
parallel to the 3 main planes +y
(X-Y, X-Z, and Y-Z)
How Cube Mapping
Works
-x
+z
-z
The images reflect

off the object to which
they have been linked +x
-y
The format for the SET CUBE MAPPING ON statement is given in FIG-39.23.
FIG-39.23
The SET CUBE
MAPPING ON
SET CUBE MAPPING ON objno , imgno1 ,
Statement
imgno2 , imgno3 , imgno4 , imgno5 , imgno6

In the diagram:
objno is an integer value giving the ID of the 3D object

to which cube mapping is to be applied.
imgno1,imgno2 are a set of integer values specifying the images

to be used on the positive and negative sides of
the x-axis respectively.
imgno3,imgno4 are a set of integer values specifying the images

the y-axis respectively.
imgno5, imgno6 are a set of integer values specifying the images

the z-axis respectively.
NOTE: The images used must have dimensions which are some power of 2 (for
example, 64, 128, 256, etc.) otherwise the statement will not work.
Often the images used will be sky. If a sky sphere is being used, the images used
here can be copied from areas of that sky sphere's own image. In FIG-39.24 we see
how Adobe ImageReady has been used to slice a sky image into five 512 by 512
pixel images.
FIG-39.24
The Original Sky Image
The sixth image, which represents the base of our 3D space can be the cobbles we
used earlier, but modified to a 512 by 512 size. All the images to be used are
identified in FIG-39.25.
imgno3
FIG-39.25
The Sections Used in the
Cube Mapping
imgno5
imgno2 imgno1 imgno6
imgno4

The program in LISTING-39.8 demonstrates the use of cube mapping, rotating a
cube which shows the reflection of the sky images around it.
LISTING 39.8 REM *** Set up screen ***

Using Cube Mapping AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Add extra light ***

MAKE LIGHT 1
POSITION LIGHT 1,20,-20,50

REM *** Load cube mapping images ***

LOAD IMAGE "sky01.jpg",1
LOAD IMAGE "cobbles04.jpg",4
REM *** Use cube mapping ***

SET CUBE MAPPING ON 22,1,2,3,4,5,6
REM *** Rotate cube using random movements ***

DO
TURN OBJECT LEFT 22,RND(200)/100.0
PITCH OBJECT UP 22,RND(200)/100.0
ROLL OBJECT RIGHT 22,RND(200)/100.0 - 1
WAIT 10
LOOP

END
Activity 39.16
Type in and test the program given in LISTING-39.8 (advanced08.dbpro).
Replace the rotating cube with a cone and see how this affects the result.
The SET ALPHA MAPPING ON Statement

We've already come across one method for making an object appear transparent
(SET GHOST ON), but true transparency can be achieved using the SET ALPHA
MAPPING ON statement which also allows us to set the degree of transparency
displayed by the object. This statement has the format shown in FIG-39.26.
FIG-39.26
The SET ALPHA SET ALPHA MAPPING ON objno , percent
In the diagram:

to which alpha mapping is to be applied.
percent is a real value giving the percentage of

opacity required. 0 - invisible; 100 - opaque.

For example, object 1 would be made semi-transparent using the line:
SET ALPHA MAPPING ON 1, 50
The program in LISTING-39.9 places a rotating cube in front of a textured plane;

the cube changes from opaque to invisible as its alpha setting moves from 100 to
0.
Using Alpha Mapping AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Create background object ***

TEXTURE OBJECT 1,1

LOAD IMAGE "grid8by8b.bmp",2
TEXTURE OBJECT 2,2
REM *** Initialise transparency setting ***

trans = 100
REM *** Rotate object ***

REPEAT
SET ALPHA MAPPING ON 2,trans
REM *** 1 in 11 chance of decreasing trans setting ***
IF RND(10) = 10
DEC trans
ENDIF
UNTIL trans = 0

WAIT KEY
END
Activity 39.17
The transparency colour
defaults to black but can Type in and test the program in LISTING-39.9 (advanced09.dbpro).
be changed using the SET
IMAGE COLORKEY Notice that any part of the textured object which is the transparency colour is
statement. invisible irrespective of the opacity setting.
Shadows
The SET SHADOW SHADING ON Statement
A 3D object can be made to cast a shadow using the SET SHADOW SHADING
ON statement. This statement has the format shown in FIG-39.27.
FIG-39.27 The SET SHADOW SHADING ON Statement
SET SHADOW SHADING ON objno

[ meshno , distance , hardflag

In the diagram:

which is to cast a shadow.
meshno is an integer value specifying any mesh used as the

shadow. When no mesh is used, this parameter
should be set to -1.
distance is an integer value specifying the maximum

distance over which the shadow should be
calculated. Objects beyond this distance will not
show shadows cast by the object. The distance
setting is ignored when the graphics hardware is
used to perform shadow calculations (see next
parameter).
hardflag 0 or 1. When set to zero, the shadow calculations

are performed by the graphics card hardware; a
setting of 1 uses the computer's own processor to
perform the calculations.
When the last three parameters are omitted, the graphics card hardware is used to
calculate shadows. Not all video cards will be able to perform this operation.
Creating shadows requires a great deal of processing power, and the frame rate will
drop significantly when this option is used.
An example of the shadow effect, a cube casting a shadow on a plane, is shown in

FIG-39.28.
FIG-39.28
The Shadow Effect
The program in LISTING-39.10 demonstrates shadowing using the simplest

version of the SET SHADOW SHADING ON statement.

Creating Shadows AUTOCAM OFF
POINT CAMERA 0,0,0

LOAD IMAGE "cobblestones.jpg",1
TEXTURE OBJECT 1,1

TEXTURE OBJECT 2,2
REM *** Switch shading on for cube ***

SET SHADOW SHADING ON 2
REM *** Rotate cube and display frame rate ***

DO
SET CURSOR 100,100
PRINT "Frame rate : ",SCREEN FPS()
LOOP

END
Activity 39.18
Modify the SET SHADOW SHADING ON statement to include the

additional parameters. Set meshno to -1, distance to 100, and hardflag to 1.
What effect does this change have on the program?
Now give the distance parameter a value of 10. How does this affect the
shadow?
Normally, the shadow cast will be the shape of the object, but it is possible to cast
a shadow of a different shape by creating a mesh and using that mesh's ID as the
meshno parameter in the SET SHADOW SHADING ON statement (see
FIG-39.29).
FIG-39.29
The Cube Casts a Round
Shadow!

Activity 39.19
In advanced10.dbpro, add the following lines immediately before the SET

SHADOW SHADER ON statement :
REM *** Create mesh ***

DELETE OBJECT 3
Change the SET SHADOW SHADER ON statement to read:
SET SHADOW SHADING ON 2,1,100,1
The SET SHADOW SHADING OFF Statement

We can turn off shadow creation for a specific object using the SET SHADOW
SHADING OFF statement which has the format shown in FIG-39.30.
FIG-39.30
The SET SHADOW SET SHADOW SHADING OFF objno
SHADING OFF
Statement
In the diagram:

whose shadow effect is to be switched off.
Activity 39.20
In your last program, switch off the cube's shadow effect when a key is
pressed.
The SET GLOBAL SHADOWS Statement

Rather than switch off the shadow of a single object, we can switch off all shadows
using the SET GLOBAL SHADOWS statement. The same statement can be used
to switch those shadows back on. The SET GLOBAL SHADOWS statement has
FIG-39.31
The SET GLOBAL
SHADOWS Statement SET GLOBAL SHADOWS
{ ON
OFF
Use SET GLOBAL SHADOW OFF to switch off shadows; use SET GLOBAL
SHADOWS ON to re-instate all shadows.
The program in LISTING-39.11 is a variation on LISTING-39.8 but allows the

cube's shadow to be switched on by pressing the "s" key and off by pressing "o".

Switching Shadows On AUTOCAM OFF
and Off POSITION CAMERA 0,10,-60
POINT CAMERA 0,0,0

TEXTURE OBJECT 1,1

TEXTURE OBJECT 2,2
REM *** Create and texture sphere ***

TEXTURE OBJECT 3,2
REM *** Point the main light ***

REM *** Switch shading on for cube and sphere ***

SET SHADOW SHADING ON 2,-1,100,1
REM *** Prepare objects for movement ***

POINT OBJECT 3,20,10,-30
movement# = 0.1
DO
REM *** Shadow on/off options ***
IF INKEY$() = "s"
SET GLOBAL SHADOWS ON
ENDIF
IF INKEY$() = "o"
SET GLOBAL SHADOWS OFF
ENDIF
REM *** Rotate cube and move sphere ***
MOVE OBJECT 2, 0.1
MOVE OBJECT 3, movement#
movement# = -0.1
ELSE
IF OBJECT POSITION Z(3) > 30
movement# = 0.1
ENDIF
ENDIF
REM *** Display frame rate ***
SET CURSOR 100,100
LOOP

END
Activity 39.21
Type in and test the program in LISTING-39.11 (advanced11.dbpro). How

does the frame rate change when shadows are switched off?

The SET GLOBAL SHADOW COLOR Statement
The colour of all shadows can be set using the SET GLOBAL SHADOW COLOR
FIG-39.32 The SET GLOBAL SHADOW COLOR Statement
SET GLOBAL SHADOW COLOR red , green , blue , alpha
In the diagram:
red is an integer value between 0 and 255 giving the

strength of the red component within the shadow
colour.
green is an integer value between 0 and 255 giving the

strength of the green component within the shadow
colour.
blue is an integer value between 0 and 255 giving the

strength of the blue component within the shadow
colour.
alpha is an integer value between 0 and 255 giving

the transparency component of the shadow.
0 - shadow invisible; 255 - shadow opaque.
For example, we could create a semi-transparent yellow shadow using the line:
SET GLOBAL SHADOW COLOR 255,255,0,100
Activity 39.22
Modify your last program so that the objects cast a semi-transparent red
shadow.
The SET GLOBAL SHADOW SHADES Statement

In the last program, as the shadows of the cube and sphere overlapped, they merged,
both shadows being the same shade. But in real life, when a second shadow moves
over a first, it will often result in an even darker shadow where the two overlap. We
can recreate this effect in our program using the SET GLOBAL SHADOW
SHADES statement which has the format shown in FIG-39.33.
FIG-39.33
SET GLOBAL SHADOW SHADES value
The SET GLOBAL
SHADOW SHADES
Statement
In the diagram:
value is an integer value specifying how many variations in

shadow shading are allowed
For example, if we knew three shadows might overlap and we wanted these overlaps
to display different degrees of darkness, we would set value to 3.

Activity 39.23
In your last program, add a line before the DO..LOOP structure to allow 2
shadow shadings.
How does this affect the frame rate?
Positioning Shadows
Of course, a shadow's position is dependent on the position of the light creating that
shadow; as we move the light, so the shadow will move.
Activity 39.24
In your last program, within the DO..LOOP structure, add the line:
POSITION LIGHT 0,LIGHT POSITION X(0)+1,100,0
Run the program and observe how the shadow moves as the light changes
position.
The SET SHADOW POSITION Statement

An alternative way of creating the same shifting shadow effect is to use the SET
SHADOW POSITION statement. In practice, this statement seems to move light
zero to a specified position (see FIG-39.34).
FIG-39.34
SET SHADOW POSITION
seems to reposition the main light
Positioning the Shadow
The SET SHADOW POSITION statement has the format shown in FIG-39.35.
FIG-39.35
The SET SHADOW SET SHADOW POSITION mode , x , y , z
POSITION Statement
In the diagram:
mode has the value -1 (other values not known).
x,y,z are real values giving the coordinates of the light

responsible for creating the shadow.

For example, in our last program we could move the shadow to the right by
specifying a light position to the left and above the cube with a line such as:
SET SHADOW POSITION -1,-65,60,0
Activity 39.25
xpos = -100
Replace the POSITION LIGHT statement added in the last Activity by code to
increment xpos and set the shadow position to the coordinates (xpos,60,0).
How does this affect the shadows?
Shadows and Models

Although it is simple enough to create a shadow for a fundamental shape such as a
cube or sphere, we have to work a little harder to create a realistic shadow for a
model such as the Hivebrain creature we looked at back in Chapter 36.
LISTING-39.12 shows the shadow effect produced by the Hivebrain model when
using the normal default settings for SET SHADOW SHADING ON.
Model Shadows AUTOCAM OFF
POINT CAMERA 0,0,0

TEXTURE OBJECT 1,1
REM *** Load, scale and position alien ***

LOAD OBJECT "H-Alien Hivebrain-Move.x",2
REM *** Position light zero ***

REM *** Show on model's shadow ***

WAIT KEY
END
Activity 39.26
As the program demonstrates, the shadow produced is far from realistic (see
FIG-39.36) .

FIG-39.36
The Default Shadow Cast
by a Complex Model.
To create a more convincing shadow, we need to turn the model into a mesh and
then specify that mesh in the SET SHADOW SHADING ON statement. This
requires the lines:

Activity 39.27
Modify your last program as described and observe how the shadow changes.
As we can see in FIG-39.37, the shadow is now an exact match for the model.
FIG-39.37
An Improved Shadow
Our next problem arises when we play an animated model - the shadow stays fixed
while the model itself changes. However, the shadow does at least move with the
model.
Activity 39.28
Add the following code just before the REM *** End program *** statement:
REM *** Play model ***

LOOP OBJECT 2

Activity 39.29
Change your program to finish with the lines:
REM *** Play model ***

LOOP OBJECT 2
z# = 0
DO
z# = z# -0.1
POSITION OBJECT 2,0,0,z#
LOOP
END
Check that the shadow moves with the model.
Other Shading Methods

The SET CARTOON SHADING ON Statement
A cartoon texturing effect can be achieved using the SET CARTOON SHADING
ON statement which has the format shown in FIG-39.38.
FIG-39.38
The SET CARTOON SHADING ON Statement
SET CARTOON SHADING ON objno , shadeimg , edgeimg
In the diagram:

which cartoon shading is to be applied.
shading is an integer value giving the ID of the image to

be used as the main texturing for the object.
edging is an integer value giving the ID of the image to

be used in edging the object texture.
Rather than texture a model in the normal way, cartoon shading colours the object
with the main colour of the shade image and slightly modifies the overall effect
using the edge image. Complex images have little impact on the final result, so the
images used with this statement are usually small, simply coloured ones.
Using a black edge image will make the object completely black; using a white edge
image allows the shade image to dominate in the final effect.
For example, we could use the images shade.bmp and edge.bmp to create a cartoon
shading on the Hivebrain model using the following lines:
LOAD IMAGE "shade.bmp",1

LOAD IMAGE "edge.bmp",2
SET CARTOON SHADING ON 1,1,2

Activity 39.30
Write a program to add cartoon shading to a plane and the Hivebrain model
used in the last program.
Try the program out with the following images:
shade1.bmp, edge1.bmp,
shade2.bmp, edge2.bmp,
shade3.bmp, edge3.bmp
FIG-39.39 shows several examples of the images used and the overall result when
using cartoon shading on a plane and the Hivebrain model.
Shade Image Edge Image Result

FIG-39.39
Examples of Cartoon
Shading A single white pixel
A single black pixel
A single white pixel

The SET RAINBOW SHADING ON Statement
The SET RAINBOW SHADING ON statement adds the colours of the rainbow to
the surface of an object. The statement has the format shown in FIG-39.40.
FIG-39.40
The SET RAINBOW
SET RAINBOW SHADING ON objno , imgno
SHADING ON Statement
In the diagram:

which rainbow shading is to be applied.
imgno is an integer value giving the ID of the image to

be used when creating the rainbow effect on the
object.
The image used needs to have an alpha channel, otherwise the overall effect is too
bright. FIG-39.41 shows the results achieved using an image with and without an
alpha channel.
Image Alpha Channel Result
FIG-39.41
Rainbow Shading and
Alpha Channels
No alpha channel

Activity 39.31
Modify your last program to use rainbow shading instead of cartoon shading.
Test the program with the files test1.tga, test2.tga and test3.tga which
correspond to the images used in FIG-39.41.
The SET REFLECTION SHADING ON Statement

In a previous chapter we used a second camera to create a mirror-like effect, but a
more realistic way of creating a mirror-like surface is to use the SET REFLECTION
SHADING ON statement. This will cause a specified object to reflect its
surroundings, but the operation requires a significant amount of processing power
and will only work if your video card is up to the job.
The SET REFLECTION SHADING ON statement has the format shown in

FIG-39.42.
FIG-39.42
SET REFLECTION SHADING ON objno
The SET REFLECTION
SHADING ON Statement
In the diagram:

which reflection shading is to be applied.
FIG-39.43 shows a plane reflecting cobbles and the Hivebrain model.

FIG-39.43
Reflections on a Plane
The program in LISTING-39.13 recreates the effect shown above.

CreateSet()
Reflections REM *** Create reflecting plane ***
REM *** Rotate plane ***
SET REFLECTION SHADING ON 3
DO
LOOP
REM *** End Program ***
END continued on next page

(continued) SET DISPLAY MODE 1280,1024,32
Reflections BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION CreateSet()
REM *** Load images used ***
REM *** Create ground ***
TEXTURE OBJECT 1,1
XROTATE OBJECT 1,-90
ENDFUNCTION
Activity 39.32
Type in and test the program in LISTING-39.13(advanced13.dbpro).
Do both sides of the plane reflect?
Modify the program to use a cube instead of a plane.
The SET SHADING OFF Statement

Any shading effect (shadow shading, rainbow shading, etc.) can be switched off
using the SET SHADING OFF statement which has the format shown in FIG-39.44.
FIG-39.44
The SET SHADING OFF SET SHADING OFF objno
Statement
In the diagram:

whose current shading effect is to be turned off.
Activity 39.33
Replace the reflective cube in your last program with a 20 by 20 plane and
allow a key press to toggle the reflective ability of the plane.
Summary
l Different types of surfaces reflect light in different ways.
l Ambient reflection refers to how a surface reflects ambient light.
l Diffuse reflection is the scattered reflection of directed light.

l Specular reflection is the mirrored reflection of directed light.
l Use SET OBJECT AMBIENT to set the ambient reflection of an object.
l Use SET OBJECT DIFFUSE to set the diffuse reflection of an object.
l Use SET OBJECT SPECULAR to set the specular reflection of an object.
l Use SET OBJECT SPECULAR POWER to specify the degree of specular

reflection from an object.
l Use SET OBJECT EMISSIVE to give the illusion that an object is emitting light.
l Use SET OBJECT LIGHT to specify how an object is to react to light.
l A light map is an image used to determine how an object reacts to light.
l Use SET LIGHT MAPPING ON to apply a light map to an object.
l A bump map is an image used to give the illusion of roughness on an object's

surface.
l Use SET BUMP MAPPING ON to apply a bump map to an object.
l Use SET SPHERE MAPPING ON to create the illusion of an object's

surroundings being reflected on its surface.
l Use SET BLEND MAPPING ON to blend two images on the surface of an

object.
l Use SET CUBE MAPPING ON to create an object reflecting its surroundings.
l Use SET ALPHA MAPPING ON to make an object transparent or invisible.
l Use SET SHADOW SHADING ON to make an object cast a shadow.
l Use SET SHADOW POSITION to position the shadow cast by an object.
l Use SET SHADOW SHADING OFF to turn off the shadow effect for a specified
object.
l Use SET GLOBAL SHADOWS to switch all allocated shadows on or off.
l Use SET GLOBAL SHADOW COLOR to set the colour-cast of all shadows.
l Use SET GLOBAL SHADOW SHADES to set the number of shades available
when shadows overlap.
l To create a realistic shadow from a model, use the mesh of the model as the
shadow.
l Use SET CARTOON SHADING ON to create a basic colouring of an object's

surface.
l Use SET RAINBOW SHADING ON to create a rainbow effect on the surface

of an object.
l Use SET REFLECTION SHADING ON to make an object's surface reflect the

objects around it.
l Use SET SHADING OFF to switch off all shading effects that have been applied
to a specific object.

Solutions
Activity 39.1 REM *** Switch off ambient reflection ***
WAIT KEY
REM *** Set up screen *** SET OBJECT AMBIENT 1,0
SET DISPLAY MODE 1280,1024,32 REM *** Switch on diffuse reflection ***
COLOR BACKDROP RGB(20,20,20) WAIT KEY
BACKDROP ON SET OBJECT DIFFUSE 1, RGB(0,255,255)
REM *** Programmed screen updates *** REM *** Add a yellow light ***
SYNC ON WAIT KEY
REM *** Position camera *** MAKE LIGHT 1
AUTOCAM OFF SET SPOT LIGHT 1,90,180
POSITION CAMERA 0,0,-60 POSITION LIGHT 1, -1,3,-5
POINT CAMERA 0,0,0 POINT LIGHT 1,0,0,0
REM *** Make cube *** COLOR LIGHT 1, RGB(255,255,0)
MAKE OBJECT SPHERE 1,10,50,50 SET OBJECT SPECULAR 1, RGB(255,255,0)
REM *** Switch off directional light *** REM *** End program ***
HIDE LIGHT 0 WAIT KEY
SYNC END
REM *** Switch off ambient reflection ***
PRINT "Change sphere's reflection of
Äambient light" Activity 39.4
SYNC REM *** Set up screen ***
WAIT KEY SET DISPLAY MODE 1280,1024,32
SET OBJECT AMBIENT 1,0 COLOR BACKDROP RGB(20,20,20)
SYNC BACKDROP ON
REM *** Switch on ambient reflection *** REM *** Position camera ***
PRINT "Change sphere's reflection of AUTOCAM OFF
Äambient light" POSITION CAMERA 0,0,-60
SYNC MAKE OBJECT SPHERE 1,10
WAIT KEY COLOR OBJECT 1, RGB(0,0,255)
SET OBJECT AMBIENT 1,1 REM *** Switch off ambient reflection ***
SYNC WAIT KEY
SYNC SET OBJECT AMBIENT 1,0
REM *** End program *** REM *** Switch on diffuse reflection ***
WAIT KEY WAIT KEY
END SET OBJECT DIFFUSE 1, RGB(0,255,255)
REM *** Add a yellow light ***
WAIT KEY
MAKE LIGHT 1
Activity 39.2 SET SPOT LIGHT 1,90,180
REM *** Set up screen *** POSITION LIGHT 1, -1,3,-5
SET DISPLAY MODE 1280,1024,32 POINT LIGHT 1,0,0,0
COLOR BACKDROP RGB(20,20,20) COLOR LIGHT 1, RGB(255,255,0)
BACKDROP ON SET OBJECT SPECULAR 1, RGB(255,255,0)
REM *** Position camera *** FOR power = 100 TO 0 STEP -1
AUTOCAM OFF SET OBJECT SPECULAR POWER 1,power
POSITION CAMERA 0,0,-60 WAIT 25
REM *** Create sphere *** NEXT power
MAKE OBJECT SPHERE 1,10 REM *** End program ***
COLOR OBJECT 1, RGB(0,0,255) WAIT KEY
REM *** Switch off ambient reflection *** END
WAIT KEY
REM *** Switch on diffuse reflection *** Activity 39.5
WAIT KEY
SET OBJECT DIFFUSE 1, RGB(0,255,255) No solution required.
WAIT KEY
END Activity 39.6
Activity 39.3 SET DISPLAY MODE 1280,1024,32
REM *** Set up screen *** BACKDROP ON
SET DISPLAY MODE 1280,1024,32 REM *** Position camera ***
COLOR BACKDROP RGB(20,20,20) AUTOCAM OFF
BACKDROP ON POSITION CAMERA 0,0,-60
REM *** Position camera *** REM *** Create sphere ***
AUTOCAM OFF MAKE OBJECT SPHERE 1,10,100,100
POSITION CAMERA 0,0,-60 REM *** Switch off lights ***
REM *** Create sphere *** WAIT KEY
MAKE OBJECT SPHERE 1,10 SET AMBIENT LIGHT 0
COLOR OBJECT 1, RGB(0,0,255) HIDE LIGHT 0

REM *** produce fake emissive effect *** WAIT KEY
WAIT KEY MAKE LIGHT 1
SET OBJECT EMISSIVE 1,RGB(0,0,255) SET POINT LIGHT 1,0,0,0
REM *** make cube *** COLOR LIGHT 1, RGB(0,0,255)
WAIT KEY REM *** Switch off cube reflection ***
MAKE OBJECT CUBE 2,10 SET OBJECT LIGHT 2,0
POSITION OBJECT 2,15,15,20 REM *** End program ***
WAIT KEY END
END
This change causes the cube 's surface to reflect a white
The cube is not affected by the blue sphere and remains light.
black.
To change the ambient colour to red, modify the line
Activity 39.7 SET AMBIENT LIGHT 0
SET DISPLAY MODE 1280,1024,32 to read
BACKDROP ON SET AMBIENT LIGHT RGB(255,0,0)
AUTOCAM OFF This change makes the cube reflect the red ambient light
POSITION CAMERA 0,0,-60 until SET OBJECT LIGHT statement for the cube is
executed. The cube's surface then returns to reflecting
REM *** Switch off lights *** white light.
WAIT KEY
SET AMBIENT LIGHT 0
HIDE LIGHT 0 Activity 39.10
REM *** produce fake emissive effect ***
WAIT KEY
REM *** make cube ***
BACKDROP ON
WAIT KEY
AUTOCAM OFF
REM *** Create light ***
POINT CAMERA 0,0,3
WAIT KEY
CreateScene()
MAKE LIGHT 1
Rem *** End program ***
WAIT KEY
END
WAIT KEY
END
REM *** Create cobbled street ***
The cube is lit by the new light. XROTATE OBJECT 1,90
LOAD IMAGE "cobbleslarge.jpg",1
TEXTURE OBJECT 1,1
Activity 39.8 POSITION OBJECT 1,0,-10,0
No solution required. MAKE OBJECT PLAIN 2,30,20
TEXTURE OBJECT 2,2
Activity 39.9 POSITION OBJECT 2,0,-1,5
REM *** Create lamppost ***
REM *** Set up screen *** MAKE OBJECT CYLINDER 3,10
SET DISPLAY MODE 1280,1024,32 SCALE OBJECT 3, 5,100,5
COLOR BACKDROP RGB(20,20,20) POSITION OBJECT 3,0,-5,0
BACKDROP ON LOAD IMAGE "lamppost.jpg",3
REM *** Position camera *** TEXTURE OBJECT 3,3
AUTOCAM OFF MAKE OBJECT SPHERE 4,2,100,100
POSITION CAMERA 0,0,-60 REM *** Make sphere appear to glow ***
REM *** Create sphere *** SET OBJECT EMISSIVE 4,RGB(255,255,255)
MAKE OBJECT SPHERE 1,10,100,100 REM *** Put light in sphere ***
REM *** Switch off lights *** MAKE LIGHT 1
WAIT KEY SET POINT LIGHT 1,0,0,0
SET AMBIENT LIGHT 0 ENDFUNCTION
HIDE LIGHT 0
REM *** produce fake emissive effect ***
REM *** make cube *** To use the new light map change the line
WAIT KEY
REM *** Create light ***
to

LOAD IMAGE "lightgrey.jpg",4 POSITION CAMERA 0,0,-50
POINT CAMERA 0,0,0
This gives a less realistic lighting effect. REM *** Add extra light ***
MAKE LIGHT 1
POSITION LIGHT 1,20,-20,50
Activity 39.12 REM *** Make cone with culling off ***
The main section of the program should read: SET OBJECT CULL 22,0
BACKDROP ON
LOAD IMAGE "cobbles04.jpg",4
AUTOCAM OFF
POINT CAMERA 0,0,3
SET CUBE MAPPING ON 22,1,2,3,4,5,6
CreateScene()
DO
LOAD IMAGE "lightgrey.jpg",4
TURN OBJECT LEFT 22,RND(200)/100.0
PITCH OBJECT UP 22,RND(200)/100.0
ROLL OBJECT RIGHT 22,RND(200)/100.0 - 1
WAIT 10
LOOP
REM *** Add bump map ***
WAIT KEY
END
LOAD IMAGE "cobblesbump.jpg",6
SET BUMP MAPPING ON 1,6
WAIT KEY
END Activity 39.17
Unfortunately, adding bump mapping appears to remove
the light mapping effect.
Activity 39.18
Activity 39.13 REM *** Set up screen ***
SET DISPLAY MODE 1280,1024,32 AUTOCAM OFF
COLOR BACKDROP RGB(155,155,0) POSITION CAMERA 0,10,-60
BACKDROP ON POINT CAMERA 0,0,0
AUTOCAM OFF REM *** Create background object ***
POSITION CAMERA 0,10,-60 MAKE OBJECT PLAIN 1,100,100
POINT CAMERA 0,0,0 LOAD IMAGE "cobblestones.jpg",1
REM *** make sphere *** TEXTURE OBJECT 1,1
MAKE OBJECT SPHERE 1,10,50,50 SCALE OBJECT TEXTURE 1,10,10
WAIT KEY XROTATE OBJECT 1,-90
REM *** Load and apply sphere map *** REM *** Create and texture cube ***
LOAD IMAGE "fish.jpg",1 MAKE OBJECT CUBE 2,20
SET SPHERE MAPPING ON 1,1 LOAD IMAGE "grid8by8.bmp",2
WAIT KEY TEXTURE OBJECT 2,2
REM *** Texture sphere *** POSITION OBJECT 2,0,20,0
LOAD IMAGE "stripes.jpg",2 REM *** Adjust light ***
TEXTURE OBJECT 1,2 SET POINT LIGHT 0,-10,100,1
REM *** Rotate sphere *** REM *** Switch cube's shadow on ***
DO SET SHADOW SHADING ON 2 ,-1,100,1
TURN OBJECT LEFT 1,1 REM *** Rotate cube & display frame rate ***
WAIT 10 DO
LOOP TURN OBJECT LEFT 2,1
REM *** End program *** SET CURSOR 100,100
END PRINT "Frame rate : ",SCREEN FPS()," "
Ä ,ord
While the texture rotates with the cube, the sphere LOOP
mapped image remains stationary. REM *** End program ***
END
Activity 39.14 The effect of the change on the author's machine was to
slightly decrease the frame rate but stop a slight flicker
No solution required. effect on the side of the cube.
Changing the distance parameter to 10 removes the shadow

Activity 39.15 since the distance from the cube to the floor plane is more
No solution required. than 10 units.

REM *** Set up screen *** REM *** Set up screen ***
SET DISPLAY MODE 1280,1024,32 SET DISPLAY MODE 1280,1024,32
AUTOCAM OFF AUTOCAM OFF

POSITION CAMERA 0,10,-60 MAKE OBJECT SPHERE 3,10,20,20
POINT CAMERA 0,0,0 TEXTURE OBJECT 3,2
REM *** Create background object *** POSITION OBJECT 3, 20,10,30
MAKE OBJECT PLAIN 1,100,100 REM *** Point the main light ***
LOAD IMAGE "cobblestones.jpg",1 SET POINT LIGHT 0,-10,50,0
TEXTURE OBJECT 1,1 REM *** Switch shading on ***
SCALE OBJECT TEXTURE 1,10,10 SET SHADOW SHADING ON 2,-1,100,1
XROTATE OBJECT 1,-90 SET SHADOW SHADING ON 3,-1,100,1
REM *** Create and texture cube *** REM *** Prepare objects for movement ***
MAKE OBJECT CUBE 2,20 POINT OBJECT 2,0,200,0
LOAD IMAGE "grid8by8.bmp",2 POINT OBJECT 3,20,10,-30
TEXTURE OBJECT 2,2 movement# = 0.1
POSITION OBJECT 2,0,20,0 REM *** Change shadow colour ***
REM *** Adjust light *** SET GLOBAL SHADOW COLOR 255,0,0,100
SET POINT LIGHT 0,-10,100,1 DO
REM *** create mesh *** REM *** Shadow on/off options ***
MAKE OBJECT SPHERE 3,30,50,50 IF INKEY$() = "s"
MAKE MESH FROM OBJECT 1,3 SET GLOBAL SHADOWS ON
DELETE OBJECT 3 ENDIF
REM *** Switch cube's shadow on *** IF INKEY$() = "o"
SET SHADOW SHADING ON 2 ,1,100,1 SET GLOBAL SHADOWS OFF
REM *** Rotate cube and display frame ENDIF
rate *** REM *** Rotate cube and move sphere ***
DO TURN OBJECT LEFT 2,1
TURN OBJECT LEFT 2,1 MOVE OBJECT 2, 0.1
SET CURSOR 100,100 MOVE OBJECT 3, movement#
PRINT "Frame rate : ",SCREEN FPS()," ", IF OBJECT POSITION Z(3) < -40
Ä
ord movement# = -0.1
LOOP ELSE
REM *** End program *** IF OBJECT POSITION Z(3) > 30
END movement# = 0.1
ENDIF
The cube now casts a circular shadow! The frame rate is ENDIF
also reduced. REM *** Display frame rate ***
SET CURSOR 100,100
LOOP
Activity 39.20 REM *** End program ***
END
The DO..LOOP structure of the program should be
changed to read:
DO
Activity 39.23
IF INKEY$() <> "" Replace the lines:
SET SHADOW SHADING OFF 2
ENDIF
REM *** Change shadow colour ***
SET GLOBAL SHADOW COLOR 255,0,0,100
SET CURSOR 100,100
PRINT "Frame rate : ",SCREEN FPS()," ",
Ä
ord with:
LOOP
REM *** Set shadow shading ***
SET GLOBAL SHADOW SHADES 2
Activity 39.21
The frame rate will probably fall. The more shades
The frame rate increases significantly when shadows are specified, the lower the frame rate.
turned off.
Activity 39.24
Activity 39.22
The DO..LOOP code should now read:
SET DISPLAY MODE 1280,1024,32 DO
AUTOCAM OFF POSITION LIGHT 0,LIGHT POSITION X(0)
POINT CAMERA 0,0,0
Ä+1,100,0
REM *** Shadow on/off options ***
IF INKEY$() = "s"
SET GLOBAL SHADOWS ON
ENDIF
TEXTURE OBJECT 1,1
IF INKEY$() = "o"
SET GLOBAL SHADOWS OFF
ENDIF
REM *** Rotate cube and move sphere ***
MOVE OBJECT 2, 0.1
TEXTURE OBJECT 2,2
REM *** Create and texture sphere ***

movement# = -0.1
ELSE
ACTIVITY 39.28
IF OBJECT POSITION Z(3) > 30 REM *** Set up screen ***
movement# = 0.1 SET DISPLAY MODE 1280,1024,32
ENDIF AUTOCAM OFF
ENDIF POSITION CAMERA 0,10,-60
REM *** Display frame rate *** POINT CAMERA 0,0,0
SET CURSOR 100,100 REM *** Create background object ***
PRINT "Frame rate : ",SCREEN FPS() MAKE OBJECT PLAIN 1, 100,100
LOOP LOAD IMAGE "cobblestones.jpg",1
TEXTURE OBJECT 1,1
Activity 39.25 XROTATE OBJECT 1,-90
xpos = -100 LOAD OBJECT "H-Alien Hivebrain-Move.x",2
DO SCALE OBJECT 2,300,300,300
INC xpos POSITION OBJECT 2,0,0,0
SET SHADOW POSITION -1,xpos,60,0 REM *** Position light zero ***
REM *** Shadow on/off options *** SET POINT LIGHT 0,-60,50,0
IF INKEY$() = "s" REM *** Show on model's shadow ***
SET GLOBAL SHADOWS ON MAKE MESH FROM OBJECT 1,2
ENDIF SET SHADOW SHADING ON 2,1,100,1
IF INKEY$() = "o" REM *** Play model ***
SET GLOBAL SHADOWS OFF LOOP OBJECT 2
ENDIF REM *** End program ***
REM *** Rotate cube and move sphere *** WAIT KEY
TURN OBJECT LEFT 2,1 END
MOVE OBJECT 2, 0.1
movement# = -0.1
Activity 39.29
ELSE REM *** Set up screen ***
IF OBJECT POSITION Z(3) > 30 SET DISPLAY MODE 1280,1024,32
movement# = 0.1 AUTOCAM OFF
ENDIF POSITION CAMERA 0,10,-60
ENDIF POINT CAMERA 0,0,0
REM *** Display frame rate *** REM *** Create background object ***
SET CURSOR 100,100 MAKE OBJECT PLAIN 1, 100,100
PRINT "Frame rate : ",SCREEN FPS() LOAD IMAGE "cobblestones.jpg",1
LOOP TEXTURE OBJECT 1,1
The shadow shifts as the light moves. XROTATE OBJECT 1, -90
Activity 39.26 SCALE OBJECT 2,300,300,300
No solution required. REM *** Position light zero ***
REM *** Show on model's shadow ***
Activity 39.27 MAKE MESH FROM OBJECT 1,2
REM *** Set up screen *** REM *** Play model ***
SET DISPLAY MODE 1280,1024,32 LOOP OBJECT 2
AUTOCAM OFF z# = 0
POINT CAMERA 0,0,0 z# = z# -0.1
REM *** Create background object *** POSITION OBJECT 2,0,0,z#
MAKE OBJECT PLAIN 1, 100,100 LOOP
LOAD IMAGE "cobblestones.jpg",1 REM *** End program ***
TEXTURE OBJECT 1,1 WAIT KEY
SCALE OBJECT TEXTURE 1,10,10 END
LOAD OBJECT "H-Alien Hivebrain-Move.x",2 Activity 39.30
POSITION OBJECT 2,0,0,0 REM *** Set up screen ***
REM *** Position light zero *** SET DISPLAY MODE 1280,1024,32
SET POINT LIGHT 0,-60,50,0 COLOR BACKDROP RGB(255,255,100)
REM *** Show on model's shadow *** BACKDROP ON
MAKE MESH FROM OBJECT 1,2 AUTOCAM OFF
SET SHADOW SHADING ON 2,1,100,1 POSITION CAMERA 0,10,-30
REM *** End program *** POINT CAMERA 0,0,0|
WAIT KEY REM *** Load images used ***
END LOAD IMAGE "shade1.bmp",1
LOAD IMAGE "edge2.bmp",2

REM *** Load, scale and position alien *** SCALE OBJECT 2,300,300,300
LOAD OBJECT "H-Alien Hivebrain-Move.x",2 POSITION OBJECT 2,0,0,0
SCALE OBJECT 2,300,300,300 ENDFUNCTION
REM *** Use cartoon shading ***
SET CARTOON SHADING ON 1,1,2 Activity 39.33
SET CARTOON SHADING ON 2,1,2
REM *** End program *** The main section is coded as:
WAIT KEY
END REM *** Set up screen ***
ScreenSetUp()
CreateSet()
Activity 39.31 REM *** Create reflecting plane ***
REM *** Set up screen *** POSITION OBJECT 3,0,5,20
SET DISPLAY MODE 1280,1024,32 XROTATE OBJECT 3, 10
COLOR BACKDROP RGB(255,255,100) REM *** Rotate plane ***
BACKDROP ON SET REFLECTION SHADING ON 3
AUTOCAM OFF reflection = 1
POINT CAMERA 0,0,0 TURN OBJECT LEFT 3,0.5
REM *** Load image used *** ch$ = INKEY$()
LOAD IMAGE "test3.tga",1 IF ch$ <> ""
REM *** Create ground *** IF reflection = 1
MAKE OBJECT PLAIN 1,200,200 SET SHADING OFF 3
XROTATE OBJECT 1,-90 reflection = 0
REM *** Load, scale and position alien *** ELSE
LOAD OBJECT "H-Alien Hivebrain-Move.x",2 SET REFLECTION SHADING ON 3
SCALE OBJECT 2,300,300,300 reflection = 1
POSITION OBJECT 2,0,0,0 ENDIF
REM *** Use rainbow shading *** REM *** Wait for key release ***
SET RAINBOW SHADING ON 1,1 WHILE INKEY$() <> ""
SET RAINBOW SHADING ON 2,1 TURN OBJECT LEFT 3,0.5
REM *** End program *** ENDWHILE
WAIT KEY ENDIF
END LOOP
END
Activity 39.32
Only one side of the plane is reflective.

ScreenSetUp()
CreateSet()
REM *** Create reflecting cube ***
REM *** Rotate plane ***
SET REFLECTION SHADING ON 3
DO
LOOP
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION CreateSet()
REM *** Load images used ***
LOAD IMAGE "test3.tga",2
TEXTURE OBJECT 1,1
REM *** Load, scale and place alien ***
LOAD OBJECT "H-Alien Hivebrain-Move.x"
Ä
,2

40
Bounding Boxes and Spheres
Object Collisions
Sliding Collisions
Static Collisions
DarkBASIC Pro: Collisions 1067

Object Collisions
Introduction
Back in Chapter 21 we saw the importance of detecting when two sprites collide -
for example, we needed to know when a bat hit a ball or a missile hit a spaceship.
That same requirement exists when dealing with a 3D world; we need to detect a
sphere's collision with a cube or a bullet with a wall.
Unfortunately, things aren't as simple in 3D as they are in 2D, and this is reflected
by the number and complexity of the commands available for 3D collision handling.
There are two main classes of collision in 3D. The first, object collision, deals with
Moveable objects are collisions between moveable 3D objects (such as spheres, cones, cubes, and loaded
often referred to as
dynamic objects. models). The second type is static collision where a moving object collides with a
fixed (static) object (for example, a character colliding with the side of a house).
Object Collision
When one 3D object collides with another, we need to detect this and make the
objects react to the event in a realistic way.
The program in LISTING-40.1 moves a sphere towards a cube.
LISTING-40.1 REM *** Setup screen ***

When 3D Objects Collide COLOR BACKDROP 0
BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Make two objects ***

REM *** Move sphere towards cube ***

DO
LOOP

END
Activity 40.1
Type in and test the program given in LISTING-40.1 (collision01.dbpro).
What happens when the sphere comes into contact with the cube?
As we saw from the Activity above, objects will not automatically react to making
contact with each other. If we want some reaction, then we must include explicit
code to deal with this in our program.
1068 DarkBASIC Pro: Collisions

The OBJECT HIT Statement
We can detect if two objects have collided using the OBJECT HIT statement which
has the format given in FIG-40.1.
FIG-40.1
objno2
The OBJECT HIT
Statement OBJECT HIT ( objno1 , )
0
integer
In the diagram:
objno1 is an integer value identifying the main object

being tested.
objno2 is an integer value identifying the second object

to be tested.
0 use this option (zero) when the ID of the second

object is unknown.
When using the first option, where both objects are specified using ID numbers, the
statement returns 1 if the objects have collided and zero if no collision has taken
place. For example, in the last program we could check for the sphere and cube
colliding using the statement:
IF OBJECT HIT(1,2) = 1
When we are unsure of which object may have been hit, we can use the second
version of this statement (with zero as the final value). This will return the ID
number of any object struck by objno1. If no object has been hit, then zero is
returned. For example, we could discover the ID of any object struck by object 1
with the code:
objectstruck = OBJECT HIT(1,0)

IF objectstruck = 0
CURSOR 100,100
PRINT "No object hit"
ELSE
CURSOR 100,100
PRINT "Object hit has ID ",objectstruck
ENDIF
Activity 40.2
Modify your last program (collision01.dbpro) so that the DO..LOOP structure

is exited when the sphere collides with the cube.
Add a second DO..LOOP structure near the end of the program to allow the
player to move the camera using the arrowkeys.
Run the modified version of the program and manually position the camera so
that you can see that the sphere has come into contact with the cube.

The OBJECT COLLISION Statement
An alternative to OBJECT HIT is OBJECT COLLISION. The two statements
perform exactly the same function. This statement has the format shown in
FIG-40.2.
FIG-40.2
objno2
The OBJECT
COLLISION Statement OBJECT COLLISION ( objno1 , )
0
integer
In the diagram:
objno1 is an integer value identifying the main object

being tested.
objno2 is an integer value identifying the second object

to be tested.
0 use this option (zero) when the ID of the second

object is unknown.
Activity 40.3
Modify your last program to use the OBJECT COLLISION statement in place
of OBJECT HIT.
The SET OBJECT COLLISION Statement

It is possible to switch off collision detection for a specific object using the SET
OBJECT COLLISION statement. The same statement can be used to switch
collision detection back on. The statement has the format shown in FIG-40.3.
FIG-40.3
ON
The SET OBJECT
COLLISION Statement SET OBJECT COLLISION objno
OFF
In the diagram:
objno is an integer value identifying the object whose

collision detection is to be modified.
When an object has its collision detection switched off using this statement, any
subsequent calls to OBJECT HIT or OBJECT COLLISION which involve this
object will always return zero.

Activity 40.4
Switch off collision detection for object 1 in your last program and check out
how this affects the results obtained.
The SET GLOBAL COLLISION Statement

While SET OBJECT COLLISION allows us to control collision detection for an
individual object, we can control collision detection of all 3D objects using the SET
GLOBAL COLLISION statement. This statement allows every object's collision
detection to be switched on or off, irrespective of any previous setting assigned to
individual objects. The SET GLOBAL COLLISION statement has the format
shown in FIG-40.4.
FIG-40.4
ON
The SET GLOBAL
COLLISION Statement SET GLOBAL COLLISION
OFF
NOTE: SET GLOBAL Activity 40.5

COLLISION ON does not
neutralise the SET Using the SET GLOBAL COLLISION statement, neutralise the effect of the
OBJECT COLLISION SET OBJECT COLLISION statement added to your program in the previous
OFF 1 statement! (DBPro Activity.
v 1.062)
How Collision Detection Works

DarkBASIC Pro automatically places an invisible box around every 3D object that
appears on the screen. This box is known as a bounding box or a collision box.
When the collision boxes of two shapes come into contact with each other, a
collision is detected.
The bounding sphere is In addition to the collision box, every object is also surrounded by a bounding or
really a polyhedron. collision sphere since, for some 3D objects, this creates a more accurate
approximation of the shape it surrounds than a box. For example, FIG-40.5 shows
the box and sphere that surround each of the objects in our previous program.
FIG-40.5
Collision Boxes
Since they are almost

perfect matches, the 3D
sphere's collision sphere
and the 3D cube's
collision box are difficult
to spot in the screen
dump.

Notice that the collision box is a perfect match for the cube object, while the
bounding sphere more accurately specifies the volume of the sphere object.
The SHOW OBJECT BOUNDS Statement

It is possible to make the bounding box and sphere that surround a specific 3D object
visible using the SHOW OBJECT BOUNDS statement which has the format shown
in FIG-40.6.
FIG-40.6
The SHOW OBJECT SHOW OBJECT BOUNDS objno , boxonlyflag
BOUNDS Statement
In the diagram:

bounding volumes are to be displayed.
boxonlyflag 0 or 1. Set to 0 to show both the box and sphere

bounding volumes; set to 1 to show only the
bounding box.
Activity 40.6
Modify your last program so that the bounding boxes of both objects are
displayed.
Activity 40.7
Change the cube in your program to a cone. How does this change affect the
collision between the two objects?
This last Activity highlights a problem: boxes and spheres do not make ideal
collision volumes for every shape.
The HIDE OBJECT BOUNDS statement

An object's bounding box and sphere can be hidden again using the HIDE OBJECT
BOUNDS statement which has the format shown in FIG-40.7.
FIG-40.7
The HIDE OBJECT HIDE OBJECT BOUNDS objno
BOUNDS Statement
In the diagram:

bounding volumes are to be hidden.
Activity 40.8
In the last program, use HIDE OBJECT BOUNDS to hide the bounding boxes
of the sphere and cone after a key is pressed and before movement starts.

We're using the term
If a 3D object is moved, its bounding volumes (box and sphere) move with it. If a
cuboid to refer to the 3D 3D object is rotated, its bounding box also rotates, but the bounding sphere, being
box object and the term symmetrical, does not need to rotate. FIG-40.8 shows the bounding volumes for a
box to refer to its sphere and cuboid after the cuboid has been rotated by 90o about the z-axis. From
surrounding bounding box. this we can see that the cuboid's bounding box has also been rotated.
FIG-40.8
A Collision Box Rotates
Along with the Object to
which it is Attached
By default, the collision boxes are always used to detect collision between two 3D
objects - even between two spheres. The program in LISTING-40.2 fires the sphere
in the direction of the rotated cuboid shown in FIG-40.8 above. Since the objects'
collision boxes do not intersect, no collision is detected.
Only a Hit Between Two COLOR BACKDROP 0
Collision Boxes is BACKDROP ON
Detected
AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Make box ***

REM *** Display its bounding volumes ***
SHOW OBJECT BOUNDS 1
REM *** Lie box on its side ***
ZROTATE OBJECT 1, 90
REM *** Create a sphere ***

REM *** Position sphere ***
POSITION OBJECT 2,0,20,-100
REM *** Move sphere towards box until they collide ***
DO
IF OBJECT HIT(1,2)
EXIT
ENDIF
LOOP
WAIT KEY
END

Activity 40.9
We can see from running this program that, although the bounding spheres of both
3D objects intersect, no collision is detected.
Modifying Collision Detection

The SET OBJECT COLLISION TO SPHERES Statement
It is possible to use the bounding spheres instead of the bounding boxes to detect
collision by using the SET OBJECT COLLISION TO SPHERES statement which
FIG-40.9
The SET OBJECT SET OBJECT COLLISION TO SPHERES objno
COLLISION TO
SPHERES Statement
In the diagram:

which is to use its bounding sphere when
detecting collisions.
Activity 40.10
Add the line
SET OBJECT COLLISION TO SPHERES 1
to your last program immediately after the cuboid object is created.
How does this change affect the program?
The SET OBJECT RADIUS Statement

When using a collision sphere, it is possible to modify the radius of that sphere from
its default size (based on the size of the object it surrounds) using the SET OBJECT
RADIUS statement, which has the format shown in FIG-40.10.
FIG-40.10
The SET OBJECT SET OBJECT RADIUS objno , radius
RADIUS Statement
In the diagram:

bounding sphere's radius is to be modified.
radius is a real number giving the radius of the collision

sphere.

Activity 40.11
Move the sphere in your previous program to a height of 40 and run the code.
Does the 3D sphere hit the cuboid's bounding sphere?
Change the radius of the cuboid's bounding sphere to 50 and re-run the
program. Does the 3D sphere collide with the modified bounding sphere?
Unfortunately, the SHOW OBJECT BOUNDS statement only displays the default
bounding box and sphere, so it does not show any changes to the bounding sphere's
size when the radius is changed.
The OBJECT COLLISION RADIUS Statement

If we need to find out the current radius of a collision sphere we can use the
OBJECT COLLISION RADIUS statement, the format for which is shown in
FIG-40.11.
FIG-40.11
The OBJECT COLLISION OBJECT COLLISION RADIUS ( objno )
RADIUS Statement
real
In the diagram:

collision sphere's radius is to be determined.
This object should previously have been set to
use a collision sphere.
The OBJECT COLLISION CENTER Statement

The centre of a collision sphere can be determined using the OBJECT COLLISION
CENTER statement which has the format shown in FIG-40.12.
FIG-40.12
X
The OBJECT
COLLISION CENTER OBJECT COLLISION CENTER Y ( objno )
Statement
real
In the diagram:
X,Y,Z Use the appropriate one of these options for the

ordinate required.

collision sphere's centre is to be determined.
This object should previously have been set to
use a collision sphere.

For example, if object 2 employs a collision sphere, then we can find that sphere's
centre with the lines:
x# = OBJECT COLLISION CENTER X(2)

y# = OBJECT COLLISION CENTER Y(2)
z# = OBJECT COLLISION CENTER Z(2)
The SET OBJECT COLLISION TO BOXES Statement

It is possible to revert to using an object's collision box with the SET OBJECT
COLLISION TO BOXES statement which has the format shown in FIG-40.13.
FIG-40.13
The SET OBJECT SET OBJECT COLLISION TO BOXES objno
COLLISION TO BOXES
Statement
In the diagram:

to use its bounding box when detecting collisions.
The SET OBJECT COLLISION TO POLYGONS Statement

There are some shapes for which boxes and spheres will just not be accurate enough
when trying to determine if a collision has occurred.
Activity 40.12
Reload your earlier program (collision01.dbpro) and check the difference

between using a box or a sphere to detect collision between the 3D sphere and
cone.
We can force the computer to detect collision with the actual polygons that make
up the 3D shape itself by using the SET OBJECT COLLISION TO POLYGONS
statement, the format for which is shown in FIG-40.14.
FIG-40.14
The SET OBJECT SET OBJECT COLLISION TO POLYGONS objno
COLLISION TO
POLYGONS Statement
In the diagram:

to use its own polygons when detecting collisions.
In LISTING-40.3 the program again fires the sphere towards the cone shape, but
this time using polygon collision detection.

Using Polygon Collision COLOR BACKDROP 0
Detection BACKDROP ON

AUTOCAM OFF
POSITION CAMERA -100,10,-20
POINT CAMERA 0,0,0

LISTING-40.3 REM *** Make cone ***
(continued) MAKE OBJECT CONE 1,40
REM *** Use polygon collision detection ***
Using Polygon Collision SET OBJECT COLLISION TO POLYGONS 1
Detection
REM *** Create and position a sphere ***
REM *** Move sphere towards cone until they collide ***
DO
IF OBJECT HIT(1,2)
EXIT
ENDIF
LOOP

WAIT KEY
END
Activity 40.13
Does the sphere come into contact with the cone?
Using polygon collision detection is very process-intensive and can slow down your
program, so don't use it unnecessarily.
The MAKE OBJECT COLLISION BOX Statement

Assuming we want to use a bounding box to detect collisions rather than spheres
or polygons, it is possible to replace the default bounding box of an object with one
of your own using the MAKE OBJECT COLLISION BOX statement. This
FIG-40.15
The MAKE OBJECT COLLISION BOX Statement
MAKE OBJECT COLLISION BOX objno ,
x1 , y1 , z1 , x2 , y2 , z2 , rotateflag
In the diagram:

bounding volumes are to be displayed.
x1,y1,z1 are a set of real values specifying the near

bottom-left vertex of the bounding box.
x2,y2,z2 are a set of real values specifying the far

top-right vertex of the bounding box.

rotateflag 0 or 1. If set to 1, the bounding box will rotate
when the object to which it is linked rotates;
when set to zero the bounding box is unaffected
by rotation of the object to which it is linked.
The diagram in FIG-40.16 shows the position of the vertices to be given when
creating a collision box.
FIG-40.16 x2,y2,z2
User-defined
The Coordinates Required collision box
When Specifying a Collision
Box
3D Object
x1,y1,z1
Normally, a collision box will rotate along with the 3D object to which it is attached.
For example, FIG-40.17 shows the position of a cone (object 1) and its
corresponding collision box after the statement
has been executed.

FIG-40.17
A Collision Box Rotates
with the Object to which
it is Attached
User-defined
collision box
3D Object
By using a setting of zero for rotateflag, the collision box will not rotate along with
the object to which it is attached. For example, if we execute the lines

MAKE OBJECT COLLISION BOX 1,-20,-20,-20,20,20,20,0
ZROTATE OBJECT 1,45
rotating the cone by 45o, then the cone and its collision box will be as shown in
FIG-40.18.
FIG-40.18 User-defined
collision box
The Collision Box Does
Not Rotate
3D Object

Although the collision box does not rotate, it will move should the object to which
it is attached be repositioned.
Unfortunately, the SHOW OBJECT BOUNDS statement is not capable of showing

a user-defined collision box, so we can't see the effects of creating a new collision
box or preventing its rotation. However, we can observe how it affects collisions in
the program given in LISTING-40.4. In this program a cuboid is given a bounding
box which is taller than the object itself. As a sphere passes above the cuboid it
collides with the bounding box and a collision is detected. FIG-40.19 shows how
the program executes.
FIG-40.19
bounding box
A Box Object and its
Collision Box sphere's trajectory
cuboid

Using MAKE OBJECT COLOR BACKDROP 0
COLLISION BOX BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0

REM *** Create a bounding box taller than 3D box ***

DO
SET CURSOR 100,100
PRINT OBJECT POSITION Z(2)," ",STATISTIC(1)
IF OBJECT COLLISION(1,2)
EXIT
ENDIF
LOOP

WAIT KEY
END

Activity 40.14
Type in and test the program in LISTING-40.4 (collision04.dbpro).
Does the sphere stop when it strikes the box?
Modify the program so that the cuboid is rotated to 90o about its z-axis after
the collision box has been created.
How does this affect the sphere?
Change the final value in the MAKE OBJECT COLLISION BOX to 1.
How is the sphere affected by this change?
The first time the cuboid is rotated, the sphere still collides with its bounding box
because the bounding box no longer rotates in conjunction with its object.
The GET OBJECT COLLISION Statement

When one object collides with a second object, and that second object has an object
collision box (created using MAKE OBJECT COLLISION BOX), and the object
collision box is designed not to rotate with the object to which it has been linked,
then it is possible to determine how far object 1 has moved into the collision box
of object 2 (see FIG-40.20).
FIG-40.20
A Sphere Moving
Through a Box
The sphere has

collided with the box...
... and moved

into the box
The amount of penetration can be determined by the GET OBJECT COLLISION

FIG-40.21
The GET OBJECT X
COLLISION Statement
GET OBJECT COLLISION Y ( )
real

In the diagram:
X,Y,Z Use one of these options to determine the

penetration of the object within the collision box
in the specified direction.
Normally, we would use all three options to get the full details of the collision:
REM *** Create a collision box round one object ***

.
.
REM *** IF collision get penetration details ***
x# = GET OBJECT COLLISION X()
y# = GET OBJECT COLLISION Y()
z# = GET OBJECT COLLISION Z()
By adding the penetration depth to the moving object's position, it will move back
outside the first object's collision box:
POSITION OBJECT 2,OBJECT POSITION X(2)+x#,

ÄOBJECT POSITION Y(2)+y#,OBJECT POSITION Z(2)+z#
The program in LISTING-40.5 demonstrates the final effect.

Using the GET OBJECT COLOR BACKDROP 0
COLLISIONS Data to BACKDROP ON
Modify the Path of an
Object
AUTOCAM OFF
POINT CAMERA 0,0,0

REM *** Create a bounding box taller than 3D box ***


DO
REM *** Move sphere 1 unit along z-axis ***
Ä OBJECT POSITION Y(2),OBJECT POSITION Z(2)+0.1
REM *** IF sphere collides with box ***
REM *** Calculate penetration values ...***
x#=GET OBJECT COLLISION X()
y#=GET OBJECT COLLISION Y()
z#=GET OBJECT COLLISION Z()
REM *** ... and use them to reset sphere's position ***
POSITION OBJECT 2,OBJECT POSITION X(2)+x#,
ÄOBJECT POSITION Y(2)+y#,OBJECT POSITION Z(2)+z#
ENDIF
LOOP
END

Activity 40.15
Type in and test the coding in LISTING-40.5 (collision05.dbpro).
What happens when the sphere comes into contact with the cuboid?
The effect becomes more interesting when the sphere is moving with an offset in
two or three axes. For example, we can make the sphere move in both the y and z
axes by changing the first line within the DO loop to read:

ÄOBJECT POSITION Y(2)-0.1,OBJECT POSITION Z(2)+0.1
Activity 40.16
Make the change suggested above to your last program. How does this affect
the movement of the sphere?
This effect is known as a sliding collision, for obvious reasons, with the sphere
appearing to slide along the surface of the cuboid until it's free to move along its
old trajectory.
The DELETE OBJECT COLLISION BOX Statement

The collision box associated with a 3D object can be deleted using the DELETE
OBJECT COLLISION BOX statement which has the format shown in FIG-40.22.
FIG-40.22
The DELETE OBJECT DELETE OBJECT COLLISION BOX objno
COLLISION BOX
Statement
In the diagram:

collision box is to be deleted.
Activity 40.17
In your previous program, delete the cuboid's bounding box. Does the sphere
halt because of collision?
The AUTOMATIC OBJECT COLLISION Statement

Another way of dealing with collisions is to use the AUTOMATIC OBJECT
COLLISION statement. Using this statement allows an object to automatically stop
moving when it collides with another object. When used, AUTOMATIC OBJECT
COLLISION always uses a bounding sphere for the object specified. The radius of
that sphere is set as part of this command. The statement has the format shown in
FIG-40.23.
FIG-40.23
The AUTOMATIC OBJECT COLLISION Statement
AUTOMATIC OBJECT COLLISION objno , radius , responseflag

In the diagram:

collision sphere is to be specified.

sphere to be used by the object.
responseflag is 0 or 1. If set to zero, the object halts as soon as

a collision is detected; if set to 1, the object will
move back to its previous position immediately
before the collision.
The program in LISTING-40.6 demonstrates the use of this statement.

Using Automatic Object COLOR BACKDROP 0
Collision BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0


REM *** Use automatic collision detection ***
AUTOMATIC OBJECT COLLISION 2,2.5,0
DO
SET CURSOR 100,100
PRINT OBJECT POSITION Z(2)
LOOP

END
Activity 40.18
The INTERSECT OBJECT Statement

Sometimes it is useful to know if one object will hit another object before the actual
collision occurs. For example, will a bolt of lightning from a wand hit the sphere
of death hurling towards our character (see FIG-40.24)?
FIG-40-24
A straight line
projected from the end
A Line Of Sight Test of the wand
How close does the

wand line come to the object?
sphere

To find out, we need some way of knowing that the direction in which the wand is
pointing intersects with the sphere. The command that achieves this is
INTERSECT OBJECT which has the format shown in FIG-40.25.
FIG-40.25
The INTERSECT
INTERSECT OBJECT ( objno ,
OBJECT Statement
x1 , y1 , z1 , x2 , y2 , z2 )
real
In the diagram:
objno is an integer value giving the target object.
x1,y1,z1 are a set of real values representing the starting point

of the projected line.
x2,y2,z2 are a set of real values representing the finishing point

of the projected line.
The statement returns the shortest distance between the target object and the line
specified. If a value of zero is returned, the line and object intersect.
For the wand and sphere example mentioned earlier, x1,y1,z1 would specify any
point on or near the front of the wand. While x2,y2,z2 would specify a point at the
end of the range of the lightning bolt. The line created by joining these two points
should be parallel to the body of the wand. objno would be the object number
assigned to the sphere. This example is shown in LISTING-40.7.
Intersection between an COLOR BACKDROP 0
Line and an Object BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0

REM *** Make and position wand ***

ZROTATE OBJECT 2,90
REM *** Projected line's start point is centre of wand ***

x1# = OBJECT POSITION X(2)
y1# = OBJECT POSITION Y(2)
z1# = OBJECT POSITION Z(2)

LISTING-40.7 REM *** The wand fires a distance of 20 units ***
(continued) REM *** so end point of line is 20 units along x-axis ***
x2# = x1# + 20
Intersection between an y2# = y1#
Line and an Object z2# = z1#
dist# = INTERSECT OBJECT(2,x1#,y1#,z1#,x2#,y2#,z2#)
DO
SET CURSOR 100,100
PRINT dist#
LOOP

END
Activity 40.19
Summary
l Use OBJECT HIT or OBJECT COLLISION to determine if two objects have
collided.
l Use SET OBJECT COLLISION to switch collision detection on or off for a

specific object.
l Use SET GLOBAL COLLISION to switch collision detection for every object
on or off.
l Collision detection uses a bounding box (also known as a collision box) or a

bounding (collision) sphere.
l Collision detection using boxes and spheres is fast but can be inaccurate.
l The default setting is to use bounding boxes for collision detection.
l Collision with the actual polygons that make up the object itself can be used to
give accurate collision detection, but requires much more complex calculations.
l Use SHOW OBJECT BOUNDS to show the bounding volumes of an object.
l Use HIDE OBJECT BOUNDS to hide the bounding volumes of an object.
l Use SET OBJECT COLLISION TO SPHERES to employ the bounding sphere

of an object rather than its bounding box when detecting collisions.
l Use SET OBJECT RADIUS to modify the size of an object's bounding sphere.
l Use OBJECT COLLISION RADIUS to find out the current radius of an object's
bounding sphere.
l Use SET OBJECT COLLISION TO BOXES to return to using the bounding

box of an object to detect collisions.
l Use SET OBJECT COLLISION TO POLYGONS to employ an object's own

polygons when detecting collisions.

l Use MAKE OBJECT COLLISION BOX to create your own bounding box for
an object.
l Use the data from GET OBJECT COLLISION to create a sliding collision effect.
l Use DELETE OBJECT COLLISION BOX to delete any bounding box

associated with a specified object.
l Use AUTOMATIC OBJECT COLLISION to have an object halt automatically

when it strikes another object.
l Use INTERSECT OBJECT to determine if a line comes into contact with a

specified object.

Static Collisions
Introduction
Some 3D objects in a game, such as buildings, walls, or trees will not move once
they have been created. This type of object will be loaded from an existing file and
become the scenery of a game. If moving objects are not to pass straight through
this scenery, then that scenery needs to have specifically created bounding boxes.
These collision boxes are known as static bounding (or collision) boxes since they
will never move. DarkBASIC Pro has a set of commands for dealing with just such
static collision boxes.
Creating and Using Static Collision Boxes

The MAKE STATIC COLLISION BOX Statement
We can create a static collision box using the MAKE STATIC COLLISION BOX
statement whose format is given in FIG-40.26.
FIG-40.26
The MAKE STATIC COLLISION
BOX Statement
MAKE STATIC COLLISION BOX x1 , y1 , z1 , x2 , y2 , z2
In the diagram:
x1,y1,z1 are a set of real values giving the coordinates of

the bottom-front, left corner of the collision box.
x2,y2,z2 are a set of values giving the coordinates of the

top-back, right corner of the collision box.
The GET STATIC COLLISION HIT Statement

The usual OBJECT HIT and OBJECT COLLISION statements don't work with
static collision boxes, so we need a new statement to detect a collision between an
object and a static collision box. This statement is GET STATIC COLLISION HIT
FIG-40.27
The GET STATIC GET STATIC COLLISION HIT
COLLISION HIT
Statement
x1 , y1 , z1 , x2 , y2 , z2
x3 , y3 , z3 , x4 , y4 , z4

In the diagram:
x1,y1,z1 area set of real values giving the coordinates of

the bottom-front, left corner of the collision box.

x3,y3,z3 are a set of real values giving the coordinates of the

bottom-front, left corner of the collision box.

Notice that the statement makes no reference to an object - so how do we use it to

detect when an object hits a static collision box?
Imagine a sphere travelling towards a static collision box as shown in FIG-40.28.

To check if the sphere hits the box we need to specify two rectangular volumes; one
for a position where the sphere had been a moment before, and one corresponding
to the sphere's current position.
FIG-40.28
Current position
Using GET STATIC of sphere’s
COLLISION HIT bounding box
Previous position
of sphere’s x4,y4,z4
bounding box x2,y2,z2
Static collision
box
Notice that the trailing x1,y1,z1

bounding box is 3 units
x3,y3,z3
behind the leading box.
The GET STATIC COLLISION HIT statement returns 1 if the leading box (that is,
the box defined using x3,y3,z3,x4,y4,z4) intersects any part of the static collision
box. If there is no intersection, zero is returned.
The program in LISTING-40.8 creates an invisible force-field, (that is, a static

collision box). When a travelling sphere hits the force field, a string of asterisks is
displayed on the screen. The line of asterisks grows with each hit detected. The ball
is not deflected from its path.

Hitting a Static Collision COLOR BACKDROP 0
Box BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Make static collision box ***

MAKE STATIC COLLISION BOX -20,0,-1,20,40,1

LISTING-40.8 REM *** Make and position sphere ***
Hitting a Static Collision
Box REM *** Record x and y ordinates of the old and new collision ***
REM *** boxes for the sphere. These don't change since the ***
REM *** sphere is moving in the z direction only ***
x1# = OBJECT POSITION X(2)-2.5
x2# = x1#+5
x3# = x1#
x4# = x2#
y1# = OBJECT POSITION Y(2)-2.5
y2# = y1#+5
y3# = y1#
y4# = y2#
REM *** Initialise string of asterisks ***

mess$ = ""
REM *** Main loop ***

DO
REM *** Move sphere ***
REM *** Get z ord of old & new bounding boxes for sphere ***
z1# = OBJECT POSITION Z(2)-3.5
z2# = z1# + 5
z3# = z1# + 1
z4# = z2# + 1
REM *** Check bounding box hit with static collision box ***
IF GET STATIC COLLISION HIT(x1#,y1#,z1#,x2#,y2#,z2#,x3#,y3#,z3#
Ä,x4#,y4#,z4#)
REM *** Add to number of asterisks ***
mess$=mess$ + "*"
ENDIF
REM *** Display message ***
SET CURSOR 100,100
PRINT mess$
LOOP

END
Activity 40.20
Type in and test the program in LISTING-40.8 (collision08.dbpro).
The GET STATIC COLLISION Statement

When the GET STATIC COLLISION HIT statement is used, the extent to which
the specified bounding cube has penetrated the static collision box is recorded. This
information can be retrieved using the GET STATIC COLLISION statement which
FIG-40.29 X
The GET STATIC
COLLISION Statement GET STATIC COLLISION Y ( )
real

In the diagram:

penetration of the bounding box within the static
collision box in the specified direction.
Normally, we would use all three options to get the full details of the collision:
x# = GET STATIC COLLISION X()

y# = GET STATIC COLLISION Y()
z# = GET STATIC COLLISION Z()
The next program (LISTING-40.9) is a modification of the last one. This time the
asterisks are replaced by the values returned by GET STATIC COLLISION
statements. Also, the display uses the SYNC statement so that we can pause the
program with a WAIT KEY statement and still see the values output by the PRINT
statement.

Understanding the Value COLOR BACKDROP 0
Returned by GET BACKDROP ON
STATIC COLLISION REM *** Use manual screen updating ***
SYNC ON

AUTOCAM OFF
POINT CAMERA 0,0,0

MAKE STATIC COLLISION BOX -20,0,-1,20,40,1

REM *** Record x and y of old and new bounding boxes ***
x2# = x1#+5
x3# = x1#
x4# = x2#
y2# = y1#+5
y3# = y1#
y4# = y2#

DO
REM *** Calculate z ordinate for old and new boxes ***
z2# = z1# + 5
z3# = z1# + 3
z4# = z2# + 3
REM *** Check for hit with static collision box ***
IF GET STATIC COLLISION HIT(x1#,y1#,z1#,x2#,y2#,z2#,x3#
Ä
,y3#,z3#,x4#,y4#,z4#)
REM *** Get collision details ***

LISTING-40.9
(continued) REM *** Display these details ***
SET CURSOR 100,100
Understanding the Value PRINT x#," ",y#," ",z#, " ",z1#," ",z2#," ",z3#,
Returned by GET
STATIC COLLISION
Ä
" ",z4#
REM *** Update screen ***
SYNC
REM *** Wait for key press ***
WAIT KEY
ENDIF
SYNC
LOOP

END
Activity 40.21
Type in and test the program in LISTING-40.9 (collision09.dbpro), observing

the figures displayed.
As we can see from the figures, as the leading bounding box enters the collision
area the value returned by GET STATIC COLLISION Z is determined by the result
of the calculation
leading bounding box's z ordinate (from the leading edge) -

static collision box's z ordinate on the entry side
which is simply:
z4 - 1
Activity 40.22
Change the first z value of the static collision box from -1 to -2 (thereby
making the static collision box 3 units thick) and run the program again.
Notice that the last value displayed by GET STATIC COLLISION Z is -0.5. Once
the mid-point of the trailing bounding box is passed the mid-point of the static
collision box, the formula for the value returned by GET STATIC COLLISION Z
changes to:
static collision box's z ordinate on the entry side -

leading bounding box's trailing edge z-ordinate
In this case that is

-2 - z3
Activity 40.23
Check that the formula still holds by
a) Changing the static collision box's z ordinates from -2 and 1 to -5 and 10.
b) Making the sphere enter from the other side of the static collision box
(that is, start the sphere at position (0,10,100) and give it a speed of -1).

Similarly, if the sphere moves parallel to the x-axis, then it is the GET STATIC
COLLISION X statement which will return a non-zero value based on the bounding
box's leading edge's x value. Approaching parallel to the y-axis, GET STATIC
COLLISION Y returns a non-zero value.
So how can we make use of the value returned by the GET STATIC COLLISION
values in our program? Well, if we use the value returned by the statement to modify
the position of our moving object, we can stop that object moving through the static
collision box area.
Activity 40.24
Modify the DO loop in your previous program to read
DO
REM *** Calculate z ordinate for old and new boxes ***
z2# = z1# + 5
z3# = z1# + 3
z4# = z2# + 3
REM *** Check for collision with static collision box ***
IF GET STATIC COLLISION HIT(x1#,y1#,z1#,x2#,y2#,z2#,
Äx3#,y3#,z3#,x4#,y4#,z4#)
REM *** Get penetration details ***
REM *** Move sphere back out of static collision box ***
POSITION OBJECT 2,0,10,OBJECT POSITION Z(2) - z#
ENDIF
SYNC
LOOP
Position the sphere at location (0,10,-100).
Run the program. What happens when the sphere hits the static box?
As you have seen, we can use the GET STATIC COLLISION value to force the
sphere to stop when it hits the static collision box.
When the moving object contains movement along a second axis, then the overall
effect is to slide the moving object along the surface of the static collision box.
We see an example of this in LISTING-40.10 where the sphere's movement has

both a y and z component. As the sphere hits the static collision box, it slides down
the side of that box until it passes underneath and then continues on its way.
Using the Value COLOR BACKDROP 0
Returned by GET BACKDROP ON
STATIC COLLISION REM *** Use manual screen updating ***
SYNC ON

LISTING-40.10 REM *** Position camera ***
(continued) AUTOCAM OFF
Using the Value POINT CAMERA 0,0,0
Returned by GET REM *** Make static collision box ***
STATIC COLLISION MAKE STATIC COLLISION BOX -20,0,-5,20,40,10

REM *** Record x for old and new bounding boxes for sphere ***
x2# = x1#+5
x3# = x1#
x4# = x2#

DO
POSITION OBJECT 2,0,OBJECT POSITION Y(2)-0.05,
Ä
OBJECT POSITION Z(2)+1
REM *** Calculate y and z ordinate for old and new boxes ***
y2# = y1#+5
y3# = y1#
y4# = y2#
z2# = z1# + 5
z3# = z1# + 3
z4# = z2# + 3
REM *** Check collision with static collision box ***
IF GET STATIC COLLISION HIT(x1#,y1#,z1#,x2#,y2#,z2#,x3#,y3#,z3#
Ä
,x4#,y4#,z4#)=1
REM *** Get penetration details ***
REM *** Move sphere back out of static collision box ***
POSITION OBJECT 2,0,OBJECT POSITION Y(2) - y#,
Ä
OBJECT POSITION Z(2) - z#
REM *** Display message ***
ENDIF
SYNC
LOOP
END
Activity 40.25
As a general rule, since we may not know through which axes an object is moving,
the values x1# to z4# should be calculated on every iteration of the main loop and
the repositioning after a hit should include all three penetration figures:
POSITION OBJECT 2,OBJECT POSITION X()-x#,

ÄOBJECT POSITION Y(2)-y#,OBJECT POSITION Z(2)-z#
The STATIC LINE OF SIGHT Statement

When we look down the barrel of a gun, we will want to know if we are going to
hit our intended target before we actually fire. DarkBASIC Pro can answer this type

of question, determining if a given trajectory will intersect with a static collision
box, using the STATIC LINE OF SIGHT statement which has the format shown in
FIG-40.30.
FIG-40.30 STATIC LINE OF SIGHT (

The STATIC LINE OF
SIGHT Statement
x1 , y1 , z1 , x2 , y2 , z2
, width , latitude )
integer
In the diagram:
x1,y1,z1 are a set of real values giving the start point of a

line.
x2,y2,z2 are a set of real values giving the end point of a line.
width is a real number giving the width of the line.
latitude is a real number affecting the degree of accuracy

of the calculation performed by this statement. A
large number represents greater degree of latitude
in the calculation.
This statement returns 1 if the line starting at point (x1,y1,z1) and passing through
the point (x2,y2,z2) eventually hits an existing static collision box; otherwise zero
is returned. In performing the calculation the line is assumed to be in the form of a
cylinder with diameter width.
Assuming a weapon's centre is at point (-40,0,0) and that the weapon has been
rotated by 90o about the z-axis, with one point along the line of sight being (-5,0,0),
then we can check if this line of sight hits a static collision box using the line:
IF STATIC LINE OF SIGHT (-40,0,0,-5,0,0.5,1,1) = 1
In this example, the line has been given a thickness of 1 and the latitude setting is
also set to 1 (thereby specifying a high degree of accuracy in the calculation).
The program in LISTING-40.11 demonstrates the use of this statement.

Using the STATIC LINE COLOR BACKDROP 0
OF SIGHT Statement BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0

LISTING-40.11 REM *** Make static collision box and visible box ***
(continued) MAKE STATIC COLLISION BOX -20,-20,-20,20,20,20
Using the STATIC LINE
OF SIGHT Statement REM *** Make and position wand ***
ZROTATE OBJECT 2,90
REM *** Discover if wand is pointing at static box ***

REM *** using two points along line of wand ***
ans = STATIC LINE OF SIGHT (-40,0,0,-5,0,0,0,1)
REM *** Display result ***

DO
set cursor 100,100
PRINT ans
LOOP

END
Activity 40.26
The STATIC LINE OF SIGHT Coordinates Statement

We can discover the approximate coordinates of the point of contact between a line
of sight and a static collision box using the STATIC LINE OF SIGHT coordinates
FIG-40.31
X
The STATIC LINE OF
SIGHT Coordinates
Statement STATIC LINE OF SIGHT Y ( )
real
In the diagram:

ordinate of the point of contact with the static
collision box for the required axis.
For example, we could record the point of contact between the line of sight and a
static collision box with the lines:
IF STATIC LINE OF SIGHT (-40,0,0,-5,-0.2,0.5,0,1)= 1

x# = STATIC LINE OF SIGHT X()
y# = STATIC LINE OF SIGHT Y()
z# = STATIC LINE OF SIGHT Z()

Activity 40.27
Modify your previous program to display the point of intersection between the
light of sight and static collision box.
Static Collision Boxes and the Camera

Back in Chapter 33 we had a look at the SET CAMERA TO FOLLOW statement.
The last parameter to this statement is a flag which determines how the camera
reacts to collision with a static collision box. Set to 0, the camera will ignore any
static collision boxes; set to 1, the camera should slide past such boxes.
The program in LISTING-40.12 creates a camera and a static collision box. The
camera is then moved to a new position using the SET CAMERA TO FOLLOW
statement and this requires the camera to come into contact with the collision box.
The SET CAMERA TO COLOR BACKDROP RGB(255,255,100)
FOLLOW Statement and BACKDROP ON
Static Collision Boxes
AUTOCAM OFF
POINT CAMERA 0,0,0

MAKE STATIC COLLISION BOX -20,-20,-3,20,20,3
REM *** Make visible object over collision box ***

REM *** Try to move camera through collision box ***

DO
SET CAMERA TO FOLLOW 0,20,15,5,2,2,100,0
WAIT 50
LOOP

END
Activity 40.28
Type in and run the program given above (collision12.dbpro).
Does the camera pass through the box?
Change the final value of the SET CAMERA TO FOLLOW statement to 1

and re-run the program.
How does the camera react this time?
Summary
l Static bounding (collision) boxes are used to create a detectable collision volume
around fixed areas.
l Use MAKE STATIC COLLISION BOX to create a static collision box.

l Use GET STATIC COLLISION HIT to detect collision with a static bounding
box.
l Use GET STATIC COLLISION to determine the penetration of an object into

a static collision box.
l Data from GET STATIC COLLISION can be used to create a sliding collision
of an object against a static bounding box.
l Use STATIC LINE OF SIGHT to determine if an imaginary line will pass

through a static bounding box.
l Use STATIC LINE OF SIGHT coordinates to determine a line of sight's contact

point with a static bounding box.
l The SET CAMERA TO FOLLOW statement should automatically slide over

any static collision box when the final parameter is set to 1.

Solutions CONTROL CAMERA USING ARROWKEYS 0, 0.1,
Activity 40.1
Ä 0.5
The sphere passes straight through the cube. LOOP
END
Activity 40.2
REM *** Setup screen *** Activity 40.5
COLOR BACKDROP 0 REM *** Setup screen ***
BACKDROP ON SET DISPLAY MODE 1280, 1024, 32
REM *** Position camera *** COLOR BACKDROP 0
AUTOCAM OFF BACKDROP ON
POSITION CAMERA 100,10,-200 REM *** Position camera ***
POINT CAMERA 0,0,0 AUTOCAM OFF
REM *** Make two objects *** POSITION CAMERA 100,10,-200
MAKE OBJECT CUBE 1, 40 POINT CAMERA 0,0,0
MAKE OBJECT SPHERE 2, 5 REM *** Make two objects ***
POSITION OBJECT 2, 0,0,-100 MAKE OBJECT CUBE 1, 40
REM *** Move sphere towards cube *** MAKE OBJECT SPHERE 2, 5
DO POSITION OBJECT 2, 0,0,-100
POSITION OBJECT 2,0,0, REM *** Switch off cube's col'sn detect ***
OBJECT POSITION Z(2)+1 SET OBJECT COLLISION OFF 1
REM *** Check for collision *** REM *** Switch on global collisions ***
IF OBJECT HIT(1,2) SET GLOBAL COLLISION ON
EXIT REM *** Move sphere towards cube ***
ENDIF DO
POSITION OBJECT 2,0,0,
Ä
LOOP
REM *** Give user camera control *** OBJECT POSITION Z(2)+1
DO IF OBJECT COLLISION(1,2)
CONTROL CAMERA USING ARROWKEYS 0, 0.1,0.5 EXIT
LOOP ENDIF
REM *** End program *** LOOP
END DO
CONTROL CAMERA USING ARROWKEYS 0, 0.1,
Ä 0.5
Activity 40.3 LOOP
The first DO..LOOP structure should be changed to read: END
DO
POSITION OBJECT 2,0,0, Activity 40.6
Ä
OBJECT POSITION Z(2)+0.1
REM *** Setup screen ***
EXIT
COLOR BACKDROP 0
ENDIF
BACKDROP ON
LOOP
AUTOCAM OFF
Activity 40.4 POINT CAMERA 0,0,0
REM *** Setup screen *** REM *** Make two objects ***
SET DISPLAY MODE 1280, 1024, 32 MAKE OBJECT CUBE 1, 40
COLOR BACKDROP 0 MAKE OBJECT SPHERE 2, 5
BACKDROP ON POSITION OBJECT 2, 0,0,-100
REM *** Position camera *** REM *** Show bounding boxes ***
AUTOCAM OFF SHOW OBJECT BOUNDS 1,1
POSITION CAMERA 100,10,-200 SHOW OBJECT BOUNDS 2,1
POINT CAMERA 0,0,0 REM *** Move sphere towards cube ***
REM *** Make two objects *** DO
MAKE OBJECT SPHERE 2, 5 Ä OBJECT POSITION Z(2)+1
POSITION OBJECT 2, 0,0,-100 IF OBJECT COLLISION(1,2)
REM *** Switch off cube's col'sn detect *** EXIT
SET OBJECT COLLISION OFF 1 ENDIF
REM *** Move sphere towards cube *** LOOP
DO DO
CONTROL CAMERA USING ARROWKEYS 0, 0.1,
Ä
OBJECT POSITION Z(2)+1 Ä 0.5
IF OBJECT COLLISION(1,2) LOOP
EXIT REM *** End program ***
ENDIF END
LOOP
DO

Activity 40.7 to
The objects should now be created using the lines: POSITION OBJECT 2,0,40,-100
REM *** Make two objects *** Change

MAKE OBJECT SPHERE 2,5 POSITION OBJECT 2,0,20,
ÄOBJECT POSITION Z(2)+1
A collision is detected long before the sphere actually
collides with the cone. to

Activity 40.8 ÄOBJECT POSITION Z(2)+1
To hide the bounds the following changes could be made:

The cube flies past the box without colliding.
REM *** Display bounding volumes ***
SHOW OBJECT BOUNDS 1,1 The version of the program needed to increase the
SHOW OBJECT BOUNDS 2,1 bounding sphere is given below:
REM *** Hide bounding boxes ***
WAIT KEY REM *** Setup screen ***
HIDE OBJECT BOUNDS 1 SET DISPLAY MODE 1280,1024,32
HIDE OBJECT BOUNDS 2 COLOR BACKDROP 0
BACKDROP ON
Activity 40.9 AUTOCAM OFF
No solution required. POINT CAMERA 0,0,0
Activity 40.10 REM *** Display its bounding volumes ***
REM *** Setup screen *** REM *** Lie box on its side ***
SET DISPLAY MODE 1280,1024,32 ZROTATE OBJECT 1, 90
COLOR BACKDROP 0 REM *** Create a sphere ***
BACKDROP ON MAKE OBJECT SPHERE 2, 5
REM *** Position camera *** SHOW OBJECT BOUNDS 2
AUTOCAM OFF REM *** Position sphere ***
POSITION CAMERA 100,10,-200 POSITION OBJECT 2,0,40,-100
POINT CAMERA 0,0,0 REM *** Use Box's bounding sphere ***
REM *** Make box *** SET OBJECT COLLISION TO SPHERES 1
MAKE OBJECT BOX 1,10,60,10 REM *** Change bounding sphere radius ***
REM *** Display its bounding volumes *** SET OBJECT RADIUS 1,50
SHOW OBJECT BOUNDS 1 REM *** Move sphere towards box ***
REM *** Lie box on its side *** DO
ZROTATE OBJECT 1, 90 POSITION OBJECT 2,0,40,
REM *** Create a sphere *** Ä
OBJECT POSITION Z(2)+1
MAKE OBJECT SPHERE 2, 5 IF OBJECT HIT(1,2)
SHOW OBJECT BOUNDS 2 EXIT
REM *** Position sphere *** ENDIF
POSITION OBJECT 2,0,20,-100 LOOP
REM *** Use Box's bounding sphere *** REM *** End program ***
SET OBJECT COLLISION TO SPHERES 1 WAIT KEY
REM *** Move sphere towards box *** END
DO
ÄOBJECT POSITION Z(2)+1 Activity 40.12
IF OBJECT HIT(1,2)
EXIT Neither method gives accurate collision detection.
ENDIF
LOOP
WAIT KEY
END The sphere correctly collides with the cone.
The 3D sphere stops when it collides with the cuboid's

bounding sphere. Activity 40.14
The sphere stops as it hits the extended collision box
Activity 40.11 above the cuboid.
Two lines need to be changed to lift the sphere to 40 units. Rotating the cuboid does not rotate the collision box and
Change therefore a collision still occurs.
POSITION OBJECT 2,0,20,-100 By changing the last value in the MAKE OBJECT

COLLISION BOX to 1, the collision box rotates along REM *** Position camera ***
with the cuboid and so the sphere passes without a AUTOCAM OFF
collision occurring. POSITION CAMERA 0,10,-100
POINT CAMERA 0,0,0
REM *** Make static collision box
REM *** and visible box ***
Activity 40.15 MAKE STATIC COLLISION BOX
The sphere halts when it collides with the cuboid. Ä-20,-20,-20,20,20,20
REM *** Make and position wand ***
Activity 40.16 SCALE OBJECT 2, 10,200,10
The sphere slides down the side of the cuboid until it
ZROTATE OBJECT 2,90
reaches the bottom and then continues on its path. REM *** If line of sight collision ***
REM *** get coordinates ***
IF STATIC LINE OF SIGHT
Activity 40.17 Ä(-40,0,0,-5,0,0,0,1) = 1
x# = STATIC LINE OF SIGHT X()
Without the bounding box, the sphere should travel y# = STATIC LINE OF SIGHT Y()
straight through the cuboid but it appears that DELETE z# = STATIC LINE OF SIGHT Z()
OBJECT COLLISION BOX does not operate correctly ENDIF
and the collision box remains (DBPro v 1.062). REM *** Display result ***
DO
SET CURSOR 100,100
Activity 40.18 PRINT "Point of intersection : ",x#," "
Ä,y#," ",z#
No solution required. LOOP
END
Activity 40.19
No solution required. Activity 40.28
The camera passes straight through the box (and collision
Activity 40.20 box).
No solution required. The camera STILL seems to pass through the collision
box!
Activity 40.21
Activity 40.22
Activity 40.23
The formula still holds in both cases.
Activity 40.24
The sphere halts when the static collision box is
encountered.
Activity 40.25
Activity 40.26
Activity 40.27
COLOR BACKDROP 0
BACKDROP ON

41
Fire Particles
Fountain Particles
Snow Particles
Setting Particle Characteristics
Using Particles
DarkBASIC Pro: Particles 1101

Particles
Introduction
If you've ever seen a sparkler in action on Guy Fawkes night (or whatever day your
country uses as an excuse to let us play with fireworks) you'll be able to imagine
the effect of the particle instructions in DarkBASIC Pro. Once you've created a
particles object it continues to create a sparkler effect until it is destroyed. A
snapshot of the basic effect is shown in FIG-41.1.
FIG-41.1
Particles in Action
The particles shoot upwards and then fall back to an imaginary ground area before
eventually fading.
There are various characteristics that can be set, such as the quantity, speed, and
duration of the particles. It will be useful if we think of a particle object as having
two components: the emitter, and the particles themselves. The particles all
originate from the emitter, which is at the centre of the particle display.
Creating Particles
The MAKE PARTICLES Statement
To create a stream of particles we'll need an image. This image is used on each
particle created, but seems to have only a minor visual effect on what you'll see on
the screen.
With an image loaded, you can now use the MAKE PARTICLES statement to create
a continual stream of particles which originate from a single point (called the
emitter), fly upwards and then fall back down before finally disappearing. The
FIG-41.2
The MAKE PARTICLES
MAKE PARTICLES partno , imgno , partflag , size
Statement
1102 DarkBASIC Pro: Particles

In the diagram:
partno is an integer value specifying the ID assigned to

this particle object.
imgno is an integer value previously assigned to a

loaded image. The image has a minor effect on
the appearance of the particles.
partflag is an integer value which affects the quantity

of particles produced. Use zero to produce no
particles; any other positive integer produces
particles.
size is a real number specifying how far out the

particles will fly from their origin and also affects
the size of the particles.
The particles object's emitter is automatically positioned at (0,0,0), but may be

moved after it has been created.
The program in LISTING-41.1 demonstrates the use of the MAKE PARTICLE

statement.
Creating a Particles COLOR BACKDROP 0
Object BACKDROP ON

AUTOCAM OFF
POSITION CAMERA 0,0,5,-10
REM *** Load image for particles ***

LOAD IMAGE "green.bmp",1
REM *** Create a particles object ***

MAKE PARTICLES 1,1,1,1

WAIT KEY
END
Activity 41.1
Type in and test the program given above (particles01.dbpro).
Change the value of size to 10 and see how this affects the particles.
The HIDE PARTICLES Statement

If we want to hide existing particles, then we can use the HIDE PARTICLES
FIG-41.3
The HIDE PARTICLES HIDE PARTICLES partno
Statement
In the diagram:

partno is the integer value specifying the ID of the
particles object to be hidden.
Activity 41.2
Modify particles01.dbpro so that the particle object disappears when a key is

pressed.
The SHOW PARTICLES Statement

If you want the particles object to reappear after being hidden, then use the SHOW
PARTICLES statement which has the format shown in FIG-41.4.
FIG-41.4
The SHOW SHOW PARTICLES partno
PARTICLES Statement
In the diagram:
partno is the integer value identifying the previously

hidden particles object which is to reappear.
The DELETE PARTICLES Statement

When you've no more need for a particles object, use the DELETE PARTICLES
statement which will free up the RAM space reserved for the object. The statement
FIG-41.5
The DELETE DELETE PARTICLES partno
PARTICLES Statement
In the diagram:
partno is the integer value identifying the particles object

which is to be deleted.
The POSITION PARTICLES Statement

Like any other 3D object, the particles object's emitter is initially placed at location
(0,0,0). To move the particles object to a different position in 3D space, use the
POSITION PARTICLES statement which takes the format shown in FIG-41.6.
FIG-41.6
The POSITION
POSITION PARTICLES partno , x , y , z
PARTICLES Statement
In the diagram:
partno is the integer value identifying the particles object

which is to be repositioned.
x,y,z are the coordinates at which the particles object's

emitter is to be placed.

This statement moves the whole particles object's emitter and currently falling
particles to a new location. The program in LISTING-41.2 demonstrates this effect.
Repositioning the COLOR BACKDROP 0
Particles BACKDROP ON

AUTOCAM OFF

LOAD IMAGE "yellow.bmp",1
REM *** Create particles object ***

REM *** Move position of particles object several times ***

post#= -10
FOR c = 1 TO 200
POSITION PARTICLES 1,post#,0,0
post# = post# + 0.1
WAIT 10
NEXT c

WAIT KEY
END
Activity 41.3
Modify the code so that the particles object is moved up the screen rather than
from left to right.
The POSITION PARTICLE EMISSIONS Statement

Whereas the POSITION PARTICLES statement moves the complete particles
object, the POSITION PARTICLE EMISSIONS statement moves only the emitter
point and leaves any existing particles in their original position, still falling. New
particles are also generated from the new emitter position. This statement has the
FIG-41.7
The POSITION POSITION PARTICLE EMISSIONS partno , x , y , z
PARTICLE
EMISSIONS Statement
In the diagram:
partno is the integer value previously assigned to the

particles object.
x,y,z are the coordinates at which the particles object's

emitter is to be placed.
The program in LISTING-41.3 only differs from the previous program by a single
line of code, but the effect on the screen is quite different.

Moving the Particles COLOR BACKDROP 0
Emitter BACKDROP ON

AUTOCAM OFF


REM *** Move position of particles object several times ***

post#= -10
FOR c = 1 TO 200
POSITION PARTICLE EMISSIONS 1,post#,0,0
post# = post# + 0.1
WAIT 10
NEXT c

WAIT KEY
END
Activity 41.4
Modify your previous program to match the code given above and check out
the difference between using POSITION PARTICLES and POSITION
PARTICLE EMISSIONS.
The ROTATE PARTICLES Statement

The emitter can be rotated about its position using the ROTATE PARTICLES
statement. This has the effect of changing the direction in which the particles are
fired and the plane in which they bounce. The format of the ROTATE PARTICLE
statement is shown in FIG-41.8.
FIG-41.8
The ROTATE ROTATE PARTICLES partno , , ,z
xangle
y yangle , zangle
PARTICLES
Statement
In the diagram:

particles object.
xangle is a real value giving the angle (in degrees)

to which the emitter is rotated about the x-axis.
yangle is a real value giving the angle (in degrees)

to which the emitter is rotated about the y-axis.
zangle is a real value giving the angle (in degrees)

to which the emitter is rotated about the z-axis.
The effects of various ROTATE PARTICLES statements are shown in FIG-41.9.

FIG-41.9 The Effects of Rotating the Particle Emitter
ROTATE PARTICLES 1,90,0,0 ROTATE PARTICLES 1,0,90,0 ROTATE PARTICLES 1,0,0,180
The emitter is now facing the camera There is no apparent change to The particles now shoot downwards
and the particles shoot towards the the display from its default setting. and return upwards to the surface.
viewer.
It's easier to get your head round this if you think of the whole particles effect as
creating a cone-shaped display, with the particles rising straight up through the
centre of the cone and then falling back round its base perimeter. Now we can show
the effects of each rotation in a more abstract way (see FIG-41.10).
FIG-41.10 How the Emitter is Rotated
y-axis y-axis y-axis y-axis

z-axis z-axis z-axis z-axis
x-axis x-axis x-axis x-axis

o o o
Normal position Rotated 90 about x-axis Rotated 90 about y-axis Rotated 180 about z-axis
The COLOR PARTICLES Statement

You can modify the colour of the particles emitted using the COLOR PARTICLES
FIG-41.11
the COLOR PARTICLES
COLOR PARTICLES partno , red
y ,z green , blue
Statement
In the diagram:

particles object.
red is an integer value specifying the intensity of the

red component in the particles.
green is an integer value specifying the intensity of the

green component in the particles.
blue is an integer value specifying the intensity of the

blue component in the particles
The particles are actually set to a colour determined by the colour in the original
image plus the colour specified here. By using a pure white image when setting up
the particles object, the COLOR PARTICLES statement causes the particles to

match the colour stated exactly.
To set the particles to yellow we could use the statement:
COLOR PARTICLES 1, 255,255,0
The program in LISTING-41.4 randomly changes the colour of the particles every
second for 10 seconds.

COLOR BACKDROP 0
Changing Particle Colour BACKDROP ON

AUTOCAM OFF
POSITION CAMERA 0, 0,0,-10
REM *** Set up particle object ***

LOAD IMAGE "white.bmp",1
REM *** Seed randomizer ***

RANDOMIZE TIMER()
REM *** Change colour 10 times ***

FOR c = 1 TO 10
COLOR PARTICLES 1, RND(255),RND(255),RND(255)
WAIT 1000
NEXT c
WAIT KEY
END
Activity 41.5
The SET PARTICLE EMISSIONS Statement

The quantity of particles produced by the emitter can be modified using the SET
PARTICLE EMISSIONS statement which has the format shown in FIG-41.12.
FIG-41.12
The SET PARTICLE
SET PARTICLE EMISSIONS partno , value
EMISSIONS Statement
In the diagram:

particles object.
value is an integer value which modifies the quantity

of particles produced by the emitter.
This statement has no effect on the height or distance over which the particles travel.
A typical statement might be:
SET PARTICLE EMISSIONS 1,12

gradually increasing the particle emissions at one second intervals.

The Effects of Particle COLOR BACKDROP 0
Emissions BACKDROP ON

AUTOCAM OFF
REM *** Set up particle object ***

REM *** Increase emissions over a 10 second interval ***

FOR c = 1 TO 10
SET PARTICLE EMISSIONS 1, c
WAIT 1000
NEXT c

WAIT KEY
END
Activity 41.6
Modify your last program to match that given above and observe the effects
produced.
The SET PARTICLE VELOCITY Statement

The velocity of particles produced by the emitter can be modified using the SET
PARTICLE VELOCITY statement which has the format shown in FIG-41.13.
FIG-41.13
The SET PARTICLE
SET PARTICLE VELOCITY partno , value
VELOCITY Statement
In the diagram:

particles object.
value is an integer value which modifies the velocity

of particles produced by the emitter.
This statement affects the height and distance over which the particles travel.
SET PARTICLE VELOCITY 1,5
Activity 41.7
Replace the SET PARTICLE EMISSIONS statement in your previous

program with the equivalent SET PARTICLE VELOCITY statement and
observe the effect this has on the particles produced.

The SET PARTICLE GRAVITY Statement
In all the examples you've seen so far, particles fly upwards and are then pulled back
to some invisible ground level. This pull, or gravity, can be modified using the SET
PARTICLE GRAVITY statement which has the format shown in FIG-41.14.
FIG-41.14
The SET PARTICLE
SET PARTICLE GRAVITY partno , pull
GRAVITY Statement
In the diagram:

particles object.
pull is an integer value which modifies the pull on

the particles produced by the emitter. A negative
value can be used and this stops the particles
falling back towards the emitter.
When a particle object is first created it has a default gravity setting of 5.
This statement affects the height and distance over which the particles travel.
A typical example of this statement might be:
SET PARTICLE GRAVITY 1, 5
Activity 41.8
Make the following modifications to your last program:
Make the FOR loop start at -5 and finish at 10.
Replace the SET PARTICLE VELOCITY statement with the equivalent SET
PARTICLE GRAVITY statement. Test the resulting program.
The SET PARTICLE CHAOS Statement

Although the particles produced from the emitter seem to follow a reasonably
predictable path, there is an element of chaotic movement in that path. Just how
erractically the particles move can be set using the SET PARTICLE CHAOS
FIG-41.15
The SET PARTICLE
SET PARTICLE CHAOS partno , factor
CHAOS Statement
In the diagram:

particles object.
factor is an integer value which modifies the chaos

factor of the particles produced by the emitter.

When a particles object is first created, it has a default chaos setting of zero.
SET PARTICLE CHAOS 1, 10
Activity 41.9
In your last program, change the FOR loop to start at -50 and finish at 50.
Replace the SET PARTICLE GRAVITY statement with the equivalent SET
PARTICLE CHAOS statement.
Reduce the wait time for each iteration from 1000 to 200.
Test the resulting program.
The SET PARTICLE SPEED Statement

A movie camera takes a series of still photographs on a roll of film. Professional
cameras take 25 pictures (known as frames) every second, so about 1/25 of a second
passes between one frame being taken and the next. However, some types of movies
need a different approach. If we want to film something that happens very slowly,
say a flower bud opening, we need to take only one frame of film every minute, but
play it back at the normal 25 frames per second.
When handling particles, it is possible to specify how much time has passed between
each screen frame using the SET PARTICLES SPEED statement. This will affect
the apparent speed of the particles and the number of particles appearing, since
some will be created and destroyed in the time between two frames.
This statement takes the format shown in FIG-41.16.

FIG-41.16
The SET PARTICLE
SET PARTICLE SPEED partno , value
SPEED Statement
In the diagram:

particles object.
value is a real value which specifies the seconds that

have elapsed between frames.
The default setting for the gap between frames is 0.005 seconds.
SET PARTICLE SPEED 1, 0.01

gradually increasing the particles' speed over several seconds.

Changing Particle Speed COLOR BACKDROP 0
BACKDROP ON

AUTOCAM OFF
REM *** Set up particles object ***

SET PARTICLE VELOCITY 1,0.01
REM *** Increase particles' speed over a 5 second period ***

WAIT KEY
speed#=0.01
FOR c = 1 TO 10
SET PARTICLE SPEED 1,speed#
speed# = speed# + 0.005
WAIT 500
NEXT c

WAIT KEY
END
Activity 41.10
Type in and test the program given in LISTING-41.6 (particles04.dbpro).
The SET PARTICLE FLOOR Statement

In all the examples you've seen so far, the particles rise and then fall back onto some
invisible ground level. This floor can be removed so that the particles continue to
fall until they fade. To do this we use the SET PARTICLE FLOOR statement which
takes the format shown in FIG-41.17.
FIG-41.17
The SET PARTICLE
SET PARTICLE FLOOR partno , floorflag
FLOOR Statement
In the diagram:

particles object.
floorflag use the value 1 if a floor is required; 0 if no floor

is required.
The initial default setting is 1.
This statement effects the height and distance over which the particles travel.
SET PARTICLE FLOOR 1,0

Activity 41.11
Write a program (act4111.dbpro) which creates a particles object, removing

its floor when any key is pressed.
Test the resulting program.
The SET PARTICLE LIFE Statement

After a particle has hit the floor and bounced, it fades away. The time between being
created and fading is known as the lifetime of the particle. This lifetime can be
modified using the SET PARTICLE LIFE statement which has the format shown
in FIG-41.18.
FIG-41.18
SET PARTICLE LIFE partno , perc
The SET PARTICLE
LIFE Statement
In the diagram:

particles object.
perc is an integer value representing the lifetime of

particles as a percentage of their normal duration.
Since the normal lifespan of a particle is 100%, we could halve this duration using
the statement
SET PARTICLE LIFE 1, 50
Activity 41.12
In your last program, add a particle life of 20%. This statement should be
added immediately after the particles object is created.
Now, test the program.
The GHOST PARTICLES ON Statement

The particles emitted can be made transparent using the GHOST PARTICLES ON
FIG-41.19
The GHOST PARTICLE
GHOST PARTICLES ON partno , mode
ON Statement
In the diagram:

particles object.
mode is an integer value representing the transparency

mode used by the particles. This can be any value
between 0 and 5.

The program shown in LISTING-41.7 creates a particles object in front of a textured
cube which helps emphasis the transparency of the particles.

Using Ghost Particles COLOR BACKDROP 0
BACKDROP ON

AUTOCAM OFF
REM *** Create background cube ***

LOAD IMAGE "spots.bmp",2
TEXTURE OBJECT 1, 2
REM *** Set up particles ***

REM *** Make particles transparent ***

GHOST PARTICLES ON 1,1

WAIT KEY
END
Activity 41.13
Place the GHOST PARTICLES ON statement within a FOR loop using the
loop counter (which should range from 0 to 5) as the last value in the GHOST
PARTICLES ON statement. Wait for a key press between each value.
The GHOST PARTICLES OFF Statement

The transparency effect created by using GHOST PARTICLES ON can be switched
off, thereby returning to normal opaque particles using the GHOST PARTICLES
OFF statement which has the format shown in FIG-41.20.
FIG-41.20
GHOST PARTICLES OFF partno
The GHOST PARTICLES
OFF Statement
In the diagram:

particles object.
Retrieving Data on a Particles Object

The PARTICLES EXIST Statement
To check that a particular particles object has been created, we can use the
PARTICLES EXIST statement which has the format shown in FIG-41.21.

FIG-41.21
PARTICLES EXIST ( partno )
The PARTICLES EXIST
Statement
integer
In the diagram:
partno is an integer value representing the ID of the

particles object being tested.
If the specified particles object does exist, then 1 is returned; otherwise zero is
returned.
A typical use of this statement might be:
IF PARTICLES EXIST(1) = 0
ENDIF
The PARTICLES POSITION Statement

The exact position of a particles emitter can be determined using the PARTICLES
POSITION statement. Three variations of the statement exist, with each variation
returning one of the object's ordinates. The statement takes the format shown in
FIG-41.22.
FIG-41.22
X
The PARTICLES
POSITION Statement
PARTICLES POSITION Y ( partno )
real
In the diagram:

if the x-ordinate of the specified object is required,
Y for the y-ordinate, and Z for the z-ordinate.
particles object.
For example, we could determine the position in space of object 1's centre using the
lines:
x = PARTICLES POSITION X(1)

y = PARTICLES POSITION Y(1)
z = PARTICLES POSITION Z(1)
DO
SET CURSOR 100,100
PRINT "Particle emitter at (",x,",",y,",",z,")"
LOOP

Particles Statements that use Vectors
Two particle-related statements make use of vectors. These are described below.
The SET VECTOR3 TO PARTICLES POSITION Statement

The position of a particles object can be recorded in a vector using the SET
VECTOR3 TO PARTICLES POSITION statement which has the format shown in
FIG-41.23.
FIG-41.23 The SET VECTOR3 TO PARTICLES POSITION Statement
SET VECTOR3 TO PARTICLES POSITION 3DVector , partno
In the diagram:
3DVector is an integer value giving the ID of the vector

in which the details are saved.
partno is an integer value giving the ID of the particles

object whose position is to be saved.
The SET VECTOR3 TO PARTICLES ROTATION Statement

The angles of rotation of a particles object about its own local axes can be recorded
in a vector using the SET VECTOR3 TO PARTICLES ROTATION statement
FIG-41.24 The SET VECTOR3 TO PARTICLES ROTATION Statement
SET VECTOR3 TO PARTICLES ROTATION 3DVector , partno
In the diagram:
3DVector is an integer value giving the ID of the vector

in which the details are saved.
partno is an integer value giving the ID of the particles

object whose rotation is to be saved.
Summary
l Using a particles object creates a sparkler effect.
l All particles are projected from a single emitter.
l The colour of the particles emitted is initially determined by the image associated
with the particles object.
l Use MAKE PARTICLES to create a particles object.

l Use HIDE PARTICLES to make a particles object invisible.
l Use SHOW PARTICLES to make a hidden particles object reappear.
l Use DELETE PARTICLES to delete a particles object from RAM.
l Use POSITION PARTICLES to move the particles object's emitter and currently
visible particles.
l Use POSITION PARTICLE EMISSIONS to move the emitter, leaving current

particles behind.
l Use ROTATE PARTICLES to rotate the particles object's emitter about its local
axes.
l Use COLOR PARTICLES to modify the colour of the particles.
l Use SET PARTICLE EMISSIONS to modify the quantity of particles produced.
l Use SET PARTICLE VELOCITY to modify the speed at which particles move
away from the emitter.
l Use SET PARTICLE GRAVITY to affect how particles fall after being emitted.
l Use SET PARTICLE CHAOS to affect the random attributes of the particles'
paths.
l Use SET PARTICLE SPEED to modify the time elapsed between each screen
update of the particles' positions.
l Use SET PARTICLE FLOOR to disable/enable use of a floor level.
l Use SET PARTICLE LIFE to set the life time of the particles.
l Use GHOST PARTICLES ON to create transparent particle effects.
l Use GHOST PARTICLES OFF to return particles to a normal state after using
GHOST PARTICLES ON.
l Use PARTICLES EXIST to determine if a specific particles object exists.
l Use PARTICLES POSITION to determine the coordinates of a particles object's

emitter.
l Use SET VECTOR3 TO PARTICLES POSITION to save a particles object's

coordinates in a vector.
l Use SET VECTOR3 TO PARTICLES ROTATION to save a particles object's

angles of rotation in a vector.

Other Types of Particles
Introduction
As well as regular particles, DarkBASIC Pro can create two other types of particles
effects: snow and fire. Only a few of the regular particles statements can be applied
to these types of particles.
The Statements
The MAKE SNOW PARTICLES Statement
Particles can be made to fall like snow from the skies using the MAKE SNOW
PARTICLES statement. Rather than originate from a single emitter point, snow
particles are created over a rectangular area and fall for a set distance before
disappearing. The overall effect is to produce a box of falling particles. The MAKE
SNOW PARTICLES statement has the format shown in FIG-41.25.
FIG-41.25
The MAKE SNOW MAKE SNOW PARTICLES partno , imgno , freq ,
PARTICLES
Statement
x , y , z , w , h , d
In the diagram:

this particles object.

loaded image.
freq is an integer value which modifies the quantity

of particles produced.
x,y,z is a set of real values giving the approximate

centre of the area containing the snow particles.
w,h,d is a set of real values giving the width (w),

height (h), and depth (d) of the box in which the
snow is to fall.
For example, assuming we have previously loaded an image and assigned it an ID

of 2, we could create a snow effect (ID 1) within a 50 x 10 x 20 rectangular area
centred on the origin using the line:
MAKE SNOW PARTICLES 1,2,100,0,0,0,50,10,20
The program in LISTING-41.8 demonstrates the effect of the MAKE SNOW

PARTICLES statement.

LISTING-41.8 REM *** Set screen resolution and backdrop ***
Using Snow Particles COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
REM *** Create snow ***
MAKE SNOW PARTICLES 1,1,2000,0,0,0,10,10,10
WAIT KEY
END
Activity 41.14
Type in and test the program given in LISTING-41.8 (particles06.dbpro).
We might reasonably suppose that the position specified in the MAKE SNOW
PARTICLES statement ((0,0,0) in the example above) represents the exact centre
of the snow box, but, in fact, this is not the case.
Activity 41.15
Modify your last program, placing a 1 unit cube at position (0,0,0). Is the cube
at the centre of the snow area?
As you can see, the cube, which has been placed at the same position as that given
in the MAKE SNOW PARTICLES statement, is not in the centre of the snow box.
In fact, the box is centred in the x and z axes but about 75% of the snow falls below
the cube's centre and about 25% above. Remember to take this into account
whenever placing snow in your scene.
Some, but not all, of the normal particles statements can be applied to snow particles.
For example, snow can be moved to a different position, rotated, shown and hidden,
but none of the SET PARTICLE statements operate.
Activity 41.16
Does the GHOST PARTICLES ON statement affect snow particles?
The MAKE FIRE PARTICLES Statement

The MAKE FIRE PARTICLES statement creates the effect of sparks rising gently
from an unseen fire. In reality it is the same effect as the MAKE SNOW
PARTICLES statement but with the particles travelling upwards rather than down!
FIG-41.26
MAKE FIRE PARTICLES partno , imgno , freq ,
The MAKE FIRE
PARTICLES
Statement
x , y , z , w , h , d

In the diagram:

this particles object.

loaded image.
freq is an integer value which modifies the quantity

of particles produced.
x,y,z is a set of real values giving the approximate

centre of the area containing the fire particles.
w,h,d is a set of real values giving the width (w),

height (h), and depth (d) of the box in which the
fire particles are to rise.
Activity 41.17
Modify particles06.dbpro to show fire particles rather than snow.
(Use the image red.bmp rather than white.bmp.)
The same restrictions apply to fire particles as snow particles with many of the
regular particles statements not operating on these particles objects.
Summary
l Use MAKE SNOW PARTICLES to create a snowing effect.
l The box in which the snow appears is based around the position (x, y, z) but,
although centred on the x and z values, is about 75% below the y value.
l Use MAKE FIRE PARTICLES to create rising spark effects.
l Only some of the standard particles object statements can be applied to snow and
fire particles.

Examples of Using Particles
Introduction
We can combine particles with other 3D elements to create some visually interesting
effects. A few possibilities are described below.
A Roman Candle
All we need is a cylinder and a particles object to create a roman candle firework
(see FIG-41.27) as coded in the program given in LISTING-41.9.
LISTING-41.9 REM *** Define constants ***
REM *** Image IDs ***
A Roman Candle #CONSTANT YellowImg 1
#CONSTANT WrapperImg 2
REM *** Object IDs ***

#CONSTANT FireworkObj 1
REM *** Particle IDs ***
#CONSTANT YellowPart 1
REM *** Set screen resolution and backdrop ***

COLOR BACKDROP 0
BACKDROP ON

FIG-41.27 AUTOCAM OFF
A Roman Candle
REM *** Load images ***
LOAD IMAGE "yellow.bmp",YellowImg
LOAD IMAGE "firework.bmp",WrapperImg
REM *** Create firework ***

MAKE OBJECT CYLINDER FireworkObj,2
TEXTURE OBJECT FireworkObj,WrapperImg
SCALE OBJECT FireworkObj, 100,300,100
REM *** Create particles ***

MAKE PARTICLES YellowPart,YellowImg,20,1
SET PARTICLE GRAVITY YellowPart,0.1
POSITION PARTICLES YellowPart, 0,3,0
SET PARTICLE FLOOR YellowPart, 0

WAIT KEY
END
Activity 41.18
Type in and test the program in LISTING-41.9 (particles07.dbpro).
Modify the program so that the firework emits both red and yellow particles.
(HINT: Add a second particles object and use the image red.bmp.)

A Spaceship
Here's a good old-fashioned 1950s Sci-Fi spaceship just about to take off (see
FIG-41.28 and LISTING-41.10)
A Spaceship Launch
COLOR BACKDROP 0
BACKDROP ON

AUTOCAM OFF
POINT CAMERA 0,0,0
REM *** Load spaceship ***

LOAD OBJECT "spaceship2.x",1
FIG-41.28
Spacecraft Launching LOAD IMAGE "yellow.bmp",1
MAKE PARTICLES 1,1,1,1.5
POSITION PARTICLES 1,0,-3.6,0
ROTATE PARTICLES 1, 0,0,180
SET PARTICLE EMISSIONS 1, 20
SET PARTICLE VELOCITY 1, 5
SET PARTICLE GRAVITY 1, -4
SET PARTICLE SPEED 1,0.02
SET PARTICLE LIFE 1,10
WAIT KEY
END
Activity 41.19
Modify the program so that the spaceship's engine blast only appears after a
key press. The rocket should then actually take off, moving out of the top of
the screen. Make the rocket also move in a z direction as it lifts - this will help
give the illusion of it moving away from the viewer.
A Dungeon Torch
It's a well-known fact that a proper dungeon always uses flaming torches to light
their long dank corridors. By combining particles and a light map, we can recreate
an illusion of that setup in DarkBASIC Pro.
The torch itself is just a cone and a particle emitter. A point light is placed at
approximately the same position as the particles. The light map used back in Chapter
34 is used to create a circle of light on the wall. Finally, to create a flickering effect
the light's range continually changes (see FIG-41.29 and LISTING-41.11).
CreateScene()
A Dungeon Torch
DO
SET LIGHT RANGE 1, RND(1000)
WAIT 10
LOOP continued on next page

LISTING-41.11 WAIT KEY
(continued) END
A Dungeon Torch FUNCTION SetUpScreen()

BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,3
ENDFUNCTION
TEXTURE OBJECT 2,2
REM *** Create torch holder ***
ZROTATE OBJECT 3,180
POSITION OBJECT 3, 0,-2,4.5
FIG-41.29 REM *** Create flames ***
A Flaming Torch MAKE PARTICLES 1,3,1,1
POSITION PARTICLES 1, 0,-1,4
ROTATE PARTICLES 1,-30,0,0
SET PARTICLE GRAVITY 1,-1
REM *** Put light in sphere ***
MAKE LIGHT 1
REM *** Add Light maps ***
ENDFUNCTION
Activity 41.20

Solutions REM *** Set up particle object ***
Activity 41.1 LOAD IMAGE "white.bmp",1
The particles become much larger with the new setting.
REM *** Increase velocity over 10 secs***
FOR c = 1 TO 10
SET PARTICLE VELOCITY 1, c
REM *** Set up screen *** NEXT c
SET DISPLAY MODE 1280,1024,32 REM *** End program ***
COLOR BACKDROP 0 WAIT KEY
BACKDROP ON END
AUTOCAM OFF
POSITION CAMERA 0,0,5,-10 Activity 41.8
LOAD IMAGE "green.bmp",1 REM *** Set up screen ***
REM *** Create a particles object *** SET DISPLAY MODE 1280,1024,32
MAKE PARTICLES 1,1,1,10 COLOR BACKDROP 0
REM *** Remove after key press *** BACKDROP ON
WAIT KEY REM *** Set up camera ***
HIDE PARTICLES 1 AUTOCAM OFF
REM *** End program *** POSITION CAMERA 0, 0,0,-10
WAIT KEY REM *** Set up particle object ***
END LOAD IMAGE "white.bmp",1
REM *** Increase gravity 15 second ***
FOR c = -5 TO 10
Activity 41.3 SET PARTICLE GRAVITY 1, c
REM *** Set up screen *** WAIT 1000
SET DISPLAY MODE 1280,1024,32 NEXT c
COLOR BACKDROP 0 REM *** End program ***
BACKDROP ON WAIT KEY
REM *** Position camera *** END
AUTOCAM OFF
REM *** Load image for particles *** Activity 41.9
REM *** Create particles object *** REM *** Set up screen ***
MAKE PARTICLES 1,1,1,1 SET DISPLAY MODE 1280,1024,32
REM *** Move particle object *** COLOR BACKDROP 0
post#= -10 BACKDROP ON
FOR c = 1 TO 200 REM *** Set up camera ***
POSITION PARTICLES 1,0,post#,0 AUTOCAM OFF
post# = post# + 0.1 POSITION CAMERA 0, 0,0,-10
WAIT 10 REM *** Set up particle object ***
NEXT c LOAD IMAGE "white.bmp",1
REM *** End program *** MAKE PARTICLES 1,1,1,1
WAIT KEY REM *** Increase chaos ***
END FOR c = -50 TO 50
SET PARTICLE CHAOS 1, c
WAIT 200
NEXT c
WAIT KEY
No solution required. END


No solution required. REM *** Set up screen ***
COLOR BACKDROP 0
REM *** Set up screen *** AUTOCAM OFF
SET DISPLAY MODE 1280,1024,32 POSITION CAMERA 0, 0,0,-10
COLOR BACKDROP 0 REM *** Set up particle object ***
BACKDROP ON LOAD IMAGE "yellow.bmp",1
REM *** Set up camera *** MAKE PARTICLES 1,1,1,2
AUTOCAM OFF SET PARTICLE GRAVITY 1,0.5

REM *** Remove floor after a key press *** WAIT KEY
WAIT KEY END
SET PARTICLE FLOOR 1,0
REM *** End program *** As you can see from the result, the cube is within the
WAIT KEY snow-bound area. It is centred in the x direction (and z)
END but about three-quarters of the way up in the y direction.

SET DISPLAY MODE 1280,1024,32 GHOST PARTICLES ON does not affect the snow
COLOR BACKDROP 0 particles.
BACKDROP ON
REM *** Set up particle object *** REM *** Set up screen ***
LOAD IMAGE "yellow.bmp",1 SET DISPLAY MODE 1280,1024,32
MAKE PARTICLES 1,1,1,2 COLOR BACKDROP 0
SET PARTICLE LIFE 1, 20 BACKDROP ON
SET PARTICLE GRAVITY 1,0.5 REM *** Position camera ***
REM *** Remove floor after a key press *** AUTOCAM OFF
WAIT KEY POSITION CAMERA 0, 0,0,-20
SET PARTICLE FLOOR 1,0 REM *** Load image for particles ***
REM *** End program *** LOAD IMAGE "red.bmp",1
WAIT KEY REM *** Create fire ***
END MAKE FIRE PARTICLES 1,1,2000,0,0,0,10,10,10
WAIT KEY
END
Activity 41.13
COLOR BACKDROP 0
BACKDROP ON The second version of the firework is coded as:
AUTOCAM OFF REM *** Define constants ***
POSITION CAMERA 0,0,0,-10 REM *** Image IDs ***
REM *** Set up particles *** #CONSTANT YellowImg 1
LOAD IMAGE "yellow.bmp",1 #CONSTANT WrapperImg 2
MAKE PARTICLES 1,1,1,2 #CONSTANT RedImg 3
REM *** Create background cube *** REM *** Object IDs ***
MAKE OBJECT CUBE 1,20 #CONSTANT FireworkObj 1
LOAD IMAGE "spot1.bmp",2 REM *** Particle IDs ***
TEXTURE OBJECT 1,2 #CONSTANT YellowPart 1
POSITION OBJECT 1,0,0,50 #CONSTANT RedPart 2
REM *** Test each ghost option *** REM *** Set up screen ***
FOR c = 0 TO 5 SET DISPLAY MODE 1280,1024,32
GHOST PARTICLES ON 1,c COLOR BACKDROP 0
WAIT KEY BACKDROP ON
NEXT c REM *** Position camera ***
REM *** End program *** AUTOCAM OFF
WAIT KEY POSITION CAMERA 0, 0,3,-20
END REM *** Load images ***
LOAD IMAGE "yellow.bmp",YellowImg
LOAD IMAGE "firework.bmp",WrapperImg
Activity 41.14 LOAD IMAGE "red.bmp",RedImg
REM *** Create firework ***
No solution required. MAKE OBJECT CYLINDER FireworkObj,2
TEXTURE OBJECT FireworkObj,WrapperImg
SCALE OBJECT FireworkObj, 100,300,100
Activity 41.15 REM *** Create yellow particles ***
MAKE PARTICLES YellowPart,YellowImg,20,1
REM *** Set up screen *** POSITION PARTICLES YellowPart, 0,3,0
SET DISPLAY MODE 1280,1024,32 SET PARTICLE GRAVITY YellowPart,0.1
COLOR BACKDROP 0 SET PARTICLE FLOOR YellowPart, 0
BACKDROP ON REM *** Create red particles ***
REM *** Position camera *** MAKE PARTICLES RedPart,RedImg,20,1
AUTOCAM OFF POSITION PARTICLES RedPart, 0,3,0
POSITION CAMERA 0, 0,0,-20 SET PARTICLE GRAVITY RedPart,0.1
REM *** Load image for particles *** SET PARTICLE FLOOR RedPart,0
LOAD IMAGE "white.bmp",1 REM *** End program ***
REM *** Create snow *** WAIT KEY
MAKE SNOW PARTICLES 1,1,2000,0,0,0,10,10,10 END

Activity 41.19
The second version of the program is coded as:

COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,10,0
LOAD OBJECT "spaceship2.x",1
MAKE PARTICLES 1,1,1,1.5
POSITION PARTICLES 1,0,-3.6,0
ROTATE PARTICLES 1, 0,0,180
SET PARTICLE EMISSIONS 1, 20
SET PARTICLE GRAVITY 1, -4
REM *** Press key to launch ***
HIDE PARTICLES 1
WAIT KEY
SHOW PARTICLES 1
z# = 0
DO
POSITION OBJECT 1,0,OBJECT POSITION Y(1)
+1, z#
POSITION PARTICLES 1,0,OBJECT POSITION
Y(1)
-3.6,z#
z# = z# +0.5
WAIT 20
LOOP
END
Activity 41.20

42
Correcting Logic Errors

Creating a 3rd Person Perspective Game
Using Previously Created Models in a Game
DarkBASIC Pro: Elevators 1127

Elevators
Introduction
This started out as a snakes-and-ladders game until it became apparent that creating
the snakes and getting the characters to climb ladders was going to prove to be too
time-consuming. So we now have elevators in place of both the snakes and the
ladders. Some elevators take you up, others bring you down.
The Equipment
The game consists of a board showing numbered squares, the player's character and
a dice. The board squares are laid out in horizontal rows, with 10 squares per row.
The squares are numbered sequentially from 1 to 100. Also on the board are a set
of elevators used to carry the player from one level to another.
The Aim
The aim of the game is to get the player's character from square 1 to square 100.
The Rules
The basic rules come from the snakes and ladders game. The player starts on square
1 and moves forward according to the value thrown on the dice. If, at the end of a
move, the character is on an elevator platform, it takes that elevator to the destination
square - sometimes upwards; sometimes down.
Creating a Computer version of the Game

User Controls
There is little interaction; the player needs only to press a key to cause the dice to
roll. The character is then moved automatically.
Game Responses
The game reacts to a dice throw by moving the player's character by the value
thrown. If the character has to move from one row of the board to the next during
a move, this is done using the lift at the end of the row. If the character stops on an
elevator square it is carried off to that elevator's destination.
Screen Layout
To create an unique look to the game, it is played in 3D with each row of the board
becoming a corridor down which the player's character is to move. The camera view
gives a 3rd person perspective, the scene being shown from a viewpoint just behind
the character's shoulder (see FIG-42.1).
1128 DarkBASIC Pro: Elevators

FIG-42.1
A Screenshot from the
Game
The Board Design

We have to begin by designing the basic board, showing the elevators and their
destinations. This can be done in an abstract way (see FIG-42.2).
FIG-42.2
The Board Layout
Showing the Elevators
Notice that all the lifts are vertical. Anything lying at an angle would just cause
more modelling problems.
The Media Used

The following 3D objects are required:
Ø Board (created as a tall, narrow box with 9 internal floors)

Ø Elevator
Ø Player's character
Ø Dice
The elevator and dice models were created in Chapter 35; the player's character
model and its texture have been taken from the The Game Creators' Dark Matter 1
model pack and the board was constructed using chUmbaLum sOft's Milkshape
modelling program.
The other images required are the textures for the 3D objects and are shown in
FIG-42.3.
FIG-42.3 The Images used to Texture 3D Objects
Data Structures
We need to set up a record structure to hold details about a single elevator. The
information required is:
Ø the elevator object's ID number

Ø the square on which the lift starts
Ø the square on which the lift terminates
Ø the number of floors covered by the lift (this will be negative for
descending elevators)
Ø the z-ordinate of the lift's position on a square
All this can be held in a record structure:
TYPE ElevatorType
ID AS INTEGER
startsquare AS INTEGER

endsquare AS INTEGER
floors AS INTEGER `(-ve for downwards)
zord AS INTEGER
ENDTYPE
There are 27 elevators in total, so we need an array to store information about each:
GLOBAL DIM lifts(27) AS ElevatorType
The board itself need only record, for each square, the presence of an elevator
starting on that square:
GLOBAL DIM board(100) AS INTEGER
All we need to know about the player's character is its position on the board:
GLOBAL playerat AS INTEGER
The 3D objects used in the game have already been textured and only need to be
assigned ID values within the game. As a general rule we should assign a name to
each ID, but since there are 27 elevators we'll assign a name to the first one only.

#CONSTANT playerobj 2
#CONSTANT diceobj 3
#CONSTANT firstliftobj 4 `27 lifts with IDs 4 - 30
So, the final code for our main data and constants is:
REM **********************************
REM *** Structures ***
REM **********************************
REM *** Elevator details ***
TYPE ElevatorType
ID AS INTEGER
zord AS INTEGER
ENDTYPE
REM **********************************
REM *** Global variables ***
REM **********************************
GLOBAL DIM Lifts(27) AS ElevatorType
REM **********************************
REM **********************************
#CONSTANT diceobj 3
#CONSTANT firstliftobj 4 `27 lifts with IDs 4 - 30
Game Logic
As usual, we want to describe the whole logic of the program in less than a dozen
lines and then look at each of these sections in more detail until we have our final
product. So, the overall game could be described as follows:

Set up the game
REPEAT
Roll the dice
Move the player
UNTIL the player reaches square 100
Show finish screen
As usual, the descriptive lines become function calls in the actual program. We'll
need to find out what value is thrown on the dice so it can be used when moving
the player, so the function for throwing the dice must return the dice value and the
function which moves the player must take that value as a parameter. This gives us
the following code for the main section:
REM **********************************
REM *** Main game logic ***
REM **********************************
SetUpGame()
REPEAT
thrown = RollDice()
MovePlayer(thrown)
UNTIL playerat = 100
EndScreen()
END
Activity 42.1
Create a new project (elevators.dbpro) and copy into it the code already given
for the data structure, global variables and constants as well as the logic of the
main section.
At the end of the code, add a REM statement containing the lines:
REM ************************************
REM *** Level 1 routines ***
REM ************************************
Under these REM statements, write test stubs for each function called. The test
stubs should be empty, containing only FUNCTION and ENDFUNCTION
statements.
The test stub for RollDice() should return the value 1.
Run the program and make sure it executes. You'll have to press Escape to exit
the REPEAT..UNTIL structure in the main section.
Adding SetUpGame()
There are many tasks to be performed by this part of the program, but we can split
those tasks into two main sections:
Ø Initialisation of the data

Ø Initialisation of the visuals
So we can code SetUpGame() as a function which does nothing more than call two
other functions:
InitialiseData()
InitialiseVisuals()
ENDFUNCTION

LISTING-42.1 These new routines will then require their own test stubs.
Elevators - Getting
Started with the Code The complete program at this point is shown in LISTING-42.1.
REM **********************************
REM **********************************
TYPE ElevatorType
ID AS INTEGER
zord AS INTEGER
ENDTYPE
REM **********************************
REM **********************************
GLOBAL DIM lifts(27) AS ElevatorType
REM **********************************
REM **********************************
#CONSTANT diceobj 3
#CONSTANT firstliftobj 4 `27 lifts IDs 4 - 30
REM **********************************
REM **********************************
SetUpGame()
REPEAT
thrown = RollDice()
MovePlayer(thrown)
EndScreen()
END
REM **********************************
REM **********************************
InitialiseData()
InitialiseVisuals()
ENDFUNCTION
FUNCTION RollDice()
ENDFUNCTION 1
FUNCTION MovePlayer(thrown)
ENDFUNCTION
FUNCTION EndScreen()
ENDFUNCTION
REM **********************************
REM **********************************
FUNCTION InitialiseData()
ENDFUNCTION
FUNCTION InitialiseVisuals()
ENDFUNCTION

Notice how the functions have been grouped; a function called by the main section
is a level 1 routine; a routine called by a level 1 function is a level 2 routine. Large
programs may have levels 3, 4 or more. This is a popular layout method for many
programmers.
Activity 42.2
Modify your program to match the code above.
Adding InitialiseData()
The main tasks of the InitialiseData() function involves us setting up the contents
of the lifts and board arrays, but we'll also need to specify the player's starting
position and seed the random number generator. So the routine can be coded as:
FUNCTION InitialiseData()
InitialiseLifts()
InitialiseBoard()
REM *** Start player at square 1 ***
playerat = 1
REM *** Seed random number generator***
RANDOMIZE TIMER()
ENDFUNCTION
Activity 42.3
Replace the InitialiseData() stub with the code given.
Add test stubs for the two new routines in a level 3 section of the program.
Adding InitialiseLifts()
We need a great deal of data to initialise the lifts array since we have to set up
information about each lift. The easiest way to set these details up is to use DATA
statements. These values can then be read into the appropriate part of the array.
The required coding is given below:
FUNCTION InitialiseLifts()
REM *** Lifts start at square ***
DATA 10,13,18,20,24,25,30,34,35
DATA 39,40,49,50,53,54,59,60,65
DATA 68,70,72,75,80,83,87,90,97
REM *** Lifts end at square ***
DATA 11,33,43,21,64,56,31,27,55
DATA 19,41,12,51,48,67,79,61,76
DATA 88,71,52,95,81,58,74,91,77
REM *** Floors covered by lifts ***
DATA 1,2,3,1,4,3,1,-1,2
DATA -2,1,-3,1,-1,1,2,1,1
DATA 2,1,-2,2,1,-3,-1,1,-2
REM *** z-ordinates of lifts' position ***
DATA 47,23,-27,-47,-17,-7,47,13,3
DATA -37,-47,33,47,23,13,-37,-47,-7
DATA 23,47,33,3,-47,-27,13,47,-17
REM *** Store lift IDs (firstliftobj to firstliftobj + 26)***
FOR c = 1 TO 27
lifts(c).ID = firstliftobj + c-1
NEXT c

REM *** Store start square for lift ***
FOR c = 1 TO 27
READ sq
lifts(c).startsquare = sq
NEXT c
REM *** Store end square for lift ***
FOR c = 1 TO 27
READ fin
lifts(c).endsquare = fin
NEXT c
REM *** Store floors covered ***
FOR c = 1 TO 27
READ fl
lifts(c).floors = fl
NEXT c
REM *** Store z-ord of lift ***
FOR c = 1 TO 27
READ z
lifts(c).zord = z
NEXT c
ENDFUNCTION
Adding InitialiseBoard()
In the board array, each cell represents one square on the board. If a square does
not contain the starting position of an elevator, we store zero in the corresponding
element of board; if a square does contain an elevator, then the corresponding cell
in board contains the elevator's number.
We have two choices when storing an elevator number; we can either store the
elevator's ID or we can store the position of that elevator's details in the lifts array.
For example, square 10 on the board has an elevator (ID 4) taking the player to
square 11, so we could either write
board(10) = 4
thereby storing the elevator's ID, or we could use
board(10) = 1
meaning that the elevator starting at square 10 has its details stored in lifts(1).
In fact, the second option is the better approach since it allows more flexibility
should we ever need to change the ID values used for the elevator objects.
As a result, the code for the InitialiseBoard() routine is:
FUNCTION InitialiseBoard()
REM ***Square empty (0) or elevator's location in lifts ***
DATA -1,-1,-1,-1,-1, -1,-1,-1,-1,1
DATA -1,-1,2,-1,-1, -1,-1,3,-1,4
DATA -1,-1,-1,5,6, -1,-1,-1,-1,7
DATA -1,-1,-1,8,9, -1,-1,-1,10,11
DATA -1,-1,-1,-1,-1, -1,-1,-1,12,13
The rather strange code needs DATA -1,-1,14,15,-1, -1,-1,-1,16,17
explaining; storing 0's in the DATA -1,-1,-1,-1,18, -1,-1,19,-1,20
DATA statements seemed to DATA -1,21,-1,-1,22, -1,-1,-1,-1,23
upset the program and so it DATA -1,-1,24,-1,-1, -1,25,-1,-1,26
was necessary to store -1's DATA -1,-1,-1,-1,-1, -1,27,-1,-1,-1
and convert them to zeros! REM *** Store this data in board array ***
FOR c = 1 TO 100
READ board(c)

IF board(c) = -1
board(c) = 0
ENDIF
NEXT c
ENDFUNCTION
Activity 42.4
Replace the stubs for InitialiseLifts() and InitialiseBoard() with the code given.
Adding InitialiseVisuals()
Now we're ready to load all the visual elements of the game and position the
cameras. The requirements of this routine can best be described as:
Set up the screen
Load the board
Load the player's character
Load the dice
Position the cameras
Set the lighting
All but the first and last of these tasks will involve a few lines of code, and it is
probably best to create a set of level 3 routines to handle the details. So,
InitialiseVisuals() can be coded as:
REM *** Set the screen resolution ***
The standard screen SET DISPLAY MODE 1280,1024,32
resolution is used and the LoadBoard()
ambient light is turned up LoadPlayerCharacter()
since the image was too LoadDice()
dark with the default PositionCameras()
setting. REM *** Increase the ambient light ***
SET AMBIENT LIGHT 50
ENDFUNCTION
Activity 42.5
Replace the stub of InitialiseVisuals() with the code given above and create
new level 3 stubs for the functions it calls.
Loading Models and Texture Files

Before continuing, it would be better if we copied the necessary model and image
files to the current directory. The following files need to be copied:
board.x the model of the board

dice.dbo the model of the dice
H-Android_Move.x the player's character model
lift.dbo the elevator model
cherry.jpg a board texture

bbrush.jpg a board texture
floor1.bmp a board texture

floorA.bmp a board texture
spot1.jpg a dice texture

All files used in this book robot.dds a player character texture
are available from our metal.bmp an elevator texture
website - steel.bmp an elevator texture
www.digital-skills.co.uk
Activity 42.6
Copy the necessary files into your folder.
Adding LoadBoard()
There are actually two parts to this routine: first we need to load and resize the board
model, but also we need to position the elevators within the board. This second task
will be performed by a separate routine. The code for LoadBoard() is:
FUNCTION LoadBoard()
REM *** Load, scale and position board ***
LOAD OBJECT "board.x",1
SCALE OBJECT 1, 1000,1000,1000
YROTATE OBJECT 1, 90
REM *** Make black areas invisible ***
REM *** These are the holes in the floor ***
REM *** through which the lifts pass ***
AddElevators()
ENDFUNCTION
Adding AddElevators()
This routine loads the elevator model (lift.dbo) and creates another 26 clones of the
model. The data in the board and lifts arrays is then used to position and resize each
elevator.
FUNCTION AddElevators()
REM *** Create the first lift ***
LOAD OBJECT "lift.dbo",firstliftobj
REM *** Create others as clones ***
FOR c = 2 TO 27
CLONE OBJECT lifts(c).ID,firstliftobj
NEXT c
REM *** Position all the lifts ***
FOR c = 1 TO 27
REM *** Rotate lift and scale to match floors ***
liftID = lifts(c).ID
YROTATE OBJECT liftID,-90

SCALE OBJECT liftID,100,10*ABS(lifts(c).floors),100
REM *** position up lifts at start square ***
IF lifts(c).floors> 0
POSITION OBJECT liftID,-4.5,
Ä
(lifts(c).startsquare-1)/10*10,lifts(c).zord
ELSE
REM *** And down ones at the finish square ***
Ä
(lifts(c).endsquare-1)/10*10,lifts(c).zord
REM *** down lifts need platform at top ***
OFFSET LIMB liftID,2,0,100,-2
ENDIF
NEXT c
ENDFUNCTION
Since this is called by a level 3 function, it needs to be placed in a level 4 section.
Adding LoadPlayerCharacter()
This routine needs to load the appropriate model, scale and rotate it. The code is:
FUNCTION LoadPlayerCharacter()
REM *** Load, scale and position player's character ***
LOAD OBJECT "H-android-move.x",playerobj
SCALE OBJECT playerobj,150,150,150
YROTATE OBJECT playerobj,180
POSITION OBJECT playerobj,2,0,-45 `square 1
REM *** Set the animation play speed ***
SET OBJECT SPEED playerobj,15
ENDFUNCTION
Notice that the routine takes the trouble to check the position of the player on the
board and rotates the character accordingly. This may seem unnecessary since the
player will obviously start at square 1 - however, while testing the program we may
want to start the player at other positions.
Adding LoadDice()
Like LoadPlayerCharacter(), LoadDice() involves loading and scaling a model:
FUNCTION LoadDice()
REM *** Load, scale and position dice ***
LOAD OBJECT "dice.dbo",diceobj
POSITION OBJECT diceobj,30,15,45
SCALE OBJECT diceobj,50,50,50
ENDFUNCTION
Adding PositionCameras()
Two cameras are used in the game. Camera 1 follows the player's character as it
moves through the corridors of the board and forms the main image on the screen;
camera 2 points at the dice and its output is shown as a picture-in-picture window
at the top right of the screen. The code for this routine is:
FUNCTION PositionCameras()
AUTOCAM OFF
REM *** Set up player's camera ***
MAKE CAMERA 1
POINT CAMERA 1,3,2,-40
REM *** Set up dice's camera ***

MAKE CAMERA 2
SET CAMERA FOV 2, 10
SET CAMERA VIEW 2,1100,20,1220,120
ENDFUNCTION
At this point, our program is as shown in LISTING-42.2.

LISTING-42.2 REM **********************************
The Game So Far
REM **********************************
TYPE ElevatorType
ID AS INTEGER
zord AS INTEGER
ENDTYPE
REM **********************************
REM **********************************
GLOBAL DIM Lifts(27) AS ElevatorType
REM **********************************
REM **********************************
#CONSTANT diceobj 3
#CONSTANT firstliftobj 4 `27 lifts IDs 4 - 30
REM **********************************
REM **********************************
SetUpGame()
REPEAT
thrown = RollDice()
MovePlayer(thrown)
EndScreen()
END
REM **********************************
REM **********************************
InitialiseData()
InitialiseVisuals()
ENDFUNCTION
FUNCTION RollDice()
ENDFUNCTION 1
ENDFUNCTION
ENDFUNCTION continued on next page

LISTING-42.2 REM **********************************
(continued) REM *** Level 2 routines ***
REM **********************************
The Game So Far FUNCTION InitialiseData()
InitialiseLifts()
InitialiseBoard()
RANDOMIZE TIMER()
REM *** Start player at square 1 ***
playerat = 1
ENDFUNCTION
LoadBoard()
LoadPlayerCharacter()
LoadDice()
PositionCameras()
ENDFUNCTION
REM **********************************
REM **********************************
FUNCTION InitialiseLifts()
DATA 10,13,18,20,24,25,30,34,35
DATA 39,40,49,50,53,54,59,60,65
DATA 68,70,72,75,80,83,87,90,97
REM *** Lifts end at square ***
DATA 11,33,43,21,64,56,31,27,55
DATA 19,41,12,51,48,67,79,61,76
DATA 88,71,52,95,81,58,74,91,77
REM *** Floors covered by lifts ***
DATA 1,2,3,1,4,3,1,-1,2
DATA -2,1,-3,1,-1,1,2,1,1
DATA 2,1,-2,2,1,-3,-1,1,-2
REM *** z-ordinate of lift's position ***
DATA 47,23,-27,-47,-17,-7,47,13,3
DATA -37,-47,33,47,23,13,-37,-47,-7
DATA 23,47,33,3,-47,-27,13,47,-17
REM *** Store lift IDs (firstliftobj to firstliftobj + 26)***
FOR c = 1 TO 27
lifts(c).ID = c + firstliftobj - 1
NEXT c
REM *** Store start square ***
FOR c = 1 TO 27
READ sq
NEXT c
REM *** Store end square ***
FOR c = 1 TO 27
READ fin
NEXT c
FOR c = 1 TO 27
READ fl
NEXT c
REM *** Store z-ord of lift ***
FOR c = 1 TO 27
READ z
lifts(c).zord = z
NEXT c
ENDFUNCTION

LISTING-42.2 FUNCTION InitialiseBoard()
(continued) `Square empty (0) or lift (lift ID)
DATA -1,-1,-1,-1,-1, -1,-1,-1,-1,1
The Game So Far DATA -1,-1,2,-1,-1, -1,-1,3,-1,4
DATA -1,-1,-1,5,6, -1,-1,-1,-1,7
DATA -1,-1,-1,8,9, -1,-1,-1,10,11
DATA -1,-1,-1,-1,-1, -1,-1,-1,12,13
DATA -1,-1,14,15,-1, -1,-1,-1,16,17
DATA -1,-1,-1,-1,18, -1,-1,19,-1,20
DATA -1,21,-1,-1,22, -1,-1,-1,-1,23
DATA -1,-1,24,-1,-1, -1,25,-1,-1,26
DATA -1,-1,-1,-1,-1, -1,27,-1,-1,-1
REM *** Store this data in board array ***
FOR c = 1 TO 100
READ board(c)
IF board(c) = -1
board(c) = 0
ENDIF
NEXT c
ENDFUNCTION
REM *** Load, scale and position board ***
LOAD OBJECT "board.x",1
SCALE OBJECT 1, 1000,1000,1000
YROTATE OBJECT 1, 90
REM *** Make black areas invisible ***
REM *** These are the holes in the floor ***
REM *** through which the lifts pass ***
AddElevators()
ENDFUNCTION
REM *** Load, scale and position player's character ***
LOAD OBJECT "H-android-move.x",playerobj
SCALE OBJECT playerobj,150,150,150
YROTATE OBJECT playerobj,180
POSITION OBJECT playerobj,2,0,-45 `square 1
REM *** Set the animation play speed ***
SET OBJECT SPEED playerobj,15
ENDFUNCTION
FUNCTION LoadDice()
REM *** Load, scale and position dice ***
LOAD OBJECT "dice.dbo",diceobj
POSITION OBJECT diceobj,30,15,45
SCALE OBJECT diceobj,50,50,50
ENDFUNCTION
AUTOCAM OFF
REM *** Set up player's camera ***
MAKE CAMERA 1
POINT CAMERA 1,3,2,-40
REM *** Set up dice's camera ***
MAKE CAMERA 2
SET CAMERA FOV 2, 10
SET CAMERA VIEW 2,1100,20,1220,120
ENDFUNCTION

LISTING-42.2 REM **********************************
(continued) REM *** Level 4 ***
REM **********************************
The Game So Far FUNCTION AddElevators()
FOR c = 2 TO 27
NEXT c
FOR c = 1 TO 27
REM *** Rotate lift and scale to match floors ***
SCALE OBJECT liftID,100,10*ABS(lifts(c).floors),100
REM *** position up lifts at start square ***
IF lifts(c).floors > 0
Ä(lifts(c).startsquare-1)/10*10,lifts(c).zord
ELSE
REM *** And down ones at the finish square ***
Ä(lifts(c).endsquare-1)/10*10,lifts(c).zord
REM *** down lifts need platform at top ***
ENDIF
NEXT c
ENDFUNCTION
Activity 42.7
Extend your program to match the code given in LISTING-42.2.
Adding RollDice()
With all the level 4, 3 and 2 functions now coded, we can return to the level 1
routines. The next of these to be coded is the RollDice() function which needs to
show the dice being rolled (this appears in camera 2's output) and also returns the
number shown on the dice.
Making what the dice shows on its top surface match the value returned by this
function adds to the complexity of the code.
FUNCTION RollDice()
REM *** Make dice spin ***
FOR c = 0 TO 100
PITCH OBJECT UP diceobj,15`RND(15)+1
TURN OBJECT LEFT diceobj,15`RND(15)+1
ROLL OBJECT LEFT diceobj,15 `RND(15)+1
WAIT 1
NEXT c
REM *** Generate the value to be returned ***
value = RND(5)+1
REM *** make the dice show the value generated ***
SELECT value
CASE 1
ROTATE OBJECT diceobj,-90,0,0
ENDCASE
CASE 2
ROTATE OBJECT diceobj,180,0,0
ENDCASE
CASE 3
ROTATE OBJECT diceobj,0,0,-90

ENDCASE
CASE 4
ENDCASE
CASE 5
ENDCASE
CASE 6
ENDCASE
ENDSELECT
WAIT KEY
ENDFUNCTION value
Activity 42.8
In your program, replace the RollDice() test stub with the code given above.
Adding MovePlayer()
Now we come to the main routine of the program. The task it has to do is quite
simple in concept; move the player forward a number of squares corresponding to
the value thrown on the dice.
Complications arise because of the elevators. If a character stops at an elevator

square, then he must be moved to the elevator's destination. Also, if, while moving,
the character arrives at the end of a row, an elevator must take the character to the
next floor. During all of this, the main camera must follow the player's character.
This time, rather than give you the completed routine, we'll build it up step-by-step,
making it more sophisticated as we go. We'll start by just trying to get the movement
along the bottom row correct. A first attempt at the logic required might be:
Start the character's animation (appears to be walking)
FOR thrown times DO
thrown is a parameter of Find the current coordinates of the character
the MovePlayer() function Move character one square
and playerat is a global ENDFOR
variable. Stop character animation and reset to frame 1
Add thrown to playerat
This can be coded as:

REM *** Animate the character ***
LOOP OBJECT playerobj
REM *** Move character a square at a time ***
FOR squares = 1 TO thrown
REM *** Find out character's current position ***
x# = OBJECT POSITION X(playerobj)
y# = OBJECT POSITION Y(playerobj)
z# = OBJECT POSITION Z(playerobj)
POSITION OBJECT playerobj,x#,y#,z#+10
WAIT 100
NEXT squares
REM *** Stop the animation ***
STOP OBJECT playerobj
SET OBJECT FRAME playerobj,1
REM *** Update character's position ***
playerat = playerat + thrown
ENDFUNCTION

Activity 42.9
Replace the MovePlayer() test stub with the code given. How good are the
results?
The biggest problem is that the character does not move smoothly. To deal with
this, we need to move the model using smaller steps.
If we change the code:
REM *** Move character a square at a time ***

POSITION OBJECT playerobj,x#,y#,z#+10
WAIT 100
NEXT squares
to
REM *** Move character one square at a time ***

FOR move# = 1 TO 10 STEP 0.5
POSITION OBJECT playerobj,x#,y#,z#+move#
WAIT 100
NEXT move#
NEXT squares
we can achieve a more realistic movement.
Activity 42.10
Make the changes described above and test your program again.
The next obvious problem is that the camera has been left behind when it should be
following the character.
This is easily solved by moving the camera in unison with the character. The
character was originally placed at position (2,0,-45) and camera 1 at (3,5,-52), so
to keep the same relative positions, the camera needs to maintain a distance of
(1,5,-7) from the character. This can be achieved using the line:
POSITION CAMERA 1,x#+1,y#+5,z#+move#-7
Activity 42.11
Add the above line at an appropriate point in your MovePlayer() routine and
test the program.
Does the camera move with the character?
What happens to the character after a few moves?

The next feature we need to add to the MovePlayer() function is to detect when the
character has reached the end of the corridor and then use the elevator to move to
the next level. This means we need to add the following logic to the function:
IF at end of a level THEN
Use elevator
ENDIF
However, we have to be careful. If we were currently in square 6 and then threw a

5, the character would move to square 11, taking the elevator up from square 10 as
part of the move. However, if we started on square 6 and threw a 4, the character
should move to square 10 and, although this is the base of an elevator, the move has
already been completed, so the lift should not be used on this turn.
We can handle these two situations if we rework the logic of MovePlayer() slightly,
changing it from:
FOR thrown times DO
Find the current coordinates of the character
Move character one square
ENDFOR
Stop character animation and reset to frame 1
Add thrown to playerat
to
FOR thrown times DO
Find the current coordinates of the character
IF at end of level THEN
Use elevator
ELSE
Add 1 to playerat
ENDIF
Move character one square
ENDFOR
Stop character animation and reset to frame 1
Notice that playerat is now incremented a square at a time and only when the
elevator is not being used (using the elevator will handle the updating of playerat
separately). This lets us know where the character is at any time during its move.
Notice also that we check for end of level before moving the character or
incrementing playerat. If you think about it, you'll see that this handles both the
situations of stopping at the end of a level without using the elevator and using the
elevator on the way to a later square during a move.
If we implement checking for the end of level and using the elevator as two new
functions, we can recode MovePlayer() as:
IF AtEndOfLevel()
UseElevator()
ELSE
INC playerat

ENDIF
POSITION OBJECT playerobj,x#,y#,z#+move#
POSITION CAMERA 1,x#+1, y#+5,move#-7
WAIT 100
NEXT move#
NEXT squares
REM *** Stop animation ***
ENDFUNCTION
The function AtEndOfLevel() will return 1 if the character is at the end of a level
(that is on square 10, 20, 30, etc.), otherwise zero will be returned.
Activity 42.12
Update your version of MovePlayer() to match the logic given.
Add a test stub for UseElevator().
Code the AtEndOfLevel() function to implement the description given above.
Being at the end of a level is not the only time when an elevator is used. Should the
character stop on some other square containing an elevator, that evelator must be
taken, moving the character to a new position on the board dependent on the
elevator's destination.
So, the MovePlayer() function should end with the logic:
IF on square with elevator AND not at the end of a level THEN
Use elevator
ENDIF
Activity 42.13
What term can be used to determine if a square contains an elevator?
Convert the logic given above to DarkBASIC Pro code.
Update your version of MovePlayer() to incorporate this logic.
Adding UseElevator()
Using the elevator requires several distinct movements of the character. These are
listed below
Ø Line the character up with the elevator. This may involve the character in
moving forward slightly.
Ø Turn the character to face the elevator.
Ø Move the character onto the elevator platform
Ø Move the elevator and character to elevator's destination.
Ø Turn the character away from the elevator.
Ø Move the character to the centre of the new square.
Ø Face the character in the direction it is to travel.
Ø Return the elevator platform to its original position.
At all stages the camera needs to be tracking the character's movements. Since the
elevator may be used part way through a move, it's probably best to halt the
character's animation at the start of this function and resume it at the end.
The various turning manoeuvres can be handled by a single routine, but for the other
stages, it would probably be best to create a function to implement each step. This
means that UseElevator() can be coded as:
FUNCTION UseElevator()
REM *** Stop character movement ***
MovePlayerToElevator()
TurnPlayer(-90)
MoveOntoPlatform()
liftnum = MoveElevator()
TurnPlayer(-180)
MoveOffPlatform()
TurnPlayer(90)
ReturnElevator(liftnum)
RepositionCamera()
REM *** Resume character movement ***
ENDFUNCTION
Notice that the lift number (liftnum) is returned by MoveElevator(). This value is
later passed to ReturnElevator() to identify which elevator's platform must be
returned to its starting position.
Activity 42.14
Replace your UseElevator() stub with the code given above and create the
appropriate new stubs in the level 3 section of the program.
Note:
MoveElevator() returns the number of the lift being used.
TurnPlayer() takes the angle through which the player is to be turned.
ReturnElevator() takes the number of the lift being returned to
its original position.
Adding MovePlayerToElevator()
We need to get the player's character level with the lift platform before turning the
character to face the lift. We can get the exact position of the lift from the lifts array.
The logic for this routine is:
Start the character's animatation
Find the coordinates of the character's current position
FOR z = current z-ordinate of character TO lift's z-ordinate DO (steps of 0.2)
Reposition character at (x,y,z)
Wait 50 milliseconds
ENDFOR
Stop the character animation and reset to frame 1

Activity 42.15
Convert the logic given above to DarkBASIC Pro code and replace the
function's stub with your code.
Check that the character moves correctly. (HINT: you may want to add a
WAIT KEY after the function is called in UseElevator(). This will give you a
chance to check out the character movement before the program continues.)
Adding TurnPlayer()
This routine takes a parameter specifying the angle through which the character is
to be turned. For anti-clockwise movement, use a negative value.
The routine has the following logic:
IF the angle is negative THEN
Set stepsize to -1
ELSE
Set stepsize to 1
ENDIF
Set startangle as the character's current rotation about its y-axis
FOR degree startangle

= startrangle+angle
TO DO (steps of stepsize
)
degree o
Rotate character to about its y-axis
Wait 1 millisecond
ENDFOR
Activity 42.16
Create the TurnPlayer() function using the logic given and test it out in your
program. (HINT: you may want to move the WAIT KEY you added
previously.)
Adding MoveOntoPlatform()
With the character facing in the correct direction, we now need to move it onto the
elevator's platform. The logic required for this is:
FOR newx = x-ord TO x-ord - 5 DO (steps of -0.1))
Wait 50 millisecond
ENDFOR
Notice that the routine's logic is very similar to MovePlayerToElevator().
Activity 42.17
Add a MoveOntoPlatform() function based on the logic given. Test the

program and ensure that the character walks onto the platform.
Adding MoveElevator()
This routine moves the platform and the character standing on it to a new square.
The routine must be general enough so that it will work for elevators of different

lengths and for ones which travel down as well as up.
Since the elevator platform is a limb, its movement will be specified as an offset
from its original position within the model. The model has the platform at the bottom
for lifts that travel upwards (that is, it starts with an offset of zero along the y-axis).
The platform will be at the top when the offset is changed to 100 (the models have
an original height of 100 units). Downward travelling elevators start with an offset
of 100 and this changes to zero to make the platform descend.
The function requires the following logic:
Get the lift's position in the lifts array from the board array
Get the lift's ID from the liftsarray>
Calculate the height of the lift as 10 * the number of floors it covers
(neg value for lifts which travel down)
IF the height is positive THEN
Set liftdirection to 1
ELSE
Set liftdirection to -1
ENDIF
IF the lift is going up THEN
Set liftendoffset to 100
ELSE
ENDIF
Set liftstartoffset to 100 - liftendoffset
Set unitsperpercent to height/100
FOR move = liftstartoffset TO liftendoffset DO (steps of liftdirection)
Offset platform limb (0,move,-2)
Position character at (x,y+unitsperperc*move,z)
Wait 10 milliseconds
ENDFOR
Set playerat to the end square of the elevator
Activity 42.18
Add a MoveElevator() function based on the logic given. Test the program and
ensure that the platform moves correctly.
Adding MoveOffPlatform()
To move the character off the platform, after having been made to face in the
appropriate direction, it is just of matter of walking the character forward. This
routine needs the following logic:
FOR newx = x-ord TO x-ord +5 DO (steps of 0.1))
Wait 50 millisecond
ENDFOR
Activity 42.19
Add a MoveOffPlatform() function based on the logic given. Test the program
and ensure that the character walks off the platform.

Adding ReturnElevator()
This routine moves the elevator's platform back to its original position. The lift's
position within the lifts array needs to be passed to the function since this
information can no longer be obtained from the board array. The routine requires
the following logic:
Get the lift's ID
Calculate the height of the lift (negative for a lift that has moved down)
IF the height is positive THEN
Set liftdirection to -1
ELSE
Set liftdirection to 1
ENDIF
IF liftdirection = -1THEN
ELSE
ENDIF
Set liftstartoffset to 100 - liftendoffset
Set unitsperperc to height / 100.0
FOR move = liftstartoffset TO liftendoffset (step by liftdirection)
Offset platform limb (2,0,move,-2)
WAIT 10 milliseconds
ENDFOR
Activity 42.20
Add a ReturnElevator() function based on the logic given. Test the program
and ensure that the platform returns to its original position.
Adding RepositionCamera()
With the character on a new level, we need to move the camera behind it, looking
down the new corridor.
The routine should implement the following logic:
Get the character's coordinates
Move the camera behind the character
Point the camera down the new corridor.
This requires the following code:
FUNCTION RepositionCamera()
REM *** Get position of character ***
# = OBJECT POSITION Y(playerobj)
# = OBJECT POSITION Z(playerobj)
REM *** Position and point camera ***
POSITION CAMERA 1, x#+1,y#+5,z#+7
POINT CAMERA 1,x#+1,y#+2,z-5
ENDFUNCTION
Activity 42.21
Add the routine given above and test out the program.
What problem occurs?

Fixing the Shortcomings
We've been so busy writing code to work on the first corridor, that we haven't taken
into account that the second corridor travels in the opposite direction.
We can either write a whole new set of routines to handle movement in this
direction, or we can make the previous routines more general so that they can handle
travel in both directions. This second method is the best approach since the changes
required are actually quite small.
We'll need a routine that can give us an indication of which direction the character
is travelling in:
FUNCTION FindDirectionOfTravel()
REM *** IF odd floor,1 else -1 ***
IF ((playerat-1)/10) mod 2 = 0
result = 1
ELSE
result = -1
ENDIF
ENDFUNCTION result
This function returns 1 if the character is on an odd floor and -1 if on an even floor.
Activity 42.22
Add the above function to your program as a Level 4 routine.
Fixing RepositionCamera()
As the player's character travels along a corridor it is moving along the z-axis. On
odd corridors (1, 3, 5, etc.) it is moving in a positive direction (its z-ordinate
becoming greater); on even corridors it is travelling in a negative direction. So, if
the camera is to be behind the character it must have a lower z-ordinate than the
character when travelling along odd corridors and a higher z-ordinate when moving
along even corridors (see FIG-42.4).
FIG-42.4 The Camera's Position and Direction of Travel
Odd level Even level
Camera 1 Camera 1
-z +z
In fact, the camera's z-ordinate should be 7 units less than the character's when
travelling along odd corridors and 7 units more when travelling along even
corridors. The direction in which the camera is pointing should be +5 z-units for
odd corridors and -5 z-units for even corridors.
So, to reposition the camera when we reach a new level, we need to take into account
what level we are on. This requires the following changes to the

RepositionCamera() function:
REM *** IF odd corridor, camera -7 z-units ***
IF FindDirectionOfTravel() = 1
POSITION CAMERA 1, x#+1,y#+5,z#-7
POINT CAMERA 1,x#+1,y#+2,z#+5
ELSE
REM *** Even corridor, +7 z-units ***
POSITION CAMERA 1, x#+1,y#+5,z#+7
POINT CAMERA 1, x#+1,y#+2,z#-5
ENDIF
ENDFUNCTION
Notice that the only difference between the two options is the sign of the 7 and 5
values when calculating the z-ordinates. Since FindDirectionOfTravel() returns 1
or -1 we could rewrite the function without the IF statement as:
POSITION CAMERA 1, x#+1,y#+5,z# -(7*FindDirectionOfTravel())
POINT CAMERA 1, x#+1,y#+2,z# + (5*FindDirectionOfTravel())
ENDFUNCTION
Activity 42.23
Modify your RepositionCamera() function as described above.
Fixing MovePlayer()
The camera and character are both moved in the MovePlayer() function, so we also
need to take the direction of travel into account in that routine.
IF AtEndOfLevel()
REM *** Move to next level ***
UseElevator()
ELSE
REM *** Update character's position ***
INC playerat
REM *** Find direction of travel ***
direction = FindDirectionOfTravel()
POSITION OBJECT playerobj,x#,y#,
Ä
z#+move#*direction
POSITION CAMERA 1,x#+1, y#+5,
Ä
z#+(move#-7*direction)

WAIT 100
NEXT move#
ENDIF
NEXT squares
REM *** IF stopped at elevator, use it ***
IF board(playerat) <> 0 AND AtEndOfLevel() = 0
UseElevator()
ENDIF
REM *** Stop animation ***
ENDFUNCTION
Activity 42.24
Update your MovePlayer() routine to match the code given.
Fixing UseElevator()
When the character uses an elevator, the angle of turn depends on which direction
the character is facing, so we need to modify the parameter of the calls to
TurnPlayer() in the UseElevator() function. Both calls should be re-coded as:
TurnPlayer(-90*FindDirectionOfTravel())
Activity 42.25
Update your UseElevator() routine to match the changes given above.
Fixing MovePlayerToElevator()
How the player moves in line with the elevator's platform is also dependent on the
direction in which the character is travelling. In MovePlayerToElevator() the line
FOR z# = OBJECT POSITION Z(playerobj) TO
Älifts(board(playerat)).zord STEP 0.2
must be changed to
FOR z# = OBJECT POSITION Z(playerobj) TO
Älifts(board(playerat)).zord STEP 0.2*FindDirectionOfTravel()
Activity 42.26
Update your MovePlayerToElevator() routine to match the change given

above.
Fixing MoveElevator()
The camera moves in synchronisation with the elevator, but which side of the
character the camera is on depends on the floor on which the elevator starts its
journey, so MoveElevator() must have the line
POSITION CAMERA 1, playerx#,OBJECT POSITION Y(playerobj)+5,

Ä playerz#-7

changed to
POSITION CAMERA 1, playerx#,OBJECT POSITION Y(playerobj)+5,

Äplayerz#-(7*FindDirectionOfTravel())
Activity 42.27
Change your MoveElevator() function as described.
Try out the game.
Adding EndGame()
Only one routine left to code! The EndGame() function will delete all objects on
the screen and then show a Rotating 3D version of the text "GAME OVER" for 5
seconds before the program terminates.
The routine requires the following code:
REM *** Delete all existing objects ***
DELETE OBJECTS 1,30
REM *** Load and position text model ***
LOAD OBJECT "endtext.x",1
REM *** Rotate text for 5 seconds ***
now = TIMER()
WHILE TIMER() - now < 5000
PITCH OBJECT UP 1,1
ENDWHILE
ENDFUNCTION
Activity 42.28
Add the EndGame() code to your program and test the completed game.
Game Review
We've created the most complex code so far - but have we created a game?
There's certainly not a great deal of player interaction in this version of the game
and there are no skills involved in achieving the game's goal, but we could do things
to improve it.
For a start we could have multiple players and then there is at least someone to
compete against.
Rather than use the dice to decide the number of squares moved, we could ask the
player up to six questions. Movement would then be based on the number of correct
questions.
We could add some animated models on some of the squares. These need not have
any purpose but would add to the overall eye candy effect.

Solutions REM **********************************
Activity 42.1
REM ********************************** REM **********************************
REM *** Structures *** GLOBAL DIM Lifts(27) AS ElevatorType
REM ********************************** GLOBAL DIM board(100) AS INTEGER
REM *** Elevator details *** GLOBAL playerat AS INTEGER
TYPE ElevatorType REM **********************************
ID AS INTEGER REM *** Constants ***
startsquare AS INTEGER REM **********************************
endsquare AS INTEGER REM *** Object constants ***
floors AS INTEGER `(-ve for downwards) #CONSTANT boardobj 1
zord AS INTEGER #CONSTANT playerobj 2
ENDTYPE #CONSTANT diceobj 3
REM ********************************** #CONSTANT firstliftobj 4 ÌDs 4-30
REM ********************************** REM **********************************
GLOBAL DIM Lifts(27) AS ElevatorType REM *** Main game logic
GLOBAL DIM board(100) AS INTEGER REM **********************************
GLOBAL playerat AS INTEGER SetUpGame()
REM ********************************** REPEAT
REM *** Constants *** thrown = RollDice()
REM ********************************** MovePlayer(thrown)
REM *** Object constants *** UNTIL playerat = 100
#CONSTANT boardobj 1 EndScreen()
#CONSTANT playerobj 2 END
#CONSTANT diceobj 3 REM **********************************
#CONSTANT firstliftobj 4 ÌDs 4-30 REM *** Level 1 routines ***
REM **********************************
REM ********************************** FUNCTION SetUpGame()
REM *** Main game logic *** InitialiseData()
REM ********************************** InitialiseVisuals()
SetUpGame() ENDFUNCTION
REPEAT
thrown = RollDice() FUNCTION RollDice()
MovePlayer(thrown) ENDFUNCTION 1
EndScreen() FUNCTION MovePlayer(thrown)
END ENDFUNCTION
REM ********************************** FUNCTION EndScreen()

REM *** Level 1 routines *** ENDFUNCTION
REM **********************************
FUNCTION SetUpGame() REM **********************************
ENDFUNCTION REM *** Level 2 routines ***
REM **********************************
FUNCTION RollDice() FUNCTION InitialiseData()
ENDFUNCTION 1 InitialiseLifts()
InitialiseBoard()
FUNCTION MovePlayer(thrown) REM *** Start player at square 1 ***
ENDFUNCTION playerat = 1
FUNCTION EndScreen() RANDOMIZE TIMER()
No solution required. REM **********************************

REM **********************************
Activity 42.3 FUNCTION InitialiseLifts()
ENDFUNCTION
REM **********************************
REM *** Structures *** FUNCTION InitialiseBoard()
REM ********************************** ENDFUNCTION
TYPE ElevatorType
ID AS INTEGER Activity 42.4
endsquare AS INTEGER Level 3 should now contain the following code:
zord AS INTEGER REM **********************************
ENDTYPE REM *** Level 3 routines ***
REM **********************************

FUNCTION InitialiseLifts() LoadBoard()
REM *** Lifts start at LoadPlayerCharacter()
DATA 10,13,18,20,24,25,30,34,35 LoadDice()
DATA 39,40,49,50,53,54,59,60,65 PositionCameras()
DATA 68,70,72,75,80,83,87,90,97 REM *** Increase ambient light ***
`Lifts end at SET AMBIENT LIGHT 50
DATA 11,33,43,21,64,56,31,27,55 ENDFUNCTION
DATA 19,41,12,51,48,67,79,61,76
DATA 88,71,52,95,81,58,74,91,77 The following code should have been added to level 3
`Floors covered
DATA 1,2,3,1,4,3,1,-1,2
DATA -2,1,-3,1,-1,1,2,1,1
DATA 2,1,-2,2,1,-3,-1,1,-2
ENDFUNCTION
`Xord
DATA 47,23,-27,-47,-17,-7,47,13,3
DATA -37,-47,33,47,23,13,-37,-47,-7
ENDFUNCTION
DATA 23,47,33,3,-47,-27,13,47,-17
REM *** Store lift IDs (31-57)***
FUNCTION LoadDice()
FOR c = 1 TO 27
ENDFUNCTION
lifts(c).ID = c + firstliftobj -1
NEXT c
REM *** Store start square ***
ENDFUNCTION
FOR c = 1 TO 27
READ sq
FUNCTION SetLights()
ENDFUNCTION
NEXT c
REM *** Store end square ***
FOR c = 1 TO 27
READ fin Activity 42.6
NEXT c No solution required.
FOR c = 1 TO 27
READ fl Activity 42.7
NEXT c No solution required.
REM *** Store x-ord of lift ***
FOR c = 1 TO 27
READ x Activity 42.8
lifts(c).zord = x
NEXT c The Level 1 code should read as follows:
ENDFUNCTION
REM **********************************
FUNCTION InitialiseBoard() REM *** Level 1 routines ***
`Square empty (0) or lift (lift ID) REM **********************************
DATA -1,-1,-1,-1,-1, -1,-1,-1,-1,1 FUNCTION SetUpGame()
DATA -1,-1,2,-1,-1, -1,-1,3,-1,4 InitialiseData()
DATA -1,-1,-1,5,6, -1,-1,-1,-1,7 InitialiseVisuals()
DATA -1,-1,-1,8,9, -1,-1,-1,10,11 ENDFUNCTION
DATA -1,-1,-1,-1,-1, -1,-1,-1,12,13
DATA -1,-1,14,15,-1, -1,-1,-1,16,17 FUNCTION RollDice()
DATA -1,-1,-1,-1,18, -1,-1,19,-1,20 REM *** Make dice spin ***
DATA -1,21,-1,-1,22, -1,-1,-1,-1,23 FOR c = 0 TO 100
DATA -1,-1,24,-1,-1, -1,25,-1,-1,26 PITCH OBJECT UP diceobj,15`RND(15)+1
DATA -1,-1,-1,-1,-1, -1,27,-1,-1,-1 TURN OBJECT LEFT diceobj,15`RND(15)+1
REM *** Store data in board array *** ROLL OBJECT LEFT diceobj,15`RND(15)+1
FOR c = 1 TO 100 WAIT 1
READ board(c) NEXT c
IF board(c) = -1 REM *** Generate return value ***
board(c) = 0 value = RND(5)+1
ENDIF REM *** Make dice show value ***
NEXT c SELECT value
ENDFUNCTION CASE 1
ENDCASE
Activity 42.5 CASE 2
In level 2, the code ENDCASE
CASE 3
FUNCTION InitialiseVisuals() ROTATE OBJECT diceobj,0,0,-90
ENDFUNCTION ENDCASE
CASE 4
should have been replaced by ENDCASE
CASE 5
FUNCTION InitialiseVisuals() ROTATE OBJECT diceobj,0,0,0
REM *** Set screen resolution *** ENDCASE
SET DISPLAY MODE 1280,1024,32 CASE 6

ROTATE OBJECT diceobj,90,0,0 The DarkBASIC Pro equivalent is
ENDCASE
ENDSELECT IF (board(playerat) <> 0) AND
WAIT KEY
ENDFUNCTION value
Ä(NOT AtEndOfLevel())
UseElevator()
ENDIF
FUNCTION MovePlayer(playerID, move)
ENDFUNCTION
Notice that extra parentheses are required to conform to
FUNCTION EndScreen() the language's syntax.
ENDFUNCTION
The new version of MovePlayer() is:
Activity 42.9 FUNCTION MovePlayer(thrown)

REM *** Find out character's current
The character moves too quickly and the camera is left position ***
behind. x# = OBJECT POSITION X(playerobj)
Activity 42.10 REM *** Animate the character ***
The code for MovePlayer() should now be: REM *** Move char 1 square at a time ***
FUNCTION MovePlayer(thrown) IF AtEndOfLevel()
REM *** Find char's current position *** UseElevator()
x# = OBJECT POSITION X(playerobj) ENDIF
y# = OBJECT POSITION Y(playerobj) FOR move# = 1 TO 10 STEP 0.5
POSITION OBJECT playerobj,x#,y#,
Ä
REM *** Animate the character *** z#+ (squares-1)*10+move#
LOOP OBJECT playerobj POSITION CAMERA 1,x#+1, y#+5,
REM *** Move char a square at a time *** Ä
z#+ (squares-1)*10+move#-7
FOR squares = 1 TO thrown WAIT 100
FOR move# = 1 TO 10 STEP 0.5 NEXT move#
POSITION OBJECT playerobj,x#,y#, REM *** Update char's position ***
Ä
z#+ (squares-1)*10+move# INC playerat
WAIT 100 NEXT squares
NEXT move# REM *** Stop animation ***
NEXT squares STOP OBJECT playerobj
REM *** Stop animation *** SET OBJECT FRAME playerobj,1
STOP OBJECT playerobj IF (board(playerat) <> 0) AND
SET OBJECT FRAME playerobj,1 Ä
(NOT AtEndOfLevel())
REM *** Update character's position *** UseElevator()
playerat = playerat + thrown ENDIF

The camera now follows the character as it moves. The new test stubs in level 3 should be:
After a few moves the character moves off the board! FUNCTION MovePlayerToLift()
ENDFUNCTION
Activity 42.12 FUNCTION TurnPlayer(angle)

ENDFUNCTIOM
The following code should be added at the end of the
FUNCTION MoveOntoPlatform()
Level 2 section.
ENDFUNCTIOM
FUNCTION AtEndOfLevel() FUNCTIONMoveElevator()
IF playerat mod 10 = 0 ENDFUNCTIOM 1
result = 1
ELSE FUNCTION MoveOffPlatform()
result = 0 ENDFUNCTIOM
ENDIF
ENDFUNCTION result FUNCTION RepositionCamera()
ENDFUNCTIOM
FUNCTION UseElevator()
ENDFUNCTION FUNCTION ReturnElevator(liftnum)
ENDFUNCTIOM
Activity 42.13
The expression required is:
Activity 42.15
The code for MovePlayerToElevator() is:
board(playerat) <> 0
FUNCTION MovePlayerToLift()

REM *** Move character to lift *** POSITION OBJECT playerobj,playerx#,
LOOP OBJECT playerobj Ä
playery#+unitsperperc#*move,
x# = OBJECT POSITION X(playerobj) Ä
playerz#
y# = OBJECT POSITION Y(playerobj) POSITION CAMERA 1, playerx#,
FOR z# = OBJECT POSITION Z(playerobj) TO Ä
playery#+unitsperperc#*move+5,
Ä
lifts(board(playerat)).zord STEP 0.2 Ä
playerz#-7
POSITION OBJECT playerobj, x#,y#,z# WAIT 10
WAIT 50 NEXT move
NEXT z# playerat = lifts(liftnum).endsquare
STOP OBJECT playerobj ENDFUNCTION liftnum
ENDFUNCTION
Activity 42.19
Activity 42.16 The routine's code is:
The code for the function is: FUNCTION MoveOffPlatform()
FUNCTION TurnPlayer(angle) x# = OBJECT POSITION X(playerobj)
IF angle < 0 y# = OBJECT POSITION Y(playerobj)
stepsize = -1 z# = OBJECT POSITION Z(playerobj)
ELSE FOR newx# = x# TO x#+5 STEP 0.1
stepsize = 1 POSITION OBJECT playerobj, newx#,y#,z#
ENDIF WAIT 50
startangle = OBJECT ANGLE Y(playerobj) NEXT moves
FOR degree = startangle TO STOP OBJECT playerobj
Ä
startangle + angle STEP stepsize SET OBJECT FRAME playerobj,1
YROTATE OBJECT playerobj, degree ENDFUNCTION
WAIT 1
NEXT angle
The routine's code is:
Activity 42.17
FUNCTION ReturnElevator(liftnum)
The code for the function is: liftID = lifts(liftnum).ID
height = lifts(liftnum).floors*10
FUNCTION MoveOntoPlatform() IF height > 0
LOOP OBJECT playerobj liftdirection = -1
x# = OBJECT POSITION X(playerobj) ELSE
y# = OBJECT POSITION Y(playerobj) liftdirection = 1
z# = OBJECT POSITION Z(playerobj) ENDIF
FOR newx# = x# TO x#-5 STEP -0.1 IF liftdirection = -1
POSITION OBJECT playerobj, newx#,y#,z# liftendoffset = 0
WAIT 50 ELSE
NEXT moves liftendoffset = 100
STOP OBJECT playerobj ENDIF
SET OBJECT FRAME playerobj,1 liftstartoffset = 100 - liftendoffset
ENDFUNCTION unitsperperc# = height / 100.0
FOR move = liftstartoffset TO
Ä
liftendoffset STEP liftdirection
Activity 42.18 OFFSET LIMB liftID,2,0,move,-2
WAIT 10
The code for the function is: NEXT move
ENDFUNCTION
FUNCTION MoveElevator()
liftnum = board(playerat)
liftID = lifts(liftnum).ID Activity 42.21
height = lifts(liftnum).floors*10
IF height > 0 The function should be added at the end of the Level 3
liftdirection = 1 code.
ELSE
liftdirection = -1
When the character moves to the next level, it is facing in
ENDIF
playerx# = OBJECT POSITION X(playerobj) the wrong direction. The camera is also incorrectly
playery# = OBJECT POSITION Y(playerobj) positioned.
playerz# = OBJECT POSITION Z(playerobj)
IF liftdirection = 1
liftendoffset = 100 Activity 42.22
ELSE
liftendoffset = 0 No soluton required.
ENDIF
liftstartoffset = 100 - liftendoffset
unitsperperc# = height / 100.0 Activity 42.23
FOR move = liftstartoffset TO
Ä
liftendoffset STEP liftdirection No soluton required.
OFFSET LIMB liftID,2,0,move,-2

InitialiseData()
Activity 42.24 InitialiseVisuals()
ENDFUNCTION
No soluton required.
FUNCTION RollDice()
REM *** Make dice spin ***
Activity 42.25 FOR c = 0 TO 100
PITCH OBJECT UP diceobj,15`RND(15)+1
No soluton required. TURN OBJECT LEFT diceobj,15`RND(15)+1
ROLL OBJECT LEFT diceobj,15 `RND(15)+1
WAIT 1
Activity 42.26 NEXT c
REM *** Generate value to be returned ***
No soluton required. value = RND(5)+1
REM *** make dice show its value ***
SELECT value
Activity 42.27 CASE 1
No soluton required. ENDCASE
CASE 2
CASE 3
The final version of the program is: ROTATE OBJECT diceobj,0,0,-90
ENDCASE
REM ********************************** CASE 4
REM *** Structures *** ROTATE OBJECT diceobj,0,0,90
REM ********************************** ENDCASE
REM *** Elevator details *** CASE 5
TYPE ElevatorType ROTATE OBJECT diceobj,0,0,0
ID AS INTEGER ENDCASE
startsquare AS INTEGER CASE 6
endsquare AS INTEGER ROTATE OBJECT diceobj,90,0,0
floors AS INTEGER `(-ve for downwards) ENDCASE
zord AS INTEGER ENDSELECT
ENDTYPE WAIT KEY
REM ********************************** ENDFUNCTION value
REM ********************************** FUNCTION MovePlayer(thrown)
GLOBAL DIM Lifts(27) AS ElevatorType REM *** Animate the character ***
GLOBAL DIM board(100) AS INTEGER LOOP OBJECT playerobj
GLOBAL playerat AS INTEGER REM *** Move char 1 square at a time ***
REM ********************************** FOR squares = 1 TO thrown
REM *** Constants *** IF AtEndOfLevel()
REM ********************************** REM *** Move to next level ***
REM *** Object constants *** UseElevator()
#CONSTANT boardobj 1 ELSE
#CONSTANT playerobj 2 REM *** Update char's position ***
#CONSTANT diceobj 3 INC playerat
#CONSTANT firstliftobj 4 REM *** Find char's position ***
`27 lifts IDs 4 - 30 x# = OBJECT POSITION X(playerobj)
REM ********************************** y# = OBJECT POSITION Y(playerobj)
REM *** Main game logic *** z# = OBJECT POSITION Z(playerobj)
REM ********************************** direction = FindDirectionOfTravel()
SetUpGame() FOR move# = 1 TO 10 STEP 0.5
REPEAT POSITION OBJECT playerobj,x#,y#,
thrown = RollDice() Ä z#+move#*direction
MovePlayer(thrown) POSITION CAMERA 1,x#+1, y#+5,
UNTIL playerat > 2 Ä z#+(move#-7)*direction
EndScreen() WAIT 100
END NEXT move#
ENDIF
FUNCTION MakeCamera3() NEXT squares
MAKE CAMERA 3 REM *** IF stopped at lift, use it ***
POSITION CAMERA 3, IF board(playerat) <> 0 AND
ÄOBJECT POSITION X(playerobj)+10, ÄAtEndOfLevel() = 0
ÄOBJECT POSITION Y(playerobj)+5, UseElevator()
ÄOBJECT POSITION Z(playerobj) ENDIF
POINT CAMERA 3, REM *** Stop animation ***
ÄOBJECT POSITION X(playerobj), STOP OBJECT playerobj
ÄOBJECT POSITION Y(playerobj), SET OBJECT FRAME playerobj,1
ÄOBJECT POSITION Z(playerobj) ENDFUNCTION
WAIT KEY
ENDFUNCTION FUNCTION EndScreen()
REM ********************************** DELETE OBJECTS 1,30
REM *** Level 1 routines *** LOAD OBJECT "endtext.x",1
REM ********************************** POSITION OBJECT 1, 0,0,100
FUNCTION SetUpGame() now = TIMER()

WHILE TIMER() - now < 5000 REM *** Store start square for lift ***
PITCH OBJECT UP 1,1 FOR c = 1 TO 27
ENDWHILE READ sq
ENDFUNCTION lifts(c).startsquare = sq
NEXT c
REM ********************************** REM *** Store end square for lift ***
REM *** Level 2 routines *** FOR c = 1 TO 27
REM ********************************** READ fin
FUNCTION InitialiseData() lifts(c).endsquare = fin
InitialiseLifts() NEXT c
InitialiseBoard() REM *** Store floors covered ***
REM *** Start player at square 1 *** FOR c = 1 TO 27
playerat = 1 READ fl
REM *** Seed random number generator *** lifts(c).floors = fl
RANDOMIZE TIMER() NEXT c
ENDFUNCTION REM *** Store z-ord of lift ***
FOR c = 1 TO 27
FUNCTION InitialiseVisuals() READ z
REM *** Set the screen resolution *** lifts(c).zord = z
SET DISPLAY MODE 1280,1024,32 NEXT c
LoadBoard() ENDFUNCTION
LoadPlayerCharacter()
LoadDice() FUNCTION InitialiseBoard()
PositionCameras() REM ***Square empty (0) or
REM *** Increase the ambient light *** Äelevator's location in lifts array ***
SET AMBIENT LIGHT 50 DATA -1,-1,-1,-1,-1, -1,-1,-1,-1,1
ENDFUNCTION DATA -1,-1,2,-1,-1, -1,-1,3,-1,4
DATA -1,-1,-1,5,6, -1,-1,-1,-1,7
FUNCTION AtEndOfLevel() DATA -1,-1,-1,8,9, -1,-1,-1,10,11
IF playerat mod 10 = 0 DATA -1,-1,-1,-1,-1, -1,-1,-1,12,13
result = 1 DATA -1,-1,14,15,-1, -1,-1,-1,16,17
ELSE DATA -1,-1,-1,-1,18, -1,-1,19,-1,20
result = 0 DATA -1,21,-1,-1,22, -1,-1,-1,-1,23
ENDIF DATA -1,-1,24,-1,-1, -1,25,-1,-1,26
ENDFUNCTION result
DATA -1,-1,-1,-1,-1, -1,27,-1,-1,-1
FUNCTION UseElevator() REM *** Store this data in board ***
REM *** Stop character movement *** FOR c = 1 TO 100
STOP OBJECT playerobj READ board(c)
SET OBJECT FRAME playerobj,1 IF board(c) = -1
MovePlayerToElevator() board(c) = 0
TurnPlayer(-90*FindDirectionOfTravel()) ENDIF
MoveOntoPlatform() NEXT c
liftnum = MoveElevator() ENDFUNCTION
TurnPlayer(-180)
MoveOffPlatform() FUNCTION LoadBoard()
TurnPlayer(-90*FindDirectionOfTravel()) REM *** Load, scale and model board ***
RepositionCamera() LOAD OBJECT "board.x",1
ReturnElevator(liftnum) SCALE OBJECT 1, 1000,1000,1000
REM *** Resume character movement *** YROTATE OBJECT 1, 90
LOOP OBJECT playerobj POSITION OBJECT 1,5,50,0
ENDFUNCTION REM *** Make black areas invisible ***
REM *** These are holes in the floor ***
REM ********************************** REM *** through which the lifts pass ***
REM *** Level 3 routines *** SET OBJECT TRANSPARENCY 1,1
REM ********************************** AddElevators()
FUNCTION InitialiseLifts() ENDFUNCTION
DATA 10,13,18,20,24,25,30,34,35 FUNCTION LoadPlayerCharacter()
DATA 39,40,49,50,53,54,59,60,65 REM *** Load, scale and place char ***
DATA 68,70,72,75,80,83,87,90,97 LOAD OBJECT "H-android-move.x",playerobj
REM *** Lifts end at square *** SCALE OBJECT playerobj,150,150,150
DATA 11,33,43,21,64,56,31,27,55 YROTATE OBJECT playerobj,180
DATA 19,41,12,51,48,67,79,61,76 POSITION OBJECT playerobj,2,0,-45
DATA 88,71,52,95,81,58,74,91,77 REM *** Set the animation play speed ***
REM *** Floors covered by lift *** SET OBJECT SPEED playerobj,15
DATA 1,2,3,1,4,3,1,-1,2 ENDFUNCTION
DATA -2,1,-3,1,-1,1,2,1,1
DATA 2,1,-2,2,1,-3,-1,1,-2 FUNCTION LoadDice()
REM *** z-ord of lifts' position *** REM *** Load, scale and place dice ***
DATA 47,23,-27,-47,-17,-7,47,13,3 LOAD OBJECT "dice.dbo",diceobj
DATA -37,-47,33,47,23,13,-37,-47,-7 POSITION OBJECT diceobj,30,15,45
DATA 23,47,33,3,-47,-27,13,47,-17 SCALE OBJECT diceobj,50,50,50
REM *** Store lift IDs ENDFUNCTION
Ä
(firstliftobj to firstliftobj + 26)***
FOR c = 1 TO 27 FUNCTION PositionCameras()
lifts(c).ID = c + firstliftobj - 1 AUTOCAM OFF
NEXT c REM *** Set up player's camera ***

MAKE CAMERA 1 IF liftdirection = 1
SET CAMERA VIEW 1, 0,0,1279,1023 liftendoffset = 100
POSITION CAMERA 1,3,5,-52 ELSE
POINT CAMERA 1,3,2,-40 liftendoffset = 0
REM *** Set up dice's camera *** ENDIF
MAKE CAMERA 2 REM *** Calculate lift start offset ***
POSITION CAMERA 2,50,43,0 liftstartoffset = 100 - liftendoffset
POINT CAMERA 2,30,15,45 REM *** Move platform,char,and camera ***
SET CAMERA FOV 2, 10 FOR move = liftstartoffset TO
SET CAMERA VIEW 2,1100,20,1220,120 Äliftendoffset STEP liftdirection
ENDFUNCTION OFFSET LIMB liftID,2,0,move,-2
POSITION OBJECT playerobj,playerx#,
FUNCTION MovePlayerToElevator() ÄOBJECT POSITION Y(playerobj)+incmove#
REM *** Move character to lift *** Ä,playerz#
LOOP OBJECT playerobj POSITION CAMERA 1, playerx#,
x# = OBJECT POSITION X(playerobj) ÄOBJECT POSITION Y(playerobj)+5,
FOR z# = OBJECT POSITION Z(playerobj)
Äplayerz#-(7*FindDirectionOfTravel())
ÄTO lifts(board(playerat)).zord STEP

WAIT 10
Ä
NEXT move
0.2*FindDirectionOfTravel() playerat = lifts(liftnum).endsquare
POSITION OBJECT playerobj, x#,y#,z# ENDFUNCTION liftnum
WAIT 50
NEXT z# FUNCTION MoveOffPlatform()
STOP OBJECT playerobj LOOP OBJECT playerobj
SET OBJECT FRAME playerobj,1 x# = OBJECT POSITION X(playerobj)
ENDFUNCTION y# = OBJECT POSITION Y(playerobj)
FUNCTION TurnPlayer(angle) FOR newx# = x# TO x#+5 STEP 0.1
IF angle < 0 POSITION OBJECT playerobj, newx#,y#,z#
stepsize = -1 WAIT 1
ELSE NEXT moves
stepsize = 1 STOP OBJECT playerobj
ENDIF SET OBJECT FRAME playerobj,1
startangle = OBJECT ANGLE Y(playerobj) ENDFUNCTION
FOR degree = startangle TO startangle
Ä+ angle STEP stepsize FUNCTION ReturnElevator(liftnum)
YROTATE OBJECT playerobj, degree liftID = lifts(liftnum).ID
WAIT 1 height = lifts(liftnum).floors*10
NEXT angle IF height > 0
ENDFUNCTION liftdirection = -1
ELSE
FUNCTION MoveOntoPlatform() liftdirection = 1
LOOP OBJECT playerobj ENDIF
x# = OBJECT POSITION X(playerobj) IF liftdirection = -1
y# = OBJECT POSITION Y(playerobj) liftendoffset = 0
z# = OBJECT POSITION Z(playerobj) ELSE
FOR newx# = x# TO x#-5 STEP -0.1 liftendoffset = 100
POSITION OBJECT playerobj, newx#,
Ä
ENDIF
y#,z# liftstartoffset = 100 - liftendoffset
WAIT 50 unitsperperc# = height / 100.0
NEXT moves FOR move = liftstartoffset TO
STOP OBJECT playerobj Äliftendoffset STEP liftdirection
SET OBJECT FRAME playerobj,1 OFFSET LIMB liftID,2,0,move,-2
ENDFUNCTION WAIT 10
NEXT move
FUNCTION MoveElevator() ENDFUNCTION
REM *** Get lift's position in lifts ***
liftnum = board(playerat) FUNCTION RepositionCamera()
REM *** Get lift's ID *** REM *** Get position of character ***
liftID = lifts(liftnum).ID x# = OBJECT POSITION X(playerobj)
REM *** Calc number of world units to be
Ä
moved *** z# = OBJECT POSITION Z(playerobj)
height = lifts(liftnum).floors*10 REM *** IF odd corridor,
REM ***Determine direction (up/down)*** Äcamera -7 z-units ***
IF height > 0 IF FindDirectionOfTravel() = 1
liftdirection = 1 POSITION CAMERA 1, x#+1,y#+5,z#-7
ELSE POINT CAMERA 1,x#+1,y#+2,z#+5
liftdirection = -1 ELSE
ENDIF REM *** Even corridor, +7 z-units ***
REM *** Get char position *** POSITION CAMERA 1, x#+1,y#+5,z#+7
playerx# = OBJECT POSITION X(playerobj) POINT CAMERA 1, x#+1,y#+2,z#-5
playery# = OBJECT POSITION Y(playerobj) ENDIF
playerz# = OBJECT POSITION Z(playerobj) ENDFUNCTION
REM *** Calc player's move per offset
Ä
unit *** REM **********************************
incmove# = height/100.0 REM *** Level 4 ***
REM *** IF lift up, final offset 100 *** REM **********************************

FUNCTION AddElevators()
FOR c = 2 TO 27
NEXT c
FOR c = 1 TO 27
REM *** Rotate lift and scale to
Ämatch floors ***
SCALE OBJECT liftID,100,
Ä10*ABS(lifts(c).floors),100
REM *** place up lifts at start
Äsquare ***
IF lifts(c).floors> 0
Ä(lifts(c).startsquare-1)/10*10,
Älifts(c).zord
ELSE
REM *** And down ones at finish
Äsquare ***
Ä(lifts(c).endsquare-1)/10*10,
Älifts(c).zord
REM *** down lifts need platform
Äat top ***
ENDIF
NEXT c
ENDFUNCTION
FUNCTION FindDirectionOfTravel()
REM *** IF odd floor,1 else -1 ***
IF ((playerat-1)/10) mod 2 = 0
result = 1
ELSE
result = -1
ENDIF
ENDFUNCTION result

43
Binary Space Partitioning Concepts
PK3 Files
Loading a BSP Model
Camera/BSP Collisions
Object/BSP Collisions
DarkBASIC Pro: Handling BSP Models 1163

Binary Space Partitioning
Introduction
A Binary Space Partitioning Tree (BSP tree) is a method of storing data about 3D
objects. It contains details about the order of the polygons that make up a complex
3D scene, storing information on how the position of polygons relate to each other.
This information is stored in a family-tree-type structure (hence the tree part of the
name) with polygons which lie in front of a specified plane on one side of the tree
and those behind the plane on the other side.
Storing the polygons that go to make up a complex 3D model in this way makes
the computer's task of culling and rendering much easier, since it can use the
information about polygon ordering that is held in the BSP structure to efficiently
determine which polygons have to be drawn onto the screen and which are hidden
behind other polygons and hence may be ignored.
We could create a BSP file containing a simple object such as a weapon or

spacecraft, but it is more typically used to contain whole worlds or levels, with
multiple buildings, corridors, rooms, etc.
Although BSP files contain the basic geometry of a world, it does not contain other
necessary details such as the images used to texture surfaces. These will be held in
separate files.
Often BSP files are stored in an archive file (that is, a file which contains a collection
of other files. ZIP and RAR files are also archive files). By using an archive file, all
the files required by the model can be stored as a single entity and extracted when
required. Such archive files have the extension .PK3.
DarkBASIC Pro has the ability to load files of this format because several major
games make use of them, notably Quake and Half-Life, to build the world in which
the player is immersed.
You may hear certain terms, such as world, map, or level, used to describe these
complex models.
Ø World refers to the complete environment in which the player's character

can wander.
Ø Map is often used interchangeably with world, but in fact, a world may
consist of several different maps, each map containing the 3D elements
of a particular section of a world.
Ø Level is a term originally derived from the old game of Dungeons and
Dragons which would often take place in a dungeon with sets of stairs
leading down to lower and lower levels. Although it is often used in the
same context, with a level consisting of rooms, corridors and stairs, it is
also used more generally to refer to different areas or different levels of
difficulty within a game.
It is possible to use the level editor which comes with a commercial game to create
your own playing environment and then import the resulting BSP file into your
DarkBASIC Pro program.
1164 DarkBASIC Pro: Handling BSP Models

Creating a BSP File
Although it's most likely that a level-editor such as the one that comes free with
Half-Life will be used to create a BSP world, any normal DirectX file (.X file) can
be converted to BSP format using the BSP compiler that comes with DarkBASIC
Pro (see FIG-43.1).
FIG-43.1
Accessing the BSP
Compiler
In the dialog box that the application creates, enter the name of the .X file to be
converted (see FIG-43.2).
FIG-43.2
Specifying the File to be
Converted
This program creates only the BSP file, not the archive file containing related texture
images used by the model.
Using BSP Files

The LOAD BSP Statement
BSP files are usually held, along with any texture files required, in a compressed
file with a .PK3 extension. To extract the BSP model from the file we use the LOAD
BSP statement which has the format shown in FIG-43.3.
FIG-43.3
The LOAD BSP
LOAD BSP pk3filename , bspfilename
Statement

In the diagram:
pk3filename is a string giving the name of the archive file

containing the BSP file. This string may contain
path information.
bspfilename is a string giving the name of the BSP file held

within the PK3 file. Only one BSP model can be
in use at any time.
For example, if the file world.pk3 contains the file castle.bsp then, assuming the
file is in the current project's folder, we could load the castle model using the line:
LOAD BSP "world.pk3","castle.bsp"
The castle model will be loaded and the textures, which should be in the world.pk3
file, will be applied automatically to the model.
When a BSP file has not been archived within a PK3 file, then pk3filename should
be an empty string. For example, if the BSP file dungeon.bsp is in the current
project's folder, we could load the model using the line:
LOAD BSP "","dungeon.bsp"
With no PK3 file to search for the texture images, DarkBASIC Pro will look for
those images in the current directory. So, if the model in dungeon.bsp uses texture
images cobblesstones.jpg and stonewall.jpg, these two files must be in the current
project folder along with dungeon.bsp.
A loaded BSP model cannot be repositioned, resized, or rotated. Only a single BSP
file can be loaded.
The program in LISTING-43.1 loads the BSP file that comes in one of the examples
packaged with DarkBASIC Pro.
LISTING-43.1
Loading a BSP Model ScreenSetUp()
REM *** Load BSP ***

LOAD BSP "","ikzdm1.bsp"
REM *** Use camera to move around ***

DO
LOOP

WAIT KEY
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
The ikzdm1.bsp file can be found in DarkBASIC Pro's Dark Basic Software\Dark
Basic Professional\Help\Tutorials\World folder.

Activity 43.1
Type in the program given in LISTING-43.1 (bsp01.dbpro).
Before you try running the program, copy the BSP file ikzdm1.bsp, and the
textures it requires, CYGRASS.bmp and CYWALLT.bmp, into the project's
folder.
What happens when you attempt to move through the wall?
The SET BSP CAMERA COLLISION Statement

To stop the camera moving through parts of our model, we need to use the SET
BSP CAMERA COLLISION statement. This statement has the format shown in
FIG-43.4.
FIG-43.4
The SET BSP CAMERA COLLISION Statement
SET BSP CAMERA COLLISION collindex , camno , radius , response
In the diagram:
collindex is an integer value giving an ID to the collision

details.
camno is an integer value identifying the camera whose

collision properties are to be set.

sphere around the camera.
response -1, 0, or 1. This value determines how the camera

reacts to a collision.
-1 : collision ignored (the camera still passes
through solid objects).
0 : the camera slides along the object hit.
1 : the camera halts when a collision occurs.
Activity 43.2
Add a SET BSP CAMERA COLLISION statement to your previous program

immediately before the DO statement.
Set the collindex value to 1; camno to zero; radius to 2; and response to 1.
What happens this time when you attempt to move the camera through the
wall (approach the wall at an angle, rather than straight on)?
Find out how the camera reacts when response is set to -1 and 0.
The SET BSP OBJECT COLLISION Statement

Just as it is necessary to specify BSP/camera collision characteristics, so we also

need to set the characteristic for BSP/object collisions. This is done using the SET
BSP OBJECT COLLISION statement which has the format shown in FIG-43.5.
FIG-43.5
The SET BSP OBJECT COLLISION Statement
SET BSP OBJECT COLLISION collindex , objno , radius , response
In the diagram:

details.

collision properties are to be set.

sphere around the object.
response -1, 0, or 1. This value determines how the object

reacts to a collision.
-1 : collision ignored (the object still passes
through solid objects).
0 : the object slides along the object hit.
1 : the object halts when a collision occurs.
The program in LISTING-43.2 shows a sphere hitting the courtyard floor for each
possible value of response.
BSP/Object Collision LOAD BSP "","ikzdm1.bsp"
FOR response = -1 TO 1
POINT OBJECT 1,0,-5,0
SET BSP OBJECT COLLISION 1,1,2,response
time = TIMER()
WHILE TIMER() - time < 7000
MOVE OBJECT 1, 1
ENDWHILE
NEXT c
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 5,0,0
ENDFUNCTION
Activity 43.3
Type in and test the program in LISTING-43.2 (bsp02.dbpro).

Of course, the collision rules apply to any 3D object including loaded 3D models.
Activity 43.4
Write a program (Act4304.dbpro) which loads ikzdm1.bsp. It should then load

the Hivebrain walking model (scaling it 200% in each dimension), placing it at
(0,-32,0).
The model should have a collision radius of 2 and implement sliding collision.
Point the alien at (-50,-32,100) and start it walking. Keep the camera on the
model at all times.
The SET BSP CAMERA COLLISION RADIUS Statement

Although we must specify the collision sphere's radius when using the SET BSP
CAMERA COLLISION statement, that sphere can be modified and distorted using
the SET BSP CAMERA COLLISION RADIUS statement. The statement has the
format given in FIG-43.6.
FIG-43.6
The SET BSP CAMERA SET BSP CAMERA COLLISION RADIUS collindex
COLLISION RADIUS
Statement
, camno , x , y , z
In the diagram:

details.
camno is an integer value identifying the camera whose

collision volume properties are to be set.
x,y,z is a sequence of real values giving the radii of the

ovoid along the x, y and z axes.
Since the radius of the collision volume can be adjusted separately in all three
dimensions, it need no longer be spherical in shape. For example, we could change
the collision volume for camera zero to an ovoid shape using the line:
SET BSP CAMERA COLLISION RADIUS 1,0,5,10,5
The SET BSP OBJECT COLLISION RADIUS Statement

Similarly, the collision volume of an object can be set using the SET BSP OBJECT
COLLISION RADIUS statement which has the format shown in FIG-43.7.
FIG-43.7
SET BSP OBJECT COLLISION RADIUS collindex
The SET BSP OBJECT
COLLISION RADIUS
Statement
, objno , x , y , z

In the diagram:

details.

collision volume properties are to be set.
x,y,z is a sequence of real values giving the radii of the

ovoid along the x, y and z axes.
The SET BSP COLLISION HEIGHT ADJUSTMENT Statement

Normally a collision sphere is centred on the camera or object to which it is attached,
but it is possible to reposition this sphere either above or below this normal position
(see FIG-43.8).
FIG-43.8
Camera Camera
A Repositioned Collision
Sphere
height
Normally, the collision

sphere is centred on
its associated object
But it can be offset
above or below that
object
Offsetting the collision sphere is achieved using the SET BSP COLLISION
HEIGHT ADJUSTMENT statement which has the format given in FIG-43.9.
FIG-43.9
The SET BSP COLLISION HEIGHT ADJUSTMENT Statement
SET BSP COLLISION HEIGHT ADJUSTMENT collindex , height
In the diagram:
collindex is an integer value of the collision ID previously

assigned.
height is a real number giving the height of the collision

sphere from the camera. A negative value means
that the collision sphere is below the camera.
For example, if we had previously set the camera's collision details using the line
SET BSP CAMERA COLLISION 1,0,2,1
we could move the camera's collision sphere far below ground level using the line
SET BSP COLLISION HEIGHT ADJUSTMENT 1,-50

and since the collision sphere is no longer anywhere near the BSP model, the camera
will be able to move through the walls once more.
Activity 43.5
Add the above statement to your program bsp01.dbpro (after the SET BSP
CAMERA POSITION statement), and check that the camera can again move
through the walls. Remove the statement after testing the program.
The SET BSP COLLISION THRESHOLD Statement

When a sliding collision causes an object or a camera to move downwards, the
sensitivity of the movement can be adjusted using the SET BSP COLLISION
THRESHOLD statement which has the format shown in FIG-43.10.
FIG-43.10
The SET BSP COLLISION SET BSP COLLISION THRESHHOLD collindex , sensitivity
THRESHOLD Statement
In the diagram:

assigned.
sensitivity is a real number giving the sensitivity required to

initiate the sliding fall. The default value is zero,
which causes falls to happen due to the slightest
change in surface height. A higher value will
delay the fall until the object/camera is further
from the new ground level.
The PROCESS BSP COLLISION Statement

Collisions are tested for automatically each time the screen is refreshed, but it is
possible for detection at an earlier point by calling the PROCESS BSP COLLISION
FIG-43.11
The PROCESS BSP PROCESS BSP COLLISION collindex
COLLISION Statement
In the diagram:

assigned.
The SET BSP COLLISION OFF Statement

Detection of a specific BSP collision (camera or objects) can be switched off using
the SET BSP COLLISION OFF statement which has the format shown in
FIG-43.12.
FIG-43.12
The SET BSP COLLISION SET BSP COLLISION OFF collindex
OFF Statement

In the diagram:
collindex is an integer value specifying the ID of the

collision to be switched off.
Activity 43.6
Modify bsp01.dbpro so that the BSP/camera collision detection is switched

off.
The BSP COLLISION HIT Statement

We can detect if a camera/object has collided with the BSP model using the BSP
COLLISION HIT statement which has the format shown in FIG-43.12.
FIG-43.12
BSP COLLISION HIT ( collindex )
The BSP COLLISION
HIT Statement
integer
In the diagram:

assigned.
The statement returns 1 if a collision has occurred, otherwise zero is returned.
The BSP COLLISION Statement

The coordinates of a collision can be discovered using the BSP COLLISION
FIG-43.13
X
The BSP COLLISION
Statement
BSP COLLISION Y ( collindex )
real
In the diagram:
X,Y,Z Use the option appropriate for the ordinate

required.

assigned.

The SET BSP CAMERA Statement
Normal cameras (as used in programs not containing BSP models) do not operate
correctly with BSP objects. Usually this won't be a problem, since camera zero is
automatically modified when a BSP model is loaded, but if we want to use a second
camera, then it needs to be converted to a BSP camera. This can be done using the
SET BSP CAMERA statement which has the format shown in FIG-43.14.
FIG-43.14
The SET BSP SET BSP CAMERA camno
CAMERA Statement
In the diagram:
camno is an integer value giving the ID of the camera to

be converted to a BSP camera.
The DELETE BSP Statement

A BSP model can be removed from RAM using the DELETE BSP statement which
FIG-43.15
The DELETE BSP DELETE BSP
Statement
Once this statement has been executed a new BSP model may be loaded.
The SET BSP MULTITEXTURING Statement

Some BSP models may have been created using multiple overlapping textures
(known as multitexturing). When recreated on screen, the visual effect created by
multitexturing can be switched on or off using the SET BSP MULTITEXTURING
{
FIG-43.16
ON
The SET BSP SET BSP MULTITEXTURING
MULTITEXTURING
Statement
OFF
Summary
l Binary Space Partitioning (BSP) is a method of storing 3D models designed to
minimise the number of polygons that have to be processed when recreating a
3D model on the screen.
l BSP files and the texture they use are often archived in a PK3 file.
l BSP files have a .BSP extension.
l DarkBASIC Pro comes with a utility for converting .X files to BSP files.
l BSP files are commonly used for creating level, map or world models.
l Use LOAD BSP to load a BSP file.

l BSP models cannot be moved, resized or rotated.
l Only one BSP model can be loaded.
l Use SET BSP CAMERA COLLISION to stop the camera moving through a
BSP model.
l Use SET BSP OBJECT COLLISION to stop another 3D object passing through
a BSP model.
l Use SET BSP CAMERA COLLISION RADIUS to set the radius of a camera's
collision sphere.
l Use SET BSP OBJECT COLLISION RADIUS to set the radius of an object's
collision sphere.
l Use SET BSP COLLISION HEIGHT ADJUSTMENT to reposition a camera or

object's collision sphere.
l Use SET BSP COLLISION THRESHOLD to set the sensitivity of a sliding

collision.
l Use PROCESS BSP COLLISION to initiate early collision detection.
l Use SET BSP COLLISION OFF to stop detection of BSP collisions.
l Use BSP COLLISION HIT to detect if a camera or object has collided with the
BSP model.
l Use BSP COLLISION to determine the coordiantes at which the collision

occurred.
l Use SET BSP CAMERA to convert additional cameras for use with a BSP
model.
l Use DELETE BSP to delete a BSP model from RAM.
l Use SET BSP MULTITEXTURING to allow the surfaces of a BSP model to

use multiple texture files.

Using a BSP Map
Introduction
This section demonstrates the use of a BSP map as the environment for a first-person
shooter. Your invisible character is equipped with a pistol and an endless supply of
bullets.
Notice how sound effects have been added to increase realism. A simple trick has
been added to give the impression of the bullet being vapourised as it hits the wall.
The Program
The main logic of the program is:
Set up screen
Initialise game
DO
Control player
IF BulletFired THEN
Move bullet
ENDIF
LOOP
This is coded as:
SetUpScreen()
InitialiseGame()
DO
ControlPlayer()
IF BulletFired()
MoveBullet()
ENDIF
LOOP
END
The main routines are:
SetUpScreen This routine initialises the screen resolution

and positions the camera. It also hides the mouse
pointer.
InitialiseGame This routine loads all the BSP maps, models, and
sounds. It also locks the gun to the screen giving
the typical FPS effect.
ControlPlayer This routine allows the player to move the camera

using the arrow keys and also calls FireGun() if
the Enter key is pressed.
FireGun This routine fires the bullet object, positioning

the bullet beside the gun and playing the firing
sound.
BulletFired This routine returns 1 if a bullet is currently

in flight; otherwise zero is returned.

MoveBullet This routine moves the bullet 2 units and
checks for a collision with the BSP world.
If a hit is detected, ShowBulletHit() is called.
ShowBulletHit This routine creates the vapourised bullet effect,

hides the bullet object and plays a bullet hit sound.
The code for the program is given in LISTING-43.3.

LISTING-43.3 REM *** Object Constants ***
#CONSTANT gunobj 1
Using a BSP Map #CONSTANT bulletobj 2
REM *** Sound constants ***
#CONSTANT firesnd 1
#CONSTANT hitsnd 2
SetUpScreen()
InitialiseGame()
DO
ControlPlayer()
IF BulletFired()
MoveBullet()
ENDIF
LOOP
END
BACKDROP OFF
REM *** Hide mouse ***
HIDE MOUSE
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION InitialiseGame()
REM *** Load BSP map ***
REM *** Set up collision parameters ***
SET BSP CAMERA COLLISION 1,0,1,0
REM *** Load the weapon ***
LOAD OBJECT "H-Pistol-Static.x",gunobj
REM *** Position object in front of camera ***
LOCK OBJECT ON gunobj
SCALE OBJECT gunobj,300,300,300
POSITION OBJECT gunobj,0,-1,2
REM *** Create hidden object to use as bullet ***
MAKE OBJECT SPHERE bulletobj,0.1
SET BSP OBJECT COLLISION bulletobj,2,0.1,0
HIDE OBJECT bulletobj
REM *** Load sound file ***
LOAD SOUND "gun 2.wav",gunobj
LOAD SOUND "mortar.wav",hitsnd
ENDFUNCTION
FUNCTION ControlPlayer()
REM *** Control camera ***
REM *** IF Enter pressed, attempt to fire ***
IF INKEY$() = CHR$(13)
FireGun()
ENDIF
ENDFUNCTION

LISTING-43.3 FUNCTION FireGun()
(continued) REM *** If bullet, visible, exit ***
IF BulletFired()
Using a BSP Map EXITFUNCTION
ENDIF
REM *** Position and show bullet ***
SHOW OBJECT bulletobj
POSITION OBJECT bulletobj,CAMERA POSITION X(),
ÄCAMERA POSITION Y()-2, CAMERA POSITION Z()
ROTATE OBJECT bulletobj, CAMERA ANGLE X(),
ÄCAMERA ANGLE Y(),CAMERA ANGLE Z()
REM *** Play sound of gun fire ***
PLAY SOUND firesnd
ENDFUNCTION
FUNCTION BulletFired()
REM *** Bullet travelling if visible ***
result = OBJECT VISIBLE(bulletobj)
ENDFUNCTION result
FUNCTION MoveBullet()
REM *** Move bullet ***
MOVE OBJECT bulletobj,2
REM *** IF it hits wall,show wall being it ***
IF BSP COLLISION HIT(bulletobj)
ShowBulletHit()
ENDIF
ENDFUNCTION
FUNCTION ShowBulletHit()
REM *** Make bullet larger and brighter ***
SCALE OBJECT bulletobj, 1000,1000,1000
SET OBJECT EMISSIVE bulletobj, RGB(255,255,255)
REM *** Show in this state for 10 millisecs ***
WAIT 10
REM *** Hide bullet and darken ***
HIDE OBJECT bulletobj
SET OBJECT EMISSIVE bulletobj, RGB(100,100,100)
REM *** Return to normal size ***
REM *** Play hit sound ***
PLAY SOUND hitsnd
ENDFUNCTION
Activity 43.7
Type in and test the program in LISTING-43.3 (bsp03.dbpro).

Solutions
The camera can move straight through the walls. The main section changes to:
ScreenSetUp()
Activity 43.2 REM *** Load BSP ***
ScreenSetUp() SET BSP CAMERA COLLISION 1,0,2,1
REM *** Load BSP *** SET BSP COLLISION HEIGHT ADJUSTMENT 1,-50
LOAD BSP "","ikzdm1.bsp" REM *** Use camera to move around ***
SET BSP CAMERA COLLISION 1,0,2,1 DO
REM *** Use camera to move around *** CONTROL CAMERA USING ARROWKEYS 0,0.1,0.1
DO LOOP
CONTROL CAMERA USING ARROWKEYS 0,0.1,0.1 REM *** End Program ***
LOOP WAIT KEY
REM *** End Program *** END
WAIT KEY
END
Activity 43.6
SET DISPLAY MODE 1280,1024,32 The main section should be coded as:
BACKDROP ON ScreenSetUp()
AUTOCAM OFF REM *** Load BSP ***
POSITION CAMERA 0,10,-30 LOAD BSP "","ikzdm1.bsp"
POINT CAMERA 0,0,0 SET BSP CAMERA COLLISION 1,0,2,1
ENDFUNCTION SET BSP COLLISION OFF 1
REM *** Use camera to move around ***
This time the camera is stopped when it meets the wall. DO
Using a value of -1 allows the camera to travel through the LOOP
wall. REM *** End Program ***
WAIT KEY
END
A value of zero creates a sliding collision effect when the
camera comes into contact with the wall.
Activity 43.7
Activity 43.4
ScreenSetUp()
REM *** Load Hivebrain ***
SET BSP OBJECT COLLISION 1,1,2,0
POINT OBJECT 1,-50,-32,100
LOOP OBJECT 1
DO
MOVE OBJECT 1,0.1
POINT CAMERA OBJECT POSITION X(1),OBJECT
POSITION Y(1),OBJECT POSITION Z(1)
LOOP
END
BACKDROP ON
AUTOCAM OFF
POSITION CAMERA -120,-20,160
ENDFUNCTION

44
Advanced Terrain Statements
Developing a Terrain-Based Game
Finding the Height of a Point on a Terrain Object
Texturing a Terrain Object
Using a Greyscale Image to Contour a Landscape
Using TERRAIN Statements to Create a Landscape
DarkBASIC Pro : Terrain 1179

Creating Terrain
Introduction
One way to create undulating, ground-like surface is to use the TERRAIN
statements. This approach allows you to use a greyscale image to determine the
contours of the terrain with black areas in the image being at height zero and white
areas in the image corresponding to the maximum height. Shades of grey in the
image create intermediate heights. An example of the effects that can be achieved
FIG-44.1
Creating Terrain
The documented terrain commands are few in number, but several newer statements
which are currently undocumented are also available. We'll start with the
documented statements.
Documented Terrain Statements

The MAKE TERRAIN Statement
A terrain object is created using the MAKE TERRAIN statement. This assigns an
integer value to the terrain and links it to a greyscale image. It is this image that
determines the contours of the terrain. The MAKE TERRAIN statement has the
format shown in FIG-44.2
FIG-44.2
The MAKE TERRAIN MAKE TERRAIN terno , filename z
Statement
In the diagram:
terno is the integer value assigned to the terrain object

being created.
filename is a string giving the name of the greyscale file

used to determine the contours of the terrain
object. The file must be square, with its width
and height being equal.
1180 DarkBASIC Pro : Terrain

The terrain’s dimensions will change to match those of the image being used as the
contour data. For example, a 512 by 512 image will create a terrain object which is
exactly 512 by 512. Initially, the bottom-left corner of the terrain is at position
(0,0,0).
The program in LISTING-44.1 uses image shown in FIG-44.3 to control the

contours of the terrain created.
FIG-44.3
Image Used to
Contour Terrain
Since the image used in this example contains only black and white (no shades of
grey) the terrain contours will be flat (black) or at maximum height (white).
The program changes the default light to a point light in order to produce shadows
over the terrain. The position of the camera can be modified using the arrow keys.
SetUpScreen()
REM *** Create terrain ***

MAKE TERRAIN 1,"contour01.bmp"
LISTING-44.1
Creating a Terrain Object DO
LOOP

END
REM *** Set display mode required ***
AUTOCAM OFF
REM *** Change to point light ***
ENDFUNCTION
Activity 44.1
Type in and test the program given above (terrain01.dbpro).

The contouring works
best if the image’s Change the contour image used to contour02.bmp.
dimensions are a power
of 2. For example, Create you own greyscale image (it must be square) and use it in your
256x256, 512x512, or program to define the contours of the terrain. (Return to using contour02.bmp
1024x1024.
once you've tested your own image.)
The DELETE TERRAIN Statement

When a terrain object is no longer required by a program, it should be deleted using

the DELETE TERRAIN statement which has the format shown in FIG-44.4.
FIG-44.4
The DELETE TERRAIN DELETE TERRAIN terno
Statement
In the diagram:
terno is the integer value previously assigned to terrain

object.
This will free up the RAM space occupied by the terrain object.
The POSITION TERRAIN Statement

Initially, the terrain is placed in 3D space with its bottom-left corner at position
(0,0,0) and the top-right corner at point (x,0,x) where x is the width of the contour
image. For example, if you are using a 512 x 512 greyscale image to supply the
contour information, then the terrain's top-right corner will be positioned at
(512,0,512).
If you need to move the terrain elsewhere within the 3D world, you can use the
POSITION TERRAIN statement which has the format shown in FIG-44.5.
FIG-44.5
The POSITION TERRAIN POSITION TERRAIN terno , x , y , z
Statement
In the diagram:
terno is the integer value previously assigned to the

terrain object.
x,y,z is a sequence of 3 real numbers giving the new

position for the top-left corner of the terrain.
For example, we could move terrain object 1 so that its top-left corner is at position
(200,100,-50) using the statement:
POSITION TERRAIN 1,200,100,-50
Activity 44.2
Modify your last program to reposition the centre of your terrain to location
(0,0,0).
(HINT: You’ll need to load the contour image as a bitmap and get its
dimensions.)
Reposition the camera so that it is at the centre of the terrain and at a height of
50.
Copy the camera.dba file we created in Chapter 33 into the current project
folder.
Change camera control to use PointCameraUsingMouse() and

MoveCameraUsingMouse().

The TERRAIN POSITION Statement
The x, y or z ordinate of the top-left corner of a terrain object is returned using the
TERRAIN POSITION statement which has the format shown in FIG-44.6.
FIG-44.6
{}
X
The TERRAIN
POSITION Statement
TERRAIN POSITION Y ( terno )
real
In the diagram:
X,Y,Z Use the appropriate option for the ordinate

required.

terrain object whose position is required.
The statement returns a real number representing the ordinate requested.
For example, to find the coordinates of the top-left corner of terrain object 1, we
could use the statement:
xord# = TERRAIN POSITION X(1)

yord# = TERRAIN POSITION Y(1)
zord# = TERRAIN POSITION Z(1)
Activity 44.3
What statement would be used to display the y ordinate of terrain object 2?
The TEXTURE TERRAIN Statement

As with other 3D objects, you can add a texture to the terrain by linking a loaded
image to the terrain. This is achieved using the TEXTURE TERRAIN statement
which takes the format shown in FIG-44.7.
FIG-44.7
The TEXTURE TEXTURE TERRAIN terno , imgno
TERRAIN Statement
In the diagram:
terno is the integer value previously assigned to terrain

object.
imgno is the integer value previously assigned to the

image object that is to be used to texture the
terrain.

For example, to use the image grid8by8.bmp to texture terrain object 1, we could
use the following code:

TEXTURE TERRAIN 1,2 `Texture terrain 1 with image 2
The result is shown in FIG-44.8.
FIG-44.8
A Textured Terrain
The image used will be tiled automatically if necessary. Make sure a seamless image
is used and the tiling will be invisible.
Activity 44.4
Add an image object (containing the file grass256.jpg) to your previous

program and use this to texture the terrain.
Can the camera move through the terrain?
The GET TERRAIN HEIGHT Statement

With other 3D objects we could use the AUTOMATIC CAMERA COLLISION
statement to prevent the camera moving through what should be solid 3D objects.
Unfortunately, that command does not work with terrain objects. So to solve this
problem we need to make sure that the camera stays above the surface at all times.
We can do this using the GET TERRAIN HEIGHT statement which returns the
height of the terrain at a specific point and adjusting the camera's position
accordingly. The GET TERRAIN HEIGHT statement has the format shown in
FIG-44.9.
FIG-44.9
The GET TERRAIN GET TERRAIN HEIGHT ( terno , x , z )
HEIGHT Statement
real
In the diagram:

terrain object.
x,z are the coordinates of the position within the

terrain whose height is to be returned.

The statement returns a real number representing the y-ordinate value at position
(x,z) on terrain terno. The bottom-left corner of the terrain is taken as position (0,0)
irrespective of the terrain object’s own position within the 3D world (see
FIG-44.10).
FIG-44.10
The Terrain Origin
This is position (0,0),

no matter where the
terrain object is placed
in the 3D world
For example, we could return the height of position (10,20) within terrain 1 using
the line:
height# = GET TERRAIN HEIGHT(1,10,20)
The program in LISTING-44.2 moves a small cube across the terrain (from left to
right). It attempts to keep the cube half buried in the terrain by using the GET
TERRAIN HEIGHT statement. The camera follows the movement.
REM *** Create a terrain object ***
Using the GET MAKE TERRAIN 1,"terra1.bmp"
TERRAIN HEIGHT REM *** Texture terrain ***
Statement LOAD IMAGE "grass.jpg",1
TEXTURE TERRAIN 1,1
REM *** Match terrain coordinates to real coordinates ***
POSITION TERRAIN 1,0,0,511
height# = GET TERRAIN HEIGHT(0,0,50)
REM *** Move cube across terrain ***
FOR c = 0 TO 511
REM *** Display current height ***
SET CURSOR 10,10
PRINT "CURRENT HEIGHT :",height#
REM *** Recalculate cube’s height and reposition camera ***
height# = GET TERRAIN HEIGHT(1,c,50)
POSITION OBJECT 1,c,height#,50
POSITION CAMERA c-5,height#+10,0
POINT CAMERA c,height#,50
WAIT 10
NEXT c
WAIT KEY
END
BACKDROP ON
AUTOCAM OFF
ENDFUNCTION

AUTHOR'S NOTE: In the
Activity 44.5
very few tests I have tried
using this statement, I find
that multiplying the value Type in and test the program given in LISTING-44.2 (terrain02.dbpro).
returned by 0.65 gives a
more accurate estimate of Is the height value accurate?
the height.
The GET TOTAL TERRAIN HEIGHT Statement

The height of the highest point on a terrain object can be discovered using the GET
TOTAL TERRAIN HEIGHT statement which has the format shown in FIG-44.11.
FIG-44.11
The GET TOTAL
GET TERRAIN HEIGHT TOTAL ( terno )
TERRAIN HEIGHT
Statement
real
In the diagram:

AUTHOR'S NOTE: terrain object.
This statement seems to
return the value 100 for
most terrain maps. For example, we could find the highest point on terrain 1 using the statement:
highest# = GET TERRAIN HEIGHT TOTAL(1)
The Advanced Terrain Statements

The advanced terrain statements (which are currently undocumented) offer not only
alternative ways of creating a terrain, but also extra statements which allow us more
control over the terrains we create. As a general rule, these newer statements should
be considered as replacements for the older terrain statements.
The MAKE OBJECT TERRAIN Statement

An alternative way of creating a terrain object is to use the MAKE OBJECT
TERRAIN statement which has the format shown in FIG-44.12.
FIG-44.12
MAKE OBJECT TERRAIN terno
The MAKE OBJECT
TERRAIN Statement
In the diagram:
terno is an integer value specifying the ID to be

assigned to the terrain object being created.
To create terrain object 1 we would use the line:
MAKE OBJECT TERRAIN 1
The terrain created in this way is a normal 3D object (just like a cube or sphere) and
can be operated on by most 3D statements such as POSITION OBJECT.
By itself, MAKE OBJECT TERRAIN is of little use, since it does not create a visible
terrain. We need to add a few more statements...

The SET TERRAIN HEIGHTMAP Statement
Once a terrain object has been created, we need to specify its contour map. This is
done using the SET TERRAIN HEIGHTMAP statement which has the format
shown in FIG-44.13.
FIG-44.13
The SET TERRAIN SET TERRAIN HEIGHTMAP terno , filename
HEIGHT MAP
In the diagram:
terno is an integer value identifying the terrain whose

contour map is to be loaded.
filename is a string giving the name of the contour map

file. The string may include path information.
Any file of 256 or more colours can be used.
For example, assuming we want to use the file terra1.bmp as our contour file for
terrain 1 and that the file is in the project directory, then we could use the statement:
SET TERRAIN HEIGHT MAP 1, "terra1.jpg"
As you know the pixels that make up an image are represented as a sequence of
numbers. It is these numbers that determine the height of any point on the terrain
object; low numbers give low points, high numbers create high points.
The SET TERRAIN SCALE Statement

The third stage in the process of creating a terrain is scaling. By default, a terrain
will be sized to match the dimensions of its contour map. Therefore, a contour map
which is 512 by 512 will create a terrain 512 units wide and 512 units deep. The
greatest height within the terrain will depend on the largest pixel values within the
contour image.
We can override the default sizes and specify a multiplying factor for each
dimension using the SET TERRAIN SCALE statement which has the format given
in FIG-44.14.
FIG-44.14 The SET TERRAIN SCALE Statement
SET TERRAIN SCALE terno , xscale , yscale , zscale
In the diagram:
terno is an integer value giving the terrain to be scaled.
xscale is a real number giving the multiplication factor

to be applied to the width of the terrain.
yscale is a real number giving the multiplication factor

to be applied to the height of the terrain.

zscale is a real number giving the multiplication factor
to be applied to the depth of the terrain.
For example, we could double the width, triple the depth and halve the height of
terrain object 1 using the statement:
SET TERRAIN SCALE 1,2,0.5,3
The SET TERRAIN TEXTURE Statement

To texture a terrain we can use the SET TERRAIN TEXTURE statement. This
statement allows us to specify two image files. One of these is used to create a grain
on the surface of the terrain, while the other supplies the basic texturing image. The
SET TERRAIN TEXTURE statement has the format shown in FIG-44.15.
FIG-44.15 The SET TERRAIN TEXTURE Statement
SET TERRAIN TEXTURE terno , grainfile , texturefile
In the diagram:
terno is an integer value identifying the terrain involved.
grainfile is a string giving the name of the file to be used

to create the underlying surface grain.
texturefile is a string giving the name of the main texture

file.
For example, we could use the file grain.bmp to create the grain on terrain 1 and
the file texture.bmp to texture it using the lines:
LOAD IMAGE "grain.bmp", 1

LOAD IMAGE "texture.bmp",2
SET TERRAIN TEXTURE 1,1,2
The BUILD TERRAIN Statement

Any terrain that we are constructing will become visible only after executing the
BUILD TERRAIN statement which has the format shown in FIG-44.16.
FIG-44.16
The BUILD TERRAIN
BUILD TERRAIN terno { , smthflag
Statement
In the diagram:
smthflag 0 or 1. 0 - gives surface smoothing (this is the

When it appears on the default); 1 - no smoothing.
screen a terrain object
has its bottom-left corner
at position (0,0,0). The program in LISTING-44.3 demonstrates the use of the statements covered so
far in this section. The images used are shown in FIG-44.17. Notice that texture.bmp
is an inverted version of contour.bmp. All images are 512 by 512.

FIG-44.17
The Images Used in
Creating the Terrain
LISTING-44.3 #INCLUDE "camera.dba"

SetUpScreen()
Creating a Terrain Using REM *** Create terrain object ***
the Undocumented MAKE OBJECT TERRAIN 1
Statements REM *** Assign a contour map ***
SET TERRAIN HEIGHTMAP 1, "contour.bmp"
REM *** Reduce the height ***
SET TERRAIN SCALE 1, 1,0.5,1
REM *** Add grain and texture ***
LOAD IMAGE "grain.bmp",1
LOAD IMAGE "texture.bmp",2
REM *** Show terrain ***
BUILD TERRAIN 1,0
DO
LOOP
END
AUTOCAM OFF
ENDFUNCTION
Activity 44.6
Type in and test the program in LISTING-44.3 (terrain03.dbpro).
Make sure the camera.dba file has been copied to the project folder.
If you look closely, you may notice that the texture file has been tiled over the
surface of the terrain. In fact, there's a copy of the image on each square unit of the
terrain.
The SET TERRAIN TILING Statement

We can change how often the texture file is tiled over the surface of the terrain using
the SET TERRAIN TILING statement which has the format shown in FIG-44.18.
FIG-44.18
The SET TERRAIN
TILING Statement
SET TERRAIN TILING terno , size

In the diagram:
size is an integer value giving the size of each tile in

world units.
For example, if our terrain is 512 units square, then we can force a single copy of
the texture image to cover the whole of terrain object 1 using the line:
SET TERRAIN TILING 1,512
Activity 44.7
Modify your last program so that a single copy of texture.bmp covers the
whole of the terrain (the terrain is 512 by 512 units).
Notice that by using an inverted texture file, the contours and textures are an exact
match, the yellow in the texture file covering the raised area of the terrain.
The SET TERRAIN LIGHT Statement

We can create a lighting effect on the terrain, controlling the position, colour and
intensity of the lights, using the SET TERRAIN LIGHT statement which has the
FIG-44.19 The SET TERRAIN LIGHT Statement
SET TERRAIN LIGHT terno , x , y , z ,
red , green , blue , shading
In the diagram:
x,y,z are a set of real values giving the position of the

light source. No actual light is created, but
shading calculations which affect the terrain's
surface assume a light is positioned at this point.
red,green,blue are a set of integer values giving the colour of the

light cast on the terrain.
shading is a real value between 0 and 1 which affects the

darkness of shaded areas of the terrain.
0 - darkest shading
1 - lightest shading
Since no light object is created by this statement, other 3D objects in the scene are
unaffected by the lighting specified.

We could simulate a blue light shining on terrain 1 from position (10,20,200),
creating the darkest shading using the line:
SET TERRAIN LIGHT 1,10,20,200,0,0,255,0
Activity 44.8
Modify the terrain's lighting in your last program so that dark shading is
created for a white light positioned at (-100,10,250).
The SET TERRAIN SPLIT Statement

Normally a terrain is created as a single mesh, but it is possible to split it into a set
of meshes using the SET TERRAIN SPLIT statement which has the format shown
in FIG-44.20.
FIG-44.20
The SET TERRAIN
SET TERRAIN SPLIT terno , value
SPLIT Statement
In the diagram:
value is an integer value giving the number of splits

along each edge of the terrain. Realistically, this
will be in the range 2 to 16.
If the split is set to 4, then the terrain will be split into a 4 by 4 set of meshes (that
is, 16 meshes). The idea behind this is that as a single mesh the terrain can put a
strain on the processor/video card, but as a set of smaller meshes only those meshes
that are currently on screen need to be processed.
We could split terrain 1 into 64 separate meshes using the statement:
SET TERRAIN SPLIT 1, 8
The GET TERRAIN GROUND HEIGHT Statement

While we have the GET TERRAIN HEIGHT for old-style terrains, new-style
terrains (those created using MAKE OBJECT TERRAIN) use the GET TERRAIN
GROUND HEIGHT statement to determine the y-ordinate of a point on the terrain.
FIG-44.21 The GET TERRAIN GROUND HEIGHT Statement
GET TERRAIN GROUND HEIGHT ( terno , x , z )
real
In the diagram:

terrain object.

x,z are the coordinates of the position within the
terrain whose height is to be returned.
The statement returns a real number representing the y-ordinate value at position
(x,z) on terrain terno. The bottom-left corner of the terrain is taken as position (0,0)
irrespective of the terrain object’s own position within the 3D world.
The program in LISTING-44.4 repeats the demonstration of moving a cube across

a terrain. Notice how much more accurate the demonstration is this time.
REM *** Create a terrain object ***
Traversing the Terrain MAKE OBJECT TERRAIN 1
REM *** Add contour map ***
SET TERRAIN HEIGHTMAP 1,"big2.bmp"
REM *** Texture terrain ***
LOAD IMAGE "grain.bmp",1
LOAD IMAGE "grass256.jpg",2
REM *** Makes texture tiles 256 units ***
SET TERRAIN TILING 1, 256
REM *** Show the terrain ***
BUILD TERRAIN 1
height# = GET TERRAIN GROUND HEIGHT(1,0,50)
REM *** Move cube across terrain ***
FOR c = 0 TO 1023
REM *** Display current height ***
SET CURSOR 10,10
PRINT "CURRENT HEIGHT :",height#
REM *** Recalculate cube’s height and reposition camera ***
height# = GET TERRAIN GROUND HEIGHT(1,c,50)
POSITION OBJECT 2,c,height#,50
POSITION CAMERA c-5,height#+10,0
POINT CAMERA c,height#,50
WAIT 1
NEXT c
WAIT KEY
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
REM *** Increase ambient light ***
ENDFUNCTION
Activity 44.9
Type in and test the program in LISTING-44.4 (terrain04.dbpro).
Before the BUILD TERRAIN statement, add a statement which splits the
terrain into 32 by 32 meshes. Does this change have any effect on the speed of
the program?

Activity 44.10
Write a program (act4410.dbpro) which creates a terrain using the files

contour1.bmp, grass256.jpg and grain.bmp. Use the default tiling setting for
the texture. Place the camera at (0,10,0). Set the scaling factors to 1,0.2,1. The
terrain should be split into 32 by 32 sections.
Using GET TERRAIN GROUND HEIGHT, MoveCameraUsingMouse() and

PointCameraUsingMouse() allow the user to move the camera about the
terrain keeping a constant 5 units above the surface (remember to copy
camera.dba to the project file).
The GET TERRAIN SIZE Statement

The dimensions of a terrain object can be obtained using the GET TERRAIN SIZE
FIG-44.22
The GET TERRAIN
SIZE Statement
GET TERRAIN SIZE
{ X
Z
( terno )
real
In the diagram:
X,Z Use the X option to find the width of the terrain.

Use the Z option to find the depth of the terrain.
terno is an integer value identifying the terrain whose

dimensions are to be found.
The function returns a real number representing the specified dimension of the
terrain in world units.
For example, to find the width and depth of terrain 1, we would use the lines:
width# = GET TERRAIN SIZE X(1)

depth# = GET TERRAIN SIZE Z(1)
We could use this information to position the centre of the terrain at the origin:
POSITION OBJECT 1, width#/2,0,depth#/2
The SAVE TERRAIN Statement

Like other 3D objects, a terrain can be saved to a file for use in other programs. This
is achieved using the SAVE TERRAIN statement which has the format given in
FIG-44.23.
FIG-44.23
The SAVE TERRAIN SAVE TERRAIN filename , terno
Statement

In the diagram:
filename is a string giving the name of the file into which

the terrain is to be saved. The filename should
have a .DBO extension.
terno is an integer value identifying the terrain which is

to be saved.
For example, we could save terrain 1 to a file named myterrain.dbo using the line:
SAVE TERRAIN "myterrain.dbo",1
Activity 44.11
In your last program, add code to save the completed terrain to a file called
myterrain.dbo when the Enter key is pressed.
The LOAD TERRAIN Statement

A previously saved terrain can be loaded into a program using the LOAD TERRAIN
FIG-44.24
The LOAD TERRAIN LOAD TERRAIN filename , terno
Statement
In the diagram:

the terrain.
terno is an integer value specifying the ID to be given

to the new terrain object being created.
The loaded terrain will automatically be contoured and textured, assuming the
relevant image files are in place. There is no need to call the BUILD TERRAIN
statement when loading a terrain from a file.
For example, we can load the previously saved myterrain.dbo using the statement:
LOAD TERRAIN "myterrain.dbo",1
Activity 44.12
Modify your last program to load the terrain used from a file.
In fact, a terrain can be loaded as a regular 3D object with a statement such as
LOAD OBJECT "myterrain.dbo",1
but it will not be recognised as a terrain and therefore none of the terrain-related
statements can be applied to the structure.

Terrains as Objects
In many respects a terrain object can be treated as a regular 3D object. For example,
we can use POSITION OBJECT to move a terrain to a new position and ROTATE
OBJECT to rotate it. However, using these statements may affect the accuracy of
calls to GET TERRAIN GROUND HEIGHT.
If the TERRAIN SPLIT statement has been used to split a terrain into several
separate meshes, these meshes become limbs and it is possible to hide or delete
these limbs.
Activity 44.13
HIDE LIMB 1,15
before the DO..LOOP structure.
What has happened to the terrain?
Other limb commands have an effect - but not always the one you might expect!
Summary
Documented Statements
l Use MAKE TERRAIN to create a landscape.
l The contours of the landscape are determined by the image associated with the
terrain object.
l The image associated with a terrain object must be square.
l The terrain object's width and depth match that of the image associated with it.
l Use DELETE TERRAIN to remove a terrain object from memory.
l When first created, the bottom-left corner of a terrain object is at (0,0,0).
l Place the terrain object within your 3D space using the POSITION TERRAIN
statement.
l Use TERRAIN POSITION to determine the current position of the top-left

terrain object.
l Add a texture image to the terrain object using the TEXTURE TERRAIN
statement.
l The texture image is tiled over the surface of the terrain.
l Determine the height of any point on the terrain object using GET TERRAIN
HEIGHT statement.

l The GET TOTAL TERRAIN HEIGHT statement returns the y-ordinate value
of the highest point on the terrain object.
Undocumented (Advanced Terrain) Statements

l Use MAKE OBJECT TERRAIN to create an advanced terrain object.
l Use SET TERRAIN HEIGHTMAP to specify the image to be used when

contouring an advanced terrain.
l Use SET TERRAIN SCALE to resize an advanced terrain in all 3 dimensions.
l Use SET TERRAIN TEXTURE to grain and texture an advanced terrain.
l Use BUILD TERRAIN to make an advanced terrain visible on the screen.
l Use SET TERRAIN TILING to specify the size of each texture tile on the
advanced terrain.
l If only one occurrence of the texture image is to be used over the whole terrain,
make the tile size equal to the terrain size.
l Use SET TERRAIN LIGHT to specify the position and colour of an assumed
light source for the terrain.
l Other objects are not affected by a terrain's light source.
l Use SET TERRAIN SPLIT to convert an advanced terrain to a set of meshes.
l Each mesh within an advanced terrain is treated as a limb of the object.
l Use GET TERRAIN GROUND HEIGHT to discover the height of a specified

point within the terrain.
l Use GET TERRAIN SIZE to find the width and depth of a terrain object.
l Use SAVE TERRAIN to save details of a terrain to a file.
l Saved terrains are held in .DBO format.
l Use LOAD TERRAIN to load a previously saved terrain.
l A loaded terrain does not require a call to BUILD TERRAIN.
l A terrain object can be manipulated using traditional object statements such as

POSITION OBJECT and ROTATE OBJECT.
l A terrain which has been split into limbs can have those limbs manipulated using
standard limb statements.

Terrain Project
Introduction
In this extended example we'll see how to use a terrain and a skybox to create the
beginnings of a treasure-seeking, first-person perspective game. FIG-44.25 gives a
screenshot of the game.
FIG-44.25
The Island
The aim of the game is to wander about the island terrain until you find the golden
orb.
Creating the Game

The game logic can, as usual, be described in a few lines:
Initialise the game

Show game introduction
REPEAT
Control player movement
UNTIL golden orb is found
Show end game
This logic being coded as:
SetUpGame()
StartGame()
REPEAT
ControlPlayer()
UNTIL treasurereached
EndGame()
END

Constants and Global Variables
The various constants and global variables required are listed below:
REM **********************************
REM **********************************
REM *** Booleans ***
#CONSTANT FALSE 0
#CONSTANT TRUE 1
REM *** Terrain ID ***
#CONSTANT IslandTerrain 1
#CONSTANT IslandTextureImg 1
#CONSTANT IslandGrainImg 2
#CONSTANT SeaImg 3
#CONSTANT OrbImg 4
#CONSTANT TitleImg 5
#CONSTANT FoundImg 6
REM *** 3D Object IDs ***
ID 1 is already used for #CONSTANT OrbObj 2
the terrain object which #CONSTANT SkyBoxObj 3
is considered to be a #CONSTANT OceanObj 4
normal 3D object. REM *** Sprites ***
#CONSTANT TitleSprt 1
#CONSTANT FoundSprt 2
REM **********************************
REM *** Globals ***
REM **********************************
GLOBAL treasurereached = FALSE Ìndicates if sphere found
GLOBAL spherex# `Coords of sphere
GLOBAL spherez#
Activity 44.14
Type in the code (treasure.dbpro) given so far and create required Level 1 test
stubs. Press Escape to terminate the program.
Adding StartUpGame()
Starting up the game involves setting the screen resolution, positioning the camera,
and creating the scene, so the logic of this routine can be described as:
Set screen resolution

Hide the mouse pointer
Position camera
Create scene
Activity 44.15
Write code for the StartUpGame() function. Position camera and Create scene
should be implemented in other functions.
The screen resolution should be set to 1280 by 1024 with 32 bit colour.
Add level 2 test stubs for routines named PositionCamera() and CreateScene().
Test your program.

Adding PositionCamera()
The plan is that the StartGame() function will show the camera sweeping in from
the sea towards the island, so this routine needs to position the camera out at sea.
The code required is:
FUNCTION PositionCamera()
AUTOCAM OFF
ENDFUNCTION
Activity 44.16
Add the code for the PositionCamera() function to your program.
Adding CreateScene()
Our game world consists of a terrain-constructed island, a sky, an ocean and the
golden orb. Each of these elements will be created by separate routines, so the code
for CreateScene() will be:
LoadTerrain()
CreateSkyBox()
LoadOcean()
PlaceOrb()
ENDFUNCTION
Activity 44.17
Add the code for the CreateScene() function to your program.
Create test stubs for LoadTerrain(), CreateSkyBox(), LoadOcean() and

PlaceOrb().
Adding LoadTerrain()
The terrain uses the files shown in FIG-44.26. All images are 1024 by 1024.
FIG-44.26 The Images used to Create the Island Terrain
The Texture Image (islandtexture.bmp)

Activity 44.18
Write the code required for LoadTerrain().
In LoadTerrain():
The terrain object should be 1024 by 1024 with a 30% height factor.
Do not allow duplicate tiling when the terrain is textured.
Use POSITION OBJECT to place the centre of the terrain at position
(0,0,0).
Use SET OBJECT TRANSPARENCY IslandTerrain,6 to make the black
areas of the terrain transparent.
Set the ambient light level to 70.
Adding CreateSkyBox()
Back in Chapter 33 we used a sky sphere to create a more realistic scene around our
gallows. However, a cube is often used as an alternative to a sphere when creating
sky and far-off scenery. We have two options when using a box to create our sky.
One option is to create a model containing 6 plane limbs and texture each separately.
We are forced to use planes rather than a cube since it is not possible to texture each
side of a cube with different images. Alternatively, we can create a cube in a 3D
modelling program and add the textures within that program (where we have more
control over what image covers each face of the cube).
Even the images used must be properly prepared if the final effect is to look seamless
and natural.
The best option is to use a package designed specifically to create skyboxes.
Luckily, DarkBASIC Pro comes with a skybox in one of its examples, so we'll use
that in this program.
The code for the function is:
FUNCTION CreateSkyBox()
REM *** Load the skybox ***
LOAD OBJECT "sb.x",SkyBoxObj
REM *** Make it large enough to circle the island ***
SCALE OBJECT SkyBoxObj,20,20,20
REM *** Tweek the texture so no seam shows ***
SET OBJECT TEXTURE SkyBoxObj,2,1
ENDFUNCTION
Activity 44.19
Copy the files required by the skybox to the current project folder. The files
are: sb.x, and sb1.dds to sb6.dds.
Add the CreateSkyBox() function to your program.
Adding LoadOcean()
The base of the skybox doesn't look too ocean-like. One way of fixing this is to
replace the texture used on the base of the skybox (one of the sb.dds files), but
another approach is to hide the skybox base with a slightly higher plane and texture

this plane with an appropriate ocean image.
It is this second approach that is used in the code below:
FUNCTION LoadOcean()
REM *** Create a plane ***
MAKE OBJECT PLAIN OceanObj,4096,4096
Texture the plane ***
LOAD IMAGE "sea.tga",SeaImg
TEXTURE OBJECT OceanObj,SeaImg
REM *** Position the plane ***
XROTATE OBJECT OceanObj,-90
POSITION OBJECT OceanObj,0,2,0
REM *** Make plane transparent ***
SET OBJECT TRANSPARENCY OceanObj,2
ENDFUNCTION
Activity 44.20
Copy sea.tga to your project folder.
Add the code for LoadOcean() to your project.
Adding PlaceOrb()
The last stage of the game initialisation is placing the golden orb somewhere on the
island. Since we'd like each game to have a different location for the orb, we'll need
to use some random positioning method.
However, we can't choose any position or the orb might end up in the water or buried
in a hillside, so we need to check that the position chosen is suitable.
Since the terrain object occupies from point (-512,-512) - bottom-left corner - to
(512,512) - top-right corner - as measured on the XZ plane, we need to generate
an x-ordinate in the range -512 to +512. The same range is required for the
z-ordinate. This can be achieved using the lines:
x# = RND(1024)-512
z# = RND(1024)-512
To discover if we have chosen a suitable point on the island, all we need to do is

check the height of the terrain at that point. If its less than say 5, then we are in the
water or by the coast and so should choose another point.
From this we can come up with an outline logic for the PlaceOrb() function:
Seed the random number generator

REPEAT
Generate random x-ordinate
Generate random z-ordinate
Set y-ordinate to ground height at position (x,z)
UNTIL y-ordinate > 5
Create orb object
Position orb at point (x, y,z)
Activity 44.21
Texture orb using From the logic given above, add the code for PlaceOrb() to your project.
gold.bmp

Adding StartGame()
The StartGame() function is going to fly the camera in over the ocean towards the
island while showing the game title at the top of the screen.
The game title is held as a sprite (loaded from file title.bmp).
This routine requires the following code:

FUNCTION StartGame()
REM *** Show title
LOAD IMAGE "title.bmp",TitleImg
SPRITE TitleSprt,400,100,TitleImg
REM *** Move camera to shore ***
FOR c = 1 TO 330
MOVE CAMERA 2
WAIT 1
NEXT c
WAIT 1000
REM *** Hide title ***
HIDE SPRITE TitleSprt
DELETE SPRITE TitleSprt
ENDFUNCTION
Activity 44.22
Add the StartGame() code to your program.
Adding ControlPlayer()
This routine will allow the player to take control of the camera using the mouse. As
a first attempt we need only use the PointCameraUsingMouse() and
MoveCameraUsingMouse() routines that we developed in a previous chapter.
However, we must adjust the height of the camera to be slightly above the terrain
it is travelling along.
The code required is:
MoveCamerausingMouse(0,1)
REM *** Lift camera above ground height ***
x# = CAMERA POSITION X(0)
z# = CAMERA POSITION Z(0)
y# = GET TERRAIN GROUND HEIGHT(IslandTerrain,x#,z#)+5
POSITION CAMERA 0,x#,y#,z#
ENDFUNCTION
Activity 44.23
Add the code for ControlPlayer() to your program.
The trouble with this simple version of ControlPlayer() is that the player can walk
on water and climb the steepest hills with ease. A more realistic version of the
routine would limit the the player to the land and stop any steep climbs or descents.
To achieve this, all we need to do is check the camera's new position after each
move and if it is too low (in the water) or is too large a change in the y-ordinate
(ascending/descending steep cliffs), then we can move the camera back to its earlier
position.

Assuming y# contains the camera's current y-ordinate and oldy# contains its
previous y-ordinate, then we can restrict movement by adding the following lines:
REM *** IF too low or large change in height ***

IF (y# < 7.5) OR (ABS(oldy#-y#)>1)
REM *** Return camera to previous position ***
POSITION CAMERA 0, oldx#,oldy#,oldz#
ENDIF
The code for the new version of the routine is :
REM *** Get camera's current position ***
oldx# = CAMERA POSITION X(0)
oldy# = CAMERA POSITION Y(0)
oldz# = CAMERA POSITION Z(0)
REM *** Move camera ***
REM *** Lift camera above ground height ***
y# = GET TERRAIN GROUND HEIGHT(IslandTerrain,x#,z#)+5
POSITION CAMERA 0, x#,y#,z#
REM *** IF too low or large change in height ***
IF (y# < 7.5) OR (ABS(oldy#-y#)>1)
REM *** Return camera to previous position ***
ENDIF
ENDFUNCTION
But even now, the routine is not quite finished. It also needs to detect when the
camera is close to the sphere so that the global variable, treasurefound, can be set
to true. This is done by ending the routine with the lines:
REM *** IF sphere reached, set variable ***
IF CameraBesideOrb()
treasurereached = TRUE
ENDIF
Activity 44.24
Add the code for ControlPlayer() to your program.
Write the code for CameraBesideOrb() which should return 1 if the x and z
ordinates of the camera and sphere are both within 5 units of each other;
Adding EndGame()
This function will flash up the message FOUND for 3 seconds before the program
shuts down. The message is created as an animated sprite (see FIG-44.27).
FIG-44.27
The Animated Sprite
Message
The code for this routine is:
FUNCTION EndGame()
REM *** Load sprite ***

CREATE ANIMATED SPRITE FoundSprt,"found.bmp",2,1,FoundImg
SPRITE FoundSprt, 400,500,FoundImg
REM *** Show it for 3 seconds ***
now = TIMER()
PLAY SPRITE FoundSprt,1,2,20
ENDWHILE
ENDFUNCTION
Activity 44.25
Add this last routine to your program and test the completed project.
Adding Testing Features

It seems to be a popular pastime to include cheats in games. If you know what keys
to press you can give yourself unlimited lives, increase your speed, gain greater
firepower, and many other options.
These features are not always included just to help out the keen player; a
programmer will often add features to a game to help with the testing. For example,
testing the island game can be time-consuming if you have no idea where the golden
orb is - or where you are, for that matter.
It would be useful if we knew the coordinates of both the sphere and the camera.
That way, we could find the orb quickly and make sure that the game ends correctly.
In the code that follows, we'll let the player find out the position of the orb by
pressing the "s" key and the position of the camera by hitting "c". To do this we'll
need to change the coding of the main section to read:
SetUpGame()
StartGame()
REPEAT
ControlPlayer()
IF INKEY$()="s"
res$ = STR$(spherex#)+" "+STR$(spherez#)
DisplayMessage(res$,200,200)
ENDIF
IF INKEY$() = "c"
res$ = STR$(CAMERA POSITION X(0))+" "
Ä
+STR$(CAMERA POSITION Z(0))
ENDIF
EndGame()
END
The new function, DisplayMessage(), is one we've used before and is coded as:
REM **********************************
REM *** Debug Routines ***
REM **********************************
FUNCTION DisplayMessage(mes$,x,y)
now = TIMER()
SET CURSOR x,y
PRINT mes$
ENDWHILE
ENDFUNCTION
Notice the function is placed in a special section for debugging (testing) routines.

Activity 44.26
Update your project to reflect these changes and use the secret keys to help
locate the orb.
Code added specifically to aid debugging would often be removed from the final
product after testing is complete, but sometimes left in to allow those in the know
to gain an advantage over others (cheat!).

Solutions MoveCameraUsingMouse(0,1)
Activity 44.1
LOOP
To use terrain02.bmp, the main section changes to: REM *** End program ***
WAIT KEY
END
SetUpScreen()
MAKE TERRAIN 1,"terrain02.bmp" The camera can pass straight through the terrain.
DO
CONTROL CAMERA USING ARROWKEYS 0,5,1 Activity 44.5
LOOP
REM *** End program *** The height value returned by GET TERRAIN
WAIT KEY HEIGHT correctly returns the value of zero when the
END terrain is at its lowest level, but returns a value of 100
for the highest level when, in fact, its height in units is
When using your own image the highlighted line only 65.
would need to be changed to contain the name of your
file.
Activity 44.6
Change the main section of the program to: Activity 44.7

#INCLUDE "camera.dba" The main section should be changed as shown below:
SetUpScreen()
REM *** Create terrain *** #INCLUDE "camera.dba"
MAKE TERRAIN 1,"terrain02.bmp" SetUpScreen()
REM *** Reposition centre to (0,0,0)*** REM *** Create terrain object ***
REM *** Find size of terrain *** MAKE OBJECT TERRAIN 1
LOAD BITMAP "terrain02.bmp",1 REM *** Assign a contour map ***
size = BITMAP WIDTH(1) SET TERRAIN HEIGHTMAP 1, "contour.bmp"
DELETE BITMAP 1 REM *** Reduce the height ***
POSITION TERRAIN 1, -size/2,0,-size/2 SET TERRAIN SCALE 1, 1,0.5,1
REM *** Move camera to centre *** REM *** Add grain and texture ***|
POSITION CAMERA 0,0,50,0 LOAD IMAGE "grain.bmp",1
REM *** Allow user to move camera *** LOAD IMAGE "texture.bmp",2
DO SET TERRAIN TEXTURE 1,1,2
PointCameraUsingMouse(0) REM *** No tiling ***
MoveCameraUsingMouse(0,1) SET TERRAIN TILING 1,512
LOOP REM *** Show terrain ***
REM *** End program *** BUILD TERRAIN 1,0
END REM *** Allow user to move camera ***
DO
Activity 44.3 MoveCameraUsingMouse(0,1)
LOOP
PRINT TERRAIN POSITION Y(2) REM *** End program ***
END
Activity 44.4
Activity 44.8
The main section should be changed to:
To add the lightig effect add the folowing lines to the
#INCLUDE "camera.dba" main section before the BUILD TERRAIN statement:
SetUpScreen()
REM *** Set up terrain lighting ***
MAKE TERRAIN 1,"terrain02.bmp"
SET TERRAIN LIGHT 1, -100,10,250,0
REM *** Reposition terrain to (0,0,0)
POSITION TERRAIN 1, 0, 0, 0
REM *** Find size of terrain ***
LOAD BITMAP "terrain03.jpg",1 Activity 44.9
size = BITMAP WIDTH(1)
DELETE BITMAP 1 The statement required to split the terrain into the
REM *** Move camera to centre *** required number of tiles is:
POSITION CAMERA 0, size/2,50,-size/2
REM *** Texture terrain *** SET TERRAIN SPLIT 1, 32
TEXTURE TERRAIN 1,2 The smoothness of the camera movement should
REM *** Allow user to move camera *** improve.
DO

END
#INCLUDE "camera.dba"
SetUpScreen()
REM *** Create a terrain object *** Activity 44.13
MAKE OBJECT TERRAIN 1
REM *** Add contour map *** Adding the HIDE LIMB statement will cause one section
SET TERRAIN HEIGHTMAP 1,"contour1.bmp" of the terrain to disappear.
REM *** Texture terrain ***
LOAD IMAGE "grain.bmp",1 Activity 44.14
SET TERRAIN TEXTURE 1,1,2 This initial core of the program is coded as:
REM *** Partition terrain ***
SET TERRAIN SPLIT 1,32 REM **********************************
REM *** Show the terrain *** REM *** Constants ***
BUILD TERRAIN 1 REM **********************************
REM *** Allow the user to move camera *** REM *** Booleans ***
DO #CONSTANT FALSE 0
PointCameraUsingMouse(0) #CONSTANT TRUE 1
MoveCameraUsingMouse(0,1) REM *** Terrain ID ***
x# = CAMERA POSITION X(0) #CONSTANT IslandTerrain 1
z# = CAMERA POSITION Z(0) REM *** Image IDs ***
y# = GET TERRAIN GROUND HEIGHT(1,x#,z#) #CONSTANT IslandTextureImg 1
POSITION CAMERA 0,x#,y#+5,z# #CONSTANT IslandGrainImg 2
LOOP #CONSTANT SeaImg 3
REM *** End program *** #CONSTANT OrbImg 4
END #CONSTANT TitleImg 5
FUNCTION SetUpScreen() REM *** 3D Object IDs ***
REM *** Set display mode required *** #CONSTANT OrbObj 2
SET DISPLAY MODE 1280,1024,32 #CONSTANT SkyBoxObj 3
COLOR BACKDROP 0 #CONSTANT OceanObj 4
BACKDROP ON REM *** Sprites ***
REM *** position camera *** #CONSTANT TitleSprt 1
AUTOCAM OFF #CONSTANT FoundSprt 2
POSITION CAMERA 0,0,10,0 REM **********************************
REM *** Increase ambient light *** REM *** Globals ***
SET AMBIENT LIGHT 50 REM **********************************
ENDFUNCTION GLOBAL treasurereached = FALSE
ÄÌndicates if sphere found
GLOBAL spherex#
Activity 44.11 Ä`Coords of sphere
GLOBAL spherez#
Change the DO..LOOP in the main section to read:
DO REM *** Main logic ***

PointCameraUsingMouse(0) SetUpGame()
MoveCameraUsingMouse(0,1) StartGame()
x# = CAMERA POSITION X(0) REPEAT
z# = CAMERA POSITION Z(0) ControlPlayer()
y# = GET TERRAIN GROUND HEIGHT(1,x#,z#) UNTIL treasurereached
POSITION CAMERA 0, x#,y#+5,z# EndGame()
REM *** Save to file on ENTER *** END
IF INKEY$() = CHR$(13)
SAVE TERRAIN "myterrain.dbo",1
ENDIF REM **********************************
LOOP REM *** Level 1 ***
REM **********************************
Activity 44.12 FUNCTION SetUpGame()

ENDFUNCTION
The main section should be changed to:
FUNCTION StartGame()
#INCLUDE "camera.dba" ENDFUNCTION
SetUpScreen()
REM *** Load terrain from file *** FUNCTION ControlPlayer()
LOAD TERRAIN "myterrain.dbo",1 ENDFUNCTION
DO FUNCTION EndGame()
PointCameraUsingMouse(0) ENDFUNCTION
z# = CAMERA POSITION Z(0) Activity 44.15
y# = GET TERRAIN GROUND HEIGHT(1,x#,z#)
POSITION CAMERA 0, x#,y#+5,z# The code for SetUpGame() should be:
LOOP

FUNCTION SetUpGame() SET TERRAIN TEXTURE IslandTerrain,
SET DISPLAY MODE 1280,1024,32 Ä
IslandGrainImg,IslandTextureImg
HIDE MOUSE SET TERRAIN TILING IslandTerrain,1024
PositionCamera() SET TERRAIN SPLIT IslandTerrain,32
CreateScene() BUILD TERRAIN IslandTerrain
ENDFUNCTION REM *** Place centre at (0,0,0) ***
The additional test stubs should be added in a new Level 2 REM *** Make black transparent ***
section of the code: SET OBJECT TRANSPARENCY IslandTerrain,6
REM ********************************** SET AMBIENT LIGHT 70
REM *** Level 2 *** ENDFUNCTION
REM **********************************
FUNCTION PositionCamera() Activity 44.19

ENDFUNCTION
A sky should now appear around the island.
ENDFUNCTION
Activity 44.20
Activity 44.16 A semi-transparent sea should now appear within the
scene.
The code given should replace the existing level 2 test
stub for PositionCamera().
Activity 44.21
Activity 44.17 The code for PlaceOrb() is:
Level 2 and Level 3 code should now be: FUNCTION PlaceOrb()

REM ********************************** RANDOMIZE TIMER()
REM *** Level 2 *** REM *** Generate a valid position ***
REM ********************************** REPEAT
FUNCTION PositionCamera() x# = RND(1024)-512
AUTOCAM OFF z# = RND(1024)-512
y# = GET TERRAIN GROUND HEIGHT
POINT CAMERA 20,8,100 Ä
(IslandTerrain,x#,z#)
ENDFUNCTION UNTIL y# > 5
REM *** Create and place sphere ***
FUNCTION CreateScene() MAKE OBJECT SPHERE OrbObj,2
LoadTerrain() LOAD IMAGE "gold.bmp",OrbImg
CreateSkyBox() TEXTURE OBJECT OrbObj,OrbImg
LoadOcean() POSITION OBJECT OrbObj,x#,y#+1,z#
PlaceOrb() REM *** Record its position ***
ENDFUNCTION spherex# = x#
spherez# = z#
REM ********************************** ENDFUNCTION
REM *** Level 3 ***
REM **********************************
FUNCTION LoadTerrain() Activity 44.22
ENDFUNCTION
The code should replace the Level 1 test stub for
FUNCTION CreateSkyBox() StartGame().
ENDFUNCTION
FUNCTION LoadOcean() Activity 44.23

ENDFUNCTION
The code should replace the Level 1 test stub for
FUNCTION PlaceOrb() ControlPlayer().
ENDFUNCTION
Remember to copy the camera.dba file to the project
folder and to add the line
Activity 44.18
The code for LoadTerrain() is: #INCLUDE "camera.dba"
FUNCTION LoadTerrain() at the start of the program.

REM *** Create terrain (1024 by 1024)***
MAKE OBJECT TERRAIN IslandTerrain
SET TERRAIN HEIGHTMAP IslandTerrain, Activity 44.24
Ä
"islandcontour.bmp"
The code for CameraBesideOrb() is:
SET TERRAIN SCALE IslandTerrain,1,0.3,1
LOAD IMAGE "islandtexture.bmp", FUNCTION CameraBesideOrb()
ÄIslandTextureImg IF ABS(CAMERA POSITION X(0)-spherex#)<5
LOAD IMAGE "rock.bmp",IslandGrainImg Ä
AND ABS(CAMERA POSITION Z(0)-spherez#)

Ä<5
result = 1 FUNCTION StartGame()
ELSE REM *** Show title
result = 0 LOAD IMAGE "title.bmp",TitleImg
ENDIF SPRITE TitleSprt,400,100,TitleImg
ENDFUNCTION result REM *** Move camera to shore ***
FOR c = 1 TO 333
MOVE CAMERA 2
NEXT c
No solution required. WAIT 1000
REM *** Hide title ***
HIDE SPRITE TitleSprt
DELETE SPRITE TitleSprt
The complete program (including debugging code) is:
REM *** Get camera's position ***
oldx# = CAMERA POSITION X(0)
REM **********************************
oldy# = CAMERA POSITION Y(0)
oldz# = CAMERA POSITION Z(0)
REM **********************************
REM *** Move camera ***
REM *** Booleans ***
#CONSTANT FALSE 0
#CONSTANT TRUE 1
REM *** Lift camera above ground ***
REM *** Terrain ID ***
#CONSTANT IslandTerrain 1
#CONSTANT IslandTextureImg
#CONSTANT IslandGrainImg
1
2
Ä(IslandTerrain,x#,z#)+5
POSITION CAMERA 0, x#,y#,z#
#CONSTANT SeaImg 3
REM *** IF too low or large change ***
#CONSTANT OrbImg 4
IF (y# < 7.5) OR (ABS(oldy#-y#)>1)
#CONSTANT TitleImg 5
REM *** Return to old position ***
REM *** 3D Object IDs ***
ENDIF
#CONSTANT OrbObj 2
REM *** If orb found, set indicator ***
#CONSTANT SkyBoxObj 3
IF CameraBesideOrb()
#CONSTANT OceanObj 4
treasurereached = TRUE
REM *** Sprites ***
ENDIF
#CONSTANT TitleSprt 1
ENDFUNCTION
#CONSTANT FoundSprt 2
REM **********************************
FUNCTION EndGame()
REM *** Globals ***
CREATE ANIMATED SPRITE FoundSprt,
REM **********************************
Ìndicates if sphere found
Ä"found.bmp",2,1,FoundImg
GLOBAL treasurereached = FALSE SPRITE FoundSprt,400,500,FoundImg
now = TIMER()
`Coords of sphere
WHILE TIMER() - now<3000
GLOBAL spherex#
GLOBAL spherez# PLAY SPRITE FoundSPrt,1,2,20
ENDWHILE
ENDFUNCTION
REM *** Main logic ***
SetUpGame()
REM **********************************
StartGame()
REPEAT REM *** Level 2 ***
ControlPlayer() REM **********************************
IF INKEY$()="s" FUNCTION PositionCamera()
res$ = STR$(spherex#)+" " AUTOCAM OFF
Ä+STR$(spherez#)
ENDFUNCTION
ENDIF
IF INKEY$() = "c"
res$ = STR$(CAMERA POSITION X(0))+
Ä" "+STR$(CAMERA POSITION Z(0))
LoadTerrain()
CreateSkyBox()
LoadOcean()
ENDIF
PlaceOrb()
ENDFUNCTION
EndGame()
END
REM **********************************
REM *** Level 3 ***
REM **********************************
REM **********************************
REM *** Level 1 ***
FUNCTION LoadTerrain()
REM **********************************
REM *** Create terrain (1024 by 1024)***
MAKE OBJECT TERRAIN IslandTerrain
SET TERRAIN HEIGHTMAP IslandTerrain,
HIDE MOUSE
PositionCamera()
Ä"islandcontour.bmp"
SET TERRAIN SCALE IslandTerrain,1,0.3,1
CreateScene()
LOAD IMAGE "islandtexture.bmp",
ENDFUNCTION
ÄIslandTextureImg

LOAD IMAGE "rock.bmp",IslandGrainImg
SET TERRAIN TEXTURE IslandTerrain,
Ä
IslandGrainImg,IslandTextureImg
SET TERRAIN TILING IslandTerrain,1024
SET TERRAIN SPLIT IslandTerrain,32
BUILD TERRAIN IslandTerrain
REM *** Position centre at (0,0,0) ***
REM *** Black areas (sea)transparent ***
SET OBJECT TRANSPARENCY IslandTerrain,6
ENDFUNCTION
FUNCTION CreateSkyBox()
LOAD OBJECT "sb.x",SkyBoxObj
SCALE OBJECT SkyBoxObj,20,20,20
SET OBJECT TEXTURE SkyBoxObj,2,1
ENDFUNCTION
FUNCTION LoadOcean()
MAKE OBJECT PLAIN OceanObj,4096,4096
LOAD IMAGE "sea.tga",SeaImg
TEXTURE OBJECT OceanObj,SeaImg
XROTATE OBJECT OceanObj,-90
POSITION OBJECT OceanObj,0,2,0
SET OBJECT TRANSPARENCY OceanObj,2
ENDFUNCTION
FUNCTION PlaceOrb()
RANDOMIZE TIMER()
REM *** Generate a valid position ***
REPEAT
x# = RND(1024)-512
z# = RND(1024)-512
Ä
(IslandTerrain,x#,z#)
UNTIL y# > 5
REM *** Create and place sphere ***
MAKE OBJECT SPHERE OrbObj,2
LOAD IMAGE "gold.bmp",OrbImg
TEXTURE OBJECT OrbObj,OrbImg
POSITION OBJECT OrbObj,x#,y#+1,z#
REM *** Record its position ***
spherex# = x#
spherez# = z#
ENDFUNCTION
FUNCTION CameraBesideOrb()
IF ABS(CAMERA POSITION X(0)-spherex#)<5
Ä
AND ABS(CAMERA POSITION Z(0)-spherez#)
Ä
<5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
REM **********************************
REM *** Debug Routines ***
REM **********************************
now = TIMER()
SET CURSOR x,y
PRINT mes$
ENDWHILE
ENDFUNCTION

45
How to Use a Matrix Object to Create a Landscape
Creating a Matrix Object
Assigning Random Heights to a Matrix
Assigning Specific Heights to a Matrix
Texturing a Matrix Object
Creating Image Tiles
Setting Matrix Transparency
Setting Matrix Drawing Priority
Scrolling the Matrix Texture
Determining the Height of a Point on the Matrix
Positioning the Matrix
DarkBASIC Pro: Using Matrices 1211

Matrices
Introduction
Although we can create ground surface using the terrain objects as described in the
previous chapter, an alternative way to create an undulating surface is to use a
matrix object.
DarkBASIC Pro uses the term matrix to describe a two-dimensional plane which
is divided into a number of rows and columns, collectively known as subdivisions.
This creates a grid pattern in the plane (see FIG-45.1).
FIG-45.1
A Flat Matrix
A 2D grid
The grid appears distorted
because of the camera angle
The intersecting subdivisions create a set of rectangular areas and each rectangle is
divided into two triangles.
The y-ordinate of individual apexes which make up the triangles within the structure
can be manipulated to create a 3D effect (see FIG-45.2).
FIG-45.2
Adding a Third
Dimension to the Matrix
The grid is now 3-dimensional
A matrix, like any other 3D object, can also have a texture applied to it (see
FIG-45.3) allowing you to create a landscape effect similar to that produced by a
terrain object.
FIG-45.3
Texturing a Matrix
The grid has been textured
The set of commands available to create and manipulate a matrix are described in
the next section.
1212 DarkBASIC Pro: Using Matrices

Creating a Matrix
The MAKE MATRIX Statement
We start by creating a flat surface lying horizontally in the XZ plane using the
MAKE MATRIX statement. This statement allows us to specify the dimensions of
the matrix grid and the number of rows and columns in the grid (that is, the number
of subdivisions along the x and z axes). The statement has the format shown in
FIG-45.4.
FIG-45.4
MAKE MATRIX matno , width , depth , xdivs , zdivs
The MAKE MATRIX
Statement
In the diagram:
matno is an integer value to be assigned to the matrix.

No two matrices may have the same ID.
width is a real number giving the width of the surface

along the x-axis.
depth is a real number giving the depth of the plane

along the z-axis.
xdivs is an integer value specifying how many

subdivisions are required along the x-axis.
zdivs is an integer value specifying how many

subdivisions are required along the z-axis.
MAKE MATRIX 1,200,50,20,5
will create the grid shown in FIG-45.5.

FIG-45.5 5 subdivisions
along the z-axis
A Typical Matrix 20 subdivisions along
the x-axis
50 units
ng
nits lo
200 u
deep
Activity 45.1
Write a statement to create a matrix grid (with an ID of 2) that is 1000 units

deep by 500 units wide.
Make each tile 100 by 100 units.
The grid is initially positioned so that its bottom-left corner is at the origin (0,0,0).

The RANDOMIZE MATRIX Statement
The quickest way to add a third dimension to a matrix grid is to use the
RANDOMIZE MATRIX statement which randomly assigns a value (within a
specified limit) to the y-ordinate of the four corners of every tile in the grid. This
FIG-45.6
The RANDOMIZE RANDOMIZE MATRIX matno , ymax
MATRIX Statement
In the diagram:
matno is an integer value giving the matrix ID.
ymax is a real value representing the maximum

positive displacement of each corner along the
y-axis.
By executing this statement, every corner in the grid will be displaced by a random
amount between 0 and ymax. For example, assuming we began by creating a four
tile matrix grid with the statement
we could randomly reposition the corners of each tile by a value of between 0 and
10.5 using the statement:
RANDOMIZE MATRIX 1,10.5
The resulting grid will, of course, vary each time the program is run, but a typical
example is shown in FIG-45.7.
FIG-45.7
The corners of
each tile have been
The Effect of the moved randomly up
RANDOMIZE MATRIX the y-axis
Statement
The UPDATE MATRIX Statement

If we make any change to a matrix grid (such as executing the RANDOMIZE
MATRIX command), that change won’t show up on the screen until the program
executes the UPDATE MATRIX statement which has the format shown in
FIG-45.8.
FIG-45.8
UPDATE MATRIX matno
The UPDATE MATRIX
Statement
In the diagram:
For example, to show any changes to grid 1 we would use the line:

UPDATE MATRIX 1
We now have enough knowledge to create our first 3D matrix grid as shown in
LISTING-45.1.
REM *** Create a grid (800 by 720) ***
Creating a Matrix Grid MAKE MATRIX 1,800,720,8,9
WAIT KEY
REM *** Randomly adjust the y ordinates ***
REM *** to a maximum of 40 ***
RANDOMIZE MATRIX 1,40
REM *** Update matrix ***
UPDATE MATRIX 1
WAIT KEY
END
REM *** Position camera for a good view of matrix ***
AUTOCAM OFF
ENDFUNCTION
Activity 45.2
Type in and test the program given in LISTING-45.1 (matrix01.dbpro).
Modify the grid to be 1000 by 500 with each rectangle being 10 by 10.
Set the random height variation to range from 0 to 20. Run the program to
check the effect of these changes.
Activity 45.3
Modify your program again so that the camera position can be altered using
the arrow keys. Start with the camera at position (0,100,-300).
Move the camera over the grid to check that the changes are operating
correctly.
The SET MATRIX HEIGHT Statement

If we want more control over the positioning of tile corners, we can use the SET
MATRIX HEIGHT statement which allows us to give an exact amount of
movement to a specific tile corner. This statement's format is shown in FIG-45.9.
FIG-45.9 The SET MATRIX HEIGHT Statement
SET MATRIX HEIGHT matno , xedge , zedge , height
In the diagram:
xedge is an integer value specifying an edge along the

x-axis The left-most edge has the value zero.

zedge is an integer value specifying an edge along the
z-axis. The closest edge has the value zero.
height is a real value representing the height to which the

specified corner is to be raised above the plane of
the grid. This cann be a positive or negative value.
The result of the following statements

SET MATRIX HEIGHT 1,3,2,10.5
UPDATE MATRIX 1
FIG-45.10 Point (3,2) has been

raised 10.5 units
Using the SET MATRIX
HEIGHT Statement
Z points
1
3
2
1
0
X points
Activity 45.4
Remove the RANDOMIZE MATRIX statement from your last program.
Add two new statements to the program which:

a) lifts the corner (8,3) 25.7 units.
b) drops corner (0,1) -12.5 units
By using nested FOR loops we can assign specific values to every point in the grid.
For example, the program in LISTING-45.2 creates a rising ramp by increasing the
size of every point as we move from the left to the right of a grid (see FIG-45.11).
FIG-45.11
A Controlled Matrix

Creating a Structured Grid REM *** Set up 200 by 100 grid (20 by 10 tiles) ***
MAKE MATRIX 1,200,100,20,10
REM *** Set initial height ***
height#=0.5
REM *** FOR each x edge DO ***
FOR x = 0 TO 20
REM *** Increase height by 30% ***
height# = height# *1.3
REM *** FOR each z edge DO ***
FOR z = 0 TO 10
REM *** Set edge to specified height ***
SET MATRIX HEIGHT 1,x,z,height#
NEXT z
NEXT x
UPDATE MATRIX 1
REM *** Move camera using cursor keys ***
DO
CONTROL CAMERA USING ARROWKEYS 0, 1, 1
LOOP
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.5
Type in the above project (matrix02.dbpro) and test it.
Modify the code so that the ramp effect goes from front to back rather than
side-to-side.
Change the height increase to 50% on each increment of the X loop.
The GET MATRIX HEIGHT Statement

If you want to discover the exact height of any tile edge, you can do this using the
GET MATRIX HEIGHT statement which has the format shown in FIG-45.12.
FIG-45.12
The GET MATRIX GET MATRIX HEIGHT ( matno , xedge , zedge )
HEIGHT Statement
real
In the diagram:
xedge is an integer value specifying a point along the

x-axis.

z-axis.
The statement returns a real number representing how high above the base of the
matrix the specified point is. For example, if we have previously executed the lines

SET MATRIX HEIGHT 1,3,2,10.5 ‘Height 10.5
UPDATE MATRIX 1
then the statement
PRINT GET MATRIX HEIGHT (1,3,12)
will display the value 10.5.
The GET GROUND HEIGHT Statement

The biggest limitation of the GET MATRIX HEIGHT statement is that it only
allows the height at specific points to be returned. There’s no way of finding out
the height half way along a tile (see FIG-45.13).
FIG-45.13 GET MATRIX HEIGHT

can find the height at the
Finding the Height of corners of the matrix tiles
any Point on the Matrix
1 but how can we find the

height of any arbitrary point?
3
2
1
0
The problem can be solved using the GET GROUND HEIGHT statement which
returns the height at any point in the matrix. The statement has the format shown in
FIG-45.14.
FIG-45.14
The GET GROUND GET GROUND HEIGHT ( matno , x , z )
HEIGHT Statement
real
In the diagram:
x,y are real numbers which represent the position in

the matrix whose height is to be returned.
These values are given in world units - not matrix
point numbers - and are relative to the bottom-left
corner of the matrix.
If we created a matrix with the statements

SET MATRIX HEIGHT 1, 3, 2, 10.5 ‘Height 10.5
UPDATE MATRIX 1
then the statement

PRINT GET GROUND HEIGHT (1,150.0,100.0)
will return the value 10.5.
Remember, there are two main differences between GET MATRIX HEIGHT and
GET GROUND HEIGHT:
Ø GET MATRIX HEIGHT can only give the height at a vertex;

GET GROUND HEIGHT can return the height of any position in the
matrix.
Ø GET MATRIX HEIGHT identifies the position using edge numbers;
GET GROUND HEIGHT identifies the position using 3D world
coordinates.
The program in LISTING-45.3 displays the height at various positions along the
back edge of a matrix.
Finding the Matrix Height MAKE MATRIX 1, 150, 100,3,2
at any Point REM *** Set the back corner to a height of 10.5 ***
SET MATRIX HEIGHT 1, 3,2, 10.5
UPDATE MATRIX 1
REM *** Get the height every 10 units ***
REM *** along the back edge of matrix ***
SYNC ON
FOR c = 0 TO 150 STEP 10
SET CURSOR 0, 10
PRINT “Height at (”,c,",",100,") = “,GET GROUND HEIGHT(1,c,100)
SYNC
WAIT 500
NEXT c
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.6
Type in and test the program given above (matrix03.dbpro).
Change the z-ordinate in the GET GROUND HEIGHT to 70.0 and check out
the values displayed.
The SET MATRIX WIREFRAME Statement

Unlike other 3D objects, a grid appears in wireframe by default. We can toggle

between wireframe and solid mode using the SET MATRIX WIREFRAME
FIG-45.15
The SET MATRIX
WIREFRAME Statement SET MATRIX WIREFRAME
{ OFF
ON
matno
In the diagram:
OFF, ON Use OFF to create a solid matrix; use ON to

return to wireframe mode.
Activity 45.7
Modify your previous program so that the grid appears solid.
The MATRIX WIREFRAME STATE Statement

We can determine if a matrix is being displayed in wireframe or solid mode using
the MATRIX WIREFRAME STATE statement. This statement has the format
shown in FIG-45.16.
FIG-45.16
The MATRIX
WIREFRAME STATE
MATRIX WIREFRAME STATE ( matno )
Statement
real
In the diagram:
matno is the integer value previously assigned to the matrix.
This statement returns the value 1 if the matrix is being displayed in wireframe
mode, otherwise zero is returned.
Adding Texture to the Matrix

The PREPARE MATRIX TEXTURE Statement
You will almost certainly want to add a texture to your grid. To do this you start by
loading the required picture into an image object, just as we did for sprites and 3D
primitives (spheres, cubes, etc.). However, the next step is a little different. The
image to be used can be split up into rectangular areas known as tiles and only one
of these tiles is used to texture the grid.
The PREPARE MATRIX TEXTURE statement is used to determine how the image
is to be split up. This statement has the format shown in FIG-45.17.

FIG-45.17 The PREPARE MATRIX TEXTURE Statement
PREPARE MATRIX TEXTURE matno , imgno , xsects , zsects
In the diagram:
imgno is an integer value giving the ID of the image that

is to be applied to the grid.
xsects is an integer value specifying the number of

vertical sections in the image.
zsects is an integer value specifying the number of

horizontal sections in the image.
For example, let's start by loading the image shown in FIG-45.18 below.
FIG-45.18
A Texture Image for the
Matrix
If we begin by loading the image with the statement

LOAD IMAGE “B.bmp”,1
and then prepare the image, without dividing it into parts, with the line
PREPARE MATRIX TEXTURE 1,1,1,1
then the image will be taken as a single tile. However, if we use the statement
then the image will be split into tiles as shown in FIG-45.19.

FIG-45.19
Each tile is assigned
Dividing an Image into a number starting with
the top-left tile
Tiles
1 2 3 Lines are included
to show how the
image would be split
into tiles
4 5 6
Activity 45.8
Reload matrix02.dbpro and add the line
at an appropriate point in the program.


Activity 45.8 (continued)
Immediately before the line

UPDATE MATRIX 1
add the statement
What effect does this have when you run your program?
Change the last line you added to

How does this change the appearance of the grid?
As you can see from the results of the last Activity, the PREPARE MATRIX
TEXTURE statement also applies one of the tiles in the image to every square in
the grid. In the first instance, since the image had not been split into several tiles,
the whole image is applied to each rectangle in the grid. In the second case, where
the image had been split into six tiles, the top-left tile (tile 1) has been applied to
every square, giving a completely red grid.
The FILL MATRIX Statement

If you’d rather fill the grid with a different tile from your image, you can use the
FILL MATRIX statement. This statement also sets the height of each corner point
in the whole grid. The statement has the format shown in FIG-45.20.
FIG-45.20
FILL MATRIX matno , height , tileno
The FILL MATRIX
Statement
In the diagram:
height is a real number specifying the y-ordinate of

every point in the grid.
tileno is an integer value specifying which tile from the

prepared image is to be used to fill every rectangle
in the grid.
This statement can only be used after a PREPARE MATRIX TEXTURE statement
has been executed to set up the image being used by the grid.
For example, we could texture the complete grid with tile 2 from B.bmp and lift the
whole grid to a height of 12.5 units using the lines:

FILL MATRIX 1,12.5,2
Since FILL MATRIX
resets the height of the
whole matrix, use this Activity 45.9
statement before using the
SET MATRIX HEIGHT
statement. Modify your previous program to use tile 5 of the B.bmp image to texture your
grid. Make sure the overall shape of the grid is unchanged.

The SET MATRIX TILE Statement
To get more control of texturing a grid we can use the SET MATRIX TILE
statement to place a single tile from the image into one or more rectangles of the
grid. This statement takes the format shown in FIG-45.21.
FIG-45.21 The SET MATRIX TILE Statement
SET MATRIX TILE matno , xsect , zsect , tileno
In the diagram:
xsect is an integer value specifying a section along

the x-axis.
zsect is an integer value specifying a section along the

z-axis.
tileno is an integer value representing the image tile to

be used as the specified rectangle’s texture.
For example, let’s assume we’ve created a 5 by 3 grid using the statement
MAKE MATRIX 1,5,3
and split an image into 3 by 2 tiles with the lines
LOAD IMAGE “B.bmp”, 1

then we could place tile 2 of the image in rectangle (4,0) using the statement
SET MATRIX TILE 1,4,0,2
The result of this is shown in FIG-45.22.
FIG-45.22 Tiled Image
The Effect of the SET 1 2 3

MATRIX TILE
Statement
4 5 6
Matrix Grid
2 Tile 2 textures
rectangle (4,0)
1
0 1 2 3 4
The program in LISTING-45.4 demonstrates the effect of this statement.

REM *** Create 5 by 3 grid ***
Texturing a Single MAKE MATRIX 1,100, 100,5,3
Rectangle REM *** Load image and tile as 3 by 2 ***
REM *** Load tile 2 into rectangle (4,0) ***
REM *** Update display ***
UPDATE MATRIX 1
DO
LOOP
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.10
Notice that, as well as loading the specified tile to the required rectangle in the grid,
the remainder of the grid is red. This is because tile 1 is always used as default to
fill the grid after the PREPARE MATRIX TEXTURE statement is executed.
Activity 45.11
Modify your last program so that tile 5 of the same image is placed in
rectangle (2,1).
Modify the program again so that all six image tiles are stored over six
rectangles. The rectangles should be placed as follows:
Tile Rectangle
1 (1,2)
2 (2,2)
3 (3,2)
4 (1,1)
5 (2,1)
6 (3,1)
To spread a single image over the whole grid without repetition of the image, all
you need to do is divide the image up so that it has one tile for every rectangle in
the grid.
So if we create the grid with the line
MAKE MATRIX 1, 200,200,20,20
then we need to create our image with the lines:
LOAD IMAGE “F.bmp”,1


Now we need to keep a tile count
tileno = 1
and then copy each tile to the correct rectangle
FOR x = 19 TO 0 STEP -1
FOR z = 0 TO 19
SET TEXTURE TILE 1,x,z,tileno
and move on to the next tile:
INC tileno
If we put all this together, we get the effect shown in FIG-45.23.

FIG-45.23
Spreading a Single
Image Over the Whole
Matrix
The complete program is shown in LISTING-45.5.
REM *** Create grid 20 by 30 rectangles ***
Spreading a Single Image MAKE MATRIX 1,200,300,20,30
Over the Whole Matrix REM *** Load image and divide into 20 by 30 tiles ***
LOAD IMAGE “F.bmp”,1,0
REM *** Copy each tile onto the grid starting at back-left ***
tileno = 1
FOR z =29 TO 0 STEP -1
FOR x = 0 TO 19
SET MATRIX TILE 1, x, z, tileno
INC tileno
NEXT x
NEXT z
REM *** Update grid ***
UPDATE MATRIX 1
REM *** Use camera to move about ***
DO
LOOP
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.12

Notice that it is necessary to start at the “back” of the grid (with z = 29) if the
image is to be correctly mapped onto the matrix.
Change the texture image used to grass.jpg and randomly modify the points in
the grid to vary up to 30 units in height.

The SET TEXTURE TRIM Statement
Very occasionally, you may not want the whole of a tile to be placed in a grid
rectangle. Using the SET TEXTURE TRIM statement you can shave a bit off the
edges of the image tile being used and thereby reduce what part of the image is
actually used to texture a rectangle in the grid. The statement has the format shown
in FIG-45.24.
FIG-45.24
The SET TEXTURE
SET TEXTURE TRIM matno , xtrim , ztrim
TRIM Statement
In the diagram:
xtrim is a real value specifying how much is to be

trimmed along the x edges of the tile image.
ztrim is a real number specifying how much is to be

trimmed along the z edges of the tile image.
This is a rather strange command to get your head round, so the program in
LISTING-45.6 demonstrates its use by incrementing the trim factor along the x edge
every time you press a key.
REM *** Set up grid with texture ***
Trimming the Texture MAKE MATRIX 1,40,40,2,2
Image LOAD IMAGE “trimmer.bmp”,1,1
REM *** Position grid ***
POSITION MATRIX 1, 0,0,0
REM *** Set xtrim value to zero ***
xoff# = 0.0
DO
REM *** Set the trim on the grid ***
SET MATRIX TRIM 1,xoff#,0.0
REM *** Fill every rectangle with same trimmed tile ***
FILL MATRIX 1,0,2
UPDATE MATRIX 1
WAIT KEY
REM *** Make the offset larger ***
xoff# = xoff# + 0.01
LOOP
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.13
Type in and test the program given above (matrix06.dbpro), observing how
the texture of the grid changes as the xoff# value is changed.
Modify the program so that both the xtrim and ztrim values are changed.

The SHIFT MATRIX Statement
Each rectangle in a grid can be shifted one position using the SHIFT MATRIX
statement. This statement has several options as shown in FIG-45.25.
{
FIG-45.25
UP
The SHIFT MATRIX
Statement
DOWN
SHIFT MATRIX matno
LEFT
RIGHT
In the diagram:
UP, DOWN, LEFT, RIGHT

Choose the keyword corresponding to the
direction in which the matrix is to be moved.
For example, the statement SHIFT MATRIX DOWN 1 has the effect of moving
every rectangle in the matrix one position backwards within the matrix. Hence,
rectangle (0,0) would move to position (0,1). The rectangles currently at the back
of the grid are moved round to the front. All the information in that grid - the height
of its corners and its texture are moved along with the rectangle. The overall effect
is to give the impression that the surface is moving under the camera. The program
in LISTING-45.7 demonstrates this effect.
SetUpScreen()
LISTING-45.7 REM *** Set up grid with texture ***
MAKE MATRIX 1,2000,3000,20,30
Shifting the Matrix LOAD IMAGE “grass.jpg”,2
Texture PREPARE MATRIX TEXTURE 1,2,20,30
tileno = 1
FOR x = 0 TO 19
INC tileno
NEXT x
NEXT z
RANDOMIZE MATRIX 1, 20
REM *** Update grid on screen ***
UPDATE MATRIX 1
REM *** Shift matrix texturing ***
DO
REM *** Shift matrix one position ***
SHIFT MATRIX DOWN 1
UPDATE MATRIX 1
WAIT 50
LOOP
END
AUTOCAM OFF
ENDFUNCTION

Activity 45.14
Modify the direction of shift to try out each of the other three options.
The MATRIX TILE COUNT Statement

The number of tiles available to a matrix - in other words, the total number of
rectangles the prepared image has been split into by the PREPARE MATRIX
TEXTURE statement, can be returned using the MATRIX TILE COUNT statement
FIG-45.26
MATRIX TILE COUNT ( matno )
The MATRIX TILE
COUNT Statement
integer
In the diagram:
For example, if we had prepared an image for matrix 1 with the statement
then the line
PRINT MATRIX TILE COUNT(1)
would display the value 600 (this being the result of 20 x 30).
The MATRIX TILES EXIST Statement

You can check that an image has been prepared for a specific matrix using the
MATRIX TILES EXIST statement which has the format shown in FIG-45.27.
FIG-45.27
MATRIX TILES EXIST ( matno )
The MATRIX TILES
EXIST Statement
integer
In the diagram:
The statement returns 1 if a PREPARE MATRIX TEXTURE statement has already

been executed for the matrix specified, otherwise zero is returned.

Positioning the Matrix in 3D Space
The POSITION MATRIX Statement
When a matrix is first created, its bottom-left corner is placed at position (0,0,0).
Although the FILL MATRIX statement allows us to move the matrix in the y-axis,
to position the whole matrix anywhere in 3D space, we need to use the POSITION
MATRIX statement which has the format shown in FIG-45.28.
FIG-45.28
The POSITION MATRIX
Statement POSITION MATRIX matno
{ ,
3Dvno
x , y , z
}
In the diagram:
x,y,z are a sequence of real values specifying where

the bottom-left corner of the grid is to be
positioned.
3Dvno is an integer value previously assigned to a 3D

vector object. The values in this object are used
to position the matrix.
POSITION MATRIX 1,50,20,-40
would place the bottom-left corner of the grid at position (50,20,-40).
The program in LISTING-45.8 creates a matrix and a cube. The cube’s centre is at
position (0,0,0) and exists only to show the position of the 3D origin as the matrix
is moved. The matrix is moved from its original position when the user presses a
key.
Positioning the Matrix REM *** Create grid 20 by 30 rectangles ***
MAKE MATRIX 1,200,300,20,30
REM *** Add random heights to matrix ***

REM *** Create reference object ***

REM *** Prepare matrix image ***

LOAD IMAGE “grass.jpg”,1
REM *** Texture matrix ***

SET MATRIX TEXTURE 1,1,1

LISTING-45.8 REM *** Move matrix to new location ***
(continued) WAIT KEY
POSITION MATRIX 1,50,20,-40
Positioning the Matrix
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.15
Type in and test the program in LISTING-45.8 (matrix08.dbpro).
The MATRIX POSITION Statement

Since a matrix object can be moved to any position within 3D space, DarkBASIC
Pro offers a way to discover that position using the MATRIX POSITION statement.
The format for the statement is shown in FIG-45.29.
FIG-45.29
{
The MATRIX POSITION X
Statement
MATRIX POSITION Y ( matno )
real
In the diagram:
X,Y,Z Choose the option appropriate for the ordinate

required.
The statement returns a real number representing the position of the bottom-left
corner of the specified matrix. For example, if the bottom-left corner of matrix 1 is
positioned at location (20,30,40) with the statement POSITION MATRIX 1,20,30,40
then the lines
PRINT “x-ord : ”,MATRIX POSITION X(1)

PRINT “y-ord : ”,MATRIX POSITION Y(1)
PRINT “z-ord : ”,MATRIX POSITION Z(1)
will display the output:
x-ord : 20
y-ord : 30
z-ord : 40

Matrix Transparency
The GHOST MATRIX ON Statement
The transparency of a matrix can be modified (assuming your graphics card can
handle this effect) using the GHOST MATRIX ON statement. This statement has
FIG-45.30
The GHOST MATRIX
ON Statement
GHOST MATRIX ON matno
[, mode
]
In the diagram:
mode is an integer value between 0 and 5 which

specifies the transparency mode to be used.
If the mode value is omitted, the matrix is displayed in standard transparency mode;
when included, the mode parameter allows different transparency effects to be
produced.
The program in LISTING-45.9 demonstrates the effects of the various mode

settings by placing a matrix representing water just above a terrain object and
switching modes each time a key is pressed. FIG-45.31 shows the display created
by various modes.
Creating a Transparent SET POINT LIGHT 0,0,100,500
Matrix REM *** Create first terrain object ***
MAKE TERRAIN 1,"terraintest.bmp"
LOAD IMAGE “grass2.jpg”,1
TEXTURE TERRAIN 1,1,
POSITION TERRAIN 1,0,0,1023
REM *** Set up matrix with water texture ***
LOAD IMAGE “030.bmp”,2
REM *** Position matrix above terrain ***
POSITION MATRIX 1,0,5,0
REM *** Use each transparency mode ***
FOR c = 0 TO 5
GHOST MATRIX ON 1,c
UPDATE MATRIX 1
WAIT KEY
NEXT c
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION

FIG-45.31 Varying Matrix Transparency
No ghosting Ghosting Mode 0 Ghosting Mode 1
Activity 45.16
The GHOST MATRIX OFF Statement

The transparency effect, if activated using GHOST MATRIX ON, can be switched
off again using the GHOST MATRIX OFF statement which has the format shown
in FIG-45.32.
FIG-45.32
The GHOST MATRIX GHOST MATRIX OFF matno
OFF Statement
In the diagram:
matrno is an integer value giving the matrix ID.

The matrix should have previously been
made transparent using the GHOST MATRIX ON
statement.
The SET MATRIX PRIORITY Statement

If a program displays several 3D objects on screen, then these are drawn one after
another to create the complete scene. Matrix objects are normally drawn first, but
there are times when this gives the wrong effect. For example, in FIG-45.33 a cube
is embedded in a transparent matrix, but you cannot see through the matrix to the
bottom section of the cube.
FIG-45.33
Transparency Issues

The effect of embedding an object within a transparent matrix is shown in
LISTING-45.10.
REM *** Create a textured, transparent grid ***
Demonstrating Problems MAKE MATRIX 1, 2000, 2000,100,100
with Transparency LOAD IMAGE “grass1.jpg”,1
GHOST MATRIX ON 1
UPDATE MATRIX 1
REM *** Embed a cube in the matrix ***
LOAD IMAGE “wall_U5_01.jpg”,2
TEXTURE OBJECT 1,2
DO
LOOP
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.17
Type in the program above (matrix10.dbpro) and observe how the cube is
displayed.
We can force a matrix object to be drawn last by using the SET MATRIX
PRIORITY statement. This will create the desired transparency effect. The SET
MATRIX PRIORITY has the format shown in FIG-45.34.
FIG-45.34
The SET MATRIX
SET MATRIX PRIORITY matno , prtyflg
PRIORITY Statement
In the diagram:
prtyflg is set to zero if the matrix is to be drawn first and

1 (the default) if the matrix is to be drawn last.
Activity 45.18
SET MATRIX PRIORITY 1,0
just before the MATRIX UPDATE statement and observe what difference this
makes to the display.

Lighting the Matrix
The SET MATRIX NORMAL Statement
As we saw in earlier chapters, every vertex in a 3D shape has an invisible line known
as a vertex normal which is used in lighting calculations (see FIG-45.35).
FIG-45.35 Vertex Normals
There is one
Vertex Normals normal for each edge
meeting at a vertex
Since several triangles within a shape may share the same vertex, a single point may
have many vertex normals. The angle and length of a normal has a direct effect on
how a 3D object is lit. By changing a normal, we create an immediate change to
how light is reflected at that point on the object to which the normal is attached.
The SET MATRIX NORMAL statement allows us to modify the length and
direction of matrix normals. The statement requires the vertex and end point of the
modified normal to be specified. The statement format is shown in FIG-45.36.
FIG-45.36 The SET MATRIX NORMAL Statement
SET MATRIX NORMAL matno , xedge , zedge , nx , ny , nz
In the diagram:
xedge is an integer value specifying an edge along the

x-axis.

z-axis.
xedge,zedge are a pair of integer values which together identify

the location of the normal that is to be modified.
nx,ny,nz are three integer values representing the new

position of the unattached end of the vertex
normal.
The program in LISTING-45.11 creates a single tile matrix and modifies a vertex
normal each time a key is pressed.

REM *** Set up 200 by 100 grid (20 by 10 tiles) ***
Modifying Matrix MAKE MATRIX 1,200, 100,1,1
Normals REM *** Change to point light ***
REM *** Load image ***
LOAD IMAGE “F.bmp”,1
REM *** Prepare image ***
REM *** Modify the normal at each vertex ***
FOR x = 0 TO 1
FOR z = 0 TO 1
SET MATRIX NORMAL 1, x, z, 0.0, 10.2, 0.0
UPDATE MATRIX 1
WAIT 1000
NEXT z
NEXT x
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION
Activity 45.19
Type in the program above (matrix11.dbpro) and find out what effect the
changes to the normal have on the way the matrix is lit.
What effects are produced by varying the nx, ny, and nz values (use both
positive and negative values)?
The SET MATRIX Statement

The SET MATRIX statement is by far the most powerful of the matrix commands,
allowing you to set several matrix characteristics in a single line. For example, we
can use it to switch wireframe mode on or off, set any black areas on the textured
grid to become transparent, and change the way the grid’s surface reflects any light.
FIG-45.37
The SET MATRIX
SET MATRIX matno , wireflg , trnsflg , cullflg
Statement
, filter , lghtflg , fogflg , ambflg
In the diagram:
wireflg 0 - displays the grid in wireframe mode

1 - displays the grid in solid mode
trnsflg 0 - grid is opaque.

1 - any area with black texturing becomes
transparent.

cullflg 0 - hidden surfaces are drawn.
1 - hidden surfaces are not drawn.
filter 0 - no texture mipmapping.

1 - no texture smoothing.
2 - uses linear filtering.
lghtflg 0 - lights reflected normally.

1 - lights reflection reduced.
fogflg 0 - fog does not affect grid.

1 - fog effects grid.
ambflg 0 - ambient light is reflected from grid.

1 - ambient light reflection is reduced.
As you can see, there are many parameters to this statement. The best way to make
sense of the effect it has on the matrix grid is to test each possible combination of
parameter values. The program in LISTING-45.12 does exactly that, allowing the
user to cycle through the values for each parameter.
Testing the SET MAKE MATRIX 1, 2000, 2000,100,100
MATRIX Statement REM *** Texture matrix ***
LOAD IMAGE “bwtexture.bmp”,1
REM *** Randomly adjust the y ordinates ***
REM *** to a maximum of 40 ***
SET MATRIX PRIORITY 1,1
UPDATE MATRIX 1
REM *** Embed a cube in the matrix ***
LOAD IMAGE “bricks.jpg”,2
TEXTURE OBJECT 1,2
REM *** Create fog ***
FOG ON
FOG DISTANCE 1000
REM *** Set up parameter values ***
wire = 0
trans = 0
cull = 0
filter = 0
light = 0
fog = 0
amb = 0
REM *** Allow user to modify settings ***
SYNC ON
DO
REM *** Display current settings ***
SET CURSOR 0,10
PRINT “ 1 - Wireframe toggle (”,wire,")"
PRINT “ 2 - Transparency toggle (”,trans,")"
PRINT “ 3 - Cull Toggle (”,cull,")"
PRINT “ 4 - Filter Increment (”,filter,")"
PRINT “ 5 - Light Toggle (”,light,")"

LISTING-45.12 PRINT “ 6 - Fog Toggle (”,fog,")"
(continued) PRINT “ 7 - Ambient Light Toggle (”,amb,")"
REM *** Get parameter to be changed ***
Testing the SET choice = VAL(GetOption())
MATRIX Statement REM *** Change appropriate parameter’s value ***
SELECT choice
CASE 1
wire = 1 - wire
ENDCASE
CASE 2
trans = 1 - trans
ENDCASE
CASE 3
cull = 1 - cull
ENDCASE
CASE 4
filter = (filter + 1) mod 3
ENDCASE
CASE 5
light = 1 - light
ENDCASE
CASE 6
fog = 1 - fog
ENDCASE
CASE 7
amb = 1 - amb
ENDCASE
ENDSELECT
REM *** Show update matrix ***
SET MATRIX 1, wire,trans,cull,filter,light,fog, amb
UPDATE MATRIX 1
SYNC
WAIT 100
LOOP
END
AUTOCAM OFF
ENDFUNCTION
FUNCTION GetOption()
result$ =""
REPEAT
result$ = INKEY$()
UNTIL result$ <>""
WHILE INKEY$() <> ""
ENDWHILE
ENDFUNCTION result$
Activity 45.20
How do the ambflg and lghtflg values interact?
The MATRIX EXIST Statement

Since matrices can be destroyed as well as created, you can check to make sure that
a specific matrix exists using the MATRIX EXIST statement as shown in
FIG-45.38.

FIG-45.38
MATRIX EXIST ( matno )
The MATRIX EXIST
Statement
integer
In the diagram:
The statement returns 1 if the specified matrix does exist, otherwise zero is returned.
Hence, we might report the non-existence of matrix 1 with the statement:
IF MATRIX EXIST(1) <> 1
PRINT “Matrix does not exist”
ENDIF
Summary
l A matrix object can be used to create a landscape.
l Use MAKE MATRIX to create a matrix object.
l A matrix is divided into a number of subdivisions along the x and z axes.
l Use RANDOMIZE MATRIX to randomly set the height of each vertex within
a matrix.
l Use UPDATE MATRIX to update a matrix on-screen after changes have been
made.
l Use SET MATRIX HEIGHT to modify the height of individual vertices within
a matrix.
l Use GET MATRIX HEIGHT to discover the height of any vertex on a matrix.
l Use GET GROUND HEIGHT to discover the height on any arbitary point on a
matrix.
l A matrix is initially shown in wireframe mode.
l Use SET MATRIX WIREFRAME to toggle between wireframe and solid

display modes.
l Use MATRIX WIREFRAME STATE to discover the current display mode of

a matrix.
l Use PREPARE MATRIX TEXTURE to split an image into 1 or more sections.

One of these sections will be used automatically to texture the matrix.
l Use FILL MATRIX to texture the matrix with a specified section of the image
specified in PREPARE MATRIX TEXTURE.
l Use SET MATRIX TILE to place a section of a prepared image anywhere on a

matrix.

l Use SET TEXTURE TRIM to fine tune the mapping of a texture to a matrix.
l Use SHIFT MATRIX to offset the rectangles that make up a grid.
l Use MATRIX TILE COUNT to discover how many sections a prepared image
has been split into.
l Use MATRIX TILES EXIST to check that an image has been prepared for a
matrix.
l Use POSITION MATRIX to reposition a matrix object. The bottom-left corner

of the matrix is placed at the position specified.
l Use MATRIX POSITION to discover the current position of a matrix object.
l Use GHOST MATRIX ON to create a transparency effect on a matrix.
l Use GHOST MATRIX OFF to switch off transparency.
l Use SET MATRIX PRIORITY to modify the order in which matrices are drawn
on the screen.
l Use SET MATRIX NORMAL to modify the lighting of a matrix at a specific

point.
l Use SET MATRIX to set several matrix properities in a single command.
l Use MATRIX EXIST statement to determine if a matrix of a specified ID exists.

Solutions
Activity 45.1 SET MATRIX HEIGHT 1, 0, 1, -12.5
MAKE MATRIX 2,1000,500,10,5 REM *** Update matrix ***

UPDATE MATRIX 1
Activity 45.2 REM *** Allow user to move camera ***

DO
REM *** Set screen resolution *** CONTROL CAMERA USING ARROWKEYS 0,5,1
SET DISPLAY MODE 800, 600, 16 LOOP
REM *** Position camera *** REM *** End program ***
POSITION CAMERA 0, 500,300,-600 END

MAKE MATRIX 1, 1000,500,100,50 Activity 45.5
REM *** Randomly adjust height *** REM *** Set screen resolution ***
REM *** to a maximum of 20 *** SET DISPLAY MODE 800,600,16
REM *** Set up 200 by 100 grid ***
REM *** Update matrix *** MAKE MATRIX 1,200, 100,20,10
UPDATE MATRIX 1
REM *** End program *** height#=0.5
WAIT KEY
END REM *** FOR each Z edge DO ***
FOR z = 0 TO 10
Activity 45.3 height# = height# *1.5
REM *** FOR each X edge DO ***
REM *** Set screen resolution *** FOR x = 0 TO 20
SET DISPLAY MODE 800, 600, 16 REM *** Set height of edge ***
REM *** Position camera *** SET MATRIX HEIGHT 1, X, Z, height#
AUTOCAM OFF NEXT x
POSITION CAMERA 0, 500,300,-600 NEXT z
REM *** Create a grid (1000 by 500) *** REM *** Update screen ***
MAKE MATRIX 1, 1000,500,100,50 UPDATE MATRIX 1
REM *** Randomly adjust height ***
REM *** to a maximum of 20 *** REM *** Position camera ***
RANDOMIZE MATRIX 1, 20 AUTOCAM OFF
REM *** Update matrix *** POSITION CAMERA 0, 0,100,-300
UPDATE MATRIX 1
REM *** Allow user to move camera *** REM *** Allow user to move camera ***
DO DO
CONTROL CAMERA USING ARROWKEYS 0,5,1 CONTROL CAMERA USING ARROWKEYS 0, 1, 1
LOOP LOOP
END WAIT KEY
END
REM *** Position camera *** Activity 45.6
AUTOCAM OFF
POSITION CAMERA 0,0,100,-300 REM *** Set screen resolution ***
ENDFUNCTION SET DISPLAY MODE 1280,1024,16

Activity 45.4 AUTOCAM OFF
SET DISPLAY MODE 800, 600, 16 REM *** Create a grid (150 by 100)***
MAKE MATRIX 1, 150,100,3,2
REM *** Position camera for a good view ***
AUTOCAM OFF REM *** Set back corner height to 10.5 ***
POSITION CAMERA 0, 500,300,-600 SET MATRIX HEIGHT 1, 3,2, 10.5
REM *** Create a grid (1000 by 500) *** REM *** Update matrix ***
MAKE MATRIX 1, 1000,500,100,50 UPDATE MATRIX 1
REM ** Lift (8,3) to 25.7 *** REM *** Get the height every 10 units ***
SET MATRIX HEIGHT 1, 8, 3, 25.7 REM *** along 70.0 line of matrix ***
REM *** Sink (0,1) to -12.5 *** SYNC ON

FOR c = 0 TO 150 STEP 10 REM *** Update screen ***
SET CURSOR 0, 10 UPDATE MATRIX 1
PRINT “Height at (”,c,",",70,") = “,
GET GROUND HEIGHT(1,c,70) REM *** Position camera ***
SYNC AUTOCAM OFF
WAIT 500 POSITION CAMERA 0, 0,100,-300
NEXT c
REM *** End program *** DO
WAIT KEY CONTROL CAMERA USING ARROWKEYS 0, 1, 1
END LOOP

END
SET DISPLAY MODE 1280,1024,16 The texture image has been split into 6 tiles (3 by 2). It is
the first of these tiles (the top-left one) which is used to
AUTOCAM OFF
texture the whole grid.
REM *** Create a grid (150 by 100)*** Activity 45.9

MAKE MATRIX 1, 150,100,3,2 REM *** Set screen resolution ***
REM *** Set back corner height to 10.5 ***
SET MATRIX HEIGHT 1, 3,2, 10.5 REM *** Load texture image ***
REM *** Make matrix solid ***
SET MATRIX WIREFRAME OFF 1 REM *** Set up 200 by 100 grid ***
MAKE MATRIX 1,200, 100,20,10
UPDATE MATRIX 1 REM *** Texture matrix ***
REM *** Get the height every 10 units *** FILL MATRIX 1, 0, 5
REM *** along 70.0 line of matrix ***
SYNC ON REM *** Set initial height ***
FOR c = 0 TO 150 STEP 10 height#=0.5
SET CURSOR 0, 10
PRINT “Height at (”,c,",",70,") = “, REM *** FOR each X edge DO ***
GET GROUND HEIGHT(1,c,70) FOR z = 0 TO 10
SYNC REM *** Increase height by 50% ***
WAIT 500 height# = height# *1.5
NEXT c REM *** FOR each Z edge DO ***
FOR x = 0 TO 20
REM *** End program *** REM *** Set height of edge ***
WAIT KEY SET MATRIX HEIGHT 1, X, Z, height#
END NEXT x
NEXT z
Activity 45.8 REM *** Update screen ***

UPDATE MATRIX 1
AUTOCAM OFF
REM *** Set up 200 by 100 grid ***
DO
MAKE MATRIX 1,200, 100,20,10
LOOP
height#=0.5
WAIT KEY
REM *** FOR each X edge DO ***
END
FOR z = 0 TO 1
height# = height# *1.5 Notice that the PREPARE MATRIX TEXTURE
REM *** FOR each Z edge DO *** statement has been moved so that the FILL MATRIX can
FOR x = 0 TO 20 be executed before SET MATRIX HEIGHT.
REM *** Set height of edge ***
SET MATRIX HEIGHT 1, X, Z, height#
NEXT x Activity 45.10
NEXT z
REM *** Texture matrix ***

First modification: REM *** Set screen resolution ***
SET DISPLAY MODE 1280, 1024, 16 REM *** Position camera ***
AUTOCAM OFF
REM *** Position camera *** POSITION CAMERA 0, 0,100,-300
AUTOCAM OFF
POSITION CAMERA 0, 200,300,-600 REM *** Create grid 20 by 30 rectangles
***
REM *** Create 5 by 3 grid *** MAKE MATRIX 1,200,300,20,30
MAKE MATRIX 1,100, 100,5,3
REM *** Set random height in matrix ***
REM *** Load image and tile as 3 by 2 *** RANDOMIZE MATRIX 1, 30
LOAD IMAGE "B.bmp",1
PREPARE MATRIX TEXTURE 1,1,3,2 REM *** Divide image into 20 x 30 tiles
***
REM *** Load tile 2 into rectangle (4,0) LOAD IMAGE "grass.jpg",1,0
*** PREPARE MATRIX TEXTURE 1,1,20,30
REM *** Copy each tile onto the grid ***
REM *** Load tile 5 into rectangle (2,1) REM *** starting at back-left ***
*** tileno = 1
SET MATRIX TILE 1,2,1,5 FOR z =29 TO 0 STEP -1
FOR x = 0 TO 19
REM *** Update display *** SET MATRIX TILE 1, x, z, tileno
UPDATE MATRIX 1 INC tileno
NEXT x
REM *** Get user control of camera *** NEXT z
DO
CONTROL CAMERA USING ARROWKEYS 0, 1, 1 REM *** Update grid ***
LOOP UPDATE MATRIX 1
REM *** End program *** REM *** Use camera to move about ***
WAIT KEY DO
END CONTROL CAMERA USING ARROWKEYS 0, 1, 1
LOOP
Second modification: REM *** End program ***
WAIT KEY
REM *** Set screen resolution *** END
REM *** Position camera *** Activity 45.13

AUTOCAM OFF
POSITION CAMERA 0, 200,300,-600 REM *** Set screen resolution ***
REM *** Create 5 by 3 grid ***
MAKE MATRIX 1,100, 100,5,3 REM *** and camera ***
AUTOCAM OFF
REM *** Load image and tile as 3 by 2 *** POSITION CAMERA 0, 0,100,-300
LOAD IMAGE "B.bmp",1
PREPARE MATRIX TEXTURE 1,1,3,2 REM *** Set up grid with texture ***
REM *** Tile 1 - (1,2) MAKE MATRIX 1,40,40,2,2
SET MATRIX TILE 1,1,2,1 LOAD IMAGE "trimmer.bmp",1,1
REM *** Load tile 2 into rectangle (4,0) PREPARE MATRIX TEXTURE 1,1,2,2
***
SET MATRIX TILE 1,4,0,2 REM *** Position grid ***
REM *** Tile 3 - (3,2) *** POSITION MATRIX 1, 0,0,0
SET MATRIX TILE 1, 3,2,3 REM *** Set xtrim and ztrim to zero ***
REM *** Tile 4 - (1,1) xoff# = 0.0
SET MATRIX TILE 1,1,1,4 zoff# = 0.0
REM *** Load tile 5 into rectangle (2,1) DO
*** REM *** Set the trim on the grid ***
SET MATRIX TILE 1,2,1,5 SET MATRIX TRIM 1,xoff#,0.0
REM Tile 6 - (3,1) REM *** Fill rectangles with tile ***
SET MATRIX TILE 1,3,1,6 FILL MATRIX 1,0,2
REM *** Update display *** UPDATE MATRIX 1
UPDATE MATRIX 1 REM *** Wait for key press ***
WAIT KEY
REM *** Get user control of camera *** REM *** Make the offset larger ***
DO xoff# = xoff# + 0.01
CONTROL CAMERA USING ARROWKEYS 0, 1, 1 zoff# = zoff# + 0.01
LOOP LOOP
REM *** End program *** REM *** End program ***
WAIT KEY WAIT KEY
END END

REM *** Set up grid with texture *** lghtflg must be 1 before ambflg is effective.
MAKE MATRIX 1,2000,3000,20,30
LOAD IMAGE "grass.jpg",2
tileno = 1
FOR x = 0 TO 19
INC tileno
NEXT x
NEXT z

UPDATE MATRIX 1
REM *** Use camera ***

AUTOCAM OFF
DO
REM *** Shift matrix one position ***
SHIFT MATRIX UP 1
UPDATE MATRIX 1
WAIT 50
LOOP

END
Change
SHIFT MATRIX UP 1
to
SHIFT MATRIX LEFT 1
and later to
SHIFT MATRIX RIGHT 1
to test other directions.
Activity 45.15
Activity 45.16
Activity 45.17
The lower part of the cube can be seen through the
transparent matrix.
Activity 45.18
The lower part of the cube can no longer be seen.
Activity 45.19
The lighting effect changes subtly for different values.
Using negative values gives a much duller finish.

46
Accessing and Modifying a Vertex's Normal
Accessing and Modifying a Vertex's UV Settings
Capturing an Object's Vertex Data
Creating New Shapes by Modifying Vertex Coordinates
Modifying a Vertex's Coordinates
Reading a Vertex's Coordinates
DarkBASIC Pro: Manipulating Vertices 1245

Manipulating Vertices
Introduction
Every 3D object is constructed from a set of triangular polygons and each of the
three vertices that make up a triangle are represented by a set of (x,y,z) coordinates
(see FIG-46.1).
FIG-46.1 Defining a Triangular Polygon
(-5,0,5) (5,0,5)
A Triangle Vertices
(0,0,0)
A triangle is the simplest polygon... ...and is defined by three vertices... ...with each vertex being defined by its
x,y,z coordinates.
The individual vetices of a 3D mesh or limb can be manipulated using a set of vertex
data statements that are included in DarkBASIC Pro. For example, we could start
with a cube and move one of its vertices to create a new shape (see FIG-46.2).
FIG-46.2
A Distorted Cube
The Statements
The LOCK VERTEXDATA FOR MESH Statement
Before we can begin to modify the vertex data of a mesh, we need to take a copy
of the coordinates of the mesh's vertices. To capture the vertex data we use the
LOCK VERTEXDATA FOR MESH statement which has the format shown in
FIG-46.3.
FIG-46.3
The LOCK LOCK VERTEXDATA FOR MESH meshno
VERTEXDATA FOR
MESH Statement
In the diagram:

whose vertex details are to be copied.
1246 DarkBASIC Pro: Manipulating Vertices

For example, we could create a simple triangle object with the statement
MAKE OBJECT TRIANGLE 1,0,0,0,-5,0,5,5,0,5
and then convert it to a mesh with the line:
From there we can save the new mesh's vertex data with the line
LOCK VERTEXDATA FOR MESH 1
The GET VERTEXDATA VERTEX COUNT Statement

A complex mesh may contain hundreds of vertices. The number of vertices in a set
of locked vertex data can be discovered using the GET VERTEXDATA VERTEX
COUNT which has the format shown in FIG-46.4.
FIG-46.4
The GET VERTEXDATA GET VERTEXDATA VERTEX COUNT ( )
INDEX COUNT Statement
integer
The statement returns the number of vertices in the locked data.
For example, we could determine the number of vertices in our locked triangle (as
given in the example above) using the line:
count = GET VERTEXDATA VERTEX COUNT()
The program in LISTING-46.1 brings together the example statements used earlier
to create a triangle, convert it to a mesh, store the triangle's data and display the
vertex count.
REM *** Create triangle ***
Finding the Number of MAKE OBJECT TRIANGLE 1,0,0,0,-5,0,5,5,0,5
Vertices in a Mesh REM *** Convert it to a mesh ***
DELETE OBJECT 1
REM *** Capture mesh vertex data ***
REM *** Find out number of vertices in mesh ***
REM *** Display number of vertices ***
DO
SET CURSOR 100,100
PRINT count
LOOP
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
ENDFUNCTION

Activity 46.1
Type in and test the program given in LISTING-46.1 (vertex01.dbpro).
The GET VERTEXDATA POSITION Statement

We can find the coordinates of each locked vertex using the GET VERTEXDATA
POSITION statement which has the format shown in FIG-46.5.
FIG-46.5
{
The GET X
VERTEXDATA
POSITION Statement
GET VERTEXDATA POSITION Y ( vertexno )
real
In the diagram:
X,Y,Z Use one of these options appropriate for the

ordinate required.
vertexno is an integer value giving the number of

the vertex whose coordinates are to be retrieved.
Note that the first vertex number is zero.
The triangle mesh we created in the first program will have vertices numbered, 0,
1, and 2. To retrieve the coordinates of vertex 0 we would use the lines:
x# = GET VERTEXDATA POSITION X(0)

y# = GET VERTEXDATA POSITION Y(0)
z# = GET VERTEXDATA POSITION Z(0)
Activity 46.2
Add the following lines at appropriate positions within your last program and
hence display the coordinates of the second vertex (vertex 1) :
REM *** Get coordinates of second vertex ***

y# = GET VERTEXDATA POSITION Y(1)
z# = GET VERTEXDATA POSITION Z(1)
SET CURSOR 100,120

PRINT STR$(x#,0)," ",STR$(y#,0)," ",STR$(z#,0)
The next program (see LISTING-46.2) stores the coordinates of a triangles vertices
in a set of arrays - one array for all the x-ordinates, one for the y-ordinates, and one
for the z-ordinates. The coordinates of the vertices are then displayed. The program
includes a routine (Format$()) to create an aligned layout.

Displaying the REM *** Create root object for model ***
Coordinates of an MAKE OBJECT SPHERE 99,0
Object's Vertices
REM *** Convert it to a mesh ***

DELETE OBJECT 1
REM *** ADD mesh to model ***

ADD LIMB 99,1,1
REM *** Copy the vertex data ***

REM *** Retrieve number of vertices in mesh ***

count = GET VERTEXDATA INDEX COUNT()
REM *** Create arrays to hold the coordinates of each vertex ***
DIM x#(count)
DIM y#(count)
DIM z#(count)
REM *** Store coords of each vertex in arrays ***

FOR c = 0 TO count-1
x#(c) = GET VERTEXDATA POSITION X(c)
y#(c) = GET VERTEXDATA POSITION Y(c)
z#(c) = GET VERTEXDATA POSITION Z(c)
NEXT c
REM *** Display vertex count and coords ***

DO
SET CURSOR 100,100
PRINT "Number of vertices in mesh : ",count
PRINT
FOR c = 0 TO count-1
SET CURSOR 100,120+c*20
PRINT "Vertex number ",c," ",Format$(x#(c),2,0)," ",
Format$(y#(c),2,0)," ",Format$(z#(c),2,0)
NEXT c
LOOP

END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
FUNCTION Format$(v#,intpart,decpl)
result$ = STR$(v#,decpl)
result$ = SPACE$(intpart+decpl-LEN(result$))+result$
ENDFUNCTION result$
Activity 46.3

The SET VERTEXDATA POSITION Statement
We can alter the coordinates of a specified vertex using the SET VERTEXDATA
POSITION statement which has the format shown in FIG-46.6.
FIG-46.6
The SET VERTEXDATA POSITION Statement
SET VERTEXDATA POSITION vertexno , x , y , z
In the diagram:
vertexno is an integer value giving the vertex number of the

vertex whose coordinates are to be changed.
x,y,z are a set of real numbers giving the new

coordinates to be assigned to the specified vertex.
For example, we could change the coordinates of vertex 0 in our triangle to (5,5,5)
using the line:
SET VERTEXDATA POSITION 0,5,5,5
However, this won't have any effect on the triangle on the screen. To do that we
must unlock the vertex values.
The UNLOCK VERTEXDATA Statement

The UNLOCK VERTEXDATA statement allows the updated vertex data to be
applied to the mesh or limb from which they were originally extracted. This
FIG-46.7
The UNLOCK UNLOCK VERTEXDATA
VERTEXDATA
Statement
The program in LISTING-46.3 adds a mesh as a limb to a model, modifies the
mesh's vertex data, and then adds the modified mesh as a second limb; finally, the
model is rotated.
Modifying a Mesh's REM *** Create model root ***

Vertices MAKE OBJECT SPHERE 99,0
REM *** Make triangle mesh ***

MAKE OBJECT TRIANGLE 1,-1,0,0,-5,0,5,5,0,5
DELETE OBJECT 1
REM *** Add triangle as first limb ***

ADD LIMB 99,1,1
REM *** Modify mesh data ***

UNLOCK VERTEXDATA

LISTING-46.3 REM *** Add modified mesh as second limb ***
(continued) ADD LIMB 99,2,1
Modifying a Mesh's REM *** Rotate model ***

Vertices DO
WAIT 1
LOOP
REM *** End program ****

END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.4
Type in and test the program in LISTING-46.3 (vertex03.dbpro).
The LOCK VERTEXDATA FOR LIMB Statement

Rather than retrieve vertex data relating to a mesh, it is also possible to capture the
vertex data of a limb (which, after all, is simply a mesh which has been added to a
model). This is done using the LOCK VERTEXDATA FOR LIMB statement which
FIG-46.8
The LOCK LOCK VERTEXDATA FOR LIMB objno , limbno
VERTEXDATA FOR
LIMB Statement
In the diagram:
objno is an integer value specifying the ID of the root

object of a model.
limbno is an integer value specifying the ID of the limb

whose vertex data is to be copied.
Once the vertex data has been copied, the usual vertex data commands such as GET
VERTEXDATA INDEX COUNT, GET VERTEXDATA POSITION and SET
VERTEXDATA POSITION can be applied. When the UNLOCK VERTEXDATA
is executed the affected limb within the model will be modified.
In the next program (see LISTING-46.4) a triangle is added as limb 1 within a model
and then modified directly.
Modifying Limb REM *** Make triangular mesh ***

vertices Directly MAKE OBJECT TRIANGLE 1,0,0,0,-5,3,5,5,0,5
DELETE OBJECT 1

LISTING-46.4 REM *** Create root object ***
Modifying Limb REM *** and add limb ***

vertices Directly ADD LIMB 1,1,1
REM *** Display current shape till key press ***

WAIT KEY
REM *** Capture triangle's vertex data ***

LOCK VERTEXDATA FOR LIMB 1,1
REM *** Modify vertex in triangle ***


UNLOCK VERTEXDATA

WAIT KEY
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.5
Modify the SET VERTEX POSITION statement so that vertex 2 changes

coordinates to (7,10,0).
How does this affect the triangle?
Actually, we can modify an object directly without having to convert it to a mesh

or a limb. The trick is to remember that the root object in a model is referred to as
limb zero. So, by specifying an object's ID number and using a limb number value
of zero, we can directly modify normal objects using the LOCK VERTEXDATA
FOR LIMB.
For example, if we create a triangle with the statement
we can then capture the triangle's vertex data with the line:
After that, it is just a matter of the usual vertex manipulation and then unlocking
the vertex data buffer.

Activity 46.6
Replace the main section of your last program with the lines
SetUpScreen()
REM *** Make triangle ***

WAIT KEY
REM *** Capture its vertex data ***
REM *** Modify vertex in triangle ***
UNLOCK VERTEXDATA
WAIT KEY
END
Does the new program create the same effect as the previous version?
The GET VERTEXDATA NORMALS Statement

A vector normal is a line originating from a vertex and is used by the program to
calculate lighting effects on a surface near that vertex. A normal has a length of
exactly one world unit and the direction in which it points is determined by its
end-point (the start point being at the vertex). The coordinates of that end-point can
be retrieved using the GET VERTEXDATA NORMALS statement which has the
FIG-46.9
{
X
The GET
VERTEXDATA
NORMALS Statement GET VERTEXDATA NORMALS Y ( vertexno )
real
In the diagram:
X,Y,Z Use one of these options appropriate for the

ordinate required.
vertexno is an integer value giving the number of the

vertex whose normal's coordinates are to be
retrieved.
For example, we could retrieve the triangle's vertex normal at vertex zero using the
lines:
x# = GET VERTEXDATA NORMALS X(0)

y# = GET VERTEXDATA NORMALS Y(0)
z# = GET VERTEXDATA NORMALS Z(0)
Activity 46.7
Modify your previous program to display the vertex normals for all three
points on the triangle using the following logic:
FOR vertexno = 0 TO 2
Get x,y, and z ordinates for vector normal
Position cursor at 100, 100 + vertexno * 20
Display x,y,z values
ENDFOR
The SET VERTEXDATA NORMALS Statement

It is also possible to modify the vertex normals and hence affect the shading of the
surface beside that normal. This is done using the SET VERTEXDATA
NORMALS statement which has the format shown in FIG-46.10.
FIG-46.10
The SET VERTEXDATA NORMALS Statement
SET VERTEXDATA NORMALS vertexno , x , y , z
In the diagram:
vertexno is an integer value giving the number of the

vertex whose vertex normal is to be changed.
x,y,z are a set of real numbers giving the new

coordinates to be assigned to the specified vertex
normal.
For example, we could change the normal at vertex 0 in our triangle to (0,0,0) using
the line:
SET VERTEXDATA NORMALS 0,0,0,0
As with any other change, the screen display will only change when the vertex data
is unlocked.
In the next program (see LISTING-46.5) we eliminate the normal at each vector by
setting their values to (0,0,0) to show how this affects the lighting.
Changing Vertex MAKE OBJECT TRIANGLE 1,0,0,0,-5,3,5,5,0,5
Normals WAIT KEY
REM ** FOR each vertex normal DO ***
FOR v = 0 TO 2
REM *** eliminate its normal ***
SET VERTEXDATA NORMALS v,0,0,0
UNLOCK VERTEXDATA
WAIT KEY
NEXT v continued on next page

LISTING-46.5 REM *** End program ***
END
Changing Vertex
Normals
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.8
The GET VERTEXDATA Statement

If a 3D object has been textured, then the UV offset at each vertex can be discovered
using the GET VERTEXDATA statement which has the format given in FIG-46.11.
FIG-46.11
The GET
VERTEXDATA UV
Statement
GET VERTEXDATA
{ U
V
( vertexno )
real
In the diagram:
U,V Choose the appropriate option for the offset data

required.
vertexno is an integer value specifying the vertex whose

UV offsets are to be discovered.
The program in LISTING-46.6 displays the UV offsets for each vertex of a triangle.
Displaying Vertex UV LOAD IMAGE "Spot1.bmp",1
Settings TEXTURE OBJECT 1,1
DO
FOR vertexno = 0 TO 2
u# = GET VERTEXDATA U(vertexno)
v# = GET VERTEXDATA V(vertexno)
SET CURSOR 100, 100 + vertexno * 20
PRINT "UV value for vertex ",vertexno," U: ",STR$(u#,2),
" V : ",STR$(v#,2)
NEXT v
LOOP
WAIT KEY

LISTING-46.6 FUNCTION SetUpScreen()
COLOR BACKDROP 0
Displaying Vertex UV BACKDROP ON
Settings AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.9
The values displayed are the default settings.
Modify the texturing of the triangle by adding the line
immediately after the triangle is textured.
How does this affect the results displayed.
The SET VERTEXDATA UV Statement

To modify the UV offset of an object's texture at a particular vertex, we can use the
SET VERTEXDATA UV statement which has the format shown in FIG-46.12.
FIG-46.12
The SET VERTEXDATA SET VERTEXDATA UV vertexno , Uoffset , Voffset
UV Statement
In the diagram:

UV offsets are to be changed.
Uoffset is a real number giving the offset to be applied

along the U axis (range: -1 to 1).
Voffset is a real number giving the offset to be applied

along the V axis (range: -1 to 1).
SET VERTEXDATA UV 0,-0.4,0.0
creates the effect shown in FIG-46.13 on a triangle textured with the white dot on
a red background image.
FIG-46.13
The Effects of
Modifying Vertex UV
Settings
Default Texture Mapping After Using the

SET VERTEXDATA UV
Statement

The program in LISTING-46.7 scrolls the texture at point zero in a triangle by
continually changing the UV offset values.
Modifying Vertex UV REM *** Load texture image ***

Values LOAD IMAGE "spot1.bmp",1
REM *** create triangle and texture ***

TEXTURE OBJECT 1,1
REM *** Capture the vertices ***

WAIT KEY
LOCK VERTEXDATA FOR LIMB 1,0,2
REM *** Modify UV offset values for vertex 0 ***

FOR offset# = -1 TO 1 STEP 0.1
SET VERTEXDATA UV 0,offset#,offset#
WAIT 100
NEXT offset#

WAIT KEY
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.10
Modify the program to offset vertices 0 and 1 by the value of offset#.
The SET VERTEXDATA DIFFUSE Statement

The diffuse setting of a vertex can be modified using the SET VERTEXDATA
DIFFUSE statement which has the format shown in FIG-46.14.
FIG-46.14
The SET
SET VERTEXDATA DIFFUSE vertexno , colour
VERTEXDATA Diffuse
Statement
In the diagram:

diffuse setting is to be changed.
colour is a integer number (usually created using RGB)

specifying the diffuse colour to be used.
For example, we could create a yellow diffuse at vertex 2 using the line:
SET VERTEXDATA DIFFUSE 2, RGB(255,255,0)

The GET VERTEXDATA DIFFUSE Statement
The current diffuse setting of a specified vertex can be discovered using the GET
VERTEXDATA DIFFUSE statement which has the format shown in FIG-46.15.
FIG-46.15
GET VERTEXDATA DIFFUSE ( vertexno )
The GET
VERTEXDATA
DIFFUSE Statement
integer

diffuse setting is to be found.
An integer value is returned. This represents the colour diffuse setting.
A typical statement would be:
vertcolour = GET VERTEXDATA DIFFUSE(2)
Handling More Complex Shapes

So far we've demonstrated this batch of commands using only a triangle, since that
is the simplest shape we can create. But now it's time to look at more complex
objects.
Next up is the plane. At first we might be fooled into thinking that a plane can be
represented by 4 vertices, but, in fact, 6 are recorded in the vertex buffer.
Two triangles are required to create a plane and each triangle consists of 3 vertices
(see FIG-46.16).
FIG-46.16 1
3 0
All Objects are Created
from Triangles
4 2
5
A plane may appear to have only four ...but it is constructed using two triangles,
vertices... so there are 6 vertices (3 for each
triangle).
Activity 46.11
How many vertices should be required to define a cube?
The program in LISTING-46.8 displays a plane in wireframe mode and lists the
coordinates of each of its vertices.

REM *** Create wireframe plane ***
Displaying Vertex MAKE OBJECT PLAIN 1,10,10
Coordinates SET OBJECT WIREFRAME 1,1
REM *** Capture vertex data ***
noofvertices = GET VERTEXDATA VERTEX COUNT()
REM *** Display vertex details ***
DO
SET CURSOR 100,100
PRINT "Vertices : ",noofvertices
FOR vertexno = 0 TO noofvertices - 1
x# = GET VERTEXDATA POSITION X(vertexno)
y# = GET VERTEXDATA POSITION Y(vertexno)
z# = GET VERTEXDATA POSITION Z(vertexno)
SET CURSOR 100, 120 + vertexno * 20
PRINT vertexno, " x : ",Format$(x#,3,2)," y : ",
Format$(y#,3,2)," z : ",Format$(z#,3,2),
NEXT vertexno
LOOP
END
FUNCTION Format$(v#,intpart,decpl)
result$ = STR$(v#,decpl)
result$ = SPACE$(intpart+decpl-LEN(result$))+result$
ENDFUNCTION result$
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.12
Change the code so that a cube (10x10x10) is created. Does the number of
vertices used match the predicted figure?
We'll discover why a cube does not use the number of vertices we predicted later.
The vertex data specifies coordinates relative to the centre of the object, not as
absolute coordinates in 3D space. We can see this from the results of the Activity
below.
Activity 46.13
In your previous program (vertex08.dbpro) immediately after the wireframe

statement, add the line:
How does this affect the vertex coordinates displayed on the screen?
Another cause for confusion is that the vertex coordinates are measured using
reversed x and z axes. This means that negative x values are to the right and negative

z values are "into the screen". For example, the coordinates of the plane's vertex
zero are given as (-5,5,0) and so we might be tempted to assume that vertex zero is
at the top-left corner. In fact, vertex zero is at the top-right corner, the -5 ordinate
being a result of the reversed x-axis measurements. LISTING-46.9 demonstrates
this point by incrementally moving vertex zero from (-5,5,0) to (-6,5,0).
Modifying a Vertex's REM *** Create wireframe plane ***

Coordinates MAKE OBJECT PLAIN 1,10,10
SET OBJECT WIREFRAME 1,1
REM *** Get position of vertex zero ***

WAIT KEY
vertexno = 0
startx# = GET VERTEXDATA POSITION X(vertexno)
REM *** Assign a new x ord for vertex ***

finx# = -6
REM *** Modify vertex position incrementally ***

FOR c# = startx# TO finx# STEP -0.1
SET VERTEXDATA POSITION vertexno, c#,y#,z#
WAIT 100
NEXT c#

WAIT KEY
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.14
Activity 46.15
Run vertex08.dbpro again.
Which vertices have the coordinates (5,5,0)?
Two other vertices have matching coordinates. What are those coordinates?
If we want to modify the position of a vertex, we may have to reposition all the
entries in the vertex data area which match that value. For example, in the plane
created in the program above.
To move a single vertex of a 3D object it may be necessary to assign the new

coordinates to several vertex data entries (see FIG-46.17)

FIG-46.17 Modifying an Plane's Shape
If we start with a plane, and wish to ... the shape shown above... ... we need to move the bottom-left
distort it to... corner to the new position as shown.
Since the original shape is constructed ... adjusting a single entry in the vertex To achieve the correct shape, both
from two triangles... data modifies a single triangle. vertex data entries must be changed.
In a complex 3D shape a single point in space can be repeated over many vertices,
so it will be useful if we can find the position of every matching entry in the vertex
data.
To do that we can create a function FindMatchingVertices(). The routine takes as

parameters the object number whose vertices are to be searched and the number of
the vertex for which all matches are to be found. Since several vertices may match
the one specified, we'll need an array to store the matching vertex numbers. In
addition, we need to keep a count of how many matches are actually found:
DIM matchvertx(12)
count AS INTEGER
Since a function cannot return an array, we'll declare the array to be global. It's
probably easiest just to make count global too:
GLOBAL DIM matchvertx(12)
GLOBAL count AS INTEGER
The routine itself has the following logic:
Capture vertex data for specified object
Get the coordinates for the specified vertex number
Set count to zero
FOR each vertex DO
IF the vertex's coords match those of specified vertex THEN
Add 1 to count
Add vertex number to matchvertx[count]
ENDIF
ENDFOR
The actual code is :

FUNCTION FindMatchingVertices(objno,vertexno)
REM *** Capture the object's vertex data ***
LOCK VERTEXDATA FOR LIMB objno,0
REM *** get coords of specified vertex ***
REM *** Set count to zero ***

count = 0
REM FOR each vertex DO ***

FOR vno = 0 TO GET VERTEXDATA VERTEX COUNT()-1
REM *** IF its coords match specified vertex coords ***
IF x# = GET VERTEXDATA POSITION X(vno) AND
y# = GET VERTEXDATA POSITION Y(vno) AND
z# = GET VERTEXDATA POSITION Z(vno)
REM *** Add to count ***
INC count
REM *** Add vertex number to array ***
matchvertx(count) = vno
ENDIF
NEXT vno
REM *** Restore data ***

UNLOCK VERTEXDATA
ENDFUNCTION
FIG-46.18 demonstrates the effect of searching for matches for vertex 2 of the
rectangle used earlier.
FIG-46.18 How to Change all Common Vertices
plane object 1
matchvertx
FindMatchingVertices(1,2)
count Object
Vertex
vertex 2
We start with the data areas in which Next, we choose the object and a Now we can call the function and
the results will be stored. vertex. search for vertices with matching coords.
FOR c = 1 TO count
matchvertx SET VERTEXDATA matchvertx(c),
-2.5,2.5,0
2 5 NEXT c
count
2
vertex 2
vertex 5
Vertex 5 has the same coords as So the function will make two entries ...and we can use this information to
vertex 2. in the data structure... move all relevant vertices and modify
the object's shape.
The program in LISTING-46.10 shows the technique in practice, adjusting the

points on a plane to create the arrow shape shown above.

LISTING-46.10 GLOBAL DIM matchvertx(12)
Modifying all Matching SetUpScreen()
Vertices REM *** Create plane ***
REM *** Find matches for vertex 2
FindMatchingVertices(1,2)
WAIT KEY
REM *** Reposition matching vertices ***
FOR c = 1 TO count
p = matchvertx(c)
SET VERTEXDATA POSITION p,3.5,3.5,0
NEXT c
UNLOCK VERTEXDATA
WAIT KEY
END
count = 0
FOR vno = 0 TO GET VERTEXDATA VERTEX COUNT()-1
IF x# = GET VERTEXDATA POSITION X(vno) AND
y# = GET VERTEXDATA POSITION Y(vno) AND
z# = GET VERTEXDATA POSITION Z(vno)
INC count
matchvertx(count) = vno
ENDIF
NEXT vno
UNLOCK VERTEXDATA
ENDFUNCTION
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.16
Modify the program so that the reshaping of the plane is performed

incrementally, creating an animated effect.
By adding the line
make the plane rotate about its y-axis while it is changing shape.
With a bit of imagination we can transform basic 3D shapes into more interesting
forms. For example, we can create a hemisphere by modifying all the vertices with
y values less than zero to have a y-ordinate of zero as shown in LISTING-46.11.

REM *** Create textured sphere ***
Creating a Hemisphere MAKE OBJECT SPHERE 1,10,50,50
TEXTURE OBJECT 1,1
REM *** Zeroise all y ords < 0 ***
FOR vertexno = 0 TO noofvertices - 1
IF y# < 0
y# = 0
SET VERTEXDATA POSITION vertexno,x#,y#,z#
ENDIF
NEXT vertexno
UNLOCK VERTEXDATA
REM *** Rotate the object ***
DO
PITCH OBJECT UP 1,1
WAIT 10
LOOP
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.17
Modify the line

y# = 0
to
y# = -y#
How does this affect the shape created?
Activity 46.18
Write a program (act4618.dbpro) which converts a cube to a pyramid (as

below) and rotates the final shape.

The ADD MESH TO VERTEXDATA Statement
Should we require it, the vertex details of a second object can be added to the vertex
data buffer. However, only data from a mesh can be added, so the object whose
details are to be appended to the buffer must first be duplicated in the form of a
mesh. For example, if we needed the vertices of a triangle object to be added to an
existing vertex data buffer, we would begin by creating the triangle
MAKE OBJECT TRIANGLE 2,-1,-1,-1,0,0,0,1,1,1
and then creating an equivalent mesh:
With the necessary mesh available, we now use the ADD MESH TO
VERTEXDATA statement to add the new vertex details to the buffer. The ADD
MESH TO VERTEXDATA statement has the format shown in FIG-46.19.
FIG-46.19
The ADD MESH TO ADD MESH TO VERTEXDATA meshno
VERTEXDATA
Statement
In the diagram:
meshno is an integer value giving the ID of the mesh

whose vertex data is to be added to the buffer.
Using this statement, we can add the details of the triangle object to the buffer with
the line:
ADD MESH TO VERTEXDATA 1 Àdd mesh 1's data to buffer
The program in LISTING-46.12 adds a triangle's data to that of a cube in the vertex
data buffer, displaying the total number of vertices within the buffer both before
and after the triangle is added.
LISTING-46.12
SetUpScreen()
Adding Data to the
Vertex Data Buffer REM *** Make cube and place its data in buffer ***
REM *** Display number of vertices in buffer ***

mes$= STR$(noofvertices)
DisplayMessage(mes$,200,200)
REM *** Create triangle and equivalent mesh ***

MAKE OBJECT TRIANGLE 2, -1,-1,-1,0,0,0,1,1,1
REM *** Add mesh to buffer ***

ADD MESH TO VERTEXDATA 1
REM *** Display number of vertices in buffer ***

mes$= STR$(noofvertices)
DisplayMessage(mes$,200,200)

WAIT KEY

COLOR BACKDROP 0
Adding Data to the BACKDROP ON
Vertex Data Buffer AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
now = TIMER()
SET CURSOR x,y
PRINT mes$
ENDWHILE
ENDFUNCTION
Activity 46.19
How many vertices are in the buffer once the triangle's data has been added?
More About the Vertex Data Buffer's Structure

The vertex data buffer is a complex structure; not only does it contain the
coordinates of every vertex that makes up a 3D object, but it also contains details
of how those vertices are linked to form the polygons that make up that object (see
FIG-46.20)
FIG-46.20 Data in the Vertex Buffer
0,12,5 3,19,2 3,9,6 0,12,5 3,8,5 0,12,5 3,19,2 3,9,6 0,12,5 3,8,5
0 1 2 3 4 0 1 2 3 4
0,1,2 3,4,5
The coordinates of
each vertex in the
3D object The first entry states
that vertices 0,1, and 2
make up the first polygon
in the 3D object
If we create a shape and then lock its ...we have access to a data structure ...but also details on how these vertices
vertex data buffer... that not only contains a list of the are joined together.
vertices within that shape...
As we can see, a second array within the vertex buffer contains details of how the
vertices are linked to form the triangular polygons that make up the structure. The
number of entries in this second array should match the number of triangles required
to construct the 3D object represented.
For example, a triangle object only needs 1 entry, a plane 2, and a cube 12 (two for
each side of the cube). To differentiate between the two sets of data, we'll call the
first area the vertex array and the second one the polygon array. By using this second
array detailing the points used to construct each polygon, the need to duplicate
vertex coordinates in the first array disappears. The reasoning behind this is shown
in FIG-46.21.

FIG-46.21 Alternative Methods of Holding an Object's Details
1 Duplication
3 0 vertex array
vertex array
5,5,0 -5,5,0 5,-5,0 -5,-5,0
5,5,0 -5,5,0 5,-5,0 -5,5,0 -5,-5,0 5,-5,0 0 1 2 3 4 5
0 1 2 3 4 5
No duplicated
0 1 2 1 2 3 vertices
Duplication
polygon array
4 2
5
When a plane is created its coordinates ...with details of all six points recorded - ...but by detailing which points link to
can be saved in the vertex buffer... even though some of these are make a polygon, duplication of vertices
duplicates... can be eliminated.
In fact, both methods are used. For simpler objects, such as triangles and planes,
only the vertex array is created, but for more complex shapes, such as cubes,
duplicate vertices are eliminated from the vertex array and details of which points
make up each polygon are recorded in the polygon array.
The GET VERTEXDATA INDEX COUNT Statement

We can discover exactly how many entries are in this polygon array of the vertex
buffer using the GET VERTEXDATA INDEX COUNT statement which has the
FIG-46.22
The GET VERTEXDATA GET VERTEXDATA INDEX COUNT ( )
INDEX COUNT Statement
integer
The statement actually returns the number of values in the polygon array, and, since
there are three values for each polygon, we'll get three times the expected amount
The program in LISTING-46.13 creates a cube and displays the number of entries
in the vertex array and the polygons used to construct the object. This last value is
arrived at by dividing the value returned by GET VERTEXDATA INDEX COUNT
by three.
REM *** Make cube and place its data in buffer ***
Displaying a Cube's MAKE OBJECT CUBE 1,10
Vertex Data LOCK VERTEXDATA FOR LIMB 1,0,2
SYNC
REM *** Display number of vertices and polygons ***
noofpolygons = GET VERTEXDATA INDEX COUNT()/3
PRINT "Vertices : ", noofvertices
PRINT "Polygons : ", noofpolygons
SYNC
WAIT KEY
END

COLOR BACKDROP 0
Displaying a Cube's BACKDROP ON
Vertex Data AUTOCAM OFF
POINT CAMERA 0,0,0
SYNC ON
ENDFUNCTION
Activity 46.20
What results are returned when the object created is:
a) a triangle
b) a plane
c) a cone
Notice that the triangle and plane return a value of zero for the polygon count. This
is because there are only three vertices in the triangle, so retaining information about
how these should be joined is unnecessary; the plane, on the other hand, retains
duplicated vertices and polygons are created by using the vertex array entries in
groups of three (that is, entries 0, 1 and 2 form the first polygon; 3,4 and 5 the
second polygon; etc.)
The GET INDEXDATA Statement

We can find out the contents of individual cells within the polygon array using the
GET INDEXDATA statement which has the format shown in FIG-46.23.
FIG-46.23
GET INDEXDATA ( sub )
The GET INDEXDATA
Statement
integer
In the diagram:
sub is an integer value specifying the element of the

polygon array to be accessed. The first element is
element zero.
We could use this statement to discover which entries in the vertex array are used
to construct a specific polygon. For example, assuming we have stored data about
a cube within the vertex buffer, we could discover which cells of the vertex array
are used in the construction of the first polygon on the first face of the cube using
the lines
v1 = GETINDEXDATA(0)
and with this information we could go on to discover the coordinates of each of the
three points:

x1 = GET VERTEXDATA POSITION X(v1)
y1 = GET VERTEXDATA POSITION Y(v1)
z1 = GET VERTEXDATA POSITION Z(v1)
Activity 46.21
Write a program which creates a 10 unit cube and displays which elements of
the vertex array are used to construct its first two polygons.
The program in LISTING-46.14 takes the program in Activity 46.21 a stage further
and displays the coordinates used in constructing the second polygon of the cube.
LISTING-46.14 TYPE CoordsType
x#
Displaying Vertex y#
Coordinates z#
ENDTYPE
DIM polycoords(3) AS CoordsType
SetUpScreen()
REM *** Make cube and place its data in buffer ***
REM *** Collect coordinates ***
FOR c = 1 TO 3
v = GET INDEXDATA(c+2)
polycoords(c).x# = GET VERTEXDATA POSITION X(v)
polycoords(c).y# = GET VERTEXDATA POSITION Y(v)
polycoords(c).z# = GET VERTEXDATA POSITION Z(v)
NEXT c
REM *** Display the data collected ***

SET CURSOR 0,820
PRINT "Coordinates of second polygon are:"
FOR c = 1 TO 3
PRINT "(",polycoords(c).x#,",",polycoords(c).y#,",",
Ä
polycoords(c).z#,")"
NEXT c

WAIT KEY
END
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.22

The SET INDEXDATA Statement
It is possible to modify the contents of the polygon array, thereby changing which
three entries in the vertex array are used to construct a polygon. This is achieved
using the SET INDEXDATA statement which has the format shown in FIG-46.24.
FIG-46.24
The SET INDEXDATA
SET INDEXDATA polyindx , vertindx
Statement
In the diagram:
polyindx is an integer value giving the element in the

polygon array which is to have its contents
changed. This should lie between 0 and
GET VERTEXDATA INDEX COUNT()-1.
vertindx is an integer value specifying the new value.

This will correspond to the index of a cell in
the vertex array. This should lie between 0 and
GET VERTEXDATA VERTEX COUNT()-1.
For example, assuming a cube's details are stored in the vertex buffer, we could
change the vertices used to construct the first polygon in the cube to those stored in
cells 7, 8, and 9 of the vertex array using the lines:
SET INDEXDATA 0,7

SET INDEXDATA 1,8
SET INDEXDATA 2,9
FIG-46.25 gives a visual interpretation of the operation.

FIG-46.25 vertex array
lower-numbered cells higher-numbered cells
The Result of Using SET 7 8 9
INDEXDATA
The polygon array
contains details of which
cells in the vertex array are to
7 8 9 1 2 3
be used as the coordinates
when constructing a polygon
polygon array
In LISTING-46.15 a cube has a vertex specification belonging to its first polygon

changed to reference a different cell in the vertex array. The modified cube is rotated
to show how this change has affected its appearance.
Modifying a Cube's REM *** Make cube and place its data in buffer ***
Structure MAKE OBJECT CUBE 1,10
REM *** Modify details of how the first polygon is constructed ***
WAIT KEY
SET INDEXDATA 0,3
REM *** Switch on culling for more detail ***

SET OBJECT CULL 1,0

LISTING-46.15 REM *** Update screen ***
UNLOCK VERTEXDATA
Modifying a Cube's REM *** Rotate cube ***
Structure DO
WAIT 10
LOOP
WAIT KEY
END
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 46.23
The DELETE MESH FROM VERTEXDATA Statement

Using ADD MESH TO VERTEXDATA we saw that it was possible to add new
models to the data in the vertex data buffer. These additions can also be removed
from the vertex buffer, but this is a more complex operation since it requires details
of which entries in the vertex array and the polygon array have to be deleted.
To delete part of the vertex data buffer we use the DELETE MESH FROM
VERTEXDATA statement which has the format shown in FIG-46.26.
FIG-46.26
The DELETE MESH DELETE MESH FROM VERTEXDATA vertindx1 , vertindx2
FROM VERTEXDATA
Statement
, polyindx1 , polyindx2
In the diagram:
vertindx1,vertindx2 are integer values representing the subscripts

of the first and last cells in the vertex array
whose contents are to be deleted.
polyindx1,polyindx2 are integer values representing the subscripts

of the first and last cells in the polygon array
whose contents are to be deleted.
For example, if we place the vertex data for a cube into the vertex buffer, 24 cells
in the vertex array will be occupied (cells 0 to 23) and 36 cells in the polygon array
(cells 0 to 35). Adding a second cube to the vertex buffer would result in an
additional 24 entries in the vertex array (cells 24 to 47) and 36 extra entries in the
polygon array (cells 36 to 71). To delete this second cube from the vertex buffer we
would use the statement:
DELETE MESH FROM VERTEXDATA 24,47,36,71

Summary
l Every 3D Object is constructed using triangles.
l By modifying the coordinates of a vertex, a 3D object can be reshaped.
l To modify an object's vertices, they must first be captured in a buffer area.
l Use LOCK VERTEXDATA FOR MESH to capture into the buffer area the
vertices that make up a mesh.
l Use GET VERTEXDATA VERTEX COUNT to discover how many vertices

are stored in the buffer area.
l Use GET VERTEXDATA POSITION to access the coordinates of a given

vertex.
l Use SET VERTEXDATA POSITION to modify the coordinates of a specific

vertex.
l Use UNLOCK VERTEXDATA to allow the modified data to be applied to the

object or mesh to which it is associated.
l Use LOCK VERTEXDATA FOR LIMB to capture the vertex data of a limb or
object.
l Use GET VERTEXDATA NORMALS to access the end coordinates of a given

vertex normal.
l Use SET VERTEXDATA NORMALS to modify the coordinates of a specific

vertex normal.
l Use GET VERTEXDATA {UV} to access the U and V offsets at a given vertex
for a textured object.
l Use SET VERTEXDATA UV to modify the U/V offsets of a specific vertex.
l Use SET VERTEXDATA DIFFUSE to modify the diffuse setting for a vertex.
l Use GET VERTEXDATA DIFFUSE to access the current diffuse setting for a
specific vertex.
l If a vertex's coordinates are modified, generally, all other vertices with identical
coordinates should also be modified.
l Manipulation of vertices allows us to create new shapes from the basic 3D

primitives. For example, a hemisphere from a sphere or a pyramid from a cube.
l Vertex coordinates are relative to the object's centre and not absolute world
coordinates.
l Use ADD MESH TO VERTEXDATA to add the details of a mesh to the data
already stored in the vertex buffer.
l The vertex data buffer contains an array (the vertex array) which holds the
coordinates of every vertex used to make up the 3D object(s) held in the buffer.

l For some simple objects such as a plane, vertex coordinates may be repeated and
vertices are used in groups of three to recreate the triangles that make up the
original object.
l For more complex shapes, each vertex's coordinates are stored only once and a
second array (the polygon array) holds links to those vertices used to construct
each polygon.
l Use GET VERTEXDATA INDEX COUNT to discover the number of entries

in the polygon array.
l Use GET INDEXDATA to discover which vertex an element of the polygon

array references.
l Use SET INDEXDATA to modify which vertex an entry in the polygon array
references.
l Use DELETE MESH FROM VERTEXDATA to remove a mesh's data from the
vertex data buffer.

Solutions
Activity 46.1
BACKDROP ON
AUTOCAM OFF
SetUpScreen() POINT CAMERA 0,0,0
ENDFUNCTION
MAKE OBJECT TRIANGLE 1,0,0,0,-5,0,5,5,0,5 The top right corner of the triangle is moved.
REM *** Convert it to a mesh ***
DELETE OBJECT 1 Activity 46.6
REM *** Capture mesh vertex data ***
LOCK VERTEXDATA FOR MESH 1 The program has the same effect as the previous one.
REM *** Find out no. of vertices in mesh ***
REM *** Get coordinates of second vertex *** Activity 46.7
y# = GET VERTEXDATA POSITION Y(1) SetUpScreen()
z# = GET VERTEXDATA POSITION Z(1) REM *** Make triangle ***
REM *** Show number of vertices *** MAKE OBJECT TRIANGLE 1,0,0,0,-5,3,5,5,0,5
REM *** & vertex 1's coords *** WAIT KEY
DO REM *** Capture its vertex data ***
SET CURSOR 100,100 LOCK VERTEXDATA FOR LIMB 1,0
PRINT count REM *** Modify vertex in triangle ***
SET CURSOR 100,120 SET VERTEXDATA POSITION 2,7,10,0
PRINT STR$(x#,0)," ",STR$(y#,0)," ", REM *** Update screen ***
ÄSTR$(z#,0) UNLOCK VERTEXDATA
LOOP LOCK VERTEXDATA FOR LIMB 1,0
REM *** End program *** DO
END FOR vertexno = 0 TO 2
normx# =
FUNCTION SetUpScreen() Ä
GET VERTEXDATA NORMALS X(vertexno)
SET DISPLAY MODE 1280,1024,32 normy# =
COLOR BACKDROP 0 Ä
GET VERTEXDATA NORMALS Y(vertexno)
BACKDROP ON normz# =
AUTOCAM OFF Ä
GET VERTEXDATA NORMALS Z(vertexno)
ENDFUNCTION SET CURSOR 100, 100 + vertexno * 20
PRINT normx#," ",normy#," ",normz#
NEXT vertexno
Activity 46.3 LOOP
No solution required. WAIT KEY
END
Activity 46.4 FUNCTION SetUpScreen()

No solution required. COLOR BACKDROP RGB(255,255,100)
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
SetUpScreen() ENDFUNCTION
REM *** Make triangular mesh ***
MAKE MESH FROM OBJECT 1,1 Activity 46.8
DELETE OBJECT 1
REM *** Create root object *** No solution required.
REM *** and add limb ***
ADD LIMB 1,1,1 Activity 46.9
WAIT KEY
REM *** Capture triangle's vertex data *** SetUpScreen()
LOCK VERTEXDATA FOR LIMB 1,1 MAKE OBJECT TRIANGLE 1,0,0,0,-5,0,5,5,0,5
REM *** Modify vertex in triangle *** LOAD IMAGE "Spot1.bmp",1
SET VERTEXDATA POSITION 2,7,10,0 TEXTURE OBJECT 1,1
REM *** Update screen *** SCROLL OBJECT TEXTURE 1,0.2,0.3
UNLOCK VERTEXDATA DO
REM *** End program *** LOCK VERTEXDATA FOR LIMB 1,0
WAIT KEY FOR vertexno = 0 TO 2
END u# = GET VERTEXDATA U(vertexno)
v# = GET VERTEXDATA V(vertexno)

SET CURSOR 100, 100 + vertexno * 20 Activity 46.13
PRINT "UV value for vertex ",
Ä vertexno," U: ",STR$(u#,2)," V : ", The vertex coordinates are unchanged. This means that the
Ä STR$(v#,2) coordinates are measured from the centre of the object and
NEXT v
not from the world origin.
UNLOCK VERTEXDATA
LOOP
WAIT KEY
END
Activity 46.14
COLOR BACKDROP 0
BACKDROP ON Activity 46.15
AUTOCAM OFF
POSITION CAMERA 0,10,-20 Vertices 1 and 3 have the coordinates (5,5,0).
POINT CAMERA 0,0,0
ENDFUNCTION Vertices 2 and 4 have the coordinates (-5,-5,0).
The UV values should now read as:

Activity 46.16
UV value for vertex 0: 0.20 0.30
Version 1 ( animated transform )
SetUpScreen()
Activity 46.10 REM *** Create plane ***
SetUpScreen() MAKE OBJECT PLAIN 1,10,10
REM *** Load texture image *** REM *** Find matches for vertex 2 ***
LOAD IMAGE "spot1.bmp",1 FindMatchingVertices(1,2)
REM *** create triangle and texture *** WAIT KEY
MAKE OBJECT TRIANGLE 1,0,0,0,-5,5,0,5,5,0 REM *** Reposition matching vertices ***
TEXTURE OBJECT 1,1 FOR post# = -5 TO 3.5 STEP 0.1
REM *** Capture the vertices *** LOCK VERTEXDATA FOR LIMB 1,0
WAIT KEY FOR c = 1 TO count
LOCK VERTEXDATA FOR LIMB 1,0,2 p = matchvertx(c)
REM *** Modify UV offset values for vertex SET VERTEXDATA POSITION p,post#,post#,0
0 *** NEXT c
FOR offset# = -1 TO 1 STEP 0.1 UNLOCK VERTEXDATA
SET VERTEXDATA UV 0,offset#,offset# WAIT 50
SET VERTEXDATA UV 1,offset#,offset# NEXT post#
WAIT 100 REM *** End program ***
NEXT offset# WAIT KEY
REM *** End program *** END
WAIT KEY
END FUNCTION FindMatchingVertices(objno,vertexno)
FUNCTION SetUpScreen() x# = GET VERTEXDATA POSITION X(vertexno)
SET DISPLAY MODE 1280,1024,32 y# = GET VERTEXDATA POSITION Y(vertexno)
COLOR BACKDROP RGB(200,200,100) z# = GET VERTEXDATA POSITION Z(vertexno)
BACKDROP ON count = 0
FOR vno = 0 TO
Ä
AUTOCAM OFF
POSITION CAMERA 0,0,-30 GET VERTEXDATA VERTEX COUNT()-1
POINT CAMERA 0,0,0 IF x# = GET VERTEXDATA POSITION X(vno)
ENDFUNCTION Ä
AND y#=GET VERTEXDATA POSITION Y(vno)
Ä
AND z#=GET VERTEXDATA POSITION Z(vno)
INC count
Activity 46.11 matchvertx(count) = vno
ENDIF
We would expect a cube to use 36 vertices. Since a plane NEXT vno
requires 6 vertices and there are 6 planes in a cube, we UNLOCK VERTEXDATA
should need a total of 6 * 6 vertices. ENDFUNCTION
Activity 46.12 COLOR BACKDROP 0
BACKDROP ON
To make a cube, change the line
AUTOCAM OFF
MAKE OBJECT PLAIN 1,10,10 POINT CAMERA 0,0,0
ENDFUNCTION
to
Version 2 ( rotating animated transform )
The cube used 24 vetices rather than the predicted 36. GLOBAL count AS INTEGER

SetUpScreen()
REM *** Create plane ***
Activity 46.18
MAKE OBJECT PLAIN 1,10,10 SetUpScreen()
REM *** Find matches for vertex 2 *** REM *** Create textured cube ***
FindMatchingVertices(1,2) MAKE OBJECT CUBE 1,10
WAIT KEY LOCK VERTEXDATA FOR LIMB 1,0
REM *** Reposition matching vertices *** LOAD IMAGE "grid8by8.bmp",1
FOR post# = -5 TO 3.5 STEP 0.1 TEXTURE OBJECT 1,1
LOCK VERTEXDATA FOR LIMB 1,0 REM *** Modify vertices ***
FOR c = 1 TO count noofvertices = GET VERTEXDATA VERTEX
p = matchvertx(c) COUNT()
SET VERTEXDATA POSITION p,post#,post#,0 FOR vertexno = 0 TO noofvertices - 1
NEXT c x# = GET VERTEXDATA POSITION X(vertexno)
UNLOCK VERTEXDATA y# = GET VERTEXDATA POSITION Y(vertexno)
WAIT 50 z# = GET VERTEXDATA POSITION Z(vertexno)
NEXT post# IF x# <> 0 AND Y# > 0
REM *** End program *** x# = 0
WAIT KEY z# = 0
END SET VERTEXDATA POSITION vertexno,
Ä
x#,y#,z#
ENDIF
LOCK VERTEXDATA FOR LIMB objno,0 NEXT vertexno
x# = GET VERTEXDATA POSITION X(vertexno) UNLOCK VERTEXDATA
y# = GET VERTEXDATA POSITION Y(vertexno) REM *** Rotate pyramid ***
z# = GET VERTEXDATA POSITION Z(vertexno) DO
count = 0 TURN OBJECT LEFT 1,1
FOR vno = 0 TO
ÄGET VERTEXDATA VERTEX COUNT()-1
WAIT 10
LOOP
IF x# = GET VERTEXDATA POSITION X(vno)
Ä
AND y#=GET VERTEXDATA POSITION Y(vno)
Ä
AND z#=GET VERTEXDATA POSITION Z(vno) END
INC count
matchvertx(count) = vno FUNCTION SetUpScreen()
ENDIF SET DISPLAY MODE 1280,1024,32
NEXT vno COLOR BACKDROP RGB(200,200,100)
UNLOCK VERTEXDATA BACKDROP ON
ENDFUNCTION AUTOCAM OFF
FUNCTION SetUpScreen() POINT CAMERA 0,0,0
SET DISPLAY MODE 1280,1024,32 ENDFUNCTION
COLOR BACKDROP 0
BACKDROP ON
POINT CAMERA 0,0,0 The cube creates 24 entries in the buffer while the
ENDFUNCTION triangle adds another 3 giving a total of 27.

The program now creates a hollow bowl shape as shown The original MAKE OBJECT CUBE statement needs to
below. be changed for each shape we require.
The triangle requires a line such as:
MAKE OBJECT TRIANGLE 1,-5,-5,-5,0,0,0,5,5,5
This gives 3 vertices and 0 polygon entries.
The plane:
vertices: 6 polygon links: 0
The cone:
vertices: 23 polygon links: 11

Activity 46.21
SetUpScreen()
REM *** Make cube and capture its data ***
SYNC
REM *** Display first 2 polygons ***
FOR poly = 0 TO 1
PRINT "Polygon ",poly," : ";
FOR vert = 0 TO 2
PRINT GET INDEXDATA(poly*3+vert)," ";
NEXT vert
PRINT
NEXT poly
SYNC
WAIT KEY
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
SYNC ON
ENDFUNCTION
Activity 46.22
Activity 46.23

47
Bitmap and Image Memblocks
Data Memblocks
Manipulating Memblocks
Mesh Memblocks
Saving and Loading Memblocks
Sound Memblocks
Using Pointers
DarkBASIC Pro: Accessing Memory 1279

Accessing Memory
Introduction
The basic storage unit in computer memory is the bit. A bit can store a single binary
digit (0 or 1). In a computer's main memory these bits are formed into groups of 8
known as bytes. Every byte is allocated an unique memory address. These addresses
are allocated in sequential order. Hence the first location is given the address 0, the
next address 1, etc.
When we create a variable with a statement such as
lives = 3
the computer reserves a few bytes of memory for that variable (see FIG-47.1) and
it's in those bytes that the variable's contents are stored.
FIG-47.1 Computer Memory

Program statement
A bit can store x = 7
1 a 1 or a 0 0
0 1 1 0 0 1 0 0
0 1 0 1 1 1 1 0 0 1 1 0 0 1 0 0
Bits are formed 0 1 0 1 1 1 1 0
0 0 1 0 0 1 0 1 0 0 1 0 0 1 0 1
into groups of 8 0 1 0 1 0 1 0 0 0 1 0 1 0 1 0 0
1 0 1 0 1 0 0 0 1 0 1 0 1 0 0 0
0 1 0 0 0 1 0 0
0 1 0 0 0 1 0 0 0 0 1 1 0 1 0 1
0 1 1 0 0 1 0 0 0 0 1 1 0 1 0 1 0 1 0 0 0 0 0 0
0 1 0 0 0 0 0 0 x is assigned 0 1 0 0 0 0 1 1
0 1 0 0 0 0 1 1 space in memory
A bit is the smallest unit of storage. Computer memory is a sequence of Every variable in a program is allocated
A byte is a group of 8 bits. bytes, each given its own address. one or more bytes somewhere in
memory.
Most of the time, finding out which locations have been allocated to a variable is
of no interest to us since we can access its contents by using the variable's name.
In fact it's not just variables that are allocated memory space; every item of data
used by a program, from simple integers to images, sounds and 3D objects will all
be allocated memory within RAM when a program is executing.
Pointers
While most variables store the traditional numeric or string values, there is a special
variable type capable of storing the address at which an item of normal data is held.
Such a variable is known as a pointer (see FIG-47.2).
FIG-47.2 A pointer variable

is designed to hold
a memory address
A Pointer Variable
0 1 1 0 0 1 0 0
0 1 0 1 1 1 1 0
2009 0 0 1 0 0 1 0 1
0 1 0 1 0 1 0 0
A pointer variable 1 0 1 0 1 0 0 0
0 1 0 0 0 1 0 0
0 0 1 1 0 1 0 1
0 1 0 0 0 0 0 0
0 1 0 0 0 0 1 1
1280 DarkBASIC Pro: Accessing Memory

Creating Pointers in DarkBASIC Pro
If we intend to use a variable to store a memory address, then it needs to be created
as a DWORD variable:
ptr AS DWORD
Assigning a Value to a Pointer

In theory we can place any integer value in a pointer
ptr = 2009
and treat that value as a memory address (after all, there's bound to be a memory
location which has that address) but, in practice, this won't work since the operating
system will not allow a program to access parts of the memory that have not been
allocated to that program.
Instead, we need to ask the program to reserve a new area of memory and then place
the address of this new area in the pointer.
The MAKE MEMBLOCK Statement

To allocate a new block of memory we need to use the MAKE MEMBLOCK
FIG-47.3
The MAKE MAKE MEMBLOCK memno , size
MEMBLOCK Statement
In the diagram:
memno is an integer value specifying the ID to be

assigned to the memory block being allocated.
size is an integer value giving the number of bytes

required in the memory block.
For example we could allocate a single byte to memory block 1 using the line:
MAKE MEMBLOCK 1,1
The GET MEMBLOCK PTR Statement

Now we need to acquire the address of the newly allocated memory. This is done
using the GET MEMBLOCK PTR statement which has the format shown in
FIG-47.4.
FIG-47.4
GET MEMBLOCK PTR ( memno )
The GET MEMBLOCK
PTR Statement
integer

In the diagram:
memno is an integer value specifying the ID of the

memory block whose address is to be found.
Using this statement, we can assign the memory block's address to our pointer:
ptr = GET MEMBLOCK PTR(1)
Using a Pointer
Now that the pointer contains the address of the memory block, we can use the
pointer to access the reserved area. To do this we dereference the pointer. This
simply means that we use the contents of the pointer to access the memory location
specified (see FIG-47.5).
FIG-47.5 Dereferencing
a pointer gives
Dereferencing a Pointer access to the memory
location specified
2009 A
ptr
If this seems a strange idea, imagine looking at a treasure map; the map does not
contain the treasure - but it tells you where to find it. A pointer variable doesn't
contain the data you're searching for, but it tells you where to find it.
In DarkBASIC Pro we put an asterisk (*) in front of the pointer variable in order to
dereference it. For example, the statement
*ptr = 65
will store the value 65 at the memory location referenced by ptr, and
PRINT *ptr
will display the value held at that location.
The program in LISTING-47.1 brings together everything we've looked at so far,

setting up a byte of memory, storing the value 65 at the location and then displaying
its contents.
LISTING-47.1 REM *** Declare the pointer variable ***
ptr AS DWORD
Using a Pointer
REM *** Reserve a single byte ***
MAKE MEMBLOCK 1,4
REM *** Copy the address of the byte to the pointer ***
ptr = GET MEMBLOCK PTR(1)
REM *** Store a value in the memory location ***

*ptr = 65
REM *** Display the contents of the memory location ***

PRINT (*ptr)
WAIT KEY
END

Activity 47.1
Type in and test the program in LISTING-47.1 (mem01.dbpro).
It is important that you understand that the term ptr refers to the contents of the
variable ptr and that the term *ptr refers to the contents of the address specified by
ptr (see FIG-47.6).
FIG-47.6 ptr 2009 A
ptr and *ptr

The term ptr The term *ptr
refers to the refers to the
contents here contents here
Using a Pointer to Return Values from a Function

By using a pointer as a parameter to a function we can modify the contents of a
memory block from within the function. The program in LISTING-47.2 creates 3
separate memory block areas, each capable of storing a single integer value. The
areas are then assigned values: 1, 2 and 3 respectively. The user can then choose
which of the three areas are to be updated by the AddTwo() function which adds 2
to the area specified in its parameter.
LISTING-47.2
REM *** Declare the pointer variable ***
ptr AS DWORD
Pointers and Parameters
REM *** Reserve 3 4-byte areas ***
MAKE MEMBLOCK 1,4
MAKE MEMBLOCK 2,4
MAKE MEMBLOCK 3,4
REM *** Store a value in each area ***

FOR c = 1 TO 3
ptr = GET MEMBLOCK PTR(c)
REM *** Store a value in each memory location ***
*ptr = c
NEXT c
To keep the code to a REM *** Get user's choice ***

minimum, the input INPUT "Which area should be updated (1,2,or 3)",area
value is not validated.
REM *** Get the address of the selected area ***
ptr = GET MEMBLOCK PTR(area)
REM *** Use function to add 2 to area ***

AddTwo(ptr)
REM *** Display the contents of the area ***

PRINT "That area now contains the value ",(*ptr)

WAIT KEY
END
REM *** Add 2 to specified address ***

FUNCTION AddTwo(p AS DWORD)
*p = (*p)+2
ENDFUNCTION.

Activity 47.2
Make sure that it works for each memory area.
Larger Memory Blocks

We can reserve as large a memory block as we want (within reason). For example,
let's say we want to store the current position of a spaceship. That will require space
for three real variables (x, y, z). Since a real number occupies 4 bytes, we'll need to
reserve a total area of 12 bytes:
MAKE MEMBLOCK 1,12
The problem now is that we need some method of storing data in the different parts
of this area (see FIG-47.7).
FIG-47.7 How Are Different Areas of a Memory Block Accessed?
x y z x y z
12 bytes
How is this area

to be accessed?
MAKE MEMBLOCK 1,12 will reserve We'd like to divide this into 3 groups ...but how are the extra bytes to be
12 bytes. each of 4 bytes to store coordinates... accessed?
The WRITE MEMBLOCK Statement

To write to part of a memory block we must use the WRITE MEMBLOCK
statement which allows us to specify exactly where within a memory block data is
to be written. The statement has the format shown in FIG-47.8.
FIG-47.8 The WRITE MEMBLOCK Statement
{
BYTE
WORD
WRITE MEMBLOCK memno , position , value
DWORD
FLOAT
In the diagram:
BYTE Use this option if a single byte is to be written to

the memory block.
WORD Use this option if an integer stored over 4 bytes
is to be written to the memory block.

DWORD Use this option if an integer stored over 8 bytes
is to be written.
FLOAT Use this option if a real number is to be written.
memno is an integer value giving the ID of the memory

block to which the data is to be written.
position is an integer value specifying the position of

the first byte within the memory block to be
used to store the value.
When the value is to be stored at the start of the
memory block, use zero.
value is the value to be stored in the memory block.

The value's type should match that used earlier
in the statement.
For example, assuming the spaceship's coordinates are (2.0,5.7,3.9), we could write
these values to memory block 1 using the statements:
WRITE MEMBLOCK FLOAT 1,0,2.0

FIG-47.9 gives a visual representation of the effect of these statements.

FIG-47.9 x y z
2.0 5.7 3.9
Storing Multiple Data
Values in a Memory
WRITE MEMBLOCK FLOAT 1,0,2.0 WRITE MEMBLOCK FLOAT 1,8,3.9
Block
The MEMBLOCK Statement

To access information already stored in a memory block, we use the MEMBLOCK
FIG-47.10
{
BYTE
The MEMBLOCK
Statement
WORD
MEMBLOCK ( memno , position )
DWORD
FLOAT
integer, DWORD or real

In the diagram:
BYTE, WORD, DWORD, FLOAT

Use the option appropriate for the type of value

being retrieved from the memory area.

memory block to be accessed.
position is an integer value specifying where within the

memory block the value to be accessed begins.
The type of value returned by this statement depends on the type of value being
read. The BYTE and WORD options both return an integer value, DWORD returns
a DWORD type, and FLOAT returns a real.
The GET MEMBLOCK SIZE Statement

We can discover how many bytes are in a memory block using the GET
MEMBLOCK SIZE statement which has the format shown in FIG-47.11.
FIG-47.11
The GET MEMBLOCK
GET MEMBLOCK SIZE ( memno )
SIZE Statement
integer
In the diagram:

memory block to be accessed.
The DELETE MEMBLOCK Statement

When it is no longer required, it is best to free up the RAM space allocated to a
memory block. To achieve this we can use the DELETE MEMBLOCK statement
FIG-47.12
The DELETE DELETE MEMBLOCK memno
MEMBLOCK Statement
In the diagram:

memory block to be deleted.
The MEMBLOCK EXIST Statement

We can check if a memory block of a specific ID exists using the MEMBLOCK
FIG-47.13
MEMBLOCK EXIST ( memno )
The MEMBLOCK
EXIST Statement
integer

In the diagram:

block to be checked.
The COPY MEMBLOCK Statement

If we set up more than one memory area, then we can copy information from a
section of one memory block to some part of the second block (see FIG-47.14).
FIG-47.14
Data from a section
of one memory block
Data from one Memory can be copied to a
Block can be Copied to second memory block
Another
Memory block 1 Memory block 2
Copying is done using the COPY MEMBLOCK statement which has the format
shown in FIG-47.15.
FIG-47.15 The COPY MEMBLOCK Statement
COPY MEMBLOCK memno1 , memno2 , offset1 , offset2 , bytes
In the diagram:
memno1 is an integer value giving the ID of the source

memory block.
memno2 is an integer value giving the ID of the destination

memory block.
offset1 is an integer value giving the position within

memno1 at which copying is to start.
offset2 is an integer value giving the starting position

within memno2 at which copied data is to be
placed.
bytes is an integer value giving the number of bytes to

be copied.
For example, if we have previously created two memory blocks using the statements
MAKE MEMBLOCK 1, 20
MAKE MEMBLOCK 2, 50
and then placed data in memory block 1, the third 10 bytes of data in memory block
1 can be copied to the fourth group of 10 bytes in memory block 2 using the
statement:
COPY MEMBLOCK 1,2,20,30,10

Strings and Memory Blocks
You may have noticed that memory blocks seem to be able to hold only numeric
values; there would appear to be no way of writing a string. Strings present a slight
difficulty since they can vary from containing no characters to containing
thousands. If we are to store strings, then we either need to have some method of
detecting the number of characters in a particular string, or we have to make all
strings the same size. Although this second option might not seem very practical,
it is, in fact, often used, particularly when many strings are being held. We'll start
by finding out how to store a single string in a memory block. First, we allow the
user to type in the string from the keyboard:
INPUT "Please enter your name ",name$
Now we need to discover how many characters are in the string so we can create a
memory block of appropriate size:
characters = LEN(name$)
MAKE MEMBLOCK 1,characters
Since we can only transfer numbers to the memory block, we'll convert each
character in the string to a numeric value and write it to the memory block:
FOR c = 1 TO characters
value = ASC(MID$(name$,c))
WRITE MEMBLOCK BYTE 1,c-1,value
NEXT c
To retrieve the information, we need to create an empty string, read each value from
the block, convert it back to a character, and add the character to the end of our
string we have just created:
result$=""
FOR c = 0 TO GET MEMBLOCK SIZE(1)-1
ch$ = CHR$(MEMBLOCK BYTE(1,c)
result$ = result$ + ch$
NEXT c
The program in LISTING-47.3 brings these statements together, storing and

retrieving a string from a memory block.
LISTING-47.3
REM *** Get string ***
Storing String Values in INPUT "Enter name ",name$
a Memory Block REM *** Create block ***
characters = LEN(name$)
MAKE MEMBLOCK 1,characters
REM *** Store name in block ***
FOR c = 1 TO characters
value = ASC(MID$(name$,c))
WRITE MEMBLOCK BYTE 1,c-1,value
NEXT c
REM *** Read name back ***
result$=""
FOR c = 0 TO GET MEMBLOCK SIZE(1)-1
ch$ = CHR$(MEMBLOCK BYTE(1,c))
result$ = result$ + ch$
NEXT c
REM *** Display string read ***
PRINT "Value retrieved was ",result$
WAIT KEY
END

Activity 47.3
In the next program we'll be a bit more ambitious. This time we'll use a memory
block to store the top five scores in a game. Each entry will contain the player's
name and his score.
This is where a fixed-size string will be useful. If we assume a name containing

exactly 25 characters, then we know that every entry in our top-five list will be
exactly 29 bytes in length; 25 for the name and 4 for the score. This, in turn, allows
us to calculate the start position of every entry in the list; 0, 29, 56, etc.
But what about names that are less than 25 characters in length? We just pad them
out with spaces to make them the required length. Names greater than 25 just won't
be accepted.
So, the new program uses the following logic:
Set up memblock for 5 entries
FOR 5 times DO
Get name
Get score
Add entry to memory block
ENDFOR
Display the memory block's contents
The logic is implemented in LISTING-47.4.

LISTING-47.4 MAKE MEMBLOCK 1,29*5
FOR c = 1 TO 5
Storing Top Score name$ = GetName$()
Details in a Memory score = GetScore()
Block AddDataToList(name$,score,c)
NEXT c
DisplayContentsOfMemBlock()
WAIT KEY
END
FUNCTION GetName$()
INPUT "Enter name ",name$
WHILE LEN(name$) > 25
INPUT "Invalid name. Please re-enter ",name$
ENDWHILE
REM *** Pad name to 25 characters ***
name$=name$+SPACE$(25 - LEN(name$))
ENDFUNCTION name$
FUNCTION GetScore()
INPUT "Enter score ",score
WHILE score < 0
INPUT "Invalid score. Please re-enter ",score
ENDWHILE
ENDFUNCTION score
FUNCTION AddDataToList(name$,score,post)
byteposition = (post-1)*29
REM *** Store name
FOR c = 1 TO 25

LISTING-47.4 value = ASC(MID$(name$,c))
(continued) WRITE MEMBLOCK BYTE 1,byteposition+c-1,value
NEXT c
Storing Top Score REM *** Store score ***
Details in a Memory WRITE MEMBLOCK DWORD 1,byteposition+25,score
Block ENDFUNCTION
FUNCTION DisplayContentsOfMemBlock()
FOR c = 1 TO 5
DisplayEntry(c)
NEXT c
ENDFUNCTION
FUNCTION DisplayEntry(post)
IF post < 1 OR post > 5
EXITFUNCTION
ENDIF
byteposition = (post-1)*29
name$=""
FOR c = byteposition TO byteposition + 24
name$ = name$ + ch$
NEXT c
score = MEMBLOCK DWORD(1,byteposition + 25)
PRINT name$," ",score
ENDFUNCTION
Activity 47.4

Use the following data:
John Smith 3712

Liz Heron 2989
Avril Tait 2508
Jack Brown 1992
James Ladd 1894
The WRITE MEMBLOCK (to file) Statement

If we really did need to hold the top scores, we'd also need to save that information
to a file so it would be available next time the game was played. To write the contents
of a memory block to file we need to use the WRITE MEMBLOCK statement. The
format for this statement is shown in FIG-47.16.
FIG-47.16
The WRITE WRITE MEMBLOCK fileno , memno
MEMBLOCK (to file)
Statement
In the diagram:
fileno is an integer value giving the ID of the file

to which the memory block's data is to be written.

block whose contents are to be written to the file.
The statement writes not only the contents of the memory block, but also its size.

Before we can output to a file, it must be opened for writing (files were covered in
Chapter 15):
OPEN TO WRITE 1,"toplist.dat"
The file will be created automatically in the current directory by the OPEN TO
WRITE statement. The whole memory block can be written to the file with the
single line:
WRITE MEMBLOCK 1,1
All that remains is to close the file:

CLOSE FILE 1
The lines have been placed in a function which also checks that the file does not
exist before it is opened:
FUNCTION SaveListToFile()
IF FILE EXIST("toplist.dat")
DELETE FILE "toplist.dat"
ENDIF
OPEN TO WRITE 1, "toplist.dat"
WRITE MEMBLOCK 1,1
CLOSE FILE 1
ENDFUNCTION
Activity 47.5
Add the function above to your last program and change the main section so
that the function is called before the program terminates.
A file produced in this way is not limited to storing data from a memory block;
additional information can be written to the file either before or after the data from
the memory block. For example, it might prove useful to write to the file details
such as the number of entries in the memory block, the size of an entry, the number
of fields in one entry and a description of the fields within an entry. For our top-five
list, that would involve saving the following information:
5 (entries in the block)

29 (the size of one entry)
2 (fields in one entry)
S25 (first field is a 25 character string)
D4 (second field is a double word which occupies 4 bytes)
The code required to write this information to the file is:
WRITE LONG 1,5

WRITE LONG 1,29
WRITE LONG 1,2
WRITE STRING 1, "S25"
WRITE STRING 1, "D4"
Activity 47.6
Modify the SaveListToFile() function so that the data described above is

written to the file before the contents of the memory block. Also, add
comments to the function.

The MAKE FILE FROM MEMBLOCK Statement
If the data to be held in the file is to be limited to only that coming from a single
memory block, then there is an even simpler way to write that data to a file; the
MAKE FILE FROM MEMBLOCK statement which has the format shown in
FIG-47.17.
FIG-47.17
The MAKE FILE FROM
MAKE FILE FROM MEMBLOCK fileno , memno
MEMBLOCK Statement
In the diagram:
fileno is an integer value giving the ID of the file which

is to store the memory block's data.

block whose data is to be copied.
We still need to open and close a file, but using this statement can be more
convenient if nothing but the contents of the memory block are to be written to the
file. Below is a second function for writing the top-five list to a file:
FUNCTION SaveListToFile2()
IF FILE EXIST("toplist2.mem")
DELETE FILE "toplist2.mem"
ENDIF
OPEN TO WRITE 1, "toplist2.mem"
MAKE FILE FROM MEMBLOCK 1,1
CLOSE FILE 1
ENDFUNCTION
Activity 47.7
Add the new function above to your last program and modify the main section
so that only this new function is used to save the memory block to a file.
The READ MEMBLOCK (from file) Statement

Of course, if we can write a memory block to a file, it makes sense that we should
be able to read that data from the file and place it back in a memory block. To do
this we need to use the READ MEMBLOCK statement which has the format shown
in FIG-47.18.
FIG-47.18
The READ MEMBLOCK READ MEMBLOCK fileno , memno
(from file) Statement
In the diagram:

containing the data to be read.

block which is to store the data read from the
file. The memory block will be created
automatically by this statement.

For the statement to operate correctly, we must ensure that the correct part of the
file containing the data is accessed. For example, in toplist.dat the memory block
is the sixth item within the file, so the other five must be read (or skipped) before
any attempt is made to execute the READ MEMBLOCK statement.
If we create flexible code, we can read back almost any memory block by
interpreting the information we find in the header.
We start by declaring global variables which will hold the header data:
GLOBAL records ` no. of records in memblock

GLOBAL recsize ` size of one record
GLOBAL fields ` fields in one record
GLOBAL DIM fieldformat$(20) ` format of each field
This allows up to 20 GLOBAL DIM fieldsize(20) ` size of each field
fields within a record.
The code below reads back all the information written to the file toplist.dat
FUNCTION ReadListFromFile()
REM *** Open file for reading ***
OPEN TO READ 1, "toplist.dat"
REM *** Read header data ***
READ LONG 1,records
READ LONG 1,recsize
READ LONG 1,fields
FOR c = 1 TO fields
READ STRING 1,field$
fieldformat$(c) = LEFT$(field$,1)
fieldsize(c) = VAL(RIGHT$(field$, LEN(field$)-1))
NEXT c
REM *** Read top 5 details ***
READ MEMBLOCK 1,1
REM *** Close the file ***
CLOSE FILE 1
ENDFUNCTION
and an individual record within the memory block can be displayed using the
following function:
FUNCTION DisplayEntry(post)
IF post < 1 OR post > records
EXITFUNCTION
ENDIF
byteposition = (post-1)*recsize
REM *** Calculate position of record within memblock ***
offset = (post-1)*recsize
REM *** Display each field in the record ***
FOR c = 1 TO fields
SELECT fieldformat$(c)
CASE "S"
text$=""
FOR k = byteposition TO byteposition +
Ä
fieldsize(c)-1
ch$ = CHR$(MEMBLOCK BYTE(1,offset))
text$ = text$ + ch$
INC offset
NEXT k
PRINT text$," ";
ENDCASE
CASE "D"
value = MEMBLOCK DWORD(1,offset)
offset = offset + 4
PRINT value," ";
ENDCASE
ENDSELECT

NEXT c
PRINT
ENDFUNCTION
Activity 47.8
Making use of the code above, create a complete program (act4708.dbpro) to

read all the data held in toplist.dat and display the top five list.
The MAKE MEMBLOCK FROM FILE Statement

If a memory block is written to a file using the MAKE FILE FROM MEMBLOCK
statement, then we can read that information back from the file, and recreate the
memory block using the MAKE MEMBLOCK FROM FILE statement which has
the format given in FIG-47.19.
FIG-47.19
The MAKE
MAKE MEMBLOCK FROM FILE fileno , memno
MEMBLOCK FROM
FILE Statement
In the diagram:

whose contents are to be loaded into a memory
block.

block which is to store the data retrieved from the
file.
For example, we could read back the contents of toplist2.mem using the code:
OPEN TO READ 1, "toplist2.mem"

MAKE MEMBLOCK FROM FILE 1,1
CLOSE FILE 1
Activity 47.9
Create a program (act4709.dbpro) that reads the contents of toplist2.mem to a

memory block and then displays the contents of that memory block.
Adding a New Top Score to our List

If a player achieves a new high score, then we'll want to add that score to the top
five list (see FIG-47.20).
We can achieve the first step in this process - moving existing entries - using the
COPY MEMBLOCK statement.
For example, to add Mary Gibson to our list, we could move entries 3 and 4 into
positions 4 and 5 using the line:
COPY MEMBLOCK 1,1,2*29,3*29,2*29

FIG-47.20 Adding a New Entry into the Top Five List
Top scores held in

a memory block
John Smith 3712 John Smith 3712 John Smith 3712
Liz Heron 2989 Liz Heron 2989 Liz Heron 2989

Mary Gibson 2963
Avril Tait 2508 Mary Gibson 2963 Avril Tait 2508 Mary Gibson 2963 Mary Gibson 2963
This entry is Avril Tait 2508 Avril Tait 2508
Jack Brown 1992 Jack Brown 1992
lost from the list Jack Brown 1992
James Ladd 1894 Jack Brown 1992
New high score James Ladd 1894
If a new high score is achieved... ...we need to move existing entries... ...to make space for the new entry.
Activity 47.10
Write a program which implements the following logic:
Read the contents of toplist2.mem into a memory block

Shift the contents of entries 3 and 4 into positions 4 and 5 in the list
Read details of a new name and score (Mary Gibson, 2963)
Place the new entry at position 3 in the top five list
Display the contents of the top five list.
Summary
l A pointer is a variable designed to contain an address.
l Use MAKE MEMBLOCK to dynamically allocate memory space.
l Use GET MEMBLOCK PTR to return the start address of a memory block.
l A pointer can be used to return values from a function.
l Use WRITE MEMBLOCK to write data to an area within a memory block.
l Use MEMBLOCK to retrieve information stored within a memory block.
l Use GET MEMBLOCK SIZE to discover the number of bytes in a memory

block.
l Use DELETE MEMBLOCK to deallocate space assigned to a memory block.
l Use MEMBLOCK EXIST to check if a specified memory block exists.
l Use COPY MEMBLOCK to copy a part of one memory block to another

memory block.
l Use WRITE MEMBLOCK to write the contents of a memory block to a file.
l The file specified in a WRITE MEMBLOCK statement may contain data other
than that from the memory block.
l Use MAKE FILE FROM MEMBLOCK to create a file containing only the
contents of a memory block.

l Use READ MEMBLOCK to read data from a file into a memory block.
l Use MAKE MEMBLOCK FROM FILE to read data from a file containing only
data copied from a memory block into a new memory block.

Media Contents and Memory Blocks
Introduction
The most interesting use of memory blocks is undoubtedly employing them to store
images, sounds, and 3D objects. Once the data representing such a component is
stored in a memory block, we can manipulate that information in ways that are not
possible using other techniques.
Bitmaps and Memory Blocks

Back in Chapter 13 we looked at bitmap objects. Bitmaps are storage areas for
images. There are 32 different areas (numbered 0 to 31). Area 0 represents the
screen - any image placed in area 0 will immediately be shown on the screen. To
load an image into a bitmap area we use a statement such as
LOAD BITMAP "bbrush.bmp",0
which loads an image file named bbrush.bmp onto the screen.
The MAKE MEMBLOCK FROM BITMAP Statement

To transfer the contents of a bitmap area to a memory block, we need to use the
MAKE MEMBLOCK FROM BITMAP statement. The memory block specified
should not already exist, since it will be created by the statement. The MAKE
MEMBLOCK FROM BITMAP statement has the format shown in FIG-47.21.
FIG-47.21 The MAKE MEMBLOCK FROM BITMAP Statement
MAKE MEMBLOCK FROM BITMAP memno , bmapno
In the diagram:
memno is an integer value giving the ID to be assigned

to the memory block created by this statement.
bmapno is an integer value giving the bitmap area that is to

be copied to the memory block.
To copy the contents of the screen to memory block 1 we would use the statement:
MAKE MEMBLOCK FROM BITMAP 1,0
If we are to manipulate the contents of the memory block, we need to know the
structure of the data held. A memory block containing a bitmap has the format
shown in FIG-47.22.
As you can see, the first 12 bytes of the memory block contain details of the bitmap's
width, height and bits per pixel. For most bitmap areas this will correspond to the
details of the actual image that has been loaded into that area, but for bitmap area
0, which is the screen, the width, height and bits per pixel are determined by the
values used in the SET DISPLAY MODE statement.

FIG-47.22 Bitmap
Memory Block Structure
The Format of a Bitmap
0
3
4
Preliminary details such 7

8
as those described for
bitmaps is often 11
generally referred to as 12
header information. Pixel Data
The program in LISTING-47.5 demonstrates the transfer of the screen's image to a

memory block and shows that the information read from the memory block matches
that used in the SET DISPLAY MODE statement.
Copying a Bitmap to a
Memory Block
REM *** Copy screen data to a memory block ***
REM *** Get details from memory block ***

width = MEMBLOCK DWORD(1,0)
height = MEMBLOCK DWORD(1,4)
bitsperpixel = MEMBLOCK DWORD(1,8)
REM *** Display details on the screen ***

PRINT "Screen is ",width," by ",height
PRINT "There are ",bitsperpixel," bits per pixel"

WAIT KEY
END
Activity 47.11
Set the screen to 800 by 600 with 16 bits per pixel and check that the details
returned match this figure.
Modify the program to display the number of bytes required by a single line of
pixels on the screen.
Activity 47.12
Write a program (act4712.dbpro) which loads the image bbrush.bmp into

There are easier way of bitmap area 1. By transferring the bitmap to a memory block, display the
discovering these details! image's width, height and bits per pixel.
(see Chapter 13)
We can make use of the data in the memory block to actually change what appears
on the screen by first changing values within the pixel data area of the memory

block and then writing the memory block back to bitmap area 0.
For example, if we assume a screen resolution of 800 by 600 and 16 bits per pixel,
we could make the data representing the top-left pixel on the screen white by
beginning with the line:
WRITE MEMBLOCK WORD 1,12,RGB(255,255,255)
The MAKE BITMAP FROM MEMBLOCK Statement

If the changes we have made to the memory block are to be reflected on the screen,
we need to copy the memory block back to the bitmap area - area 0 in this case. This
can be done using the MAKE BITMAP FROM MEMBLOCK statement whose
format is shown in FIG-47.23.
FIG-47.23
The MAKE BITMAP MAKE BITMAP FROM MEMBLOCK bmapno , memno
FROM MEMBLOCK
Statement
In the diagram:
bmapno is an integer value giving the bitmap area into

which the memory block is to be copied.

block to be copied.
So, if we've changed the contents of memory block 1, we can write the updated
version back to the screen using the line:
MAKE BITMAP FROM MEMBLOCK 0,1
The program in LISTING-47.6 starts by loading the black screen image into a
memory block and changes the first pixel to white before rewriting the memory
block's data to the screen.
Manipulating a Bitmap's REM *** Copy screen to memory block ***
Data MAKE MEMBLOCK FROM BITMAP 1,0
PRINT "Screen captured"
WAIT KEY
REM *** Change first pixel in block to white ***
WRITE MEMBLOCK DWORD 1,12,RGB(255,255,255)
PRINT "Memblock changed"
WAIT KEY
REM *** Rewrite block to screen ***
PRINT "Screen updated"
WAIT KEY
END
Activity 47.13
Make sure that your
monitor is adjusted to Type in and test the program in LISTING-47.6 (mem06.dbpro).
show all pixels that make
up the screen output. Modify the program so that it is the 10th pixel on the first row that is made
white.

Mapping a Screen Position to a Memory Block Location
If we are to have absolute control over the manipulation of an image, we need to be
able to calculate the position in the memory block of any pixel. Let's start by working
out the position of any pixel on the first row.
Pixels are addressed using a column and row system, starting with column zero,
row zero. So, pixel (0,0) has its data stored at position 12 in the memory block. If
we are using a 32 bits / pixel setup, then pixel (1,0) starts at position 16, pixel (2,0)
at position 20, etc. More generally, we can write that:
pixel(x,0)) starts at position x

*4 + 12
If the image is 1280 pixels wide (as the screen is in the earlier examples) then the
last pixel on the first line - pixel (1279,0) will start at position 1279*4 + 12 = 5128
and therefore, the first pixel in the next row (0,1) must start at position 5132.
Activity 47.14
Modify your last program to make the last pixel of the first row and the first
pixel of the second row white.
If each pixel requires 4 bytes of storage, and there are 1280 pixels in a row, then 1
row requires 5120 pixels. Now we can extend our formula to calculate the position
of any pixel's data within the memory block to be:
pixel(col,row) = col*4 + row*5120 + 12
Activity 47.15
Modify your last program to change the bottom right pixel on the screen -
pixel (1279,1023) - to white
Our formula is still too specific - it assumes a resolution of 1280 by 1024 and 32
bits per pixel. A more general formula would be:
pixel(col,row) = col * bytesperpixel + row *pixelsperrow*bytesperpixel + 12
The program in LISTING-47.7 loads an image, bbrush.jpg, onto the screen and then
creates a negative version of the image by inverting the value of every pixel in the
image.

Calculating a Data Item's REM *** Load image ***
Position in a Memory LOAD BITMAP "bbrush.jpg",0
Block REM *** Copy screen to memory block ***
WAIT KEY
REM *** Modify every pixels' data ***

FOR row = 0 TO 1023
FOR col = 0 TO 1279
post = col*4 + row*5120 + 12
currentvalue = MEMBLOCK DWORD(1,post)

LISTING-47.7 WRITE MEMBLOCK DWORD 1, post, NOT currentvalue
(continued) NEXT col
NEXT row
Calculating a Data Item's PRINT "Memblock changed"
Position in a Memory WAIT KEY
Block
REM *** Copy new data to screen ***
WAIT KEY
END
Activity 47.16
Activity 47.17
Write a program (act4717.dbpro) which performs the following logic:
Load the image cactus.jpg into bitmap area 1

Copies the image to a memory block
Exclusive ORs (XOR) the value 0X212121 to every pixel's data
Copies the memory block's data to the screen.
Mapping the Mouse Position to a Memory Block Location

We can read the mouse pointer's position on the screen using MOUSEX and
There are other ways to
MOUSEY (see Chapter 27). By using the formula we developed earlier, we could
achieve the same effect. use the mouse pointer to modify parts of the screen. The program in LISTING-47.8
changes the pixel at the mouse pointer's position to yellow using memory blocks.

How the Mouse Position
Corresponds to the REM *** Copy screen to memory block ***
Position of a Data Item in MAKE MEMBLOCK FROM BITMAP 1,0
a Memory Block
DO
col = MOUSEX()
row = MOUSEY()
REM *** Change matching area to yellow ***
WRITE MEMBLOCK DWORD 1,row*5120 + col*4 + 12,RGB(255,255,0)
REM *** Copy memory block back to screen ***
LOOP
END
Activity 47.18
Modify the program so that a mouse button must be pressed before a pixel
turns yellow.

Images and Memory Blocks
What we can do with bitmaps, we can also do to images; copying them to a memory
block, changing the information held there, and then updating the image object with
the modified data.
The MAKE MEMBLOCK FROM IMAGE Statement

To copy an image to a memory block we use the MAKE MEMBLOCK FROM
IMAGE statement which has the format shown in FIG-47.24.
FIG-47.24
The MAKE MAKE MEMBLOCK FROM IMAGE memno , imgno
MEMBLOCK FROM
IMAGE Statement
In the diagram:

imgno is an integer value giving the ID of the image to be

copied to the memory block.
The memory block created by this statement has exactly the same format as that
created with a bitmap, with the first 12 bytes giving the width, height and bits per
pixel, before the image's own data is reached.
We could load an image into a memory block with statements such as:
LOAD IMAGE "bbrush.jpg",1

MAKE MEMBLOCK FROM IMAGE 1,1
The MAKE IMAGE FROM MEMBLOCK Statement

To reverse the process - copying a memory block to an image - we use the MAKE
IMAGE FROM MEMBLOCK statement which has the format shown in
FIG-47.25.
FIG-47.25
The MAKE IMAGE MAKE IMAGE FROM MEMBLOCK imgno , memno
FROM MEMBLOCK
Statement
In the diagram:

object to be created by this statement.

block to be copied.
The image ID specified in this statement should not be in use when the statement
is executed.
In the program in LISTING-47.9 the bottlebrush image is used to texture a rotating

cube. The image is then negated and reapplied to the cube.

Manipulating an Image
REM *** Create textured object ***
LOAD IMAGE "bbrush.jpg",1
TEXTURE OBJECT 1,1
REM *** Copy image to memory block ***

MAKE MEMBLOCK FROM IMAGE 1,1
REM *** Negate image ***
bytes = MEMBLOCK DWORD(1,8)/8
FOR row = 0 TO height -1
FOR col = 0 TO width-1
post = col*bytes + row*bytes*width + 12
currentvalue = MEMBLOCK DWORD(1,post)
WRITE MEMBLOCK DWORD 1, post, NOT currentvalue
NEXT col
NEXT row

DO
REM *** Rotate object
REM *** IF a key pressed, texture cube with new image ***
IF INKEY$() <> ""
DELETE IMAGE 1
MAKE IMAGE FROM MEMBLOCK 1,1
TEXTURE OBJECT 1,1
ENDIF
WAIT 5
LOOP
END
Activity 47.19
Sounds and Memory Blocks

The next media object that can be stored and manipulated in a memory block is a
sound object.
The MAKE MEMBLOCK FROM SOUND Statement

To transfer a loaded sound to a memory block, we need to use the MAKE
MEMBLOCK FROM SOUND statement which has the format shown in
FIG-47.26.
FIG-47.26
The MAKE MEMBLOCK MAKE MEMBLOCK FROM SOUND memno , sndno
FROM SOUND Statement
In the diagram:


sndno is an integer value giving the ID of the sound to be
The header information included at the start of a sound object is much more complex
than that of a bitmap or image. Details such as the sound format being used, number
of channels and sample rate are included. The actual structure or this header data is
an agreed international standard and is shown in FIG-47.27.
Sound
FIG-47.27 Memory Block Structure
The Data Structure Used

for a Sound 0
format code DWORD
3
4
channels DWORD
7
8
samples DWORD
11 per second
12
average bytes DWORD
15 per second
16
block alignment DWORD
19
20
bits per DWORD
23 sample
24
extras size DWORD
27
28
extra data
27+ extras size
28+ extras size
Sound Data
Each field in the diagram is explained below:
format code Many sound file formats use this same core
data structure, but with different data compression
algorithms. The actual type of compression used
is identified here using a coded value registered
with Microsoft.
channels Identifies the number of sound channels used.

For monaural the value here would be 1; for
stereo, 2.
samples per second Identifies how often the original sound has been
sampled per second during the digitising session.
average bytes per second Specifies the average data transfer rate in bytes.
Normally, this will be
samples per second x block alignment
block alignment This value gives the size of each data block
sample. It should be equal to
channels x bits per sample

bits per sample Specifies the number of bits used when sampling
a single channel. Normally, 8 or 16.
extras size Specifies the size in bytes of any information

that has been added to this header section before
the sound data is reached.
extra data This is the area where any extra header

information has been added. This area must be
extra size bytes long.
The MAKE SOUND FROM MEMBLOCK Statement

Once sound data has been manipulated within a memory block it will need to be
transferred back to a sound object, and this is achieved using the MAKE SOUND
FROM MEMBLOCK statement which has the format shown in FIG-47.28.
FIG-47.28
The MAKE SOUND MAKE SOUND FROM MEMBLOCK sndno , memno
FROM MEMBLOCK
Statement
In the diagram:
sndno is an integer value giving the ID of the sound


block to be copied.
The program in LISTING-47.10 loads a sound file and displays the information
held in its header area.
LISTING-47.10 REM *** Load and play sound ***

LOAD SOUND "welcome.wav",1
Memory Blocks and PLAY SOUND 1
Sound
REM *** Copy sound to memory block ***
MAKE MEMBLOCK FROM SOUND 1,1
REM *** Get data ***

channels = MEMBLOCK DWORD(1,4)
rate# = MEMBLOCK DWORD(1,8)/1000.0
avBperS = MEMBLOCK DWORD(1,12)
block = MEMBLOCK DWORD(1,16)
bits = MEMBLOCK DWORD(1,20)
extra = MEMBLOCK DWORD(1,24)
REM *** Display the information ***

PRINT "Channels used : ", channels
PRINT "Sample rate : ", rate#," kilohertz"
PRINT "Bytes per sec : ", avBperS
PRINT "Block size : ", block," bytes"
PRINT "Bits per sample: ", bits
PRINT "Extra data : ", extra," bytes"

WAIT KEY
END

Activity 47.20
Modify the program so that within the memory block the sample rate is set to
44100 and the bytes per second to 88200. Delete sound 1 and recreate a sound
1 object from memory block 1. Play the sound. How have the changes altered
the sound playback?
3D Objects and Memory Blocks

It is even possible to modify a mesh by transferring it to a memory block. However,
the data structure used to represent mesh data is a complex one and we need to take
care when manipulating its contents.
Mesh data is held in a format known as Flexible Vertex Flags or FVF which contains
details of a vertex's coordinates, its normal, its diffuse value and UV texture offset.
The MAKE MEMBLOCK FROM MESH Statement

To convert a mesh to a memory block we need to use the MAKE MEMBLOCK
FROM MESH statement which has the format shown in FIG-47.29.
FIG-47.29
The MAKE MAKE MEMBLOCK FROM MESH memno , meshno
MEMBLOCK FROM
MESH Statement
In the diagram:

meshno is an integer value giving the ID of the mesh to be

The structure of the header information for a mesh is shown in FIG-47.30.

FIG-47.30 Mesh
Memory Block Structure
The Basic Header
Structure for a Mesh 0
FVF Format
3
4
FVF Size
7
8 number of
vertices
11
12
Vertex Data
The fields in the header information are:
FVF Format There are several possible FVF formats. The

integer value given here identifies the format used.

FVF Size This field gives the size in bytes of each entry in
the data area.
number of vertices Specifies the number of vertices in the mesh. This

corresponds to the number of entries in the Vertex
Data area of the memory block.
The program in LISTING-47.11 creates a mesh from a cube, stores the data in a
memory block, and displays the block's header details.
LISTING-47.11 #CONSTANT cubeobj 1

#CONSTANT meshobj 1
Placing a Cube's Data in a #CONSTANT memobj 1
Memory Block REM *** Manual screen updating ***
SYNC ON
MAKE OBJECT CUBE cubeobj,10
SYNC
REM *** Convert cube to mesh ***
MAKE MESH FROM OBJECT meshobj,cubeobj
REM *** Copy mesh to memory block ***
MAKE MEMBLOCK FROM MESH memobj,meshobj
REM *** Get and display details ***
FVFformat = MEMBLOCK DWORD (1,0)
entrysize = MEMBLOCK DWORD (1,4)
noofvertices = MEMBLOCK DWORD (1,8)
PRINT "FVF format : ", FVFformat
PRINT "Entry size : ", entrysize
PRINT "No. of vertices : ", noofvertices
SYNC
WAIT KEY
END
Activity 47.21
Most DarkBasic Pro models seem to use format FVF 274 which holds, for each
vertex, the information shown in FIG-47.31.
Vertex Entry Structure
FIG-47.31
Length
The Detailed Header (bytes)
Structure for a Mesh
4 x-ord FLOAT
4 y-ord FLOAT
4 z-ord FLOAT
4 x-normal FLOAT
4 y-normal FLOAT
4 z-normal FLOAT
4 U offset FLOAT
4 V offset FLOAT

The fields in the vertex block are:
x-ord, y-ord, z-ord The coordinates of the vertex.
x-normal, y-normal, z-normal

The coordinates of the end point of the vertex's
normal. These measurements are relative to the
position of the vertex. A normal will usually be
exactly 1 unit in length.
U offset, V offset The UV offsets to be applied at that point when

mapping a texture to the object.
The program in LISTING-47.12 displays the details of each vertex of the cube we
created previously.
#CONSTANT meshobj 1
Accessing Vertex
#CONSTANT memobj 1
Details from a Memory
Block
REM *** Manual screen updating ***
SYNC ON
REM *** Create mesh from cube ***

SYNC
REM *** Create memory block from mesh ***

REM *** Get header details ***

FVFformat = MEMBLOCK DWORD(1,0)
entrysize = MEMBLOCK DWORD(1,4)
noofvertices = MEMBLOCK DWORD(1,8)
REM *** Get and display details of each vertex ***

FOR c = 0 TO noofvertices-1
post = 12 + c*entrysize
x# = MEMBLOCK FLOAT(1,post)
y# = MEMBLOCK FLOAT (1,post+4)
z# = MEMBLOCK FLOAT(1,post+8)
nx#= MEMBLOCK FLOAT(1,post+12)
ny# = MEMBLOCK FLOAT (1,post+16)
nz# = MEMBLOCK FLOAT(1,post+20)
U# = MEMBLOCK FLOAT(1,post+24)
V# = MEMBLOCK FLOAT (1,post+28)
PRINT "Coords (",STR$(x#,0),",",STR$(y#,0),",",STR$(z#,0),")"
PRINT "Normal (",STR$(nx#,0),",",STR$(ny#,0),","
Ä,STR$(nz#,0),")"
PRINT "UV offsets U: ",U#," V: ",V#
IF c mod 10 = 0
SYNC
WAIT KEY
ENDIF
NEXT c
SYNC
REM *** end program ***

WAIT KEY
END

Activity 47.22
Type in and test the program given in LISTING-47.12 (mem12.dbpro).
Another common format for 3D objects is FVF format 338. This structure is similar
to 274 but includes a DWORD diffuse value in each vertex data block between the
normal's z-ordinate and the U offset value.
Each group of three vertices in the object are used to create a triangular polygon;
the collection of polygons form the object.
The MAKE MESH FROM MEMBLOCK Statement

We can modify data in a memory block and then create a mesh from this new data.
To create the new mesh from the memory block we use the MAKE MESH FROM
MEMBLOCK statement which has the format shown in FIG-47.32.
FIG-47.32
The MAKE MESH
MAKE MESH FROM MEMBLOCK meshno , memno
FROM MEMBLOCK
Statement
In the diagram:


block to be copied.
The program in LISTING-47.13 stores a cube's mesh in a memory block, modifies

the coordinates of some of the vertices and then recreates a mesh and object from
the modified data.

#CONSTANT rootobj 2
Modifying a Mesh using #CONSTANT meshobj 1
a Memory Block #CONSTANT memobj 1
SetUpScreen()
REM *** Create memory block from cube ***

DELETE OBJECT cubeobj
DELETE MESH meshobj
REM *** Get details of object ***

REM *** Get coords of each vertex ***

post = 12 + c*entrysize
x# = MEMBLOCK FLOAT(1,post)
y# = MEMBLOCK FLOAT (1,post+4)
z# = MEMBLOCK FLOAT(1,post+8)
REM *** Modify any vertices above y = 0 ***
IF y# > 0

LISTING-47.13 WRITE MEMBLOCK FLOAT 1,post,0
(continued) WRITE MEMBLOCK FLOAT 1,post+4,0
WRITE MEMBLOCK FLOAT 1,post+8,0
Modifying a Mesh using ENDIF
a Memory Block NEXT c
REM *** Make mesh from memory block ***

MAKE MESH FROM MEMBLOCK meshobj, memobj
REM *** Add mesh as limb to make it visible ***

MAKE OBJECT SPHERE rootobj,0
ADD LIMB rootobj,1,meshobj
REM *** Rotate object created ***

DO
PITCH OBJECT UP 2,1
WAIT 10
LOOP

END
AUTOCAM OFF
ENDFUNCTION
Activity 47.23
Modify the program so that, rather than change the coordinates of specific
vertices, the program changes all vertex normal coordinates to (1,1,1).
How does this effect the cube?
The CHANGE MESH FROM MEMBLOCK Statement

Rather than delete the original mesh and then recreate it using MAKE MESH
FROM MEMBLOCK, we can simply update the existing mesh using the CHANGE
MESH FROM MEMBLOCK statement which has the format shown in FIG-47.33.
FIG-47.33
The CHANGE MESH CHANGE MESH FROM MEMBLOCK meshno , memno
FROM MEMBLOCK
Statement
In the diagram:

object to be updated by this statement.

block to be copied.
Activity 47.24
Modify your last program so that the cube's mesh is updated rather than
deleted and recreated.

Summary
l Bitmaps, images, sounds and meshes can be converted to or from memory
blocks.
l Use MAKE MEMBLOCK FROM BITMAP to convert a bitmap's data to a

memory block.
l A bitmap's data starts with header information detailing width, height and colour
depth of the bitmap.
l Use MAKE BITMAP FROM MEMBLOCK to convert a memory block

(containing the appropriate data) to a bitmap.
l Use MAKE MEMBLOCK FROM IMAGE to convert an image's data to a

memory block.
l Use MAKE IMAGE FROM MEMBLOCK to convert memory block data to an

image.
l Use MAKE MEMBLOCK FROM SOUND to convert a sound object to a

memory block.
l A sound object's data starts with header information such as compression format,
channels, and sample rate data.
l Use MAKE SOUND FROM MEMBLOCK to convert memory block data to a

sound object.
l Use MAKE MEMORY BLOCK FROM MESH to convert a mesh's data to a

memory block.
l A mesh's data starts with header information detailing the FVF format and the
number of vertices in the mesh.
l Following the mesh header are details about each vertex within the mesh.
l Vertex data varies with the FVF format being used, but includes the vertex's
spatial coordinates, normal coordinates and UV offset information.
l Use MAKE MESH FROM MEMBLOCK to convert memory block data to a

mesh.
l Use CHANGE MESH FROM MEMBLOCK to update an existing mesh.

Solutions WAIT KEY
Activity 47.1 END
Activity 47.8
Activity 47.2 GLOBAL size
GLOBAL records
No solution required. GLOBAL recsize
GLOBAL fields
GLOBAL DIM fieldformat$(20)
Activity 47.3 GLOBAL DIM fieldsize(20)
No solution required. ReadListFromFile()

WAIT KEY
Activity 47.4 END
No solution required. FUNCTION ReadListFromFile()

REM *** Open file for reading ***
OPEN TO READ 1, "toplist.dat"
Activity 47.5 REM *** Read header data ***
READ LONG 1,records
The main section should be changed to: READ LONG 1,recsize
READ LONG 1,fields
MAKE MEMBLOCK 1,29*5 FOR c = 1 TO fields
FOR c = 1 TO 5 READ STRING 1,field$
name$ = GetName$() fieldformat$(c) = LEFT$(field$,1)
score = GetScore() fieldsize(c) = VAL(RIGHT$(field$,
AddDataToList(name$,score,c) Ä
LEN(field$)-1))
NEXT c NEXT c
DisplayContentsOfMemBlock() REM *** Read top 5 details ***
SaveListToFile() READ MEMBLOCK 1,1
WAIT KEY REM *** Close the file ***
END CLOSE FILE 1
ENDFUNCTION
Activity 47.6 FUNCTION DisplayContentsOfMemBlock()

FOR c = 1 TO records
The code for function SaveListToFile() should be changed DisplayEntry(c)
to: NEXT c
ENDFUNCTION
FUNCTION SaveListToFile()
REM *** IF file exists, destroy it *** FUNCTION DisplayEntry(post)
IF FILE EXIST("TopList.dat") IF post < 1 OR post > records
DELETE FILE "Toplist.dat" EXITFUNCTION
ENDIF ENDIF
REM *** Create file for writing *** byteposition = (post-1)*recsize
OPEN TO WRITE 1, "Toplist.dat" REM *** Calculate position of entry ***
REM *** Write header info to file *** offset = (post-1)*recsize
WRITE LONG 1,5 FOR c = 1 TO fields
WRITE LONG 1,29 SELECT fieldformat$(c)
WRITE LONG 1,2 CASE "S"
WRITE STRING 1, "S25" text$=""
FOR k = byteposition TO
WRITE STRING 1, "D4"
REM *** Write memblock to file *** Ä
byteposition + fieldsize(c)-1
ch$ = CHR$(MEMBLOCK
Ä
WRITE MEMBLOCK 1,1
REM *** Close file *** BYTE(1,offset))
CLOSE FILE 1 text$ = text$ + ch$
ENDFUNCTION INC offset
NEXT c
PRINT text$," ";
The updated version of the main section is: CASE "D"

value = MEMBLOCK DWORD(1,offset)
offset = offset + 4
MAKE MEMBLOCK 1,29*5
PRINT value," ";
FOR c = 1 TO 5
ENDCASE
name$ = GetName$()
ENDSELECT
score = GetScore()
NEXT c
AddDataToList(name$,score,c)
PRINT
NEXT c
ENDFUNCTION
SaveListToFile2()

Activity 47.9 Ä
byteposition+25,score
ENDFUNCTION
OPEN TO READ 1, "toplist2.mem"
MAKE MEMBLOCK FROM FILE 1,1 FUNCTION DisplayContentsOfMemBlock()
CLOSE FILE 1 FOR c = 1 TO 5
DisplayContentsOfMemBlock() DisplayEntry(c)
WAIT KEY NEXT c
END ENDFUNCTION
FUNCTION DisplayContentsOfMemBlock() FUNCTION DisplayEntry(post)

FOR c = 1 TO 5 IF post < 1 OR post > 5
DisplayEntry(c) EXITFUNCTION
NEXT c ENDIF
ENDFUNCTION byteposition = (post-1)*29
FUNCTION DisplayEntry(post) name$=""
IF post < 1 OR post > 5 FOR c = byteposition TO byteposition+24
EXITFUNCTION ch$ = CHR$(MEMBLOCK BYTE(1,c))
ENDIF name$ = name$ + ch$
byteposition = (post-1)*29 NEXT c
REM *** Read name back *** score = MEMBLOCK DWORD(1,byteposition+25)
name$="" PRINT name$," ",score
FOR c = byteposition TO byteposition + 24 ENDFUNCTION
name$ = name$ + ch$
NEXT c Activity 47.11
score = MEMBLOCK DWORD(1,byteposition +
Ä
25) To use the display mode requested, the line
PRINT name$," ",score
ENDFUNCTION SET DISPLAY MODE 1280,1024,32
The code is much simpler than that in Activity 47.8, but is should be changed to:
designed to handle a specific memblock structure only.
Activity 47.10 The modified version of the program is:

OPEN TO READ 1, "TopList2.mem"
MAKE MEMBLOCK FROM FILE 1,1
CLOSE FILE 1
REM *** Copy screen data to memory block ***
REM *** Move entries 3 and 4 ***
COPY MEMBLOCK 1,1,2*29,3*29, 2*29
name$ = GetName$()
score = GetScore()
AddDataToList(name$,score,3)
REM *** Display details on the screen ***
WAIT KEY
PRINT "Screen is ",width," by ",height
END
PRINT "There are ",bitsperpixel,
FUNCTION GetName$() Ä" bits per pixel"
PRINT "A line of pixels requires ",width *
INPUT "Enter name",name$
WHILE LEN(name$) > 25 Äbitsperpixel / 8, " bytes"
INPUT "Invalid name. Please REM *** End program ***
Ä
re-enter",name$ WAIT KEY
END
ENDWHILE
REM *** Pad name to 25 characters ***
name$=name$+SPACE$(25 - LEN(name$))
ENDFUNCTION name$ Activity 47.12
FUNCTION GetScore()
INPUT "Enter score ",score
REM *** Load bitmap ***
WHILE score < 0
LOAD BITMAP "bbrush.jpg",1
INPUT "Invalid score. Please
Ä
re-enter",score
REM *** Copy screen data to memory block ***
ENDWHILE REM *** restore screen as default bitmap ***
ENDFUNCTION score SET CURRENT BITMAP 0
FUNCTION AddDataToList(name$,score,post) width = MEMBLOCK DWORD(1,0)
byteposition = (post-1)*29 height = MEMBLOCK DWORD(1,4)
REM *** Store name bitsperpixel = MEMBLOCK DWORD(1,8)
FOR c = 1 TO 25 REM *** Display details on the jpeg ***
value = ASC(MID$(name$,c)) PRINT "Picture is ",width," by ",height
WRITE MEMBLOCK BYTE 1,
Ä
byteposition+c-1,value
PRINT "There are ",bitsperpixel,
Ä" bits per pixel"
NEXT c REM *** End program ***
REM *** Store score *** WAIT KEY
WRITE MEMBLOCK DWORD 1, END

Activity 47.13 REM *** Get bitmap details ***
Only the offset value must be changed (and the comment):
REM *** Modify every pixel's data ***
FOR row = 0 TO height-1
REM *** Copy screen to memory block ***
FOR col = 0 TO width-1
post = col*4 + row*4*width + 12
value = MEMBLOCK DWORD(1,post)
WAIT KEY
REM *** Change pixel 10 to white ***
ÄXOR 0X212121
WRITE MEMBLOCK DWORD 1,post,value
NEXT col
NEXT row
WAIT KEY
WAIT KEY
REM *** Copy new data to screen ***
WAIT KEY
END
WAIT KEY
END
Activity 47.14
REM *** Set screen resolution *** Activity 47.18
WAIT KEY
REM *** (1279,0) and (0,1) to white ***
DO
col = MOUSEX()
row = MOUSEY()
WAIT KEY
REM *** On mousedown, show yellow ***
IF MOUSECLICK() <> 0
WRITE MEMBLOCK DWORD 1,row*5120 +
Äcol*4 + 12,RGB(255,255,0)
REM *** Copy memblock to screen ***
WAIT KEY
END
ENDIF
LOOP
Activity 47.15 END
REM *** Copy screen to memory block *** Activity 47.19
PRINT "Screen captured" No solution required.
WAIT KEY
REM *** Change (1279,1023) to white ***
WRITE MEMBLOCK DWORD 1,1279*4 + 1023*5120, Activity 47.20
ÄRGB(255,255,255)
REM *** Load and play sound ***
WAIT KEY
PLAY SOUND 1
REM *** Get data ***
MAKE MEMBLOCK FROM SOUND 1,1
channels = MEMBLOCK DWORD(1,4)
rate# = MEMBLOCK DWORD(1,8)/1000.0
WAIT KEY
avBperS = MEMBLOCK DWORD(1,12)
END
block = MEMBLOCK DWORD(1,16)
bits = MEMBLOCK DWORD(1,20)
extra = MEMBLOCK DWORD(1,24)
Activity 47.16 REM *** Display the information ***
PRINT "Channels used : ", channels
No solution required. PRINT "Sample rate : ", rate#,"
kilohertz"
PRINT "Bytes per sec : ", avBperS
Activity 47.17 PRINT "Block size : ", block," bytes"
REM *** Set screen resolution *** PRINT "Bits per sample: ", bits
SET DISPLAY MODE 1280,1024,32 PRINT "Extra data : ", extra," bytes"
REM *** Load image *** REM *** Modify sound ***
LOAD BITMAP "cactus.jpg",1
REM *** Return output to screen *** WRITE MEMBLOCK DWORD 1,8,44100
SET CURRENT BITMAP 0 WRITE MEMBLOCK DWORD 1, 12,88200
REM *** Copy bitmap to memory block *** DELETE SOUND 1

should be removed from the program and the line
REM *** Reload modified sound ***
MAKE SOUND FROM MEMBLOCK 1,1 MAKE MESH FROM MEMBLOCK meshobj, memobj
WAIT KEY
PLAY SOUND 1
changed to
WAIT KEY CHANGE MESH FROM MEMBLOCK meshobj, memobj
END
Playback sounds speeded up.
Activity 47.21
Activity 47.22
Activity 47.23
#CONSTANT cubeobj 1
#CONSTANT rootobj 2
#CONSTANT meshobj 1
#CONSTANT memobj 1
SetUpScreen()
REM *** Create memory block from cube ***
DELETE OBJECT cubeobj
DELETE MESH meshobj
REM *** Get details of object ***
REM *** Get coords of each vertex ***
REM *** Calc start of normal data ***
post = 12 + c*entrysize + 12
REM *** Modify all normals to (1,1,1) ***
WRITE MEMBLOCK FLOAT 1,post,1
NEXT c
REM *** Make mesh from memory block ***
MAKE MESH FROM MEMBLOCK meshobj, memobj
REM *** Add mesh as limb to make it visible
***
MAKE OBJECT SPHERE rootobj,0
ADD LIMB rootobj,1,meshobj
REM *** Rotate object created ***
DO
PITCH OBJECT UP 2,1
WAIT 10
LOOP
END
AUTOCAM OFF
ENDFUNCTION
The cube changes shade as it rotates.
Activity 47.24
The line
DELETE MESH meshobj

48
Adding Physical Laws to a 3D Object
Applying Friction
Applying Gravity
Dynamic and Static ODE Volumes
Linear and Angular Velocities
Surface and Contact Properties
DarkBASIC Pro: ODE 1317

Using ODE
Introduction
In virtual space objects stay still until they are moved. Things don't fall, bounce,
skid or get damaged. There's no gravity and no friction. Objects don't have mass; a
large object won't push a smaller one out of the way as they collide. In fact, unless
we implement collision checks, objects can pass straight through each other without
any interaction.
Although this lack of real-world laws can give us great freedom in what happens in
our 3D world, it also makes simulations of real life difficult.
So to help things along, DarkBASIC Pro has some basic commands built-in which
give us a hint at what is possible when the laws of physics are applied to our objects.
The ODE statements in DarkBASIC Pro are undocumented and although they are
still present in DarkBASIC Pro version 1.062, they need to be loaded separately.
The details given in this chapter are based on the results obtained from various
DarkBASIC Pro (v1.059) test programs and from other ODE documentation. If
you're impressed by these features, get hold of The Game Creator's DarkPhysics
add-on for really professional results.
Basic ODE Statements

Every visual 3D object that we want to react in a realistic fashion must be linked to
its own ODE object (see FIG-48.1).
FIG-48.1
Invisible ODE objects
are linked to Normal
3D Objects
Real objects are

linked to ODE objects
While our own 3D objects are visible, ODE objects are always invisible. But when
an ODE object reacts to physical laws, it causes the visible object to which it is
attached to react in the same way.
The ODE CREATE DYNAMIC BOX Statement

The physics box to be linked to a normal 3D object is created using the ODE
CREATE DYNAMIC BOX statement which has the format shown in FIG-48.2.
FIG-48.2
The ODE CREATE DYNAMIC BOX Statement
ODE CREATE DYNAMIC BOX objno

[ , width , height , depth
]
1318 DarkBASIC Pro: ODE
In the diagram:

to which the ODE box is to be linked. Only one
ODE box can be linked to each 3D object.
width is a real number giving the width of the ODE

box.
height is a real number giving the height of the ODE

box.
depth is a real number giving the depth of the ODE

box.
For example, if we create a normal 3D cube using the line
CREATE OBJECT CUBE 1, 10
we could link an ODE box to that cube using the line

ODE CREATE DYNAMIC BOX 1
The ODE box would be exactly the same size as the cube which it surrounds.
Should we want to make the ODE box twice the size of the cube, then we could use
the line
ODE CREATE DYNAMIC BOX 1, 20,20,20
making the ODE box 20 units in all three dimensions.
The ODE START Statement

The ODE physics engine does a massive amount of calculations to ensure that all
ODE boxes (and other shapes - see later) react in a realistic way. The engine needs
to be initialised and this is done using the ODE START statement which has the
FIG-48.3
The ODE START ODE START
Statement
This statement must appear in your program before any other ODE related
statement.
The ODE END Statement

When we no longer require the physics engine to operate, the ODE END statement
should be executed. This statement has the form shown in FIG-48.4.
FIG-48.4
The ODE START ODE END
Statement
The ODE UPDATE Statement

The physics engine needs to update the screen in order to reposition the 3D models

as a result of gravitational pull or other forces acting on the objects. This is achieved
by using the ODE UPDATE statement which has the format shown in FIG-48.5.
FIG-48.5
The ODE UPDATE
ODE UPDATE
Statement
Now we're ready to make a start on using ODE. In LISTING-48.1 a cube falls from
above, under the influence of gravity.
ScreenSetUp()
LISTING-48.1 REM *** Make cube ***
Getting Started with ODE POSITION OBJECT 1,0,25,0
WAIT KEY
REM *** Start Physics engine ***
ODE START
REM *** Link ODE box to cube ***
DO
ODE UPDATE
LOOP
ODE END
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 48.1
Type in and test the program given in LISTING-48.1 (ODE01.dbpro).
The ODE SET WORLD GRAVITY Statement

The ODE world comes with gravity built in - as we saw in the last program. But
the force of gravity can be changed (and even made to act in other directions) using
the SET WORLD GRAVITY statement which has the format shown in FIG-48.6.
FIG-48.6
The SET WORLD GRAVITY Statement
ODE SET WORLD GRAVITY xforce , yforce , zforce
In the diagram:
xforce is a real number giving the gravitational force

applied along the x-axis.
yforce is a real number giving the gravitational force

applied along the y-axis.

zforce is a real number giving the gravitational force
applied along the z-axis.
The default gravitational force is down the y-axis only and is equivalent to the
statement:
ODE SET WORLD GRAVITY 0,-9.81,0
Activity 48.2
In your last program, add the following line immediately after ODE START
ODE SET WORLD GRAVITY 0,-2,0
Check out the following other settings:

xforce yforce zforce
-2 0 0
0 0 2
0 -2 2
-2 -2 2
The ODE CREATE STATIC BOX Statement

If we create a plane to represent the ground, perhaps we can get the falling box to
stop when it hits the ground's surface. We'll test this by adding the lines

to the previous program.
Activity 48.3
Add the above lines near the start of your last program.
Return the gravity to a setting of 0,-2,0.
Add a WAIT KEY statement before starting the ODE engine. Does the cube
stop when it hits the ground?
The cube goes straight through the ground plane. Of course, this is because the plane
object does not have an ODE box.
Activity 48.4
Add the line
at the appropriate point in your program. Does the cube stop when it hits the
ground?
And so another problem becomes apparent! Putting a dynamic box around any
object means it is affected by gravity just like the cube.

When we want an object to remain in a fixed position and not be affected by normal
forces and yet still be part of the ODE universe (so that other objects will react to
coming into contact with it), we need to build a static ODE box around the fixed
object rather than a dynamic box. This is done using the ODE CREATE STATIC
BOX statement which has the format shown in FIG-48.7.
FIG-48.7
The ODE CREATE ODE CREATE STATIC BOX objno
STATIC BOX Statement
In the diagram:

to which the ODE box is to be linked.
For example, we could link a static collision box to object 2 using the line:
ODE CREATE STATIC BOX 2
Activity 48.5
Modify the last program to link a static box to the plane object.
Does the ground and cube react correctly now?
Modify the dynamic box around the cube so that it is has width, height and
depth of 20 units.
What happens when the cube falls?
The ODE CREATE DYNAMIC SPHERE Statement

To make a collision appear accurate, the shape of the ODE volume should match
as closely as possible that of the 3D object to which it is attached. For example,
linking a dynamic box to a 3D sphere will not give accurate collision results under
some circumstances. To handle this there are several other dynamic volumes which
can be linked to a 3D object. A dynamic sphere can be linked using the ODE
CREATE DYNAMIC SPHERE statement. This statement has the format shown in
FIG-48.8.
FIG-48.8
The ODE MAKE ODE CREATE DYNAMIC SPHERE objno
DYNAMIC SPHERE
Statement
In the diagram:

to which the ODE sphere is to be linked.
Activity 48.6
Change the ODE volume around the 3D cube to a sphere. How does this affect
the collision?
The ODE CREATE DYNAMIC CYLINDER Statement

Another option when creating a dynamic volume is the cylinder. This is created

using the ODE CREATE DYNAMIC CYLINDER statement which has the format
shown in FIG-48.9.
FIG-48.9
The ODE CREATE ODE CREATE DYNAMIC CYLINDER objno
DYNAMIC CYLINDER
Statement
In the diagram:

to which the ODE cylinder is to be linked.
The program in LISTING-48.2 creates a cube, a sphere and a cylinder, then links
each object to the appropriate ODE dynamic volume before letting all three drop to
the ground.
Dropping Multiple REM *** Make cube ***

Objects MAKE OBJECT CUBE 1,10
REM *** Make cylinder ***
REM *** Make and texture ground ***

TEXTURE OBJECT 4,1
REM *** Wait before starting ODE ***
WAIT KEY
REM *** Start ODE ***

ODE START
REM Add ODE volumes ***

ODE CREATE DYNAMIC SPHERE 2
ODE CREATE DYNAMIC CYLINDER 3
REM *** Continually update physics ***

DO
ODE UPDATE
LOOP
ODE END
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION

Activity 48.7
The ODE CREATE DYNAMIC TRIANGLE MESH Statement

If a complex model has to be linked to an ODE volume, this can be done with the
standard shapes already mentioned, but alternatively a more accurate volume can
be produced using the ODE CREATE DYNAMIC TRIANGLE MESH statement
FIG-48.10
The ODE CREATE ODE CREATE DYNAMIC TRIANGLE MESH objno
DYNAMIC TRIANGLE
MESH Statement
In the diagram:

to which the ODE mesh is to be linked.
In LISTING-48.3 the Hivebrain model uses an ODE mesh to detect its collision
with the ground.
REM *** Load model ***
Using a Triangle Mesh LOAD OBJECT "H-Alien Hivebrain-Move.x",1
REM *** Make ground ***

TEXTURE OBJECT 4,1
REM *** Wait before starting ODE ***

WAIT KEY

ODE START
REM *** Use triangle mesh on model ***

ODE CREATE DYNAMIC TRIANGLE MESH 1
REM *** Use box on ground ***

DO
ODE UPDATE
LOOP
ODE END
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION

Activity 48.8
Type in and test the program in LISTING-48.3 (ODE03.dbpro).
How accurate is the collision detection?
Change the ODE volume used on the model to a box. Is this more accurate?
The ODE SET WORLD STEP

Every time the physics engine recalculates the position of the ODE volumes, it
assumes a specific amount of time has passed since the last calculation. To give an
accurate simulation, this time gap must be small - just a fraction of a second. To set
this time gap, we can use the ODE SET WORLD STEP statement which has the
FIG-48.11
The ODE SET WORLD ODE SET WORLD STEP time
STEP Statement
In the diagram:
time is a real number representing the time interval

between each calculation. Values between 0.04
and 1.7 work.
The effect of this statement is to speed up or slow down movement (since the time
scale is changed) as well as affecting the accuracy of the movements.
Activity 48.9
In your last program add the line

ODE SET WORLD STEP 0.05
immediately after the ODE START statement.
Change the step size to 0.2.
How does this change the overall effect?
The ODE CREATE STATIC TRIANGLE MESH Statement

The program in LISTING-48.4 has two static objects; the ground and a cone placed
on the ground. A cube then falls onto the cone.
Adding a Realistic ODE REM *** Make objects ***

Volume for the Cone MAKE OBJECT CUBE 1,7
TEXTURE OBJECT 4,1

LISTING-48.4 REM *** Wait to start ODE ***
Adding a Realistic ODE ODE START

Volume for the Cone
REM *** Build ODE volumes ***

DO
ODE UPDATE
LOOP

ODE END
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 48.10
As we see, the cube stops when it hits the cone. It doesn't tumble, because the cone
is using a box collision volume. A more accurate result would be obtained if a static
mesh was used on the cone. This is done using the ODE CREATE STATIC
TRIANGLE MESH statement which has the format shown in FIG-48.12.
FIG-48.12
The ODE CREATE
ODE CREATE STATIC TRIANGLE MESH objno
STATIC TRIANGLE
MESH Statement
In the diagram:

to which the ODE mesh is to be linked.
Activity 48.11
In your last program, change the ODE volume for object 2 to use a static mesh.
The ODE SET WORLD ERP Statement

When two objects collide, the physics engine handles this by creating a temporary
ERP stands for Error link (known as a joint) between the two objects. This joint is used to calculate how
Reduction Parameter. the two objects should react to the collision.
In performing this calculation some rounding errors will occur. How these errors

are handled can be set using the ODE SET WORLD ERP statement which has the
FIG-48.13
The ODE SET WORLD ODE SET WORLD ERP value
ERP Statement
In the diagram:
value is a real number which should lie between 0 and

1. A value of zero will result in no corrective
adjustments; 1 will give maximum correction.
For example, we could switch off corrective action using the line:
ODE SET WORLD ERP 0.0
In fact, this statement can have a major effect on how objects react to collisions. In
the next Activity we'll see just what effects it can have on the cube as it falls onto
the cone.
Activity 48.12
ODE SET WORLD ERP 0.0
immediately after the ODE START statement.
How does this affect the falling cube?
Try running the program again with first a value of 1.0 and then 0.5 for the
ERP.
The ODE SET WORLD CFM Statement

CFM stands for
Constraint Force
Falling onto a concrete base is going to hurt a lot more than falling onto a trampoline
Mixing. - it's all to do with bounce. The more give there is in a material the more things
bounce. We can set the bounce factor for all the dynamic objects in our ODE
universe using the ODE SET WORLD CFM statement which has the format shown
in FIG-48.14.
FIG-48.14
ODE SET WORLD CFM value
The ODE SET WORLD
CFM Statement
In the diagram:
value is a real number which should lie between 0 and

5. A value of zero gives a very rigid (no bounce )
effect; 5 gives maximum bounce.
For example, a large bounce can be effected using the line:
ODE SET WORLD CFM 5

Activity 48.13
Add the line

ODE SET WORLD CFM 2.0
immediately after the ERP statement in your last program.
The ODE SET CONTACT FDIR1 Statement
FDIR1 stands for You should have seen how the cube slides along the surface after it lands. This is
Friction DIRection 1 due to a lack of friction - as if the block was made of melting ice. We can set the
(there's a second friction friction of a dynamic object using the ODE SET CONTACT FDIR1 statement
direction).
FIG-48.15
The ODE SET ODE SET CONTACT FDIR1 objno , friction
CONTACT FDIR1
Statement
In the diagram:

which a friction coefficient is to be applied.
friction is a real number setting the friction coefficient

of the object. A value of zero results in no friction
(the default for all objects) while 100 gives
maximum friction.
To add a friction coefficient to our falling cube we could use the line:
ODE SET CONTACT FDIR1 1,50
Activity 48.14
Add the above statement to your last program. Insert the line immediately
after the ODE volume is assigned to the cube.
The ODE SET LINEAR VELOCITY Statement

We can add a velocity to an object using the ODE SET LINEAR VELOCITY
FIG-48.16
The ODE SET LINEAR VELOCITY Statement
ODE SET LINEAR VELOCITY objno , xcomp , ycomp , zcomp
In the diagram:

which a velocity is to be applied.
xcomp is a real value giving the value of the x

component of the velocity.

ycomp is a real value giving the value of the y
zcomp is a real value giving the value of the z

For example, we could make object 1 move horizontally using the line:
ODE SET LINEAR VELOCITY 1, 20,0,0
A larger value will result in faster movement.
The program in LISTING-48.5 moves a cube from left to right above the ground
plane. Notice that gravitational forces are set to zero.
REM *** Make objects ***
Giving an Object a MAKE OBJECT CUBE 1,7
Velocity POSITION OBJECT 1, -20,20,0
TEXTURE OBJECT 4,1
REM *** Wait to start ODE ***
WAIT KEY
ODE START
REM *** No gravity ***
ODE SET WORLD GRAVITY 0,0,0
REM *** Add a velocity to the cube ***
ODE SET LINEAR VELOCITY 1,20,0,0
DO
ODE UPDATE
LOOP
ODE END
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 48.15
Find out the effect of each of the following velocity settings:
xcomp ycomp zcomp

100 0 0
-10 0 0
0 -5 0
0 0 20
0 -10 20

If we switch gravity back on, this will affect the path taken by the object to which
a velocity has been applied.
Activity 48.16
Return the velocity settings of the cube to 0,0,20.
Remove the ODE SET GRAVITY statement, thereby returning the

gravitational force to its default setting.
Run the program again. How is the cube's path affected?
In the next program (see LISTING-48.6), two cubes travel towards each other and
collide.
Colliding Bodies REM *** Make cubes ***
POSITION OBJECT 1, -20,20,0
REM *** Wait to start ODE ***

WAIT KEY
ODE START

REM *** No gravity ***

REM *** Add a velocity to the cubes ***

ODE SET LINEAR VELOCITY 2,-2,0,0
DO
ODE UPDATE
LOOP

ODE END
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 48.17
What happens when the two cubes collide?

The force applied to each cube is equal but opposite, so the collision results in these
energies cancelling each other out. Of course, if the velocities of the two cubes were
to be different, then the result obtained would change.
Activity 48.18
In your last program, what effect is achieved for each of the following settings
in the ODE SET LINEAR VELOCITY statement for object 1:
xcomp ycomp zcomp
4 0 0
4 -1 0
The ODE SET ANGULAR VELOCITY Statement

We can set an object spinning by assigning it an angular velocity. This is achieved
using the ODE SET ANGULAR VELOCITY statement which has the format
shown in FIG-48.17.
FIG-48.17
The ODE SET ANGULAR VELOCITY Statement
ODE SET ANGULAR VELOCITY objno , xcomp , ycomp , zcomp
In the diagram:

which an angular velocity is to be applied.
xcomp is a real value giving the spin force about the

x-axis.
ycomp is a real value giving the spin force about the

y-axis.
zcomp is a real value giving the spin force about the

z-axis.
To keep the spin to a reasonably slow rate, xcomp, ycomp, and zcomp values should
be less than 1.
For example, to create a slow spin of object 1 with rotation about the x-axis being
half the speed of rotation about the y-axis, we could use the line:
ODE SET ANGULAR VELOCITY 1, 0.1,0.2,0
Activity 48.19
Modify your last program by including the statement given above immediately
after the linear velocity of object 1 has been set.
How is the cube affected?
Move the ODE SET ANGULAR VELOCITY statement placing it within the
program's DO..LOOP structure.

The ODE SET BODY ROTATION Statement
It is also possible to get an object to rotate to a specific position about its own local
axes using the ODE SET BODY ROTATION statement. Although we already have
a ROTATE OBJECT statement, it is important to use the ODE version of the
statement if we want the ODE box (sphere, cylinder, etc.) associated with the object
also to be rotated.
The ODE SET BODY ROTATION statement has the format shown in FIG-48.18.
FIG-48.18
The ODE SET BODY ROTATION Statement
ODE SET BODY ROTATION objno , xangle , yangle , zangle
In the diagram:

which the rotation is to be applied.
xangle is a real value giving the object's angle of

rotation about its x-axis.
yangle is a real value giving the object's angle of

rotation about its y-axis.
zangle is a real value giving the object's angle of

rotation about its z-axis.
For example, we could rotate object 1 45o about its y-axis using the statement:
ODE SET BODY ROTATION 1,0,45,0
Activity 48.20
Replace the ODE SET ANGULAR VELOCITY statement in your last

program with the ODE SET BODY ROTATION statement given above.
Notice that the cube is set to a specific angle and is then fixed. Since this
statement creates a single movement of the object, it need not be within the
DO..LOOP structure.
Move the ODE SET BODY ROTATION statement to a position before the
loop structure and check that the overall effect remains the same.
The ODE SET BODY MASS Statement

In the real world, objects have weight - or, rather, mass. In ODE objects only have
a mass if they are explicitly assigned one. This can be done using the ODE SET
BODY MASS statement which has the format given in FIG-48.19.
FIG-48.19
The ODE SET BODY ODE SET BODY MASS objno , value
MASS Statement

In the diagram:

which a mass is to be assigned.
value is a real value giving the mass to be assigned to

the object. This value does not represent any
specific units of measurement.
Activity 48.21
Modify your last program so that the mass of object 1 is set to 8 and that of
object 2 to 100.
How does this effect the cubes?
Set gravity to -0.1. How are the cubes affected this time?
The program in LISTING-48.7 shows the effects produced by throwing a ball at a

wall of cubes (see FIG-48.20).
FIG-48.20
Tumbling Blocks
Tumbling Blocks REM *** Load image used to texture objects ***
REM *** Make, texture and position cubes ***

objno = 1
FOR row = 0 TO 7
FOR col = 0 TO 7
MAKE OBJECT CUBE objno,7
TEXTURE OBJECT objno,1
POSITION OBJECT objno,col*7-28,row*7+6,0
INC objno
NEXT col
NEXT row
REM *** Make ball


TEXTURE OBJECT 66,1
XROTATE OBJECT 66,-90 continued on next page

LISTING-48.7 REM *** Start ODE ***
(continued) ODE START
Tumbling Blocks REM *** Static box around floor ***
REM *** Build ODE volumes around cubes ***

objno = 1
FOR row = 0 TO 7
FOR col = 0 TO 7
ODE CREATE DYNAMIC BOX objno
ODE SET CONTACT FDIR1 objno,80
INC objno
NEXT col
NEXT row
REM *** ODE volume around sphere ***

ODE CREATE DYNAMIC SPHERE 65
REM *** Give sphere a push ***

REM *** Engage physics after key press ***

WAIT KEY
DO
ODE UPDATE
LOOP

ODE END
END
COLOR BACKDROP 0
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 48.22
Modify the program so that the camera can be moved about within the scene.
(HINT: include the camera.dba routines we created in Chapter 33.)
The ODE DESTROY OBJECT Statement

The ODE object linked to a regular 3D object can be deleted using the ODE
DESTROY OBJECT statement which has the format shown in FIG-48.21.
FIG-48.21
The ODE DESTROY
ODE DESTROY OBJECT objno
OBJECT Statement
In the diagram:

to which the ODE volume is attached.

The ODE object attached to the specified 3D object is deleted; the 3D object itself
is not affected by the operation, but its movement is no longer controlled by the
physics engine.
Activity 48.23
Inside the DO..LOOP structure of your last program, add the following lines:
IF SPACEKEY()
ODE DESTROY OBJECT 65
ENDIF
This will destroy the ODE volume attached to the sphere.
Test the new program by destroying the ODE volume before the ball hits the
blocks. How is the program affected?
The ODE GET BODY LINEAR VELOCITY Statement

We can discover the x, y and z components of an object's velocity using the ODE
GET BODY LINEAR VELOCITY statement which has the format shown in
FIG-48.22.
FIG-48.22 The ODE GET BODY LINEAR VELOCITY Statement
{
X
ODE GET BODY LINEAR VELOCITY Y

( objno )
Z
real
In the diagram:
X,Y,Z Use X to retrieve the x component of the velocity;

use Y to retrieve the y component of the velocity;
use Z to retrieve the z component of the velocity.

whose velocity is to be determined.
Activity 48.24
Modify the program in ODE06.dbpro to display both objects' linear velocity at

all times.
The ODE GET BODY HEIGHT Statement

The ODE GET BODY HEIGHT returns the height of a specified object. Although
we might be tempted to think that an object will have whatever height we set it to
when the object is first created, an irregularly shaped object will change in height
as it is rotated. The format for this statement is given in FIG-48.23.

FIG-48.23
The ODE GET BODY
ODE GET BODY HEIGHT ( objno )
HEIGHT Statement
real
In the diagram:

whose height is to be determined.
The statement returns the length of the object along the world y-axis.
The ODE COLLISION MESSAGE EXISTS Statement

When a collision occurs, details of that collision are placed on a stack. We can check
if that stack currently contains any messages using the ODE COLLISION
MESSAGE EXISTS statement which has the format shown in FIG-48.24.
FIG-48.24
The ODE COLLISION ODE COLLISION MESSAGE EXISTS ( )
MESSAGE EXISTS
Statement
integer
The statement returns 1 if at least one message exists in the stack; otherwise zero is
returned.
The ODE COLLISION GET MESSAGE Statement

The top message in the ODE collision stack can be retrieved using the ODE
COLLISION GET MESSAGE statement (format shown in FIG-48.25).
FIG-48.25
The ODE COLLISION ODE COLLISION GET MESSAGE
GET MESSAGE
Statement
The details in the value retrieved from the stack can then be accessed using other
commands.
Typical code for retrieving collision details from the stack would be:
IF ODE COLLISION MESSAGE EXISTS()

ODE COLLISION GET MESSAGE
ENDIF
The ODE GET OBJECT Statement

To discover what two objects have been in collision, we need to use the ODE GET
OBJECT statement which takes the format shown in FIG-48.26.

FIG-48.26
The ODE GET OBJECT
Statement
ODE GET OBJECT
{ A
B
( )
integer
In the diagram:
A,B Use A to retrieve the ID of the first object

involved in a collision; use B to retrieve the ID
of the second object.
This statement should be used after a message has been retrieved from the collision
stack.
For example, we could see which objects have collided using the code:

obj1 = ODE GET OBJECT A()
obj2 = ODE GET OBJECT B()
REPEAT
SET CURSOR 100,100
PRINT "Objects collided: ",obj1," ",obj2
ENDIF
Activity 48.25
Reload ODE06.dbpro. Create a function, DisplayCollision(), from the lines

above and modify the DO..LOOP to display only the output from this
function. Run the program and check that the objects colliding are identified.
The ODE GET OBJECT VELOCITY Statement

The x, y and z components of a colliding object's velocity can be determined using
the ODE GET OBJECT VELOCITY statement whose format is shown in
FIG-48.27.
FIG-48.27
{
X
{
The ODE GET A
OBJECT
VELOCITY ODE GET OBJECT VELOCITY Y ( )
Statement B
Z
real
In the diagram:


X,Y,Z Use X to retrieve the x component of the velocity;
use Y to retrieve the y component of the velocity;
use Z to retrieve the z component of the velocity.
Activity 48.26
Update DisplayCollision() in your last program to display the velocity offsets

of each colliding object.
The ODE GET OBJECT ANGULAR VELOCITY Statement

The x, y and z components of a colliding object's angular velocity can be determined
using the ODE GET OBJECT ANGULAR VELOCITY statement whose format is
shown in FIG-48.28.
FIG-48.28 The ODE GET OBJECT ANGULAR VELOCITY Statement
{
X
ODE GET OBJECT

{ A
B
ANGULAR VELOCITY Y
Z
( )
real
In the diagram:

X,Y,Z Use X to retrieve the x component of the angular

velocity; use Y to retrieve the y component of the
angular velocity; use Z to retrieve the z component
of the angular velocity.
The ODE ADD FORCE Statement

An additional force can be applied to an object using the ODE ADD FORCE
statement. The statement allows the magnitude and direction of the force to be
specified as well as the point on the object at which the force is to be applied. By
applying the force off-centre, we add a spin to the object. The format for the ODE
ADD FORCE statement is given in FIG-48.29.
FIG-48.29 The ODE ADD FORCE Statement
ODE ADD FORCE objno , xf , yf , zf , x , y , z
In the diagram:

which the force is to be applied.

xf,yf,zf are a set of real values giving the strength of the
force along each axis.
x,y,z are a set of real values giving the position on the

object at which the force is to be applied.
The program in LISTING-48.8 allows the user to apply a force to a cube by pressing
the space bar.
Applying a Force
REM *** Force to be applied ***
force# = 0.01
REM *** Force already applied indicator ***

applied = 0
REM *** Start ODE and eliminate gravity ***

ODE START
REM *** Load dice model and add mass ***

LOAD OBJECT "dice.dbo",2
ODE SET BODY MASS 2, 5
REM *** Engage physics engine ***

DO
ODE UPDATE
REM *** First time space bar pressed, apply force ***
IF SPACEKEY()=1 AND (NOT applied)
ODE ADD FORCE 2,force#,0,0,0,0,0
REM *** Remember force has been applied ***
applied=1
ENDIF
LOOP

ODE END
END
BACKDROP ON
AUTOCAM OFF
POINT CAMERA 0,0,0
ENDFUNCTION
Activity 48.27
Try modifying the value of force# and applying it to the y and z directions
instead of x.
Check out the effects of changing the point at which the force is applied.

Surface Contact Statements
When two objects collide, a contact joint is created. This contact joint contains
details of how the two objects should react to that collision (see FIG-48.30).
FIG-48.30 A Contact
Joint Object 1
Point of A contact normal

contact vector
Object 2
When two objects collide with each A temporary contact joint is created at The contact joint contains a contact
other... the point of contact. normal detailing a repelling force used
by the two objects after colliding.
Fdir1 Fdir2
If the two objects approach each other The friction between the two surfaces These friction vectors and the softness
at a narrow angle one may pass over is determined by two friction vectors of an object's structure can be set using
the surface of the other. Fdir1 and Fdir2. various surface contact statements.
There are a series of statements allowing the programmer to fine-tune how the
collision should be dealt with and the forces involved. We can set factors such as
the bounciness of the objects, and the surface friction components along both
directions (Fdir1 and Fdir2).
Exactly which options are to be used must first be selected using a set of ODE
SETSURFACE MODE CONTACT statements which are designed to
select/deselect various characteristics. These statements and a brief explanation of
what effect is being selected are shown in TABLE-48.1.
TABLE-48.1 Statement Parameters Description
SETSURFACE o d e s e t s u r f a c e mo d e c o n t a c t b o u n c e objno:int, state:int When state = 0, objno, will create

CONTACT MODE a bouncey collision
Statements o d e s e t s u r f a c e mo d e c o n t a c t s o f t e r p objno:int, state:int When state = 0, objno's ERP
can be set. Useful for creating
soft objects.
o d e s e t s u r f a c e mo d e c o n t a c t s o f t c f m objno:int, state:int When state = 0, objno's CFM can
be set. Useful when creating
soft objects.
o d e s e t s u r f a c e mo d e c o n t a c t mo t io n 1 objno:int, state:int When state = 0, objno's surface
is assumed to be moving
independently (like a conveyer
belt) in direction Fdir1
o d e s e t s u r f a c e mo d e c o n t a c t mo t io n 2 objno:int, state:int When state = 0, objno's surface
is assumed to be moving
independently in direction Fdir2
o d e s e t s u r f a c e mo d e c o n t a c t s l ip 1 objno:int, state:int When state = 0, a force-
dependent slip in direction Fdir1
can be applied.
o d e s e t s u r f a c e mo d e c o n t a c t s l ip 2 objno:int, state:int When state = 0, a force-
dependent slip in direction Fdir2
can be applied.
o d e s e t s u r f a c e mo d e c o n t a c t a p p r o x 11 objno:int, state:int When state = 0 calculate friction
using a friction-pyramid
approximation in direction Fdir1
and Fdir2

Once the required options have been selected, the strengths of actual forces involved
can be set using various ODE SET CONTACT statements. These are described
briefly in TABLE-48.2.
TABLE-48.2 Statement Parameters Description

ODE SET CONTACT
o d e s e t c o n t a c t f d ir 1 objno: int, value:real Sets the degree of friction along Fdir1 on
Statements
objno to value. value range 0 to 100.
0 = no friction 100 = max friction
o d e s e t c o n t a c t mu 2 objno: int, value:real Sets the degree of friction along Fdir2
o d e s et c o n t ac t bo u n c e objno: int, value:real Sets the amount of elasticity and bounce
of objno to value. value range 0 to 1.
o d e s e t c o n t a c t v e l o c it y objno: int, value:real Sets minimum velocity required to create
a bounce effect.
o d e s et c o n t ac t s o f t er p objno: int, value:real Sets ERP for objno to value
o d e s et c o n t ac t s o f t c f m objno: int, value:real Sets CPM for objno to value
o d e s e t c o n t a c t mo t io n 1 objno: int, value:real Sets the movement factor along Fdir1
for objno to value
o d e s e t c o n t a c t mo t io n 2 objno: int, value:real Sets the movement factor along Fdir2
for objno to value
o d e s e t c o n t a c t s l ip 1 objno: int, value:real Sets the slip value to be applied along
direction Fdir1 to value
o d e s e t c o n t a c t s l ip 2 objno: int, value:real Sets the slip value to be applied along
direction Fdir1 to value
The program in LISTING-48.9 demonstrates the bounce effect, bouncing a falling

dice off the ground. All the statements necessary to create a bounce effect are held
in a separate function, CreateBounce(). Notice that the effects are applied only to
the dynamic object.
ScreenSetUp()
LISTING-48.9
Creating Bounce
ODE START
ODE SET WORLD STEP 0.05
ODE SET WORLD ERP (0.2)*2.5
ODE SET WORLD CFM (10^-5)*2.5
ODE SET STEP MODE 0

TEXTURE OBJECT 1,1
REM *** Load dice ***

LOAD OBJECT "dice.dbo",2
REM *** Add ODE components to objects ***

REM ** Give the dice weight ***

ODE SET BODY MASS 2,100
CreateBounce(2)
DO
ODE UPDATE
LOOP

ODE END
END

AUTOCAM OFF
Creating Bounce POSITION CAMERA 0,30,-150
ENDFUNCTION
FUNCTION CreateBounce(objno)
ODE SETSURFACE MODE CONTACT BOUNCE objno,0
ODE SET CONTACT BOUNCE objno,0.7
ODE SETSURFACE MODE CONTACT SOFT ERP objno,0
ODE SET CONTACT SOFT ERP objno, 0.8
ODE SETSURFACE MODE CONTACT SOFT CFM objno, 0
ODE SET CONTACT SOFT CFM objno,200.5
ENDFUNCTION
Activity 48.28
Try varying the parameters of the statements in the new function and check
out what effect this has.
The next routine tests out the conveyer-belt effect created using the CONTACT
MOTION1 and CONTACT MOTION2 statements:
FUNCTION ConveyerBelt(objno)
ODE SET CONTACT FDIR1 objno,50
ÒDE SETSURFACE MODE CONTACT MOTION1 objno,0
ÒDE SET CONTACT MOTION1 objno,10
ÒDE SETSURFACE MODE CONTACT MOTION2 objno,0
ÒDE SET CONTACT MOTION2 objno,10
ENDFUNCTION
Activity 48.29
Add the code for ConveyerBelt() to your last program (notice that most of the
lines have been commented out at this stage).
In the main section of your program, comment out the call to CreateBounce()
and replace it with the lines:
ODE ADD FORCE 2,4,0,0,0,0,0

ConveyerBelt(2)
Run the program and observe how the dice moves.
Now remove the comments from lines 2 and 3 in ConveyerBelt() and run the
program again. What happens to the dice this time?
Next, re-comment out lines 2 and 3 and uncomment lines 4 and 5. How does
the dice react ?
Remove all comments in ConveyerBelt() and run the program again. What
path does the dice take?

Summary
l The Open Dynamics Engine (ODE) is an undocumented part of DarkBASIC
Pro.
l By using ODE we can make basic laws of physics apply to 3D objects.
l ODE places invisible volumes around 3D objects.
l Static volumes should be placed around 3D objects which are stationery.
l Dynamic volumes should be placed around objects which can be moved.
l Use ODE CREATE DYNAMIC BOX to place a box volume around a moveable
object.
l Use ODE CREATE DYNAMIC SPHERE to place a spherical volume around a

moveable object.
l Use ODE CREATE DYNAMIC CYLINDER to place a cylindrical volume

around a moveable object.
l Use ODE CREATE DYNAMIC TRIANGLE MESH to place a mesh-based

volume around a moveable object.
l Use ODE CREATE STATIC BOX to place a box volume around an immovable
object.
l Use CREATE STATIC TRIANGLE MESH to place a mesh-based volume

around an immovable object.
l Use ODE START before any other ODE statements in your program. This
initialises the ODE.
l Use ODE END to end use of the ODE.
l Use ODE UPDATE in a loop structure to continually update the calculations and
movements performed by the ODE.
l By default, ODE assumes a gravitational pull down the y-axis.
l Use ODE SET WORLD GRAVITY to modify the gravitational force used in a
program.
l Use ODE SET WORLD STEP to set the time interval assumed to have passed
between each ODE update.
l Use ODE SET WORLD ERP to specify the error production parameter.
l Use ODE SET WORLD CFM to set the bounciness of every ODE object.
l Use ODE SET CONTACT FDIR1 to set the surface friction of an object.
l Use ODE SET LINEAR VELOCITY to set the velocity of an object.
l Use ODE SET ANGULAR VELOCITY to set the angular velocity of an object.

l Use ODE SET BODY ROTATION to rotate an object to a specific angle.
l Use ODE SET BODY MASS to assign a mass to an object.
l Use ODE DESTROY OBJECT to destroy the ODE volume surrounding a

standard 3D object.
l Use ODE GET BODY LINEAR VELOCITY to determine the linear velocity
of an object.
l Use ODE GET BODY HEIGHT to determine the height of an object. This may
change as an asymmetrical shape is rotated.
l Use ODE COLLISION MESSAGE EXISTS to determine if a collision has taken

place between two objects and the corresponding details placed on a collision
stack.
l Use ODE COLLISION GET MESSAGE to retrieve the latest entry in the
collision stack.
l Use ODE GET OBJECT to determine which two objects have collided. This
statement is used after an entry has been retrieved from the collision stack.
l Use ODE GET OBJECT VELOCITY to retrieve the velocities of the two objects
that have collided. This statement is used after an entry has been retrieved from
the collision stack.
l Use ODE GET OBJECT ANGULAR VELOCITY to retrieve the angular

velocities of the two objects that have collided. This statement is used after an
entry has been retrieved from the collision stack.
l Use ODE ADD FORCE to apply a force to a specified object.
l Surface contact statements are used to set the characteristics of a collision.

Solutions
Activity 48.1 ODE CREATE DYNAMIC BOX 2
No solution required. to

Activity 48.2
The ground stays still and the cube stops when it hits the
The cube fall varies in speed and direction with each ground.
setting.
Change
Activity 48.3 ODE CREATE DYNAMIC BOX 1
The code is:
to
ScreenSetUp()
REM *** Make cube *** ODE CREATE DYNAMIC BOX 1,20,20,20
POSITION OBJECT 1,0,25,0 With a larger surrounding volume, the cube stops well
REM *** Make ground *** above ground level.
XROTATE OBJECT 2,90
ODE START Since a sphere is not an accurate enclosure for a cube, the
cube again stops above ground level.
REM *** Link ODE box to cube ***
DO Activity 48.7
ODE UPDATE
LOOP
ODE END
END Activity 48.8
Neither volume seems to be very accurate.
The cube passes straight through the ground plane.
Activity 48.9
Activity 48.4
The whole process runs much more slowly at 0.05. Using
The code is: 0.2 speeds things up.
ScreenSetUp()
Activity 48.10
XROTATE OBJECT 2,90
WAIT KEY
Activity 48.11
ODE START
Change
REM *** Link ODE box to cube and plane *** ODE CREATE STATIC BOX 2
ODE CREATE DYNAMIC BOX 2 to
DO ODE CREATE STATIC TRIANGLE MESH 2
ODE UPDATE
LOOP Although it has a static volume, the cone sinks into the
ground! The cube tilts on hitting the top of the cone.
ODE END
END
Activity 48.12
The plane which is meant to represent ground level also
falls! The ERP values have the following effects:
0.0 the cube slides

Activity 48.5 1.0 the cube bounces
Change the line 0.5 the cube tilts over the cone

REM *** Slight gravity ***
Activity 48.13 ODE SET WORLD GRAVITY 0,-0.1,0
This change adds a slight bounce before the cube slides REM *** Set cubes' characteristics ***
ODE SET LINEAR VELOCITY 1,4,-1,0
off. ODE SET BODY MASS 1, 8
ODE SET BODY ROTATION 1,0,45,0
ODE SET LINEAR VELOCITY 2,-2,0,0
Activity 48.14 ODE SET BODY MASS 2, 100
This stops the cube sliding off along the ground. DO
ODE UPDATE
LOOP
ODE END
The effects are: END
100,0,0 cube moves quickly to the right Since object 2 has a much larger mass, it has a greater
-10,0,0 cube moves slowly to the right effect on object 1's trajectory.
0,5,0 cube moves up
0,0,20 cube moves off into the screen With a slight gravity, the cubes fall as they move.
0,-10,20 cube moves down and in, then moves in
only
Activity 48.22
Activity 48.16 To move the camera about the scene, the DO..LOOP
structure should be changed to:
As the cube moves away it also falls, eventually hitting
the ground and sliding on until it falls over the edge of the DO
ground. ODE UPDATE
MoveCameraUsingMouse(0,0.5)
Activity 48.17 LOOP
The cubes stop moving when they collide. Also, to include the necessary code, we need to add the line
Activity 48.18
When object 1's velocity is set to 4,0,0, it pushes object 2 at the start of the program.
to the right.
Activity 48.23
When object 1's velocity is set to 4,-1,0, it tilts and halts
object 2. The main loop should now be:
DO
Activity 48.19 ODE UPDATE
IF SPACEKEY()
Object 1 tilts slightly, but does not continue to rotate. ODE DESTROY OBJECT 65
ENDIF
When the statement is moved within the DO..LOOP, the PointCameraUsingMouse(0)
cube continues to rotate as it moves. MoveCameraUsingMouse(0,0.5)
LOOP
Activity 48.20 When the ODE volume is destroyed, the sphere is no

longer influenced by the physics engine and stops moving.
Activity 48.24
Activity 48.21
To keep the main section of the program simple, it is
The final version of the main section of the program probably best if we do the hard work in two routines:
should now be:
FUNCTION DisplayVelocities()
ScreenSetUp() SET CURSOR 100,100
REM *** Make cubes *** PRINT "Velocity (object 1):",
MAKE OBJECT CUBE 1,7 Ä
VelocityDetails$(1)
POSITION OBJECT 1, -20,20,0 SET CURSOR 100,130
MAKE OBJECT CUBE 2,7 PRINT "Velocity (object 2):",
POSITION OBJECT 2,20,20,0 Ä
VelocityDetails$(2)
REM *** Wait to start ODE *** ENDFUNCTION
WAIT KEY
ODE START FUNCTION VelocityDetails$(objno)
REM *** Build ODE volumes *** x# = ODE GET BODY LINEAR VELOCITY
ODE CREATE DYNAMIC BOX 1 Ä
X(objno)
ODE CREATE DYNAMIC BOX 2 y# = ODE GET BODY LINEAR VELOCITY

ÄY(objno) Ä,obj1,": ",
z# = ODE GET BODY LINEAR VELOCITY ÄCollisionVelocityDetails$("A")
ÄZ(objno) SET CURSOR 100,160
result$ = "X: "+STR$(x#)+" Y: "+STR$(y#) PRINT "Collision velocity "
Ä+" Z: "+STR$(z#) Ä
,obj2,": ",
ENDFUNCTION result$ Ä
CollisionVelocityDetails$("B")
The DO..LOOP only needs one new line: ENDIF
ENDFUNCTION
DO
ODE UPDATE
DisplayVelocities() Activity 48.27
LOOP
By increasing the value of the force, the cube speeds up.
By changing the direction, the cube moves in another
Activity 48.25 direction. By moving the position at which the force is
applied, the cube turns about one or more axes.
The code for the function is:
FUNCTION DisplayCollisions() Activity 48.28

ODE COLLISION GET MESSAGE No solution required.
REPEAT Activity 48.29
SET CURSOR 100,100
PRINT "Objects collided: ",obj1, On the first run, the dice slides off to the right.
Ä
" ",obj2
UNTIL INKEY$() <> "" On the second run, with lines 2 and 3 of ConveyerBelt()
ENDIF
ENDFUNCTION
included, the cube slides off to the left.
The DO..LOOP should now be: On the third run, with lines 4 and 5 of ConveyerBelt()
included, the cube slides off into the screen.
DO
ODE UPDATE On the last run, with lines 2, 3, 4 and 5 of ConveyerBelt()
DisplayCollisions() included, the cube reverts to sliding off to the left.
LOOP
Activity 48.26
Again, we start by adding a helper function to perform all
the work
FUNCTION CollisionVelocityDetails$(op$)
IF op$ = "A"
x# = ODE GET OBJECT A VELOCITY X()
y# = ODE GET OBJECT A VELOCITY Y()
z# = ODE GET OBJECT A VELOCITY Z()
result$ = "X: "+STR$(x#)+" Y: "
Ä
+STR$(y#)+" Z: "+STR$(z#)
ELSE
x# = ODE GET OBJECT B VELOCITY X()
y# = ODE GET OBJECT B VELOCITY Y()
z# = ODE GET OBJECT B VELOCITY Z()
result$ = "X: "+STR$(x#)+" Y: "
Ä
+STR$(y#)+" Z: "+STR$(z#)
ENDIF
ENDFUNCTION result$
and then updating DisplayCollisions() to print the extra

details:
FUNCTION DisplayCollisions()
REPEAT
SET CURSOR 100,100
PRINT "Objects collided: ",obj1,
Ä" ",obj2
SET CURSOR 100,130
PRINT "Collision velocity "

49
3D Vectors
4D Vectors
Matrices for Shaders
DarkBASIC Pro: Vectors and Matrices 1349

3D Vectors
Introduction
Back in Chapter 25 we looked at how 2D vectors could be used when manipulating
sprites. To perform the same sort of manipulations on a 3D object we need to employ
3D vectors. Many of the operations performed on 3D vectors are identical in nature
to those employed in 2D vectors. But since the basic concepts are covered in Volume
1, which not everyone may have, we'll quickly run over the terminology.
A Mathematical Description of 3D Vectors

Whereas a single value (such as 12, or -3) is known as a scalar, the term vector is
used to denote a list of values. The values within a vector are termed the elements
of the vector and the number of elements determines the vector's dimensions. So a
vector containing three elements is a three dimensional vector. Vectors can be
shown horizontally, [12.3,4.6,1.7] and are known as row vectors, or vertically
12.3
4.6
1.7
in which case they are termed column vectors.
A vector is often assigned an identifying name, normally a lowercase letter shown

in bold. For example,
a = [12.3, 4.6, 1.7]
The individual elements within a 3D vector are named x, y and z. To refer to an

element with a named vector, we use the vector name followed by the element name
in subscript. For example,
ax = 12.3
In a graphical context, the term vector refers to a directed line that has both
magnitude and direction (see FIG-49.1).
FIG-49.1 A Visual Representation of a Vector
Visualisation of a Vector The direction is

denoted by the angle of rotation
and the arrow at the head
of the vector
The tail of
the vector
magnitude is denoted
by the length of the line The head of
the vector
The details of a vector can be stored as three numbers giving the position of the
head of the vector relative to its tail, as shown in FIG-49.2.
1350 DarkBASIC Pro: Vectors and Matrices

FIG-49.2 12.3 (x)
How a 3D Vector is 4.6 (y)

Specified
1.7 (z)
From the offset values held within the vector, its magnitude can be determined. For
vector a, its magnitude (written as ||a||) can be calculated using the formula:
||a|| = ax2 + ay2 + az2
What is a 3D Vector in DarkBASIC Pro?

A 3D vector is nothing more than a record structure capable of holding three real
numbers (see FIG-49.3). What makes it a little different is that DarkBASIC Pro
contains a great number of commands for manipulating 3D vectors.
FIG-49.3 A 3D Vector is a
data structure which stores
A 3D Vector 3 real numbers
12.3 4.6 1.7
Why do we need 3D Vectors?

A 3D vector is simply a convenient storage area for the coordinates of a point in 3D
space (see FIG-49.4).
FIG-49.4
The coordinates of
Storing a Point's a point in 3D space...
Coordinates in a 3D
Vector
12.3 4.6 1.7
...can be stored in
a 3D vector
It can also be used to store an object's velocity.
3D Vector Statements
The MAKE VECTOR3 Statement
To create a 3D vector we need to use the MAKE VECTOR3 statement which has
FIG-49.5
The MAKE VECTOR3 MAKE VECTOR3 ( vectno )
Statement
integer

In the diagram:
vectno is an integer value which is to be the ID of the

vector being created.
This statement returns 1 if the vector is created successfully; otherwise zero is

returned. Normally, we wouldn't bother checking the returned value, so we'd create
a 3D vector with a statement such as:
MAKE VECTOR3(1)
To help identify vectors within a program it's probably best, in all but the shortest
programs, to use a constant for the vector ID:
#CONSTANT myvector
MAKE VECTOR3(myvector)
The fields within a 3D vector are identified as x, y and z.
The SET VECTOR3 Statement

To assign values to the fields within a 3D vector, we use the SET VECTOR3
FIG-49.6
The SET VECTOR3 SET VECTOR3 vectno , xvalue , yvalue , zvalue
Statement
In the diagram:
vectno is an integer value giving the ID of the 3D vector

to be assigned a value.
xvalue is a real number specifying the value to be

assigned to the x field within the 3D vector.
yvalue is a real number specifying the value to be

assigned to the y field within the 3D vector.
zvalue is a real number specifying the value to be

assigned to the z field within the 3D vector.
For example, we could assign the values 12.3, 4.6 and 1.7 to myvector using the
lines:
#CONSTANT myvector
MAKE VECTOR3(myvector)
SET VECTOR3 myvector,12.3,4.6,1.7
Retrieving Data from a 3D Vector

To retrieve one of the values stored within a 3D vector, we use the VECTOR3

{
FIG-49.7
X
The VECTOR3 Statement
y VECTOR3 ( vectno )
real
In the diagram:
X,Y,Z Use the option appropriate for the value you

wish to retrieve.
vectno is an integer value giving the ID of the vector

whose data is to be accessed.
For example, using the 3D vector we set up in the example above, the line
PRINT Y VECTOR3(myvector)
would display the value 4.6.
The DELETE VECTOR3 Statement

When finished with a 3D vector, we can delete it from RAM using the DELETE
VECTOR3 statement which has the format shown in FIG-49.8.
FIG-49.8
The DELETE
DELETE VECTOR3 ( vectno )
VECTOR3 Statement
integer
In the diagram:
vectno is an integer value giving the ID of the vector to

be deleted.
If the vector is successfully deleted, this statement returns 1; otherwise zero is

returned.
The COPY VECTOR3 Statement

The contents of one 3D vector can be copied to another using the COPY VECTOR3
FIG-49.9
The COPY VECTOR3 COPY VECTOR3 destvectno , srcvectno
Statement
In the diagram:
destvectno is an integer value giving the ID of the destination

3D vector.

srcvectno is an integer value giving the ID of the source
3D vector.
For example, we could copy the contents of vector 2 to vector 1 using the line
COPY VECTOR3 1,2
The MULTIPLY VECTOR3 Statement

Each value within a 3D vector can be multiplied by a specified factor using the
MULTIPLY VECTOR3 statement which has the format shown in FIG-49.10.
FIG-49.10
The MULTIPLE MULTIPLY VECTOR3 vectno , multiplier
VECTOR3 Statement
In the diagram:
vectno is an integer value giving the ID of the 3D vector

which is to be used in the calculation.
multiplier is a real number giving the value by which the

elements of the vector are to be multiplied.
The contents of vectno are modified by this operation.
For example, 3D vector 1 could have its contents multiplied by 1.3 using the line:
MULTIPLE VECTOR3 1, 1.3
The SCALE VECTOR3 Statement

The SCALE VECTOR3 statement is similar to MULTIPLE VECTOR3 but this
time the result is stored in a different vector and the original vector left unchanged.
The format for the SCALE VECTOR3 statement is given in FIG-49.11.
FIG-49.11
The SCALE VECTOR3 SCALE VECTOR3 destvectno , srcvectno , multiplier
Statement
In the diagram:
destvectno is an integer value giving the ID of the 3D vector

in which the result is to be stored.
srcvectno is an integer value giving the ID of the vector

whose contents are to be used in the calculation.
multiplier is a real number giving the value by which the

elements of the vector are to be multiplied.
The contents of srcvectno are not altered by this operation.
We could use this statement to store the result of multiplying the contents of vector
2 by 1.3 in vector 1 using the line:
SCALE VECTOR3 1,2,1.3

The DIVIDE VECTOR3 Statement
We can divide a vector's contents by a specified amount using the DIVIDE
VECTOR3 statement whose format is shown in FIG-49.12.
FIG-49.12
The DIVIDE VECTOR3 DIVIDE VECTOR3 vectno , divisor
Statement
In the diagram:

which is to be used in the operation.
divisor is a real number giving the value by which the

contents of the vector are to be divided.
The contents of vectno are altered by this operation.
To divide the contents of 3D vector 1 by 2, we would use the line:
DIVIDE VECTOR3 1,2
Activity 49.1
Write a program (act4901.dbpro) which reads in 3 real numbers from the

keyboard and stores them in a 3D vector.
When the up-arrow key is pressed, the contents of the vector should be
increased by 10% and when the down-arrow key is pressed, the vector's
contents should be decreased by 10%. Display the vector's contents each time
it is changed.
The LENGTH VECTOR3 Statement

If a 3D vector is being used to represent direction and speed, the speed is represented
by the magnitude of the vector. That magnitude can be calculated from the details
held within a 3D vector using the LENGTH VECTOR3 statement whose format is
shown in FIG-49.13.
FIG-49.13
The LENGTH VECTOR3 LENGTH VECTOR3 ( vectno )
Statement
real
In the diagram:

whose magnitude is to be determined.
Activity 49.2
Add to your previous program so that the vector is used to set the velocity of a
cube. The speed of the cube should be displayed.

The SQUARED LENGTH VECTOR3 Statement
While LENGTH VECTOR3 returns the magnitude of a vector, the square of the
magnitude can be discovered using the SQUARED LENGTH VECTOR3 statement
FIG-49.14
The SQUARED SQUARED LENGTH VECTOR3 ( vectno )
LENGTH VECTOR3
Statement
real
In the diagram:

for which the square of the magnitude is to be
determined.
The ADD VECTOR3 Statement

We can add two 3D vectors using the ADD VECTOR3 statement. The format for
this statement is shown in FIG-49.15.
FIG-49.15
The ADD VECTOR3 ADD VECTOR3 destvectno , vectno1 , vectno2
Statement
In the diagram:
destvectno is an integer value giving the ID of the vector

which is to contain the result of the calculation.
vectno1 is an integer value giving the ID of the first

vector involved in the calculation.
vectno2 is an integer value giving the ID of the second

For example, the statements
#CONSTANT v1 1
#CONSTANT v2 2
#CONSTANT v3 3
ADD VECTOR3 v3,v1,v2
will result in the following:
v3.x = v1.x + v2.x

v3.y = v1.y + v2.y
v3.z = v1.z + v2.z
The obvious question at this point is; Why might we want to add two vectors?
Simply, if two thrust vectors are added, it gives us the resulting thrust (see
FIG-49.16).

FIG-49.16 Why we need to Add Vectors
A A A
B
A spaceship has fired its main engines It now uses its thrusters to rotate about It now fires its main engines again
giving it velocity A. its own axes, but continues to travel creating a thrust represented by B.
along the previous trajectory.
A
B B B
C
A A
To calculate the ship's new velocity This can be achieved visually by A line drawn from the start of A to the
we need to add velocities A and B. moving the start of velocity B to the end end of B represents the ship's new
of velocity A... velocity, C.
The SUBTRACT VECTOR3 Statement

We can subtract the contents of one vector from another and store the result in a
third using the SUBTRACT VECTOR3 statement which has the format shown in
FIG-49.17.
FIG-49.17
The SUBTRACT
SUBTRACT VECTOR3 destvectno , vectno1 , vectno2
VECTOR3 Statement
In the diagram:



If we assume vectors a and b contain coordinates of two points rather than velocities,
then the SUBTRACT VECTOR3 statement can be used to calculate the position of
point b relative to point a by subtracting vector a from vector b.
The DOT PRODUCT VECTOR3 Statement

When two vectors, a and b, are multiplied, this is written as a.b, the dot being used

to indicate multiplication. The calculation performed by this operation is:
axbx + ayby + azbz
We can perform this operation in DarkBASIC Pro using the DOT PRODUCT
FIG-49.18
The DOT PRODUCT DOT PRODUCT VECTOR3 ( vectno1 , vectno2 )
VECTOR3 Statement
real
In the diagram:
vectno1,vectno2 are integer values giving the IDs of the vectors

involved in the multiplication.
This operation is used when we need to discover the angle between two vectors (see
FIG-49.19).
FIG-49.19 b
The Angle Between

Two Vectors
θ a
The angle θ (theta) between vectors a and b can be calculated using the formula:
θ = acos
( a.b
||a|| ||b||
The NORMALIZE VECTOR3 Statement

The formula for calculating the angle between two vectors becomes simpler if both
vectors a and b have a magnitude of 1, the formula then reducing to:
θ = ab
acos( . )
It is possible to reduce a vector's magnitude to 1 without affecting its direction using

the NORMALIZE VECTOR3 statement which has the format shown in FIG-49.20.
FIG-49.20
The NORMALIZE NORMALIZE VECTOR3 destvectno , srcvectno
VECTOR3 Statement
In the diagram:
destvectno is an integer value giving the ID of the destination

3D vector.
srcvectno is an integer value giving the ID of the source

3D vector.

The IS EQUAL VECTOR3 Statement
We can check if two 3D vectors contain identical values using the IS EQUAL
FIG-49.21
The IS EQUAL
IS EQUAL VECTOR3 ( vectno1 , vectno2 )
VECTOR3 Statement
integer
In the diagram:
vectno1, vectno2 are integer values giving the IDs of the two
vectors being compared.
If the contents of the two vectors are identical, the statement returns 1; otherwise
zero is returned.
The MAXIMIZE VECTOR3 Statement

We can construct a vector using the maximum values from two existing vectors
with the MAXIMIZE VECTOR3 statement. The concept behind the statement is
demonstrated visually in FIG-49.22.
FIG-49.22
Vector 1 4.6 5.1 -4.6
How MAXIMIZE y
x z
VECTOR3 Operates
Vector 2 7.9 1.7 -1.5
x y z
Max Max Max

X Y Z
Resulting Vector 7.9 5.1 -1.5

x y z
The format for the MAXIMIZE VECTOR3 statement is given in FIG-49.23.

FIG-49.23
The MAXIMIZE MAXIMIZE VECTOR3 destvectno , vectno1 , vectno2
VECTOR3 Statement
In the diagram:




The MINIMIZE VECTOR3 Statement
A new vector constructed from the minimum values from two other vectors can be
constructed using the MINIMIZE VECTOR3 statement whose format is given in
FIG-49.24.
FIG-49.24
The MINIMIZE
MINIMIZE VECTOR3 destvectno , vectno1 , vectno2
VECTOR3 Statement
In the diagram:



The CROSS PRODUCT VECTOR3 Statement

We can determine the expression for a vector which is at right-angles to two existing
3D vectors using the CROSS PRODUCT VECTOR3 statement which has the
FIG-49.25 The CROSS PRODUCT VECTOR3 Statement
CROSS PRODUCT VECTOR3 destvectno , vectno1 , vectno2
In the diagram:



For example, using the vectors [0,0,100] and [0,10,0] (which are parallel to the z
and y axes respectively) the CROSS PRODUCT VECTOR3 statement would create
a new vector at right angles to these - that is a vector parallel to the x-axis. This is
demonstrated in LISTING-49.1.
LISTING-49.1
Creating a Cross Product r = MAKE VECTOR3(1)
Vector SET VECTOR3 1, 0,0,100
r = MAKE VECTOR3(2)
SET VECTOR3 2, 0,10,0
r = MAKE VECTOR3(3)
CROSS PRODUCT VECTOR3 3,1,2
x# = X VECTOR3(3)

LISTING-49.1 y# = Y VECTOR3(3)
(continued) z# = Z VECTOR3(3)
PRINT "VECTOR 3", x#," ",y#," ",z#
Creating a Cross Product WAIT KEY
Vector END
Activity 49.3
Type in and test the program given in LISTING-49.1 (vector01.dbpro).
Summary
l 3D vectors are structures which hold 3 real numbers.
l The fields within a vector object are identified as x, y, and z.
l A 3D vector can be used to store the coordinates of a point in 3D space or the

velocity of an object in 3D space.
l Use MAKE VECTOR3 to create a 3D vector object.
l Use SET VECTOR3 to assign values to a 3D vector.
l Use X VECTOR3 to retrieve the x value within a 3D vector. Use Y VECTOR3

and Z VECTOR3 to retrieve the other values.
l Use DELETE VECTOR3 to delete a 3D vector from RAM.
l Use COPY VECTOR3 to copy the contents of one vector into another.
l Use MULTIPLY VECTOR3 to multiply each element within a vector by a fixed

value. The result is stored in the named vector.
l Use SCALE VECTOR3 to multiply each element within a vector by a fixed

amount and store the result in a second vector.
l Use DIVIDE VECTOR3 to divide the elements of a vector by a fixed value. The
result is stored in the named vector.
l Use LENGTH VECTOR3 to calculate the magnitude of a velocity vector. This

is only appropriate when the vector is used to store a velocity rather than a point's
coordinates.
l Use SQUARED LENGTH VECTOR3 to calculate the square of the magnitude

of a velocity.
l Use ADD VECTOR3 to add the contents of two vectors and store the result in
a third vector.
l Use SUBTRACT VECTOR3 to subtract the contents of one vector from another
and store the result in a third vector.
l Use DOT PRODUCT VECTOR3 to calculate the dot product of two vectors.
l Use NORMALIZE VECTOR3 to calculate a vector with a magnitude of 1 but

pointing in the same direction as an existing vector.

l Use IS EQUAL VECTOR3 to check if two vectors are exactly equal.
l Use MAXIMIZE VECTOR3 to create a new vector from the maximum values
in two other vectors.
l Use MINIMIZE VECTOR3 to create a new vector from the minimum values in
two other vectors.
l Use CROSS PRODUCT VECTOR3 to create a vector which is at right angles

to two existing vectors.

4D Vectors
Introduction
Sorry, but DarkBASIC Pro doesn't allow you to create 4-dimensional objects -
despite the title of this section. But we do need vectors which contain 4 elements.
These vectors are used to store 3D coordinates or vectors, but a fourth element (w)
is added (see FIG-49.26) because some of the mathematics we need to perform
demand 4 element vectors, not 3.
FIG-49.26
A 4D Vector A 4D Vector is a
data structure which stores
4 real numbers
12.3 4.6 1.7 1
x y z w
The new element within the vector is known as w. We are free to assign any value
we wish to w.
To convert the contents of a 3D vector, a, to the equivalent 4D in vector b, the

following formula is used:
bx = ax * w
by = ay * w
bz = az * w
bw = w
To convert from 4D to 3D the formula is:
ax = bx / w
ay = by / w
az = bz / w
These calculations become trivial if w is set to 1.
Coordinates stored in this fashion, with an extra w component, are known as

homogeneous coordinates.
Activity 49.4
Convert the point (30,9,-12) to homogeneous coordinates in a 4D vector (w=1).
Convert the homogeneous vector [12, 3, -18, 2] to the equivalent 3D vector.
Many of the commands available for 4D vectors perform the same operations as
those already covered for 2 and 3 dimensional vectors, so rather than describe each
one separately they are listed below in TABLE-49.1.

TABLE-49.1 4D Vector Statements
ADD VECTOR4 VR, V1, V2 VR set to V1 + V2

COPY VECTOR4 VR, V VR set to V
DELETE VECTOR4 V num(0 or 1) deletes V returns 1 if OK
DIVIDE VECTOR4 VR, num divides VR by num
IS EQUAL VECTOR4 V1, V2 num (0 or 1) returns 1 if V1 = V2
LENGTH VECTOR4 V len len set magntude of V
MAKE VECTOR4 V num (0 or 1) creates V; returns 1 if OK
MAXIMIZE VECTOR4 VR, V1, V2 VR set to max values from V1 and V2
MINIMIZE VECTOR4 VR, V1, V2 VR set to min values from V1 and V2
MULTIPLY VECTOR4 VR, num multiples VR by num
NORMALIZE VECTOR4 VR,V VR set to normalised version of V
SCALE VECTOR4 VR, V, num VR set to V*num
SET VECTOR4 VR, xn, yn, zn, wn VR set to [xn,yn,zn,wn]
SQUARED LENGTH VECTOR4 V len len set to magnitude of V squared
SUBTRACT VECTOR4 VR, V1, V2 VR set to V1 - V2
W VECTOR4 V num num set to Vw
X VECTOR4 V num num set to Vx
Y VECTOR4 V num num set to Vy
Z VECTOR4 V num num set to Vz

Matrices
Introduction
Don't confuse the term matrix as it is used here with the matrix grids we created
back in Chapter 45. Here we use the term matrix to mean a grid of values - just like
a two-dimensional array. An example of a matrix is shown below:
[ 4
7
10
-42
8
The matrix in the example is a 2 by 3 matrix (2 rows by 3 columns).
Like vectors, matrices are often assigned names, usually shown in uppercase bold:
M=
[ 4
7
10
-42
8
Just like arrays, the individual elements within a matrix are identified by their row
and column positions. Some texts start subscripting from 1, hence the top-left
element is referred to as 1,1; others start subscripts at 0 and the top-left element is
then 0,0. We will subscript from 0,0 since DarkBASIC Pro arrays begin
subscripting at zero.
M=
[ m00
m10
m01
m11
m02
m12
DarkBASIC Pro has statements for creating and manipulating only 4 by 4 matrices.
The matrices in DarkBASIC Pro are really intended for use with shaders (see
Chapter 50) and hence have very limited accessibility. They can be created and
manipulated but are not designed to be examined in any way.
Matrix Statements
The MAKE MATRIX4 Statement
To create a 4 by 4 matrix we use the MAKE MATRIX4 statement which has the
FIG-49.27
MAKE MATRIX4 ( matno )
The MAKE MATRIX4
Statement
integer
In the diagram:
matno is an integer value giving the ID to be assigned

to the matrix being created.

The statement returns 1 if the matrix was created successfully; otherwise zero is
returned.
You could be forgiven for thinking you can now set up any values you want within
the matrix you've just created, but, in fact, DarkBASIC Pro matrix objects are
designed to hold only specific values which are loaded into the matrix using
tailor-made statements.
The SET IDENTITY MATRIX4 Statement

When dealing with scalar values, the number 1 is a rather important value; it is the
only number that has no effect when we multiply or divide by it. A similar value
exists for matrices and is called the identity matrix. The 4 by 4 identity matrix is:
[
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
Notice the diagonal line of 1s - this is the sign of an identity matrix, no matter what
size the matrix is, although it must be square.
We can set the contents of a matrix to the identity matrix using the SET IDENTITY
MATRIX4 statement which has the format shown in FIG-49.28.
FIG-49.28
The SET IDENTITY SET IDENTITY MATRIX4 matno
MATRIX4 Statement
In the diagram:
matno is an integer value specifying the matrix to be

set to the identity matrix value.
A typical statement would be:
SET IDENTITY MATRIX4 1
The IS IDENTITY MATRIX4 Statement

We can check to find out if a matrix contains the identity matrix using the IS
IDENTITY MATRIX4 statement whose format is shown in FIG-49.29.
FIG-49.29
IS IDENTITY MATRIX4 ( matno )
The IS IDENTITY
MATRIX4 Statement
integer
In the diagram:
matno is an integer value specifying the matrix to be

tested.

Other Matrix Assignment Statements
As a 3D object's details are sent to your graphics card, it is possible to pick up the
coordinates of each polygon and store these in a 4D matrix. The coordinates can be
accessed in various formats: world coordinates (the position of vertices in 3D
space), view coordinates (coordinates within the viewport), or projection
coordinates (coordinates on screen). A statement exists to load each of these options
(see TABLE-49.2).
TABLE-49.2
Other Matrix
Assignment Statements PROJECTION MATRIX4 MR MR set to projection values
VIEW MATRIX4 MR MR set to view values
WORLD MATRIX4 MR MR set to world values
For example, we could load the world coordinates of a polygon into matrix 1using
the line:
WORLD MATRIX4 1
The COPY MATRIX4 Statement

The contents of one matrix can be copied to another using the COPY MATRIX4
FIG-49.30
The COPY MATRIX4 COPY MATRIX4 destmatno , srcmatno
Statement
In the diagram:
destmatno is an integer value giving the ID of the destination

4D matrix.
srcmatno is an integer value giving the ID of the source

4D matrix.
The IS EQUAL MATRIX4 Statement

To check if the contents of two matrices are exactly equal, we can use the IS EQUAL
FIG-49.31
The IS EQUAL IS EQUAL MATRIX4 ( matno1 , matno2 )
MATRIX4 Statement
integer
In the diagram:
matno1, matno2 are integer values giving the IDs of the two
matrices being compared.
If the contents of the two matrices are identical, the statement returns 1; otherwise
zero is returned.

The ADD MATRIX4 Statement
The contents of two 4 by 4 matrices can be added using the ADD MATRIX4
FIG-49.32
The ADD MATRIX4 ADD MATRIX4 destmatno , matno1 , matno2
Statement
In the diagram:
destmatno is an integer value giving the ID of the matrix

matno1 is an integer value giving the ID of the first

matrix involved in the calculation.
matno2 is an integer value giving the ID of the second

Two 4 by 4 matrices are added using the formula shown below:
[ [ [
a00 a01 a02 a03 b00 b01 b02 b03 a00+b00 a01+b01 a02+b02 a03+b03
a10 a11 a12 a13 b10 b11 b12 b13 a10+b10 a11+b11 a 12+b12 a13+b13
A= B= A +B =
a20 a21 a22 a23 a20+b20 a21+b21 a22+b22 a23+b23
b20 b21 b22 b23
a30 a31 a32 a33 a30+b30 a31+b31 a32+b32 a33+b33
b30 b31 b32 b33
The SUBTRACT MATRIX4 Statement

The contents of two 4 by 4 matrices can be subtracted using the SUBTRACT
FIG-49.33
The SUBTRACT SUBTRACT MATRIX4 destmatno , matno1 , matno2
MATRIX4 Statement
In the diagram:



The DIVIDE MATRIX4 Statement

We can divide the contents of a matrix by a scalar value using the DIVIDE
MATRIX4 statement whose format is shown in FIG-49.34.
FIG-49.34
The DIVIDE MATRIX4
Statement
DIVIDE MATRIX4 destmatno , divisor

In the diagram:

divisor is a real value. Each element within the matrix

will have its original value divided by this amount.
The MULTIPLY MATRIX4 Statement

We can multiply a matrix by a scalar value or, alternatively, multiply two matrices
using the MULTIPLY MATRIX4 statement whose format is shown in FIG-49.35.
{
FIG-49.35
, matno1 , matno2
The MULTIPLE MULTIPLY MATRIX4 destmatno
MATRIX4 Statement
, multiplier
In the diagram:

Option 1:


The contents of matno1 will be multiplied by

the contents of matno2 and the result stored in
destmatno.
Option 2:
multiplier is a real number.
Every element in destmatno will be multiplied by

multiplier.
Although multiplication by a scalar value is quite straight-forward, multiplying two

matrices is much more complex. FIG-49.36 shows the principle involved in the
FIG-49.36 operation.
Multiplying Matrices
Arow 0 * B col 0 Arow 0 * B col 1
a00*b00 + a01*b10 + a02*b20 + a03*b30
[ [ [
a00 a01 a02 a03 b00 b01 b02 b03 c00 c01 c02 c03
a10 a11 a12 a13 b10 b11 b12 b13 c10 c11 c12 c13
A= B= C =
a20 a21 a22 a23 c20 c21 c22 c23
b20 b21 b22 b23
a30 a31 a32 a33 c30 c31 c32 c33
b30 b31 b32 b33

Activity 49.5
What row and column would be used to calculate the value at position c21?
The INVERSE MATRIX4 Statement

The inverse of a scalar value is simple enough to derive; just divide it into 1. For
example, the inverse of 2 is 1/2. Another property of an inverse is that when
multiplied by the original value, we always get a result of 1 ( 2 * 1/2 = 1).
Calculating the inverse of a given matrix is more complex, but when that inverse is
multiplied by the original matrix, the result is the identity matrix. Just to make life
a little more complicated, not every matrix has an inverse!
We can attempt to calculate the inverse of a specified matrix using the INVERSE
MATRIX4 statement whose format is shown in FIG-49.37.
FIG-49.37
The INVERSE
INVERSE MATRIX4 ( matno1 , matno2 )
MATRIX4 Statement
integer
In the diagram:
matno1 is an integer value specifying the ID of the matrix

in which the result is to be stored.
matno2 is an integer value specifying the ID of the matrix

whose contents are used in calculating the inverse
matrix.
The statement returns 1 if an inverse matrix is successfully calculated; otherwise

zero is returned.
The SCALE MATRIX4 Statement

Remembering that the contents of a matrix will be a triangle from a 3D object, it is
possible to expand that triangle (increase its width, height and depth), using the
SCALE MATRIX4 statement which has the format shown in FIG-49.38.
FIG-49.38
The SCALE MATRIX4 SCALE MATRIX4 matno , x , y , z
Statement
In the diagram:
matno is an integer value specifying the ID of the matrix

to be scaled.
x,y,z are real values specifying the scaling factor for

width (x), height (y) and depth (z).
For example, if matrix 1 held the coordinates of a triangle, we could double its size
in all dimensions using the line:

SCALE MATRIX 1, 2.0,2.0,2.0
The TRANSLATE MATRIX4 Statement

Rather than scale a triangle, we can move it using the TRANSLATE MATRIX4
FIG-49.39
The TRANSLATE TRANSLATE MATRIX4 vectno , x , y , z
MATRIX4 Statement
In the diagram:

to be translated.
x,y,z are real values specifying the distance the

coordinates are to be shifted in each dimension.
The ROTATE MATRIX4 Statement

The final effect we can apply to the triangle's coordinates is rotation. We can rotate
the vertices recorded in the matrix about any axis using the ROTATE MATRIX4
FIG-49.40
{{
The ROTATE
X
MATRIX4 Statement
y MATRIX4 matno , angle
ROTATE
z
YPR MATRIX4 , yaw , pitch , roll
In the diagram:
Option 1:
X,Y,Z Use the option corresponding to the axis about

which the rotation is required.

to be rotated.
angle is a real value specifying the angle of rotation.
Option 2:
YPR Use this option to rotate about more than one axis.
yaw, pitch, roll are the angles of rotation about the y, x and z
axes respectively (note the order!).

The TRANSPOSE MATRIX4 Statement
A transposed matrix is one in which the values making up each row are switched
to make up the columns instead. A matrix A and its transform AT , are shown below:
[ [
1 2 3 4 1 5 9 13
5 6 7 8 2 6 10 14
A= AT =
9 10 11 12 3 7 11 15
13 14 15 16 4 8 12 16
To transform a matrix we use the statement TRANSFORM MATRIX4 whose

format is shown in FIG-49.41.
FIG-49.41
The TRANSPOSE TRANSPOSE MATRIX4 destvectno , matno
MATRIX4 Statement
In the diagram:

which is to contain the result of the calculation

containing the original version of the matrix data.
The DELETE MATRIX4 Statement

A matrix can be deleted from RAM using the DELETE MATRIX4 statement whose
format is given in FIG-49.42.
FIG-49.42
The DELETE MATRIX4 DELETE MATRIX4 ( matno )
Statement
integer
In the diagram:

to be deleted.
The statement returns 1 if the matrix is successfully deleted, otherwise zero is

returned.
As stated already, the sole purpose of the matrix data structure is to be passed to a
shader to modify the overall appearance of the 3D objects displayed on the screen.
Summary
l A mathematical matrix is a grid of values.
l DarkBASIC Pro has statements for creating 4 by 4 matrices only.

l The identity matrix has a diagonal row of 1's from top-left to bottom-right.
l The matrices in DarkBASIC Pro are intended to be used as parameters for

shaders.
l Use MAKE MATRIX4 to create a matrix object.
l Use SET IDENTITY MATRIX4 to set the contents of a matrix to the identity
matrix.
l Use IS IDENTITY MATRIX4 to determine if a matrix is an identity matrix.
l Use PROJECTION MATRIX4, VIEW MATRIX4 or WORLD MATRIX4 to

assign a set of transformed coordinates from a polygon to a matrix.
l Use COPY MATRIX4 to copy the contents of one matrix to another.
l Use IS EQUAL MATRIX4 to check if two matrices have identical contents.
l Use ADD MATRIX4 to add two matrices and store the result in a third matrix.
l Use SUBTRACT MATRIX4 to subtract one matrix from another and store the
result in a third matrix.
l Use DIVIDE MATRIX4 to divide a matrix by a specified value.
l Use MULTIPLY MATRIX4 to multiply two matrices and store the result in a
third matrix or to multiply a matrix by a specified value.
l Use INVERSE MATRIX4 to create the inverse of an existing matrix.
l Use SCALE MATRIX4 to scale the coordinate values held within a matrix.
l Use TRANSLATE MATRIX4 to add to the coordinate values within a matrix.
l Use ROTATE MATRIX4 to rotate the coordinate values within a matrix.
l Use TRANSPOSE MATRIX4 to switch the rows and columns within a matrix.
l Use DELETE MATRIX4 to delete a matrix from RAM.

Solutions Activity 49.3
Activity 49.1
GLOBAL v1#,v2#,v3# No solution required.
SetUpScreen()
Read3Values()
REM *** Set up Vector *** Activity 49.4
r = MAKE VECTOR3(1)
SET VECTOR3 1,v1#,v2#,v3# [30, 9, -12, 1]
DO
IF UPKEY() [6, 1.5, -9]
ModifyVector(1,10)
ENDIF
IF DOWNKEY() Activity 49.5
ModifyVector(1,-10)
ENDIF C21 = Arow 2 * Bcol 1
DisplayVector(1)
WAIT 200
LOOP
WAIT KEY
END
ENDFUNCTION
FUNCTION Read3Values()
INPUT "Enter first value : ", v1#
INPUT "Enter second value : ", v2#
INPUT "Enter third value : ", v3#
ENDFUNCTION
FUNCTION DisplayVector(vector)
x# = X VECTOR3(vector)
y# = Y VECTOR3(vector)
z# = Z VECTOR3(vector)
PRINT "VECTOR ",vector," : ", x#,
Ä
" ",y#," ",z#
ENDFUNCTION
FUNCTION ModifyVector(vector, perc#)

multiplier# = 1 + perc#/100
MULTIPLY VECTOR3 vector, multiplier#
ENDFUNCTION
Activity 49.2
The main section of the program should be
changed to:
GLOBAL v1#,v2#,v3#
SetUpScreen()
Read3Values()
REM *** Set up Vector ***
r = MAKE VECTOR3(1)
SET VECTOR3 1,v1#,v2#,v3#
POINT OBJECT 1, X VECTOR3(1),
ÄY VECTOR3(1),Z VECTOR3(1)
DO
IF UPKEY()
ModifyVector(1,10)
ENDIF
IF DOWNKEY()
ModifyVector(1,-10)
ENDIF
speed# = LENGTH VECTOR3(1)
MOVE OBJECT 1,speed#
SET CURSOR 100,100
PRINT "Speed : ",speed#
LOOP
WAIT KEY
END

50
Checking Your Video Card Capabilities
Pixel Shader Statements
Using FX Files
Vertex Shader Statements
DarkBASIC Pro: Shaders 1375

Shaders and FX Files
Introduction
Although DarkBASIC Pro allows us to create some spectacular effects with just a
single command, more complex manipulation of an object's texture, lighting or
shape needs to be done using a shader. A shader contains a set of instructions on
how an object is to be manipulated. By using shader files, we can create
photo-realistic effects that cannot be achieved by any other method. However,
Your graphics card
should support pixel
shaders can require a large amount of processing power and are dependent upon
shader Version 2 or compliant graphics cards.
above.
There are three stages required to use a shader:
Ø Create the shader

Ø Load the file into your program
Ø Apply the shader to the object
Shader files are written in a C-like language. Direct X uses the language HLSL
(High-Level Shader Language) to create shaders, but other graphic-card-specific
languages are also used. The compiler for HLSL can be downloaded from the
Microsoft website. To use it you'll also need Visual Studio C++.NET.
Creation of a shader is beyond the scope of this text since it involves learning a
whole new language, but DarkBASIC Pro has commands to load and apply these
files.
There are two main types of shader files:
Ø vertex shaders
Ø pixel shaders
Several shaders can also be placed in a special effects file (FX file) so that their
effects can be combined.
Vertex Shader
Graphical data sent to your graphics card goes through a great deal of processing
before the final image appears on the screen. One of the earliest stages is the vertex
shader which is a set of commands that manipulates the vertices of the the polygons
that make up a 3D shape. By manipulating these vertices, a 3D object can be
transformed, deformed or morphed. Lighting and texturing can also be altered
within a vertex shader.
Pixel Shader
Once the 3D objects which make up the visible part of your world have been
remapped onto a 2D plane for display on your monitor, a pixel shader can modify
the colour of each individual pixel on the screen. This allows further lighting and
texturing effects to be applied as well as other options such as bump mapping. As
1376 DarkBASIC Pro: Shaders

you can imagine this requires a great deal of processing which has to be repeated
every time the screen is refreshed.
FX Files
An FX file is a Microsoft DirectX effects file and it not only contains both vertex
and pixel shaders, but also can define constants, parameters and texture files used
by those shaders.
Graphics Card Check Statements

The GET MAXIMUM VERTEX SHADER VERSION Statement
One of the most important things we need to discover is exactly which version of
the pixel shader and vertex shader your graphics card supports. The first of these
can be determined using the GET MAXIMUM VERTEX SHADER VERSION
FIG-50.1
The GET MAXIMUM GET MAXIMUM VERTEX SHADER VERSION ( )
VERTEX SHADER
VERSION Statement
real
The statement returns a real number giving the version number of the vertex shader
used by your graphics card. We can display that version number using the line:
PRINT GET MAXIMUM VERTEX SHADER VERSION()
The GET MAXIMUM PIXEL SHADER VERSION Statement

The current pixel shader version can be determined using the GET MAXIMUM
PIXEL SHADER VERSION statement which has the format shown in FIG-50.2.
FIG-50.2
The GET MAXIMUM GET MAXIMUM PIXEL SHADER VERSION ( )
PIXEL SHADER
VERSION Statement
real
The statement returns a real number giving the version number of the pixel shader
used by your graphics card. To display the version number use the line:
PRINT GET MAXIMUM PIXEL SHADER VERSION()
Activity 50.1
Write a short program (act5001.dbpro) to display the vertex and pixel shaders'
version numbers available on your graphics card.

If you do not have version 2 or above for both shaders, then there may be problems
trying to execute an FX file.
FX Statements
The easiest way to get started with shaders is to make use of FX files which already
contain all the necessary details to execute vertex and pixel shaders.
The LOAD EFFECT Statement

To load an FX file, we use the LOAD EFFECT statement which has the format
shown in FIG-50.3.
FIG-50.3
The LOAD EFFECT LOAD EFFECT filename , effno , textureflag
Statement
In the diagram:
filename is a string giving the name of the FX file to be

loaded. This must be a DirectX compatible shader
which would normally have the file extension .fx.
effno is an integer value giving the ID to be assigned to

the effect once loaded from the file.
textureflag 0 or 1. If set to zero, the object to which the shader

is to be applied continues to use its own textures
(as specified earlier in the program). If set to 1,
the object will be textured using images detailed
in the shader.
For example, we could load the effects file bubble.fx and use the textures it defines
with the line:
LOAD EFFECT "bubble.fx",1,1
The EFFECT EXIST Statement

There can be problems loading an effects file. Other than the obvious file not found,
the contents of a file may be invalid (like any program). We can check that the
effects file has been successfully loaded using the EFFECT EXIST statement which
FIG-50.4
The EFFECT EXIST EFFECT EXIST ( effno )
Statement
integer
In the diagram:
effno is an integer value giving the ID of the effects object

to be checked.

The statement returns 1 if a valid effects object of the ID specified has be created;
The PERFORM CHECKLIST FOR EFFECT ERRORS Statement

If an error does occur when loading an effects file, a detailed list of the problems
encountered can be produced using the PERFORM CHECKLIST FOR EFFECT
FIG-50.5 ERRORS statement which has the format shown in FIG-50.5.
The PERFORM
CHECKLIST FOR
PERFORM CHECKLIST FOR EFFECT ERRORS effno
EFFECT ERRORS
Statement
After this statement has been executed, we need to use the standard checklist
statements to access the contents of the list.
A typical check for a successful load would be:

IF EFFECT EXIST(1) = 0
PERFORM CHECKLIST FOR EFFECT ERRORS
NEXT c
WAIT KEY
END
ENDIF
The SET OBJECT EFFECT Statement

With the effect object created, it can then be applied to an object using the SET
OBJECT EFFECT statement which takes the format shown in FIG-50.6.
FIG-50.6
The SET OBJECT SET OBJECT EFFECT objno , effno
EFFECT Statement
In the diagram:

to which the effect is to be applied.
effno is an integer value giving the ID of the effect.
The program in LISTING-50.1 demonstrates the use of an effects file applied to a

sphere.
Demonstrating an Effect REM *** Create texture object ***
File MAKE OBJECT SPHERE 1, 10
LOAD IMAGE "lava1.bmp",1
TEXTURE OBJECT 1,1
REM *** Load effect ***


LISTING-50.1 FOR c = 1 TO CHECKLIST QUANTITY()
(continued) PRINT CHECKLIST STRING$(c)
NEXT c
Demonstrating an Effect WAIT KEY
File END
ENDIF
REM *** Apply effect after key press ***

WAIT KEY
SET OBJECT EFFECT 1,1

WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION
Activity 50.2
Type in and test the program in LISTING-50.1 (shader01.dbpro).
The SET EFFECT ON Statement

The tasks performed by the LOAD EFFECT and SET OBJECT EFFECT statements
are combined in the SET EFFECT ON statement which has the format shown in
FIG-50.7.
FIG-50.7
SET EFFECT ON objno , filename , textureflag
The SET EFFECT ON
Statement
In the diagram:

to which the effect is to be applied.

the effect to be applied.
textureflag 0 or 1. If set to zero, the object to which the shader

is to be applied continues to use its own textures
(as specified earlier in the program). If set to 1,
the object will be textured using images detailed
in the shader.
Activity 50.3
Modify your last program to use the SET EFFECT ON statement in place of
LOAD EFFECT and SET OBJECT EFFECT.
Although the SET EFFECT ON statement might appear to be the easier method of
applying an effect it is, in fact, more inefficient since a new copy of the effect needs
to be loaded for each object to which it is applied, whereas using LOAD EFFECT

and SET OBJECT EFFECT applies the same copy of the effect to each object.
The DELETE EFFECT Statement

A loaded effect can be deleted using the DELETE EFFECT statement. Obviously,
an effect should not be deleted while any objects to which it has been applied still
exist. The format of the DELETE EFFECT statement is shown in FIG-50.8.
FIG-50.8
The DELETE EFFECT DELETE EFFECT effno
Statement
In the diagram:
effno is an integer value giving the ID of the effect to be

deleted.
The SET LIMB EFFECT Statement

Not only can we apply an effect to an object, but we can also apply an effect to an
individual limb within a model using the SET LIMB EFFECT statement which has
FIG-50.9
The SET LIMB EFFECT SET LIMB EFFECT objno , limbno , effno
Statement
In the diagram:
objno is an integer value giving the ID of the limb's root

object.
limbno is an integer value giving the ID of the limb.
effno is an integer value giving the ID of the effect to be

applied.
The program in LISTING-50.2 loads the ladder model we created in Chapter 35

and applies bubble.fx to one of its limbs.
REM *** Create texture object ***
Applying an Effect to a LOAD OBJECT "ladder.dbo", 1
Limb REM *** Apply effect after key press ***
WAIT KEY
SET LIMB EFFECT 1,2,1
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION

Activity 50.4
Type in and test the program in LISTING-50.2 (effects02.dbpro).
The PERFORM CHECKLIST FOR EFFECT VALUES Statement

Some effects can be supplied with values in much the same way that a parameter
can be passed to a function. We can discover names of any values which can be
supplied to the effect using the PERFORM CHECKLIST FOR EFFECT VALUES
FIG-50.10
The PERFORM PERFORM CHECKLIST FOR EFFECT VALUES effno
CHECKLIST FOR
EFFECT VALUES
Statement We can then list the names of these values with code such as:
PERFORM CHECKLIST FOR EFFECT VALUES

NEXT c
Activity 50.5
Write a program (act5005.dbpro) to display the values that may be set in the
bubble.fx effects file.
Actually, we need more information about these named values within the FX file.
Specifically, we need to know each value's type (integer, real, etc.) and if it can have
its value changed from within a DarkBASIC Pro program.
By using CHECKLIST VALUE A we can discover a value's type. The statement

returns a coded integer value. The code and the type it represents are shown in
TABLE-50.1.
TABLE 50.1 Code Data Type
CHECKLIST VALUE A 0 undefined

Return Codes 1 Boolean
2 integer
3 float
4 vector
5 matrix
For example, if CHECKLIST STRING$(6) = ticks and CHECKLIST VALUE

A(6) = 3, then we know that variable ticks is a float (real) value.
Not all variables can actually be modified. Those that can, have their CHECKLIST
VALUE B set to 1. And even this is not an end of the limitations; DarkBASIC Pro
does not contain a statement which will allow us to assign a new value to values of
an undefined type.
Activity 50.6
Modify your last program to display only those values which are not of an
undefined type and can be modified.

The SET EFFECT CONSTANT Statement
To modify one of the parameters of an FX file, we need to use the SET EFFECT
CONSTANT statement whose format is shown in FIG-50.11.
FIG-50.11 The SET EFFECT CONSTANT Statement
{
BOOLEAN
INTEGER
SET EFFECT CONSTANT FLOAT effno , string , value
VECTOR
MATRIX
In the diagram:
BOOLEAN,INTEGER,FLOAT,VECTOR,MATRIX
Use the option appropriate to the type of value
being assigned.
effno is an integer value giving the ID of the effect to

which a value is to be sent.
string is a string giving the name of the parameter which

is to have its value set.
value is the value to be assigned to the parameter. This

value's type must match the option chosen
(Boolean, integer, float, etc.).
For example, the bubbles.fx file has a float parameter named ticks. This could have
Actually, it's not very its value set to 8.5 using the lines:
smart to change ticks
since it represents the LOAD EFFECT "bubble.fx",1,0
time, but the instruction SET EFFECT CONSTANT FLOAT 1,"ticks",8.5
itself is valid.
The SET EFFECT TECHNIQUE Statement

Some FX files contain a selection of labelled techniques (we can see these if we
look at the contents of the FX file). We can choose to execute a specific technique
from the file using the SET EFFECT TECHNIQUE statement which has the format
shown in FIG-50.12.
FIG-50.12
The SET EFFECT SET EFFECT TECHNIQUE effno , name
TECHNIQUE Statement
In the diagram:
effno is an integer value giving the ID of the effect in

which a technique is to be specified.

name is a string giving the name of the technique within
the FX file which is to be used.
Often an FX file will contain multiple techniques with one suited to

high-performance graphics cards and another for more run-of-the-mill cards. Using
this command we can select the technique we wish to have executed in our program.
The SET EFFECT TRANSPOSE Statement

Some shaders expect the matrices supplied to it to be transposed, others do not. The
SET EFFECT TRANSPOSE statement can be used to automatically transpose all
matrices set as parameters to any shaders within an FX file. The format for the
statement is given in FIG-50.13.
FIG-50.13
The SET EFFECT SET EFFECT TRANSPOSE effno , flag
TRANSPOSE Statement
In the diagram:
effno is an integer value identifying the loaded effect to

which transposed matrices are to be supplied.
flag 0 or 1. Set to 1, all matrices sent to the FX file

will be transposed; set to 0, matrices remain
unchanged.
Vertex Shader Statements

Although using an FX file is probably the easiest way of creating a special effect,
it is possible to use a vertex shader on its own without it being embedded within an
FX file. This involves creating a vertex shader object loaded from a vertex shader
file.
The CREATE VERTEX SHADER FROM FILE Statement

The DarkBASIC Pro examples include a single vertex shader file named
vshader.vsh. This is a text file and can be examined using Notepad, but its unlikely
to make much sense unless you have a knowledge of HLSL.
To make use of a shader, we must load the file into a vertex shader object using the
CREATE VERTEX SHADER FROM FILE statement which has the format shown
in FIG-50.14.
FIG-50.14 The CREATE VERTEX SHADER FROM FILE Statement
CREATE VERTEX SHADER FROM FILE vshno , filename
In the diagram:
vshno is an integer value giving the ID to be assigned to

the vertex shader object being created.

the vertex shader. Vertex shader files have a .vsh
extension.
The SET VERTEX SHADER ON Statement

Once a vertex shader has been loaded, it still has to be applied to a specific 3D object
and this is done using the SET VERTEX SHADER ON statement which takes the
FIG-50.15
The SET VERTEX SET VERTEX SHADER ON vshno , objno
SHADER ON Statement
In the diagram:
vshno is an integer value giving the ID of the vertex

shader object to be applied.

to which the vertex shader is to be applied.
The SET VERTEX SHADER OFF Statement

When we no longer wish a vertex shader to be applied to an object, we can switch
off the effect using the SET VERTEX SHADER OFF statement which has the
FIG-50.16
The SET VERTEX SET VERTEX SHADER OFF objno
SHADER OFF Statement
In the diagram:

whose vertex shader is to be switched off.
The DELETE VERTEX SHADER Statement

A vertex shader can be deleted using the DELETE VERTEX SHADER statement
FIG-50.17
The DELETE VERTEX DELETE VERTEX SHADER vshno
SHADER Statement
In the diagram:
vshno is an integer value giving the ID of the vertex

shader to be deleted from RAM.
A vertex shader may require data in the form of a vector or a matrix. Such
information can be sent to the vertex shader using the following statements.

Other Vertex Shader Statements
There are a few other vertex shader statements designed to pass values of various
types to the loaded shader objects, but without detailed knowledge of the shader's
requirements these are impossible to use. The statements are:
SET VERTEX SHADER VECTOR Passes a vector to the shader.

SET VERTEX SHADER STREAM Sets a vertex shader stream.
SET VERTEX SHADER STREAMCOUNT
Sets the vertex shader stream count.
CONVERT OBJECT FVF Converts a 3D object's data to a

specific FVF format.
Pixel Shader Statements

A small set of statements also exist for creating pixel shader objects. These are:
CREATE PIXEL SHADER FROM FILE Creates a pixel shader from a

specified file.
DELETE PIXEL SHADER Deletes a previously created pixel
shader.
SET PIXEL SHADER ON Activates the pixel shader.
SET PIXEL SHADER OFF Deactivates the pixel shader.
SET PIXEL SHADER TEXTURE Specifies an image to be used by
the shader when texturing.
Summary
l Shaders are used to modify the final image that appears on the screen.
l Vertex shaders can modify the coordinates, shading and texturing of individual
vertices.
l Pixel shaders can modify the colour of pixels before they appear on screen.
l Shaders are most often written in HLSL (High-Level Shader Language).
l FX files can contain both vertex and pixel shaders as well as other code.
l The shader capabilities of your system are determined by your graphics card.
l Use GET MAXIMUM VERTEX SHADER VERSION to determine the version

of vertex shader available on your machine.
l Use GET MAXIMUM PIXEL SHADER VERSION to determine the version

of pixel shader available on your machine.
FX Files
l Use LOAD EFFECT to load an FX file.
l Use EFFECT EXIST to check that an effect is available.

l Use PERFORM CHECKLIST FOR EFFECT ERRORS to create a list of errors
which have been detected when attempting to load an effects file.
l Use SET OBJECT EFFECT to apply an effect to a specific object.
l Use SET EFFECT ON to load and apply an effect.
l Use DELETE EFFECT to delete an effect from RAM.
l Use SET LIMB EFFECT to apply an effect to a limb.
l Use PERFORM CHECKLIST FOR EFFECT VALUES to determine which

values within an effect can be influenced by the program in which the effect is
running.
l Use SET EFFECT CONSTANT to assign a value to an effect parameter.
l Use SET EFFECT TECHNIQUE to choose which technique within an effects

file is to be executed.
l Use SET EFFECT TRANSPOSE to transpose a matrix being used as a parameter

to an effect.
Shader Files
l Vertex and pixel shaders can exist independently of an effects file.
l Vertex and pixel shaders can be loaded and applied to specific objects.

Solutions FOR c = 1 TO CHECKLIST QUANTITY()
Activity 50.1 IF CHECKLIST VALUE B(c)=1
PRINT "Vertex shader version : ", PRINT CHECKLIST STRING$(c)
ÄGET MAXIMUM VERTEX SHADER VERSION() ENDIF
PRINT "Pixel shader version : ", NEXT c
ÄGET MAXIMUM PIXEL SHADER VERSION() REM *** end program ***
WAIT KEY
WAIT KEY END
END
AUTOCAM OFF
No solution required. ENDFUNCTION
Activity 50.3
LOAD IMAGE "lava1.bmp",1
TEXTURE OBJECT 1,1
REM *** Load and apply effect ***
SET EFFECT ON 1,"bubble.fx",0
NEXT c
WAIT KEY
END
ENDIF
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION
Activity 50.4
Activity 50.5
SetUpScreen()
REM *** Load effect file ***
REM *** Display parameters ***
PERFORM CHECKLIST FOR EFFECT VALUES 1
NEXT c
WAIT KEY
END
AUTOCAM OFF
ENDFUNCTION
Activity 50.6
SetUpScreen()
REM *** Load effect file ***
REM *** Display parameters ***
PERFORM CHECKLIST FOR EFFECT VALUES 1

51
Basic Networking Concepts

Choosing a Connection Method
Game Clients
Hosting a Game
DarkBASIC Pro: Network Programming 1389

Networked Games
Introduction
Back in Chapter 26 we created a two-player game, but that had both players using
the same keyboard! A much better approach would be to allow two or more players
to participate in the same game while sitting at different computers.
This setup gives us a lot more scope for a game; we can hide information from other
players, the players can be half a world apart, and playing against another human
being is always more fun than trying to best a few thousand IF statements.
Hardware Requirements
To get computers to communicate with each other, they need to be connected in
some way. If only two machines are involved and they are located in the same room,
then a simple cable connecting the machines' serial ports will be sufficient. Over a
longer distance two machines could communicate using a phone line and modems.
But when more machines are involved, then they need to be part of a network.
A local area network (LAN) is one where the computers are in close physical
proximity to each other; perhaps in the same room, building or campus.
A wide area network (WAN) is one where the linked computers are spread over
greater distances, possibly even in different countries.
In setting up a network there are choices to be made in how the machines are to
communicate with each other.
In a peer-to-peer network, each machine has equal status. Information from one
machine is set along a common connection and collected by the machine for which
it was destined. Every machine in the network has its own unique address and the
addresses of both the source computer and destination computer are sent as part of
the information transmitted.
In a client/server setup, one machine acts as a server and the others as clients. The
server has links to every client. While the server can communicate directly with any
client, communication between clients must be routed through the server.
The client machines send requests for data to the server and the server sends back
the necessary information.
The programs in this chapter have been tested on a LAN using a peer-to-peer setup.
Peer-to-peer is the default setup for Microsoft Windows.
Getting Started
During the early testing stages, it is likely that we will be running our programs
over machines located in the same room, since this will make testing much easier -
especially if you're working alone! We need to start by making sure that all the
computers that are to be used are switched on and communicating with each other.
1390 DarkBASIC Pro: Network Programming

The PERFORM CHECKLIST FOR NET CONNECTIONS Statement
The PERFORM CHECKLIST FOR NET CONNECTIONS statement looks for all
possible connections to other computers that are available to the program and
populates a checklist with the connections found. It will detect not only normal LAN
connections but also connections between two machines using the serial port,
modem connections, and connections to the Internet. The statement has the format
shown in FIG-51.1.
FIG-51.1
The PERFORM PERFORM CHECKLIST FOR NET CONNECTIONS
CHECKLIST FOR NET
CONNECTIONS
Statement
Once created, the contents of the checklist can be displayed using standard checklist
statements.
The program in LISTING-51.1 checks for connections available on your machine.
LISTING-51.1 REM *** Check for links ***

PRINT "Searching for connections..."
Listing the Connection PERFORM CHECKLIST FOR NET CONNECTIONS
Options Available
REM *** Display links found ***
PRINT CHECKLIST QUANTITY()," Connections found"
FOR c = 1 to CHECKLIST QUANTITY()
PRINT c,". ",CHECKLIST STRING$(c)
NEXT c

WAIT KEY
END
Activity 51.1
Type in and test the program in LISTING-51.1 (host01.dbpro).
FIG-51.2 shows a screen dump of a typical output from the program above.
FIG-51.2
The Output Produced by
a Connections Check
The IPX connection is usually installed when you have a network card (Ethernet)
installed.
The TCP/IP connection will be available if you have an Internet connection.
When we choose which connection we will be using in our program, we use the
corresponding number given in the listing. For example, if a machine created the
list shown above, then we would use option 4 to create a TCP/IP connection between
the machines. However, TCP/IP will not be listed as option 4 on every computer;
perhaps the IPX option is not available on your machine; in this case, TCP/IP will
be listed as option 3.

DarkBASIC Pro makes use of the DirectPlay component within DirectX to handle
the networking aspects of the program.
A connection can be in a peer-to-peer or client/server setup. In either case we begin

the game on one machine known as the host.
TCP/IP
Transmission Control Protocol/Internet Protocol is nothing more than an agreed
set of standards for transmitting digital information on the Internet. The same
standard is also used when transmitting data over most networks, even when there
is no access to the Internet itself.
The two main parts of this standard are:
Ø TCP - which ensures the correct delivery of data from one machine to
another. It can detect errors in transmission and cause retransmission
until the data is successfully received.
Ø IP - which is responsible for moving data from machine to machine. The
data is organised into packets with each packet including 4 bytes of
information giving the address of the machine to which that packet is to
be sent (it's a bit like an address on an envelope). Every machine on a
network will have a different IP address.
Activity 51.2
In this Activity you are going to discover the IP address of your own machine.
On one of the machines you are using, create a command line window
(START:RUN: Enter cmd).
At the command prompt, enter ipconfig.
In the resulting display, under the line

Ethernet adapter Local Area Connection
there should be a line giving an IP address. It will be something similar to
192.168.0.1. Write down the IP address displayed.
The SET NET CONNECTION Statement

The computer that is to start up the game (the host or server) must specify which
type of connection is being used (chosen from the list given by PERFORM
CHECKLIST FOR NET CONNECTIONS). To do that, it needs to execute the SET
NET CONNECTION statement which has the format shown in FIG-51.3.
FIG-51.3
The SET NET
SET NET CONNECTION connecttype
[ , address
CONNECTION
Statement
In the diagram:
connecttype is an integer value (1 to 4) specifying which

type of connection is to be used between the
computers.

address is a string, the value of which is dependent on the
type of connection chosen:
(serial) use "" ( a blank string).
(modem) use the phone number to be
accessed.
(IPX) this parameter is not used.
(TCP/IP) the IP or URL address.
In our examples we will be using a TCP/IP connection, so we might be tempted just

to write:
This assumes the
machine's IP address is SET NET CONNECTION 4,"192.168.0.1"
192.168.0.1
However, since TCP/IP will not be option 4 on every machine, a more fool-proof
approach would be to create a function to find the TCP/IP option's number:
FUNCTION GetTCPIPNumber()
result = 0
REM *** Create list of available connections ***
PERFORM CHECKLIST FOR NET CONNECTIONS
REM *** Search for TCP/IP connection ***
IF LEFT$(CHECKLIST STRING$(c),8) = "Internet"
result = c
ENDIF
NEXT c
REM *** IF not found, terminate program ***
IF result = 0
PRINT "No connection available. Press any key."
WAIT KEY
END
ENDIF
ENDFUNCTION result
Now the main section of the program can create a connection using just 2 lines:
TCPIP = GetTCPIPNumber()
SET NET CONNECTION TCPIP,"192.168.0.1"
The program in LISTING 51.2 creates a connection, displaying messages at each

stage.
LISTING-51.2 REM *** Use manual screen updating ***
SYNC ON
Making a Connection
REM *** Get TCP/IP connection number ***
PRINT "Checking connections available ..."
SYNC:SYNC
PRINT "Connections found"
SYNC
REM *** Set up connection ***

PRINT "Creating Connection "
SYNC
PRINT "Connection created"
SYNC

WAIT KEY
END

LISTING-51.2 FUNCTION GetTCPIPNumber()
(continued) result = 0
Making a Connection FOR c = 1 to CHECKLIST QUANTITY()
result = c
ENDIF
NEXT c
IF result = 0
SYNC
WAIT KEY
END
ENDIF
ENDFUNCTION result
When you run the
program, Windows may
throw up a dialog box Activity 51.3
informing you that it has
blocked the program's Type in and test the program in LISTING-51.2 (host02.dbpro).
access to the network. If
this happens click on the
button allowing access.
REMEMBER TO REPLACE THE IP ADDRESS IN THE CODE WITH
YOUR OWN MACHINE'S IP ADDRESS.
The CREATE NET GAME Statement

Once we've specified the connection details, we need to announce that this machine
is going to host a program with which others may interact. This is done using the
CREATE NET GAME statement whose format is shown in FIG-51.4.
FIG-51.4 The CREATE NET GAME Statement
CREATE NET GAME game , player , max

[ , setup
In the diagram:
game is a string specifying the name to be given to the

program being hosted. Any name can be used.
Despite its name, this
statement is used to player is a string giving the name of the user at the host
activate any machine. In a game situation, this will be your
multi-terminal software - character's name.
not just a game.
max is an integer value giving the maximum number

of players allowed to access the program at any
one time. The maximum value is 256.
setup is 1 or 2 and specifies the type of network setup:

1 peer-to-peer
2 client/server
1 is the default value.
The max value determines how many machines (including the host) can be part of
the link at any one time. Although a maximum of 256 is possible, a much smaller
figure will probably be more practical. The maximum number of players will be
determined by the speed of the game. This, in turn, is determined by the speed of

the connection between the machines, the amount of data that needs to be sent
between machines, and how frequently that data must be sent. You may find for a
real-time 3D game that you are limited to as little as 8 players.
In our first attempt at multi-player programming we're going to start by sending

messages between two machines.
Activity 51.4
In your last program, just before the end of the main section, add the lines
CREATE NET GAME "Passing Messages","John",2,1
PRINT "Program hosted"
SYNC
Writing Code for the Client Machine

So far we've created program designed to run on the machine hosting our program,
now we have to create the software for the client machine.
Like the host, the client will have to detect what connection options are available
and choose a TCP/IP connection, but it will be joining rather than creating a game,
so it must start by discovering what games are available (that is, what games are
being hosted on other machines).
The PERFORM CHECKLIST FOR NET SESSIONS Statement

We can check to see what hosted programs are available on our network using the
PERFORM CHECKLIST FOR NET SESSIONS statement which has the format
shown in FIG-51.5.
FIG-51.5
The PERFORM PERFORM CHECKLIST FOR NET SESSIONS
CHECKLIST FOR NET
SESSIONS Statement
Again, we need to use the standard checklist statements to gain access to the contents
of the checklist created.
Only programs hosted on other machines will be listed - programs hosted on the
machine executing this statement do not appear in the list.
The program needed to detect the hosted applications available to us is given in

LISTING-51.3. The program starts by creating a TCP/IP link and then displaying
which net sessions are available.
LISTING-51.3 REM *** Get TCP/IP connection number ***
Working on a Client PRINT "Connections found"
Machine
REM *** Select TCP/IP connection ***
REM *** List hosted programs ***

ListNetSessions()
WAIT KEY
END

LISTING-51.3 FUNCTION GetTCPIPNumber()
(continued) result = 0
Working on a Client FOR c = 1 to CHECKLIST QUANTITY()
Machine IF LEFT$(CHECKLIST STRING$(c),8) = "Internet"
result = c
ENDIF
NEXT c
REM *** If no connection found, stop ***
IF result = 0
NOTE: The SYNC PRINT "No connection available. Press any key."
statement is not used in WAIT KEY
this program. END
ENDIF
ENDFUNCTION result
FUNCTION ListNetSessions()
PRINT "Sessions available : "
PERFORM CHECKLIST FOR NET SESSIONS
PRINT c,". ",CHECKLIST STRING$(c)
NEXT c
PRINT "List complete"
ENDFUNCTION
Notice that the SET NET CONNECTION statement gives the IP address of the host
machine, not the client machine's own IP address.
Activity 51.5
On your second machine (the client machine) type in and run the program in
LISTING-51.3 (client01.dbpro).
MAKE SURE host02.dbpro IS ALREADY RUNNING ON THE HOST

MACHINE.
The listing produced includes two numbers at the end of each entry in the list. The
entry beside Passing Messages should be (1/2). This tells us that the program can
have up to 2 participants and that it already has 1 (the host).
The JOIN NET GAME Statement

Now that the client machine is aware of the Passing Messages program, it needs to
start communicating with the program using the JOIN NET GAME statement which
FIG-51.6
The JOIN NET GAME JOIN NET GAME session , player
Statement
In the diagram:
session is an integer value giving the program (or session)

that the client machine wishes to join. This value
will be taken from the list produced by the
statement.

player is a string giving the name the user on the client
machine wishes to be known by within the game.
For example, assuming the client machine has listed Passing Messages as the first
item in its net sessions list, then the user at the client machine can participate in that
program (using the name Rog) by executing the line:
JOIN NET GAME 1,"Rog"
The PERFORM CHECKLIST FOR NET PLAYERS Statement

We can discover the names (and other details) assigned to each participant in a
session using the PERFORM CHECKLIST FOR NET PLAYERS statement which
FIG-51.7
The PERFORM PERFORM CHECKLIST FOR NET PLAYERS
CHECKLIST FOR NET
PLAYERS Statement
The details recorded are listed in TABLE-51.1.
TABLE-51.1 Information Accessed by Typical Value
User's name CHECKLIST STRING$() Rog

Player Details Player's Local ID CHECKLIST VALUE A() 1
Player's Universal ID CHECKLIST VALUE B() 42120305
IsUser CHECKLIST VALUE C() 1 (0 or 1)
IsHost CHECKLIST VALUE D() 0 (0 or 1)
User's Name is the name the player used when joining the
session.
Player's Local ID is the ID used to identify a user from the

perspective of the machine producing this listing.
For example, when a player's list is displayed on
John's machine, he will have a local ID of 1 and
Rog will be 2. But when the player's list is shown
on Rog's machine, Rog's local ID will be 1 and
John's 2.
Player's Universal ID is a machine-independent ID. This ID is the

same irrespective of which machine lists the
player's details. Every player has an unique
universal ID.
IsUser is 1 if the player details are the user's own details;

it is 0 if the details are for another player.
On John's machine, the entry for John will have
an IsUser value of 1 and the entry for Rog
will be zero. On Rog's machine the entry for Rog
will be 1 and for John, zero.
IsHost is 1 on the entry for the host machine and zero

for client machines.
FIG-51.8 shows the player details produced for a session with John (host), Rog, Liz
and Mary.

FIG-51.8
Player Details as Listed
at Each Machine in a
Host : John Client : Rog Client : Liz Client : Mary
Session
Player List Player List Player List Player List
Mary 4 81270319 0 0 Mary 4 81270319 0 0 Mary 4 81270319 0 0 Mary 1 81270319 1 0
Liz 3 71241068 0 0 Liz 3 71241068 0 0 Liz 1 71241068 1 0 Liz 4 71241068 0 0
Rog 2 42120305 0 0 Rog 1 42120305 1 0 Rog 3 42120305 0 0 Rog 3 42120305 0 0
John 1 1 1 1 John 2 1 0 1 John 2 1 0 1 John 2 1 0 1
Activity 51.6
On your client machine, modify client01.dbpro, so that the user joins Passing
Messages using the name Rog.
Add a function, ListPlayersDetails(), that lists the details of all players in the
current session.
Add a call to this function near the end of the main section and test the
program. What numbers are assigned in the list to the two players?
MAKE SURE THE HOST IS CURRENTLY EXECUTING host02.dbpro.
Using a Single Machine as Both Host and Client

Are you finding it rather awkward running between two different machines? Well,
we can actually make one machine act like several different machines by using a
special IP address - 127.0.0.1.
Activity 51.7
We're going to make the programs Windows-based, since it will be easier to

switch between the two programs that way.
Creating Window
Applications was covered
in Chapter 14. Load client01.dbpro.
At the start of the program, add the lines:
SET WINDOW ON
SET WINDOW SIZE 500,900
SET WINDOW POSITION 501,0
SET WINDOW TITLE "CLIENT"
Change the line giving the IP address to use "127.0.0.1".

Save the program.
Load the program host02.dbpro.

At the start of the program add the lines:
SET WINDOW ON
SET WINDOW TITLE "HOST"
Change the line giving the IP address to use "127.0.0.1".

Save the program and then run it.
Switch back to DarkBASIC Pro, load client01.dbpro and run it.

Combining the Host/Client Requirements
Life is going to be complicated if we have to write two different programs; one to
create a network connection on the host machine and another to create a connection
on client machines. To get round this problem we need to write a single routine
which will handle both situations. Our new function, EstablishConnection(), will
use the following logic:
Get the TCP/IP connection number
Get the IP address of the host machine
Create a network connection
Get host/client choice
IF
host:
Get user's session name
Get maximum users
Get session name
Host the program
client:
List sessions currently available
Get session to be joined
Get user's session name
Join session
List chosen session's users
ENDIF
Return host/client choice
This can be coded as:
FUNCTION EstablishConnection()
TCPIPno = GetTCPIPNumber()
IPaddress$ = GetIPAddress$()
SET NET CONNECTION TCPIPno,IPaddress$
option = HostOrClient()
IF option = HOST
REM *** Host program ***
INPUT "Enter the name you wish to use : ",name$
INPUT "Enter maximum users : ",max
INPUT "Enter session name : ", session$
CREATE NET GAME session$,name$,max,1
ELSE
ListNetSessions()
session = GetSessionToJoin()
INPUT "Enter the name you wish to use : ",name$
JOIN NET GAME session,name$
ListPlayersDetails()
ENDIF
ENDFUNCTION option
The code makes use of the following constants
#CONSTANT HOST 1
#CONSTANT CLIENT 2
and several routines:
FUNCTION GetTCPIPNumber()
result = 0
result = c
ENDIF
NEXT c

IF result = 0
WAIT KEY
END
ENDIF
ENDFUNCTION result
PRINT c," ",CHECKLIST STRING$(c)
NEXT c
ENDFUNCTION
FUNCTION ListPlayersDetails()
PRINT "Players in session "
PERFORM CHECKLIST FOR NET PLAYERS
PRINT c," Name: ",CHECKLIST STRING$(c)," Local ID : ",
Ä
CHECKLIST VALUE A(c)," Universal ID : ",
Ä
CHECKLIST VALUE B(c)," Me? : ",CHECKLIST VALUE C(c),
Ä
" Host? ",CHECKLIST VALUE D(c)
NEXT c
ENDFUNCTION
FUNCTION HostOrClient()
PRINT "Choose from the following options : "
PRINT " 1 - Host a new session"
PRINT " 2 - Join an existing session"
INPUT "Enter option : ",option
WHILE option < 1 OR option > 2
PRINT "Enter 1 or 2"
INPUT "Enter option : ",option
ENDWHILE
ENDFUNCTION option
FUNCTION GetIPAddress$()
INPUT "Enter the IP address of the HOST machine
Ä(xxx.xxx.xxx.xxx) : ",IPaddress$
ENDFUNCTION IPaddress$
FUNCTION GetSessionToJoin()
INPUT "Enter session you wish to join : ",session
WHILE session < 1
PRINT "Enter the session number from the list given"
INPUT "Enter option : ",session
ENDWHILE
ENDFUNCTION session
Activity 51.8
Type in the code given for EstablishConnection(), constants and called

routines.
Save the file as NetConnect.dbpro.
We'll include this later in other network-based programs.

Communicating
The SEND NET MESSAGE Statement (Version 1)
With the two users linked, we now need to exchange information between the two
machines. If we wish to send standard data (integer, real, or string), then we can use
version 1 of the SEND NET MESSAGE statement whose format is shown in
FIG-51.9.
FIG-51.9
{
INTEGER
The SEND NET
MESSAGE
Statement SEND NET MESSAGE FLOAT player , value
STRING
In the diagram:
INTEGER, FLOAT, STRING

being sent.
player is an integer value specifying the local ID of the

player to receive the message.
If a value of zero is used, every member of the
session (other than the person sending the
message) will receive the message.
value is the value to be sent. The value's type should

match the keyword used in the statement (integer,
real (float) or string.
For example, if we assume the host machine's player (John) wants to send the
message "Hello there" to Rog's client machine, and that Rog has, when listed on
John's machine, the local ID of 2, then the required statement that should be added
to the host program is:
SEND NET MESSAGE 2,"Hello there, Rog"
On the other hand, if Rog wants to send John a message, since, from Rog's
perspective, he has a local ID of 1 and John's ID is 2, a message from Rog to John
would also use 2 as the first parameter:
SEND NET MESSAGE 2,"Hi, John"
The GET NET MESSAGE Statement

If one machine sends a message (using the SEND NET MESSAGE statement), the
other machine needs to read that message.
Every message received by a machine is placed in a queue. To make the first

message in the queue available for reading, we must start by executing the GET
NET MESSAGE statement whose format is given in FIG-51.10.

FIG-51.10
GET NET MESSAGE
The GET NET
MESSAGE Statement
The NET MESSAGE EXISTS Statement

The big problem for any machine waiting for a message is knowing if that message
has arrived. We can discover if a message has arrived using the NET MESSAGE
EXISTS statement which has the format shown in FIG-51.11.
FIG-51.11
NET MESSAGE EXISTS ( )
The NET MESSAGE
EXISTS Statement
integer
The statement returns 1 if a message is available, otherwise zero is returned.
If we want a program to wait for a message arriving ,we can use the following code:
REPEAT
GET NET MESSAGE
UNTIL NET MESSAGE EXISTS()
Once we have received a message, we need to read it.
The NET MESSAGE Statement (Version 1)

Reading a received message is done using the NET MESSAGE statement which
{
FIG-51.12 INTEGER
The NET MESSAGE
Statement NET MESSAGE FLOAT ( )
STRING$
value
In the diagram:
INTEGER, FLOAT, STRING$

being read. The type must match the type of value
sent by the other machine.
The type of value returned by this statement will match the type name used. That
is, the statement will return an integer, real (float), or string.

The program in LISTING-51.4 sends messages between the host and client
machines.
LISTING-51.4 #INCLUDE "NetConnect.dba"
REM *** Use a window ***
Sending and Receiving SET WINDOW ON
Messages SET WINDOW SIZE 500,900
r = EstablishConnection()
REM *** Send and receive messages ***
DO
REM *** Send message ***
INPUT "Choose number of player to receive message ",playerno
INPUT "Enter your message ", mes$
SEND NET MESSAGE STRING playerno,mes$
REM *** Display any received messages ***

GET NET MESSAGE
WHILE NET MESSAGE EXISTS()
PRINT NET MESSAGE STRING$()
GET NET MESSAGE
ENDWHILE
LOOP
END
Activity 51.9
Type in and test the program given in LISTING-51.4 (net04.dbpro) on the

host machine.
Copy the .exe version of the program onto the client machine and check that
the two users can communicate with each other.
The NET MESSAGE PLAYER FROM Statement

When a message is received, we can discover who has sent it using the NET
MESSAGE PLAYER FROM statement whose format is shown in FIG-51.13.
FIG-51.13
The NET MESSAGE NET MESSAGE PLAYER FROM ( )
PLAYER FROM
Statement
integer
The statement returns the local ID (as seen from the receiver's perspective) of the
user that sent the message which has been received.
The NET MESSAGE PLAYER TO Statement

It is also possible to discover the local ID of the message recipient using the NET
MESSAGE PLAYER TO statement. This would seem a rather pointless enquiry
since, if you've received a message, it was obviously meant for you and your own
local ID is always 1! Even if the sender broadcasts the message to all other users
(using the zero option in the SEND NET MESSAGE statement), the NET
MESSAGE PLAYER TO statement still returns a value of 1. The NET MESSAGE
PLAYER TO statement has the format shown in FIG-51.14.

FIG-51.14
NET MESSAGE PLAYER TO ( )
The NET MESSAGE
PLAYER TO
Statement
integer
Activity 51.10
Modify your last program to display the local ID of the sender of each
message.
The SEND NET MESSAGE Statement (Version 2)

As well as being able to send simple data values, it is also possible to send much
more complex structures such as images, sounds, 3D objects and memory blocks.
This is achieved using the second version of the SEND NET MESSAGE statement
whose format is given in FIG-51.15.
FIG 51.15 The SEND NET MESSAGE Statement
{
BITMAP
IMAGE
SEND NET MESSAGE SOUND player , id , send
MESH
MEMBLOCK
In the diagram:
BITMAP, IMAGE, SOUND, MESH, MEMBLOCK

Use the option appropriate for the type of message
being sent.
player is an integer value specifying the player to

receive the message. The number allocated to a
player can be found using the checklist produced
by the PERFORM CHECKLIST FOR NET
PLAYERS statement.
If a value of zero is used, every member of the
session (other than the person sending the
message) will receive the message.
id is an integer value giving the ID of the object

being sent. The object must already exist.
send 0 or 1. With items which may be split into several

data packets for transmission over the network,
there is a chance that data will not be correctly
received or may not be sent at all because of
backlogs on a slow network. By setting send to

1 the network is forced to transmit the data and
ensure it is correctly received. Setting send to
zero means there are no guarantees of correct
transmission.
For example, to transmit the details of a cube to all other machines in a session, we
could use the following lines:

MAKE MESH FROM LIMB 1,0
SEND NET MESSAGE MESH 0,1,1
The NET MESSAGE Statement (Version 2)

To receive large data items we need to use a second version of the NET MESSAGE
statement, the format of which is given in FIG-51.16.
FIG-51.16
{
BITMAP
The NET MESSAGE
Statement
IMAGE
NET MESSAGE SOUND id
MESH
MEMBLOCK
value
In the diagram:
BITMAP, IMAGE, SOUND, MESH, MEMBLOCK

Use the option appropriate for the type of message
being received.
id is an integer value giving the ID to be assigned to

the object being received.
In the next program (LISTING-51.5) we are going to send images between the host
and client.
LISTING-51.5
#INCLUDE "NetConnect.dba"
Sending an Image REM *** Use a window ***
SET WINDOW ON
REM *** Load images ***
LOAD IMAGE "image1.jpg",1
LOAD IMAGE "image2.bmp",2
DO
REM *** Get recipient and image ***

LISTING-51.5
(continued) INPUT "Which image 1 or 2", imageno
Sending an Image SEND NET MESSAGE IMAGE playerno,imageno,1
REM *** Get message ***
GET NET MESSAGE
NET MESSAGE IMAGE 3
PASTE IMAGE 3,0,0
GET NET MESSAGE
ENDWHILE
LOOP
END
Activity 51.11
Type in and test the program in LISTING-51.5 (net05.dbpro).
The NET MESSAGE TYPE Statement

So far, we've written programs that know what type of data to expect (string, image,
etc.), but a more flexible program needs to be able to receive and handle any type
of message. Using the NET MESSAGE TYPE we can discover the type of message
that has been received. This statement has the format shown in FIG-51.17.
FIG-51.17
NET MESSAGE TYPE ( )
The NET MESSAGE
TYPE Statement
integer
This statement returns an integer value based on the type of the latest message to
be received using the GET MESSAGE statement. The value returned is coded as
shown in TABLE-51.2.
TABLE-51.2 Code Message Type
The Codes Returned by 1 integer

the NET MESSAGE 2 real (float)
TYPE Statement 3 string
4 memory block
5 image
6 bitmap
7 sound
8 mesh
Using this statement, we can create a much more intelligent communications

program. First we'll create a routine which can send any type of message. This
function, SendMessage(), takes as parameters the recipient's ID, the message type
(1 to 8) and the message value as a string. For objects, the message value will be
the ID of the object to be sent. The function then determines which type of message
is to be sent and uses the appropriate version of the SEND NET MESSAGE
statement. The code for the routine is:
FUNCTION SendMessage(recipient, messagetype, value$)

SELECT messagetype
CASE 1 ìnteger
value = VAL(value$)
SEND NET MESSAGE INTEGER recipient,value

ENDCASE
CASE 2 `real
value# = VAL(value$)
SEND NET MESSAGE FLOAT recipient, value#
ENDCASE
CASE 3 `string
SEND NET MESSAGE STRING recipient,value$
ENDCASE
CASE 4 `memory block
value = VAL(value$)
SEND NET MESSAGE MEMBLOCK recipient,value,1
ENDCASE
CASE 5 ìmage
value = VAL(value$)
SEND NET MESSAGE IMAGE recipient,value,1
ENDCASE
CASE 6 `bitmap
value = VAL(value$)
SEND NET MESSAGE BITMAP recipient,value,1
ENDCASE
CASE 7 `sound
value = VAL(value$)
SEND NET MESSAGE SOUND recipient,value,1
ENDCASE
CASE 8 `mesh
value = VAL(value$)
SEND NET MESSAGE MESH recipient,value,1
ENDCASE
ENDSELECT
ENDFUNCTION
Next we need a function to receive a message, recognise the type of message and
display (or play) the message. This function, ReceiveMessage(), is coded as:
FUNCTION ReceiveMessage()
messagetype = NET MESSAGE TYPE()
messagefrom = NET MESSAGE PLAYER FROM()
SELECT messagetype
CASE 1 ìnteger
PRINT NET MESSAGE INTEGER()
ENDCASE
CASE 2 `real
PRINT NET MESSAGE FLOAT()
ENDCASE
CASE 3 `string
PRINT NET MESSAGE STRING$()
ENDCASE
CASE 4
NET MESSAGE MEMBLOCK 2
ENDCASE
CASE 5
NET MESSAGE IMAGE 2
PASTE IMAGE 2,0,0
DELETE IMAGE 2
ENDCASE
CASE 6
NET MESSAGE BITMAP 2
COPY BITMAP 2,0
DELETE BITMAP 2
ENDCASE
CASE 7
NET MESSAGE SOUND 2
PLAY SOUND 2
WAIT KEY
DELETE SOUND 2
ENDCASE
CASE 8

NET MESSAGE MESH 2
ADD LIMB 10,1,2
WAIT KEY
DELETE OBJECT 10
DELETE MESH 2
ENDCASE
ENDSELECT
ENDFUNCTION
Activity 51.12
Add the two routines given above to NetConnect.dba.
In the next program (see LISTING-51.6) we'll demonstrate more flexible message
passing where the users can choose the type of message that is to be transmitted.
Passing Messages of
Various Types REM *** Use a window ***
SET WINDOW ON
EstablishConnection()
REM *** Get recipient and message ***
PRINT "Loading..."
LOAD IMAGE "image1.jpg",1
PRINT "bottlebrush"
LOAD BITMAP "image2.jpg",1
SET CURRENT BITMAP 0
PRINT "Cactus"
PRINT "Sound"
DO
INPUT "Which type of message (1..8) :", messagetype
INPUT "Message value : ",value$
SendMessage(playerno,messagetype,value$)
GET NET MESSAGE
ReceiveMessage()
GET NET MESSAGE
ENDWHILE
LOOP
WAIT KEY
END
Activity 51.13
Type in and test the program in LISTING-51.6 (net06.dba).
The NET BUFFER SIZE Statement

If messages are being sent frequently, there's always the chance that a queue of
messages begin to build up, awaiting their turn to be transmitted. Too large a build
up will result in delays and maybe even the loss of messages. We can discover how
many items are waiting in that queue using the NET BUFFER SIZE statement which

FIG-51.18
The NET BUFFER
NET BUFFER SIZE ( )
SIZE Statement
integer
If the statement returns zero, this means that no messages are currently awaiting
transmission. So, we could check how many messages are currently in the queue
using a statement such as:
PRINT "Messages awaiting transmission : ",NET BUFFER SIZE()
Session Dynamics
If you were writing a multi-player on-line game, you'd expect new players to arrive
and existing players to leave during the game. To handle this we need statements
which detect and handle such events.
The NET PLAYER CREATED Statement

We can detect if a new player has joined a session using the NET PLAYER
CREATED statement which returns the local ID value of any new player joining a
session. The statement has the format shown in FIG-51.19.
FIG-51.19
NET PLAYER CREATED ( )
The NET PLAYER
CREATED Statement
integer
If no new player has joined, the statement returns the value zero.
Every time a player joins a session, his details will be added as the first entry in the
checklist created by the PERFORM CHECKLIST FOR NET PLAYERS statement.
The NET PLAYER DESTROYED Statement

When a player leaves a session, his local ID number can be discovered using the
NET PLAYER DESTROYED statement which has the format shown in FIG-51.20.
FIG-51.20
NET PLAYER DESTROYED ( )
The NET PLAYER
DESTROYED Statement
integer
The statement returns zero if no one has left the session.
The program in LISTING-51.7 maintains a list of the players in a session, recording

their local ID numbers and names. Every time a new player joins, a message and

the latest list of all players are displayed.
REM *** Structure used to hold player's details ***
Maintaining a List of TYPE PlayerDetailsType
Connected Users id AS INTEGER
name$ AS STRING
ENDTYPE
REM *** Details of all players ***
GLOBAL DIM players(10) AS PlayerDetailsType
*** Number of players currently in session ***
GLOBAL noofentries AS INTEGER
REM *** Use a window ***
SET WINDOW ON
DO
REM *** If new player, display message and updated list ***
IF NET PLAYER CREATED() <> 0
PRINT CHECKLIST STRING$(1)," has joined the session"
MaintainPlayerList()
DisplayPlayerList()
ENDIF
LOOP
WAIT KEY
END
FUNCTION MaintainPlayerList()
noofentries = CHECKLIST QUANTITY()
FOR c = 1 TO noofentries
players(c).id = CHECKLIST VALUE A(c)
players(c).name$ = CHECKLIST STRING$(c)
NEXT c
ENDFUNCTION
FUNCTION DisplayPlayerList()
PRINT players(c).id," ",players(c).name$
NEXT c
ENDFUNCTION
Activity 51.14
Type in the program in LISTING-51.7 (net07.dba).
Run the program creating four players:
John (host)
Rog (client)
Liz (client)
Mary (client)
Modify the program to display the name of any player leaving the session. The
player list should also be updated and displayed when a player leaves.
To test the new version of the program, create the same four users (as listed
earlier in the question) and have them leave the session (by closing the
appropriate Window) in the following order: Rog, Mary, Liz.

The NET GAME NOW HOSTING Statement
In the last Activity we deliberately retained the host player (John). But what happens
if the host should leave a session? The software is ready for such a situation and
simply transfers the hosting duties to one of the other members of the session. We
can discover if a machine has become the host using the NET GAME NOW
HOSTING statement which has the format shown in FIG-51.21.
FIG-51.21
The NET GAME NOW
NET GAME NOW HOSTING ( )
HOSTING Statement
integer
This statement will return 1 if the current user has become the new host for a session;
Activity 51.15
Modify your last program so that the message "You are new host" is displayed
by any user who becomes the new host for a session.
To test the program, create John, Rog, Liz, and Mary as before.
Close Rog, then John.
Who becomes the host?
Close Liz.
The FREE NET GAME Statement

Actually, rather than just terminate a program handling one of the players in a
session, we should really close that player's link in a more organised way, freeing
up any resources used by the program. To do this we can use the FREE NET GAME
FIG-51.22
FREE NET GAME
The FREE NET GAME
Statement
Activity 51.16
In your last program, add the lines
IF ESCAPEKEY()
EXIT
ENDIF
FREE NET GAME
at the appropriate points so that resources are freed before a player's

connection to a session is terminated.

The CREATE NET PLAYER Statement
We can add pseudo-players to a session using the CREATE NET PLAYER
statement. This allows us to create a new player in a session even though that player
does not represent a real user. This can be useful if a game requires a non-playing
character (NPC) to be treated in the same way as real players. The format for the
CREATE NET PLAYER is shown in FIG-51.23.
FIG-51.23
The CREATE NET
CREATE NET PLAYER ( player )
PLAYER Statement
integer
In the diagram:
player is a string giving the name to be assigned to

the new session member.
The statement returns the local ID assigned to the new player.
The FREE NET PLAYER Statement

Should we want to remove a pseudo-player from the session list, we can do so using
the FREE NET PLAYER statement which has the format shown in FIG-51.24.
FIG-51.24
The FREE NET PLAYER FREE NET PLAYER id
Statement
In the diagram:
id is an integer value giving the local ID assigned

to the player.
Although it is unlikely that you'll need to use CREATE NET PLAYER or FREE
NET PLAYER, we can demonstrate their use by adding the following code to the
main section of our previous program:
REM *** Add new player (Enter key) ***
IF RETURNKEY()
id = FindPlayerName("Avril")
IF id = 0
CREATE NET PLAYER ("Avril")
ENDIF
ENDIF
REM *** Remove new player (space bar)***

IF SPACEKEY()
id = FindPlayerName("Avril")
IF id <> 0
FREE NET PLAYER id
ENDIF
ENDIF
The new function called by the code above is:

FUNCTION FindPlayerName(name$)
result = 0
IF players(c).name$ = name$
result = players(c).id
EXIT
ENDIF
NEXT c
ENDFUNCTION result
Activity 51.17
Using the code given above, modify your previous program and check that the
new user can be successfully added and removed.
The NET GAME EXISTS Statement

If a session is being run on a client/server setup, closing the server will cause the
whole game to be closed down. We can check that a game is still being hosted on
the server using the NET GAME EXISTS statement which has the format shown
in FIG-51.25.
FIG-51.25
The NET GAME
NET GAME EXISTS ( )
EXISTS Statement
integer
The statement returns 1 if the current game is still being hosted on a server;
The NET GAME LOST Statement

We can detect when a game is no longer being hosted using the NET GAME LOST
FIG-51.26
NET GAME LOST ( )
The NET GAME LOST
Statement
integer
The statement returns 1 if the current session no longer has a host; otherwise zero
is returned.
Summary
l Networking involves communication between two or more machines.
l A Local Area Network (LAN) is a set of linked machines in close physical

proximity.

l A Wide Area Network (WAN) is a set of linked machines spread over a large
physical area.
l Networks can be set up in a peer-to-peer or client/server configuration.
l Use PERFORM CHECKLIST FOR NET CONNECTIONS to create a checklist

of connections available on your system.
l Use SET NET CONNECTION to create a connection to a machine with a

specified IP address.
l Use CREATE NET GAME to host a program.
l Use PERFORM CHECKLIST FOR NET SESSIONS to create a checklist of

programs currently being hosted on other machines.
l Use JOIN NET GAME to join a hosted program as a client.
l Use PERFORM CHECKLIST FOR NET PLAYERS to create a checklist of

other members of the program you have joined.
l Use SEND NET MESSAGE to send data to another machine in the current
session.
l Use NET MESSAGE EXISTS to check if a message has been received.
l Use NET MESSAGE to read a message which has been received.
l Use NET MESSAGE PLAYER FROM to get the ID of the player from whom
the latest message was received.
l Use NET MESSAGE PLAYER TO to get the ID of the player to whom the last
message was directed.
l Use NET MESSAGE TYPE to determine what type of data is in the latest
message to be received.
l Use NET BUFFER SIZE to determine the number of messages waiting to be

transmitted.
l Use NET PLAYER CREATED to detect a new user joining the current session.
l Use NET PLAYER DESTROYED to detect a user leaving a session.
l Use NET GAME NOW HOSTING to discover if your machine is the host of
the current session.
l Use FREE NET GAME to withdraw from the current session.
l Use CREATE NET PLAYER to create a pseudo-user in the current session.
l Use FREE NET PLAYER to withdraw a pseudo-user from the current session.
l Use NET GAME EXISTS to check that the server is still active in a client/server
setup.
l Use NET GAME LOST to detect when a session no longer has a host.

A Networked Game
Introduction
Gomuku is a Japanese game similar to Noughts and Crosses (Tick-Tack-Toe) but
instead of having to get 3 X's or O's in a row, you need to get 5.
The board is a 19 by 19 grid. On each turn a player writes an X or an O into one of

the squares. The first player to create a continuous line of 5 symbols (horizontally,
diagonally, or vertically) is the winner. A screen dump of the game is shown in
FIG-51.27.
FIG-51.27
Gomuku
The game will be implemented using sprites. The board will be a sprite, as will each
X and O symbol placed on the board.
A Non-Networked Version
We'll start by creating a non-networked version of the game and then add in the
networking requirements later once we have the game itself working.
Program Data
We need to set up named constants for such things as the images and sprites used
in the game. Also we need a record structure defined to store details about a player
- a player ID number and the ID of the image used by that player (X image or O
image). Finally, we need a set of global variables to hold the current state of the
board, the position of the latest move and the ID to be used for the next sprite placed
on the screen.
The code for all this is:

REM *********************************
REM *********************************
REM *** Numeric ***
#CONSTANT X 1
#CONSTANT O 2
REM *** Images ***
#CONSTANT ximg 1
#CONSTANT oimg 2
REM *** Sprites ***
#CONSTANT boardspr 1
REM *********************************
REM *** Types ***
REM *********************************
TYPE PlayerType
value
imgID
ENDTYPE
REM *********************************
REM *** GLOBALS ***
REM *********************************
GLOBAL DIM board(19,19)
GLOBAL nextsprite = 2
GLOBAL totalO
GLOBAL totalX
GLOBAL me AS PlayerType
GLOBAL you AS PlayerType
GLOBAL move
Game Logic
The game requires the following logic:
Set up details of each player
Set up screen
Set up board
REPEAT
Get X's move
IF X has not won THEN
Get O's move
ENDIF
UNTIL someone wins
End Game
This is coded as:
REM *** Main logic ***

SetUpPlayerDetails(1)
SetUpScreen()
SetUpBoard()
REPEAT
won = -GetMove(X)
IF NOT won
won = GetMove(O)
ENDIF
UNTIL won
EndGame(won)
END
Notice that the value returned from the first call to GetMove() is negated. This means
won will be -1 if X wins and 1 if O wins. We can make use of this to choose which
message should be displayed by EndGame().

Activity 51.18
Create a basic program from the code given, adding level 1 test stubs for each
function called.
Adding SetUpPlayerDetails()
The two global variables, me and you, are designed to contain details about each
player. The value field is used to assign a value to the board array in the cell
equivalent to where the player has placed his symbol (X or O); the imgID field is
used to identify which image is to be used when creating a sprite on the screen
representing the player's symbol.
The routine takes a value representing the parameter for me.value - all other
assignments can be calculated from this. The code for the routine is:
FUNCTION SetUpPlayerDetails(r)
REM *** Record my details ***
me.value = r
me.imgID = r
REM *** Record opponents details ***
you.value = 3-r
you.imgID = 3-r
ENDFUNCTION
Adding SetUpScreen()
The SetUpScreen() function is a typical routine for setting the screen resolution and
colour. Its code is:
BACKDROP ON
ENDFUNCTION
Adding SetUpBoard()
The SetUpBoard() function loads the images used in the game (board, X symbol
and O symbol) and displays the board using a sprite. The code is:
FUNCTION SetUpBoard()
LOAD IMAGE "board2.bmp",boardimg
LOAD IMAGE "x.bmp",ximg
LOAD IMAGE "o.bmp",oimg
SPRITE boardspr,200,200,boardimg
ENDFUNCTION
Adding GetMove()
The GetMove() function takes as a parameter an indication of which player is to add
a symbol to the board. It then calls either GetMyMove() or GetOpponentsMove() as
appropriate. It also checks for a winning line of 5 symbols. The code for GetMove()
is:
FUNCTION GetMove(player)
IF me.value = player
GetMyMove()

ELSE
GetOpponentsMove()
ENDIF
result = CheckForWin()
ENDFUNCTION result
Adding GetMyMove()
The GetMyMove() function gets the player's "move" and updates the screen by
adding an X or O sprite on the board and updating the board array.
Reading the mouse pointer's position to determine where on the board a symbol is
to be added, is actually performed by another function, GetSquare(), which returns
the row and column of the square selected as a single integer value.
FUNCTION GetMyMove()
REM *** Get a valid move ***
REPEAT
move = GetSquare()
row = move /100
col = move mod 100
result = board(row,col)
UNTIL result = 0
REM *** Add sprite to screen ***
SPRITE nextsprite,col*33+203,row*33+203,me.imgID
REM *** Update board array ***
board(row,col) = me.value
REM *** next sprite ID to be used is incremented ***
INC nextsprite
ENDFUNCTION
Adding GetSquare()
The GetSquare() function reads the position of the mouse pointer over the board.
Each square on the board is 33 pixels by 33 pixels, so the mouse coordinates are
rounded to a multiple of 33. The REPEAT..UNTIL structure ensures that the mouse
has been clicked within the board area.
The coordinates are converted to row and column values and these are combined
into a single value. For example, if the player clicked in row 12 column 8, this would
be converted to the single value 1208. It is this single value which is returned by
the function.
FUNCTION GetSquare()
REM *** Get mouse position within board ***
REPEAT
WAIT MOUSE
x1 = (MOUSEX()-200)/33 * 33
y1 = (MOUSEY()-200)/33 * 33
UNTIL InRange(x1,0,600) AND InRange(y1,0,600)
REM *** Convert to row and column ***
col = x1/33
row = y1/33
REM *** Convert row and column to a single value ***
result = row*100+col
ENDFUNCTION result
Adding InRange()
The InRange() function checks to see if a value is within a given range. If it is, 1 is
returned, otherwise zero is returned.

FUNCTION InRange(no,min,max)
IF no >= min AND no <= max
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
Adding GetOpponentsMove()
The GetOpponentsMove() function is almost identical to GetMyMove(), but uses
the you variable rather than me.
FUNCTION GetOpponentsMove()
REM *** Get a valid move ***
REPEAT
move = GetSquare()
row = move /100
col = move mod 100
UNTIL result = 0
REM *** Add sprite ***
SPRITE nextsprite,col*33+203,row*33+203,you.imgid
REM *** Update board ***
board(row,col)= you.value
REM *** Another sprite used ***
INC nextsprite
ENDFUNCTION
Activity 51.19
Add all the code given so far and any additional test stubs required.
Test the program to make sure it accepts each player's move correctly.
Adding CheckForWin()
We need to check for a winning line of 5 symbols. This can be in a horizontal,
vertical or diagonal direction (see FIG-51.28).
FIG-51.28
A line can be Created in
any One of Four
Directions

The main checking routine, CheckForWin() will make use of other routines to check
each of these directions separately.
FUNCTION CheckForWin(player)
result = 0
result = HorizontalCheck(player)
IF result = 0
result = VerticalCheck(player)
ENDIF
IF result = 0
result = BLTRDiagonalCheck(player)
ENDIF
IF result = 0
result = TLBRDiagonalCheck(player)
ENDIF
ENDFUNCTION result
Adding the Other Search Routines

We need not go into the details of how the other routines work, other than to say
that each returns 1 if a row of 5 symbols is found, otherwise 0 is returned.
FUNCTION HorizontalCheck(player)
REM *** Calc row and col of last move ***
row = move / 100
col = move mod 100
REM *** Upper and lower limit of search area ***
start = col - 4
IF start < 0
start = 0
ENDIF
finish = col + 4
IF finish > 18
finish = 18
ENDIF
REM *** search ***
total = 0
FOR col = start TO finish
IF board(row,col)= player
INC total
IF total = 5
EXIT
ENDIF
ELSE
total = 0
ENDIF
NEXT col
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
FUNCTION VerticalCheck(player)
row = move / 100
col = move mod 100
start = row - 4
IF start < 0
start = 0
ENDIF
finish = row + 4

IF finish > 18
finish = 18
ENDIF
REM *** search ***
total = 0
FOR row = start TO finish
INC total
IF total = 5
EXIT
ENDIF
ELSE
total = 0
ENDIF
NEXT row
REM *** Determine result ***
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
FUNCTION BLTRDiagonalCheck(player)
row = move / 100
col = move mod 100
REM *** Upper and lower limit of search area (row)***
srow = row
scol = col
FOR p = 1 TO 4
IF srow = 18 OR scol = 0
EXIT
ENDIF
INC srow
DEC scol
NEXT p
frow = row
fcol = col
FOR p = 1 TO 4
IF frow = 0 OR fcol = 18
EXIT
ENDIF
DEC frow
INC fcol
NEXT p
range = fcol-scol+1
FOR p = 1 TO range
IF board(srow,scol)= player
INC total
IF total = 5
EXIT
ENDIF
ELSE
total = 0
ENDIF
DEC srow
INC scol
NEXT p
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result

FUNCTION TLBRDiagonalCheck(player)
row = move / 100
col = move mod 100
srow = row
scol = col
FOR p = 1 TO 4
EXIT
ENDIF
DEC srow
DEC scol
NEXT p
frow = row
fcol = col
FOR p = 1 TO 4
EXIT
ENDIF
INC frow
INC fcol
NEXT p
range = fcol-scol+1
FOR p = 1 TO range
INC total
IF total = 5
EXIT
ENDIF
ELSE
total = 0
ENDIF
INC srow
INC scol
NEXT p
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
Adding EndGame()
A winning message is supplied by the EndGame() function which uses the
parameter won to decide on which message is to be displayed.
FUNCTION EndGame(won)
REM *** Find out who won ***
IF won = -1
winner$ = "X"
ELSE
winner$ = "O"
ENDIF
REM *** Display message for 5 seconds ***
SET TEXT SIZE 40
now = TIMER()
SET CURSOR 1000,500
PRINT winner$," wins!!!"
ENDWHILE
ENDFUNCTION

Activity 51.20
Add the search and end game functions to your code and test the final program.
Networking the Game

Updating the main section
Perhaps, surprisingly, we don't have to make too many changes to turn our game
into a networked version.
Before we start modifying our code, we need to copy the NetConnect.dba file we
created earlier in the chapter into our game folder.
Now the main logic needs to be changed so that host and client connections are
established and that the game waits for both players to be connected before
continuing. The new code is:
WaitForSecondPlayer()
SetUpPlayerDetails(r)
SetupScreen()
SetUpBoard()
REPEAT
won = -GetMove(X)
IF NOT won
won = GetMove(O)
ENDIF
UNTIL won
EndGame(won)
END
Adding WaitForSecondPlayer()
The WaitForSecondPlayer() function causes the program to pause until the number
of players connected to the session equals 2. So the player is aware that the program
hasn't stalled, a full stop is displayed every 3 seconds. The routine's code is:
FUNCTION WaitForSecondPlayer()
PRINT "Waiting for second player ";
WHILE NumberOfPlayers() <> 2
PRINT ".";
WAIT 3000
ENDWHILE
ENDFUNCTION
Adding NumberOfPlayers()
The NumberOfPlayers() routine returns the number of entries in the checklist after
executing a PERFORM CHECKLIST FOR NET PLAYERS statement. The code
for NumberOfPlayers() is:
FUNCTION NumberOfPlayers()
result = CHECKLIST QUANTITY()
ENDFUNCTION result

Modifying the Call to SetUpPlayerDetails()
The strange thing about network programming is trying to get your head round the
fact that the same code is being run on 2 or more machines and yet that code must
be able to give each player an unique identity. In our game, one machine must assign
its player the X symbol, while the other machine assigns the O symbol.
We get round this problem by assigning the me and you variables different values
on the different machines. The value assigned depends on whether the player is
using the host machine; in which case, me.value is set to 1, or the client machine
where me.value is set to 2. The you variable is then set to the opposite of me. These
values are then used to decide on which sprite is displayed when the player makes
a move.
The modification to the call to SetUpPlayerDetails() is simply to use the value

returned by EstablishConnection() as the parameter - and hence the value assigned
to me.
Modifying GetMyMove()
As before GetMyMove() needs to get the player's move from the mouse click,
display a symbol and update the board array. But now it must also transmit the
move chosen to the other machine using the line:
SEND NET MESSAGE INTEGER 2,move
The updated version of the routine is:
REPEAT
move = GetSquare()
row = move /100
col = move mod 100
UNTIL result = 0
INC nextsprite
REM *** Send details of move to other player ***
ENDFUNCTION
Modifying GetOpponentsMove()
The changes to the GetOpponentsMove() function are more fundamental since the
opponent's move will be made on the other machine. The routine needs to receive
details of that move over the network and then use these details to add a symbol and
update board. This gives us the code:
REM *** Read move from network ***
REPEAT
GET NET MESSAGE
move = NET MESSAGE INTEGER()
row = move/100
col = move mod 100

INC nextsprite
ENDFUNCTION
Modifying EndGame()
The only change to EndGame() is that we should free the network connection before
finishing, so we should add the line
FREE NET PLAYER 1
at the end of the routine.
Activity 51.21
Make the changes suggested and test out your game over a network.
A Complete Listing
A complete listing of the network version of the game is given in LISTING-51.8.
LISTING-51.8
Gomuku - the Networked
Version REM *********************************
REM *********************************
REM *** Numeric ***
#CONSTANT X 1
#CONSTANT O 2
REM *** Images ***
#CONSTANT ximg 1
#CONSTANT oimg 2
REM *** Sprites ***
#CONSTANT boardspr 1
REM *********************************
REM *** Types ***
REM *********************************
TYPE PlayerType
value
imgID
ENDTYPE
REM *********************************
REM *** GLOBALS ***
REM *********************************
GLOBAL DIM board(19,19)
GLOBAL nextsprite = 2
GLOBAL totalO
GLOBAL totalX
GLOBAL me AS PlayerType
GLOBAL you AS PlayerType
GLOBAL move

LISTING-51.8 REM *** Main logic ***
(continued) r = EstablishConnection()
WaitForSecondPlayer()
Gomuku - the Networked SetUpPlayerDetails(r)
Version SetupScreen()
SetUpBoard()
REPEAT
won = -GetMove(X)
IF NOT won
won = GetMove(O)
ENDIF
UNTIL won
EndGame(won)
END
REM ******************************
REM *** Level 1 ***
REM ******************************
FUNCTION WaitForSecondPlayer()
PRINT "Waiting for second player ";
WHILE NumberOfPlayers() <> 2
PRINT ".";
WAIT 3000
ENDWHILE
ENDFUNCTION
FUNCTION SetUpPlayerDetails(r)
me.value = r
me.imgID = r
you.value = 3-r
you.imgID = 3-r
ENDFUNCTION
BACKDROP ON
ENDFUNCTION
LOAD IMAGE "board2.bmp",boardimg
LOAD IMAGE "x.bmp",ximg
LOAD IMAGE "o.bmp",oimg
SPRITE boardspr,200,200,boardimg
ENDFUNCTION
IF me.value = player
GetMyMove()
ELSE
GetOpponentsMove()
ENDIF
result = CheckForWin(player)
ENDFUNCTION result
REM *** Find out who won ***
IF won = -1
winner$ = "X"
ELSE
winner$ = "O"
ENDIF
REM *** Display message for 5 seconds ***
SET TEXT SIZE 40

LISTING-51.8 now = TIMER()
(continued) WHILE TIMER() - now < 5000
SET CURSOR 1000,500
Gomuku - the Networked PRINT winner$," wins!!!"
Version ENDWHILE
FREE NET PLAYER 1
ENDFUNCTION
REM ******************************
REM *** Level 2 ***
REM ******************************
FUNCTION NumberOfPlayers()
result = CHECKLIST QUANTITY()
ENDFUNCTION result
REPEAT
move = GetSquare()
row = move /100
col = move mod 100
UNTIL result = 0
INC nextsprite
REM *** Send details of move to other player ***
ENDFUNCTION
REM *** Read move from network ***
REPEAT
GET NET MESSAGE
move = NET MESSAGE INTEGER()
row = move/100
col = move mod 100
INC nextsprite
ENDFUNCTION
FUNCTION CheckForWin(player)
result = 0
result = HorizontalCheck(player)
IF result = 0
result = VerticalCheck(player)
ENDIF
IF result = 0
result = BLTRDiagonalCheck(player)
ENDIF
IF result = 0
result = TLBRDiagonalCheck(player)
ENDIF
ENDFUNCTION result

LISTING-51.8 REM ******************************
(continued) REM *** Level 3 ***
REM ******************************
Gomuku - the Networked
Version FUNCTION GetSquare()
REM *** Get mouse position ***
REPEAT
WAIT MOUSE
x1 = (MOUSEX()-200)/33 * 33
y1 = (MOUSEY()-200)/33 * 33
UNTIL InRange(x1,0,600) AND InRange(y1,0,600)
col = x1/33
row = y1/33
result = row*100+col
ENDFUNCTION result
FUNCTION HorizontalCheck(player)
row = move / 100
col = move mod 100
start = col - 4
IF start < 0
start = 0
ENDIF
finish = col + 4
IF finish > 18
finish = 18
ENDIF
REM *** search ***
total = 0
FOR col = start TO finish
INC total
IF total = 5
EXIT
ENDIF
ELSE
total = 0
ENDIF
NEXT col
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
FUNCTION VerticalCheck(player)
row = move / 100
col = move mod 100
start = row - 4
IF start < 0
start = 0
ENDIF
finish = row + 4
IF finish > 18
finish = 18
ENDIF
REM *** search ***
total = 0
FOR row = start TO finish
INC total

LISTING-51.8 IF total = 5
(continued) EXIT
ENDIF
Gomuku - the Networked ELSE
Version total = 0
ENDIF
NEXT row
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
FUNCTION BLTRDiagonalCheck(player)
row = move / 100
col = move mod 100
REM *** Upper and lower limit of search area (row)***
srow = row
scol = col
FOR p = 1 TO 4
EXIT
ENDIF
INC srow
DEC scol
NEXT p
frow = row
fcol = col
FOR p = 1 TO 4
EXIT
ENDIF
DEC frow
INC fcol
NEXT p
range = fcol-scol+1
FOR p = 1 TO range
INC total
IF total = 5
EXIT
ENDIF
ELSE
total = 0
ENDIF
DEC srow
INC scol
NEXT p
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
FUNCTION TLBRDiagonalCheck(player)
row = move / 100
col = move mod 100
srow = row
scol = col

LISTING-51.8 FOR p = 1 TO 4
(continued) IF srow = 0 OR scol = 0
EXIT
Gomuku - the Networked ENDIF
Version DEC srow
DEC scol
NEXT p
frow = row
fcol = col
FOR p = 1 TO 4
EXIT
ENDIF
INC frow
INC fcol
NEXT p
range = fcol-scol+1
FOR p = 1 TO range
INC total
IF total = 5
EXIT
ENDIF
ELSE
total = 0
ENDIF
INC srow
INC scol
NEXT p
IF total = 5
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
FUNCTION Minimum(a,b)
IF a = min AND no <= max
result = 1
ELSE
result = 0
ENDIF
ENDFUNCTION result
Suggested Enhancements
As usual, there is plenty of scope for enhancing the program. An obvious addition
would be a message indicating who's turn it is. Perhaps the number of moves made,
or the time elapsed could be displayed. Or what about being able to have the program
replay all the moves in a game?

Solutions Ä" ",CHECKLIST VALUE D(c)
Activity 51.1
NEXT c
No solution required. ENDFUNCTION

No solution required. The main section of client01.dbpro should be changed to:
SET WINDOW ON
Activity 51.3 SET WINDOW SIZE 500,900
No solution required. SET WINDOW TITLE "CLIENT"
Activity 51.4 PRINT "Connections found"
No solution required. SET NET CONNECTION TCPIP,"127.0.0.1"
REM *** List hosted programs ***
ListNetSessions()
JOIN NET GAME 1,"Rog"
Activity 51.5 ListPlayersDetails()
WAIT KEY
END
Activity 51.6
host02.dbpro's main section should be:
TCPIP = GetTCPIPNumber() REM *** Use manual screen updating ***
PRINT "Connections found" SET WINDOW ON
REM *** Select TCP/IP connection *** SET WINDOW SIZE 500,900
SET NET CONNECTION TCPIP,"192.168.0.1" SET WINDOW POSITION 0,0
REM *** List hosted programs *** SET WINDOW TITLE "Host"
ListNetSessions() SYNC ON
JOIN NET GAME 1,"Rog" REM *** Get TCP/IP connection number ***
ListPlayersDetails() PRINT "Checking connections available ..."
REM *** End program *** SYNC:SYNC
WAIT KEY TCPIP = GetTCPIPNumber()
END PRINT "Connections found"
SYNC
FUNCTION GetTCPIPNumber() REM *** Set up connection ***
result = 0 PRINT "Creating Connection "
PERFORM CHECKLIST FOR NET CONNECTIONS SYNC
FOR c = 1 to CHECKLIST QUANTITY() SET NET CONNECTION TCPIP,"127.0.0.1"
IF LEFT$(CHECKLIST STRING$(c),8) = PRINT "Connection created"
Ä
"Internet" SYNC
result = c CREATE NET GAME "Passing Messages","John",
ENDIF Ä2,1
NEXT c PRINT "Program hosted"
REM *** If no connection found, stop *** SYNC
IF result = 0 REM *** End program ***
PRINT "No connection available. WAIT KEY
Ä
Press any key." END
WAIT KEY
END
ENDIF Activity 51.8
ENDFUNCTION result
PERFORM CHECKLIST FOR NET SESSIONS Activity 51.9
PRINT c,". ",CHECKLIST STRING$(c) No solution required.
NEXT c
PRINT "List complete"
FUNCTION ListPlayersDetails() #INCLUDE "NetConnect.dba"
PRINT "Players in session " REM *** Use a window ***
PERFORM CHECKLIST FOR NET PLAYERS SET WINDOW ON
FOR c = 1 to CHECKLIST QUANTITY() SET WINDOW SIZE 500,900
PRINT c," ",CHECKLIST VALUE A(c), SET WINDOW POSITION 0,0
Ä
" ",CHECKLIST VALUE B(c), SET WINDOW TITLE "Host"
Ä
" ",CHECKLIST VALUE C(c), r = EstablishConnection()

REM *** Send and receive messages *** IF players(c).id = id
DO result = c
REM *** Send message *** EXIT
INPUT "Choose number of player to ENDIF
Äreceive message ",playerno NEXT c
INPUT "Enter your message : ",mes$ ENDFUNCTION result
SEND NET MESSAGE STRING playerno,mes$
REM *** Display any received messages ***
GET NET MESSAGE Activity 51.15
PRINT NET MESSAGE PLAYER FROM(), The DO..LOOP of the program should be coded as:
Ä
" : ",NET MESSAGE STRING$()
GET NET MESSAGE DO
ENDWHILE IF NET PLAYER CREATED()
LOOP PERFORM CHECKLIST FOR NET PLAYERS
REM *** End program *** PRINT CHECKLIST STRING$(1),
END Ä
" has joined the session"
DisplayPlayerList()
Activity 51.11 ENDIF
playerleft = NET PLAYER DESTROYED()
No solution required. IF playerleft
PRINT
Äplayers(FindPlayerID(playerleft)).name$
Activity 51.12 Ä
," has left the session"
No solution required. DisplayPlayerList()
ENDIF
IF NET GAME NOW HOSTING()
Activity 51.13 PRINT "You are new host"
ENDIF
No solution required. LOOP
When John leaves the session, Liz becomes the host.

Activity 51.14 When Liz leaves the session Mary becomes the host.
The main section should become:
Activity 51.16
The last part of the main section should now be:
TYPE PlayerDetailsType
id AS INTEGER DO
name$ AS STRING IF NET PLAYER CREATED()
ENDTYPE PERFORM CHECKLIST FOR NET PLAYERS
PRINT CHECKLIST STRING$(1),
GLOBAL DIM players(10) AS PlayerDetailsType
GLOBAL noofentries AS INTEGER
Ä
REM *** Use a window *** DisplayPlayerList()
SET WINDOW ON ENDIF
SET WINDOW SIZE 500,900 playerleft = NET PLAYER DESTROYED()
SET WINDOW POSITION 0,0 IF playerleft
r=EstablishConnection() PRINT
DO Äplayers(FindPlayerID(playerleft)).name$
IF NET PLAYER CREATED()
Ä
," has left the session"
PRINT CHECKLIST STRING$(1),
Ä
DisplayPlayerList()
ENDIF
MaintainPlayerList() IF NET GAME NOW HOSTING()
DisplayPlayerList() PRINT "You are new host"
ENDIF ENDIF
playerleft = NET PLAYER DESTROYED() IF ESCAPEKEY()
IF playerleft EXIT
PRINT
Ä
players(FindPlayerID(playerleft)).name$
ENDIF
Ä
LOOP
," has left the session" REM *** End program ***
MaintainPlayerList() FREE NET GAME
DisplayPlayerList() END
ENDIF
LOOP
END
The complete program is coded as:
The new routine is:
FUNCTION FindPlayerID(id)

#INCLUDE "NetConnect.dba" ENDIF
NEXT c
TYPE PlayerDetailsType ENDFUNCTION result
id AS INTEGER
name$ AS STRING FUNCTION FindPlayerName(name$)
ENDTYPE result = 0
GLOBAL DIM players(10) AS PlayerDetailsType IF players(c).name$ = name$
GLOBAL noofentries AS INTEGER result = players(c).id
REM *** Use a window *** EXIT
SET WINDOW ON ENDIF
SET WINDOW SIZE 500,900 NEXT c
SET WINDOW POSITION 0,0 ENDFUNCTION result
r=EstablishConnection()
DO
IF NET PLAYER CREATED() Activity 51.18
PRINT CHECKLIST STRING$(1), REM *********************************
Ä " has joined the session" REM *** Constants ***
MaintainPlayerList() REM *********************************
DisplayPlayerList() REM *** Numeric ***
ENDIF #CONSTANT X 1
playerleft = NET PLAYER DESTROYED() #CONSTANT O 2
IF playerleft REM *** Images ***
PRINT #CONSTANT ximg 1
Ä players(FindPlayerID(playerleft)) #CONSTANT oimg 2
Ä .name$," has left the session" #CONSTANT boardimg
REM *** Sprites
3
***
DisplayPlayerList() #CONSTANT boardspr 1
ENDIF REM *********************************
IF NET GAME NOW HOSTING() REM *** Types ***
PRINT players(1).name$, REM *********************************
Ä " now hosting session" TYPE PlayerType
value
ENDIF
IF ESCAPEKEY() imgID
EXIT ENDTYPE
ENDIF
REM *** Create new player (Enter key)*** REM *********************************
IF RETURNKEY() REM *** GLOBALS ***
id = FindPlayerName("Avril") REM *********************************
IF id = 0 GLOBAL DIM board(19,19)
CREATE NET PLAYER ("Avril") GLOBAL nextsprite = 2
ENDIF GLOBAL totalO
ENDIF GLOBAL totalX
REM *** Remove new player (space bar)*** GLOBAL me AS PlayerType
IF SPACEKEY() GLOBAL you AS PlayerType
id = FindPlayerName("Avril") GLOBAL move
IF id <> 0
FREE NET PLAYER id REM *** Main logic ***
ENDIF SetUpPlayerDetails(1)
ENDIF SetUpScreen()
LOOP SetUpBoard()
REM *** End program *** REPEAT
FREE NET GAME won = -GetMove(X)
END IF NOT won
won = GetMove(O)
FUNCTION MaintainPlayerList() ENDIF
PERFORM CHECKLIST FOR NET PLAYERS UNTIL won
noofentries = CHECKLIST QUANTITY() EndGame(won)
FOR c = 1 TO noofentries END
players(c).id = CHECKLIST VALUE A(c)
players(c).name$ = REM *********************************
ÄCHECKLIST STRING$(c) REM *** Level 1 ***
REM *********************************
NEXT c
ENDFUNCTION FUNCTION SetUpPlayerDetails(r)
PRINT "SetUpPlayerDetails"
FUNCTION DisplayPlayerList() ENDFUNCTION
PRINT players(c).id, FUNCTION SetUpScreen()
Ä" ",players(c).name$
PRINT "SetUpScreen"
ENDFUNCTION
NEXT c
ENDFUNCTION
PRINT "SetUpBoard"
FUNCTION FindPlayerID(id)
ENDFUNCTION
IF players(c).id = id
result = c
EXIT

PRINT "GetMove"
ENDFUNCTION 1
PRINT "EndGame"
ENDFUNCTION
Activity 51.19
The only test stub required is for CheckForWin()
Activity 51.20
Activity 51.21
The complete program is given in LISTING-51.8.

52
Checking FTP Connection Status
Creating an FTP Connection
Downloading and Uploading Files Using FTP
Retrieving File Information Using FTP
DarkBASIC Pro: Using FTP 1435

Internet File Transfers
Introduction
A frequent requirement of any Internet based software is to transfer files from a web
site to your own machine or vice versa.
A common method of file transfer is File Transfer Protocol (FTP) and DarkBASIC
Pro has a set of commands specifically designed to handle this type of situation.
These instructions allow us to connect to a web site, examine and download files
from the site and even delete and add files.
Obviously, you need to be connected to the Internet (or have created your own
Intranet) to use these statements.
The Instructions
The FTP CONNECT Statement
Step 1 in accessing the files on a web site is to establish a connection between your
machine and that web site. This is done using the FTP CONNECT statement which
FIG-52.1
The FTP CONNECT FTP CONNECT url , name , password
Statement
In the diagram:
url is a string giving the web address of the site to be

accessed.
name is a string giving the user name required for

accessing the site's files.
password is a string giving the user's password for accessing

the files. Use an empty string if no password is
required.
Many FTP sites require you to register with them and hence the need for a name
and password. However, it is also common to allow general access to many sites,
in which case the name is often "anonymous" and there is no password.
For example, we could access the FTP site, "ftp:\\ftp.darkbasic.co.uk" using the line
FTP CONNECT "ftp:\\ftp.darkbasic.co.uk","anonymous",""
The GET FTP FAILURE Statement

There are many reasons why an attempted connection might fail; the wrong address,
the site is down, your Internet connection may be faulty. So, after executing an FTP
CONNECT statement, a program should check for a successful connection. This
1436 DarkBASIC Pro: Using FTP

can be done using the GET FTP FAILURE statement which has the format shown
in FIG-52.2.
FIG-52.2
The GET FTP
GET FTP FAILURE ( )
FAILURE Statement
integer
This statement returns 1 if an attempted FTP connection is unsuccessful; otherwise
zero is returned.
For example, we could indicate failure to connect using the lines:
FTP CONNECT "ftp://ftp.darkbasic.co.uk","anonymous",""

IF GET FTP FAILURE()
PRINT "Failed"
ENDIF
The GET FTP ERROR$ Statement

We can find out more about a failure to connect using GET FTP ERROR$ which
will return a string detailing the reason for a failure to connect. The format for this
statement is given in FIG-52.3.
FIG-52.3
GET FTP ERROR$ ( )
The GET FTP ERROR$
Statement
string
For example, rather than simply display the word failed, as in the last example, we
could display the reason for a failure to connect using the code:
FTP CONNECT "ftp://ftp.darkbasic.co.uk","anonymous",""

PRINT GET FTP ERROR$()
ENDIF
The GET FTP STATUS Statement

Even if a connection is successfully made, that does not mean that problems with
the connection may not arise later. When working on an FTP site, you should
continually check for connection problems. This can be done using the GET FTP
STATUS statement which has the format shown in FIG-52.4.
FIG-52.4
GET FTP STATUS ( )
The GET FTP STATUS
Statement
integer
This statement returns 1 if the connection is still valid, and zero if the connection
fails.

The FTP SET DIR Statement
Although we might have included directory details in the FTP CONNECT
statement, it is also possible to move to a different directory within an FTP site after
a connection has been made using the FTP SET DIR statement, which has the format
shown in FIG-52.5.
FIG-52.5
FTP SET DIR path
The FTP SET DIR
Statement
In the diagram:
path is a string specifying the new directory to be

accessed. This will become the current directory.
For example, we could change to the members folder within the site
ftp:\\ftp.digital-skills.co.uk using the lines:
FTP CONNECT "ftp:\\ftp.digital-skills.co.uk","anonymous",""

FTP SET DIR "members"
It is possible to move one level up the folder hierarchy using the path "..". So, if the
current folder was members we could return to the original level using the line:
FTP SET DIR ".."
The GET FTP DIR$ Statement

We can discover which folder is the current directory using the GET FTP DIR$
FIG-52.6
GET FTP DIR$ ( )
The GET FTP DIR$
Statement
string
This statement will return a string detailing the current folder and path. A typical
statement would be:
currentfolder$ = GET FTP DIR$()
The FTP FIND FIRST Statement

Once within a folder, we can select a specific file in various ways. One option is to
move through the list of entries within the folder until the required file is selected.
We can select the first file within the current folder using the FTP FIND FIRST
FIG-52.7
FTP FIND FIRST
The FTP FIND FIRST
Statement
For example, we could move to the first file within the members folder using the
lines:

FTP FIND FIRST
The FTP FIND NEXT Statement

We can select the next file within the current folder using the FTP FIND NEXT
statement whose format is given in FIG-52.8.
FIG-52.8
FTP FIND NEXT
The FTP FIND NEXT
Statement
The GET FTP FILE TYPE Statement

The currently selected file's type can be determined using the GET FTP FILE TYPE
FIG-52.9
GET FTP FILE TYPE ( )
The GET FTP FILE
TYPE Statement
integer
The statement returns one of the values shown in TABLE-52.1.

TABLE-52.1 Code File Type
Possible Return Values -1 No more files

from FTP GET FILE 0 Normal file
1 Sub-folder
TYPE
The GET FTP FILE NAME$ Statement

The name of the currently selected file can be discovered using the GET FTP FILE
NAME$ statement which has the format shown in FIG-52.10.
FIG-52.10
GET FTP FILE NAME$ ( )
The GET FTP FILE
NAME$ Statement
string
The GET FTP FILE SIZE Statement

The size of the currently selected file can be determined using the GET FTP FILE
SIZE statement which has the format shown in FIG-52.11.
FIG-52.11
The GET FTP FILE
GET FTP FILE SIZE ( )
SIZE Statement
integer

The size of the file is given in bytes.
With the following code we could display the details of every file within the
members folder:

FTP FIND FIRST
WHILE GET FTP FILE TYPE()<>-1
PRINT "TYPE:";GET FTP FILE TYPE()
PRINT "NAME:";GET FTP FILE NAME$()
PRINT "SIZE:";GET FTP FILE SIZE()
FTP FIND NEXT
ENDWHILE
The FTP DISCONNECT Statement

To terminate connection to an FTP site, we need to use the FTP DISCONNECT
FIG-52.12
The FTP DISCONNECT
FTP DISCONNECT
Statement
The program in LISTING-52.1 lists the details of the files set up by The Game
Creators' site specifically for testing FTP access.
LISTING-52.1 REM *** Connect to FTP site ***

FTP CONNECT "www.darkbasic.com","",""
Display Details of the REM *** IF there's a problem, show message and stop ***
Files Available IF GET FTP FAILURE()
PRINT "Failed"
WAIT KEY
END
ENDIF
REM *** Find the first file ***
FTP FIND FIRST
REM *** WHILE not at end of files list ***
WHILE GET FTP FILE TYPE()<>-1
REM *** Display the file's details ***
PRINT "TYPE: ",GET FTP FILE TYPE()
PRINT " NAME: ",GET FTP FILE NAME$()
PRINT " SIZE: ",GET FTP FILE SIZE()
REM *** Move on to next file ***
FTP FIND NEXT
ENDWHILE
REM *** Disconnect ***
FTP DISCONNECT
WAIT KEY
END
Activity 52.1
Type in and test the program in LISTING-52.1 (ftp01.dbpro).
The FTP GET FILE Statement

What we've covered so far allows us look at file details, but if we want access to
the contents of a specific file, we need to use the FTP GET FILE statement which

FIG-52.13
The FTP GET FILE
FTP GET FILE filename [[ , localname [ , blocksize
Statement
In the diagram:
filename is a string giving the name of the file to be

accessed.
localname is a string giving the file name to be used for

the copy of the file made on your own hard drive.
blocksize is an integer value giving the size of data blocks

to be transmitted. Typically, this will be 512,
1024, 2048, etc.
For example, we could download the welcome.msg text file from the Game Creators'
site and save it on our own machine as temp.txt using the program in LISTING-52.2.
REM *** Connect to site ***

LISTING-52.2 FTP CONNECT "www.darkbasic.com","",""
Downloading a File PRINT "Failed"
WAIT KEY
END
ENDIF
REM *** Down load file ***

FTP GET FILE "welcome.msg","temp.txt"
IF GET FTP FAILURE()=0
PRINT "File downloaded"
ELSE
PRINT "Download failed"
ENDIF
REM *** Disconnect ***

FTP DISCONNECT

WAIT KEY
END
Activity 52.2
Type in and test the program in LISTING-52.2 (ftp02.dbpro).
Use Notepad to read the file you have downloaded.
The FTP PROCEED Statement

Although FTP GET FILE works well on small files, larger files need to be fetched
a block at a time. In that situation we use the FTP PROCEED statement to download
each block of the file. This statement has the format shown in FIG-52.14.
FIG-52.14
The FTP PROCEED
Statement
FTP PROCEED

The GET FTP PROGRESS Statement
Once we start to download a file using the FTP GET FILE statement, we can check
on the progress of the download using the GET FTP PROGRESS statement which
FIG-52.15
GET FTP PROGRESS ( )
The GET FTP
PROGRESS Statement
integer
The statement returns the percentage of the file's contents which has been
downloaded to your machine. When the download is complete, -1 is returned.
For example, let's assume that the Game Creators' area contains a file named
panorama.bmp (the file doesn't actually exist), then we could download it using the
following code:
FTP GET FILE "panorama.bmp","photo.bmp",1024

WHILE GET FTP PROGRESS()<>-1
CLS
PRINT "Downloading..."
PRINT "PROGRESS:",GET FTP PROGRESS()
FTP PROCEED
ENDWHILE
ELSE
PRINT "Download failed"
ENDIF
The FTP TERMINATE Statement

Should we need to halt a download before it has been completed, then we can use
the FTP TERMINATE statement which has the format shown in FIG-52.16.
FIG-52.16
FTP TERMINATE
The FTP TERMINATE
Statement
For example, we could allow the user to halt a download by pressing any key using
the lines
IF INKEY$() <> ""

FTP TERMINATE
ENDIF
in the loop structure given on the previous page.
The FTP DELETE FILE statement

Assuming we have been assigned the appropriate rights of access, we can also delete
a file on an FTP site using the FTP DELETE FILE statement which has the format
shown in FIG-52.17.

FIG-52.17
The FTP DELETE FILE FTP DELETE FILE filename
Statement
In the diagram:
filename is a string giving the name of the file to be deleted.
The FTP PUT FILE Statement

A file can also be uploaded to an FTP site (assuming we have the appropriate
privileges) using the FTP PUT FILE statement which has the format shown in
FIG-52.18.
FIG-52.18
FTP PUT FILE filename
The FTP PUT FILE
Statement
In the diagram:
filename is a string giving the name of the file to be copied

to the FTP site.
The code below demonstrates how a file (myfile.dat) might be uploaded to an FTP
site:
FTP CONNECT "ftp://ftp.yoursite.com","",""

PRINT "Could not connect"
PRINT "Error:";GET FTP ERROR$()
WAIT KEY
END
ENDIF
FTP PUT FILE "myfile.dat"
PRINT "File uploaded"
ELSE
PRINT "Upload failed"
ENDIF
Summary
l FTP stands for File Transfer Protocol.
l An FTP web site is designed to hold files and allow those with the appropriate
rights to store and retrieve files from that site.
l Use FTP CONNECT to connect to an FTP site.
l Use GET FTP FAILURE to check if an attempted connection to an FTP site has
failed.
l Use GET FTP ERROR$ to access a string specifying why an attempted FTP
connection has failed.
l Use GET FTP STATUS to check that an FTP connection continues to operate
correctly.

l Use FTP SET DIR to change the directory being accessed on an FTP site.
l Use GET FTP DIR$ to discover which directory within an FTP site is currently
being accessed.
l Use FTP FIND FIRST to select details of the first entry in the directory being
accessed.
l Use FTP FIND NEXT to select details of the next entry in the directory being
accessed.
l Use GET FTP FILE TYPE to discover the type of the currently selected entry.
l Use GET FTP FILE NAME to get the name of the currently selected entry.
l Use GET FTP FILE SIZE to get the size of the currently selected entry.
l Use FTP DISCONNECT to disconnect from the FTP site being accessed.
l Use FTP GET FILE to download the contents of a file from the FTP site.
l Use FTP PROCEED to get a file's next block of data from an FTP site. This is
used when large files are being downloaded.
l Use GET FTP PROGRESS to discover what percentage of a large file has been
downloaded so far.
l Use FTP TERMINATE to halt the downloading of a large file.
l Use FTP DELETE FILE to delete a file from the FTP site.
l Use FTP PUT FILE to upload a file to an FTP site.
No solutions are required for this chapter.

53
Adding Help Pages to the DarkBASIC Pro IDE

Creating New DarkBASIC Pro Statements
Creating a String Table
Using Standard DLLs
Writing DLLs in C++
DarkBASIC Pro: DLLs 1445

Creating New DBPro Statements
Introduction
If you are fluent in another programming language such as C, C++, or Delphi, it is
possible to create new program language statements which can be embedded in your
copy of DarkBASIC Pro. You can then use these new commands in any future
DarkBASIC Pro project that you write.
Obviously, this is not the place to start teaching you another programming language,
but if you already know an appropriate language, then expanding DarkBASIC Pro
to meet your own specific needs is a very powerful feature.
A Dynamic Link Library (DLL)

It is possible to write a collection of functions and place them in a single library file.
In fact, we've already done the same sort of thing back in Chapter 8 when we created
a collection of string-handling functions.
Once created, a library can be linked to any new application being written and the
function within that library can then be used by the application. One way to create
such a library of functions is in a Dynamic Link Library or DLL.
In fact, all the commands in DarkBASIC Pro are held in a series of DLLs. By adding
our own DDLs we can expand the statements available to us in DarkBASIC Pro.
Creating a DLL
As we stated earlier, it is possible to create a DLL library using many different
programming languages, but here we are going to make use of Visual C++ .NET.
Obviously, if you know C or C++ what follows will be much more meaningful, but
for others, a step-by-step guide of how to add your own instructions should be useful
in the future.
Starting Up Visual Studio

Visual Studio offers many options on start up, but we are going to create C++
Windows 32-bit code, so the options to be selected are those shown in FIG-53.1.
Specifically, we need to create a DLL (rather than an application), so we need to

select the option shown in the dialog box that appears (see FIG-53.2).
At last we have created the project we require. By double clicking on the source
code file listed in the folder-structure on the left-hand side of the screen, we can
bring the file's contents into the edit window (see FIG-53.3). Notice that a few lines
have been created automatically by Visual Studio.
Activity 53.1
Following the instructions given in FIG-53.1 to FIG-53.3, create a DLL

named StringFunctions.
1446 DarkBASIC Pro: DLLs

FIG-53.1
Choosing the Basic
Options
1
Click on
New Project
2
Select
Win32 Project
3
Enter a name 4
for your DLL Choose the folder
in which the DLL is to
be saved
5
Click on OK
FIG-53.2 1
Select
Specifying the DLL Application Settings
Option 2
Select DLL
3
Click on Finish
FIG-53.3
1
The DLL Source File Double click on
The DLL name
2
This prewritten
code will appear in
the edit area

Adding the Code for New Statements
When the StringFunctions.cpp file is opened by double clicking, the editor contains
the following lines:
// StringFunctions.cpp : Defines the entry point for the DLL

Ä application.
//
#include "stdafx.h"
BOOL APIENTRY DllMain ( HANDLE hModule,
DWORD ul_reason_for_call,
LPVOID lpReserved
)
{
return TRUE;
}
Routines which are to be used in other applications need to begin with the rather
strange phrase:
__declspec(dllexport)
but since this is rather awkward to type, we can create a substitute term in a #define
statement:
#define DarkDLL __declspec(dllexport)
Now when we write our code, we can use the term DarkDLL rather than
__declspec(dllexport). Just before compilation begins, the compiler will replace all
occurrences of the term DarkDLL with the correct term, __declspec(dllexport).
Since we will not be using the additional features of C++, it will be easier if we tell
the compiler to treat our code as C rather than C++. To do this we need to add the
line
extern "C"
braces are the symbols and then enclose everything we write within a set of braces.
{}
Next we need to write a function for every command we intend to add to
DarkBASIC Pro. In this example, we are going to add POS which will return the
position of a specified character within a string and OCCURS which will return the
number of times a character occurs within a string. Both of these functions were
described in Chapter 8.
The code required for the two functions is:
DarkDLL int pos(LPSTR s1, LPSTR ch)

{
int result = -1;
for(int c = 0; c < strlen(s1);c++)
if (s1[c] == ch[0])
{
result = c+1;
break;
}
return result;
}
DarkDLL int occurs(LPSTR s1, LPSTR ch)

{
int result = 0;

if (s1[c] == ch[0])
result++;
return result;
}
Activity 53.2
Under the existing lines of code within StringFunctions.cpp, add the #define
line followed by the code for the two functions, which must be enclosed in an
extern "C" block.
Click on Build|Build Solution and make sure the code compiles correctly.
Adding a String Table

Next we need to include information which the DarkBASIC Pro compiler can read.
This information tells DarkBASIC Pro the function names, the type of parameters
they require, the types of values returned, and the term to be used in DarkBASIC
to call up the function. All of this information is held in a string table. To create a
string table we start by selecting Add Resource (see FIG-53.4).
FIG-53.4
Adding a Resource
1
Right click on
StringFunctions
2
Choose Add... 3
...then click on
Add Resource
Clicking on Add Resource will produce a dialog box in which we select the String
Table option (see FIG-53.5).
FIG-53.5
Selecting a String Table
2
Click on New
1
Select String Table

Once String Table has been selected, the project files are updated and we need to
add an entry to the string table for each of our functions that are included in the DLL
(see FIG-53.6).
FIG-53.6
An Entry in the String
Table 5
1 Click in the
A .rc resource file Caption area of the first line
has been added to the 3 to create an edit box
4
project... 2 A String Table
We need to enter details
...within this is the entry contains 3 fields:
in the Caption area
String Table ID, Value and Caption
Constructing the Caption

The caption is made up of several parts with a percent symbol (%) separating each
part. These parts are:
The name of the new DarkBASIC Pro statement

This is whatever word or words are to be used to
invoke execution of the function. This need not
be the same as the function's name. If the function
returns a value, the name should end with an
opening square bracket ( [ ).
Parameter list The type of the function's return value and

each of its parameters are coded as a letter.
The letters used are shown in TABLE-53.1.
TABLE-53.1 Code Parameter Type
Parameter List Codes L integer

F real (float)
S string
O double
R longint
D others
0 void
If you haven't enclosed your
functions in an extern "C"
block, C++ creates mangled The function's name The name of the function.
function names and it's these
mangled names that must be In fact, C++ makes up its own name for a function,
used in the string table. You but by using the extern "C" statement we have
can find out what the mangled ensured that the original function name is used.
name is by searching the DLL For example, without extern the function occurs
file that we will create for the
original file name. could be called ?occurs@@YAKH.

The parameters in English
This part lists the parameters of the function in
English and can be used to add help for the new
instruction in the DarkBASIC Pro environment.
Now we're ready to attempt the caption for our first entry in the string table. Since
the first function is called pos and we want to use the same term in DarkBASIC Pro,
then our statement name will be
POS
but, since the function returns a value, we need to change this to
POS[
Next we have a separator, %, the return type, and the parameter list. The function
returns an integer (L) and takes a string (S) and character (S) as parameters, so the
caption now becomes:
POS[%LSS
Another separator and the function's C++ name give us:
POS[%LSS%pos
And, finally, a separator and the parameter list in English:
POS[%LSS%pos%string,string
Activity 53.3
Write down the caption required for the occurs() function. Enter the captions
for both functions into the string table and close the table.
Adding the New Statements to DarkBASIC Pro

With the string table complete, we need to rebuild our code (Build|BuildSolution).
This creates a file called StringFunctions.dll which is placed in a sub-folder named
debug.
DarkBASIC Pro contains a folder named Compiler, and within this is a sub-folder
named plugins-user. We need to copy StringFunctions.dll into this folder. Once this
FIG-53.7 is done, the new statements, POS and OCCURS will be recognised by DarkBASIC
Pro (see FIG-53.7).
Copying the DLL
A$ = "ABCDEFBBBFRGB"
B$ = "B"
PRINT B$," appears ",OCCURS(A$,B$)," times"
StringFunctions.dll PRINT B$," first appears at position ", POS(A$,B$)
WAIT KEY
END
StringFunctions.dll
Building the code within C++ will This file must be copied into Now the new statements can be used
create a file named StringFunctions.dll Plugins-user in the DarkBASIC Pro's within DarkBASIC Pro.
in the Debug sub-folder. Compiler folder

Activity 53.4
Carry out the steps explained above and test out the OCCURS and POS
statements in a short DarkBASIC Pro program.
Adding Help
Our new commands aren't fully integrated into DarkBASIC Pro because, even
though they function correctly, they are not highlighted in blue and don't bring up
a help window when you move the cursor over the statement and press F1. But,
with a bit more effort, we can have all that happen too.
The blue highlight is created by having a command listed in one of several special
.ini files. These files are in DarkBASIC Pro's Keywords folder within the Editor
(see FIG-53.8).
FIG-53.8
The .ini Files
Only statements listed in one of these files will appear with a blue highlight in the
edit window, so if we want our commands to be highlighted, they must either be
added to one of the existing .ini files, or placed in a new .ini file.
In this case, we'll create a new .ini file - that way it will be easier to remove all our
changes should that be required.
Activity 53.5
Create a new file in Notepad and add the following lines:
OCCURS=commands\extras\OCCURS.htm=(searched,searchedfor)
POS=commands\extras\POS.htm=(searched,searchedfor)
Save the file as myfunctions.ini.
Load up the DarkBASIC Pro program you created in Activity 53.4. Are the
terms OCCURS and POS now highlighted in blue?
Although we've created the highlighting, we still need to create a help page for each
statement. In fact, the .ini file actually lists details of the help file name and where
it is to be found. For example, in the line
OCCURS=commands\extras\OCCURS.htm=(searched,searchedfor)
the term commands\extras\OCCURS.htm tells DarkBASIC Pro that the help page
for the OCCURS statement is in a file named OCCURS.htm which is to be found

in the extras sub-folder within the commands folder which is itself within
DarkBASIC Pro's Help folder (see FIG-53.9)
FIG-53.9
The extras Folder
Activity 53.6
The extras folder doesn't yet exist, so it needs to be created.
Create an extras sub-folder within Help's commands folder as shown above.
Now we need to create each of the help files. These are just HTML files giving
details of a command. The code for OCCURS.htm is shown below:
<BODY COLOR=black BACKGROUND="..\gfx/backdrop.jpg"

BGPROPERTIES=FIXED>

OCCURS

 

This command will return the number of times a specified character occurs within a
string.


Syntax

Return integer=OCCURS(string,character) 
<ul>
</ul>

Related Information


<a href="..\extras.htm">New Commands Menu 
</a>
<a href="..\..\index.htm">Index</a> 


Example Code


<code>
PRINT OCCURS("BADFBBASB","B")
</code>
 
 The above statement would display 4 since B occurs 4 times within BADFBBASB



</BODY>
This will create the page shown in FIG-53.10.

FIG-53.10
The Help Page for
OCCURS
The code for the second file, POS.htm, is:
<BODY COLOR=black BACKGROUND="..\gfx/backdrop.jpg"

BGPROPERTIES=FIXED>

POS

 

This command will return the first position at which a specified character occurs within a
string.
If the character does not appear within the string, the value -1 is returned.


Syntax

Return integer=POS(string,character) 
<ul>
</ul>

Related Information


<a href="..\extras.htm">New Commands Menu 
</a>
<a href="..\..\index.htm">Index</a> 


Example Code


<code>
PRINT POS("BADFBBASB","B")
</code>
 
 The above statement would display 1 since B first occurs at position 1 within
BADFBBASB.
 
<code>
<pre>
A$ = "ABCDEF"
B$ = "X"
result = POS(A$,B$)
</pre>
</code>

result would be set to -1 since X does not appear within ABCDEF.

</BODY>
Activity 53.7
Create both OCCURS.htm and POS.htm, containing the code given and save
them in the extras folder.
It is usual for all the instructions that belong to a specific group (for example, camera
statements, lighting statements, etc.) to have an HTML file listing each of the
commands in that group with links to their individual help pages. In fact, our two
new pages already have links to the extras command list page - even though we
haven't created it yet!
This last help-related file, extras.htm, has the following code:
<BODY COLOR=black BACKGROUND="gfx/backdrop.jpg"

BGPROPERTIES=FIXED><CENTER>
EXTRAS</CENTER> 
 
<PRE>
<a href="extras\OCCURS.htm">OCCURS</a> : Return integer =
OCCURS(searched,searchedfor)
<a href="extras\POS.htm">POS</a> : Return integer = POS(searched,searchedfor)

</PRE> 
CLICK <A HREF="..\main.htm">HERE</A> TO RETURN TO THE MAIN MENU 
 
</BODY>
and creates the page shown in FIG-53.11.
FIG-53.11
A Group Help File
Activity 53.8
Create extras.htm and save it in the DarkBASIC Pro's commands folder.
At last we have a completely integrated set of new commands.

Adding More New Commands
Some new commands are going to prove to be slightly more difficult to create than
OCCURS and POS.
Functions that Return Real Values

Functions which return real values should in theory be written something like:
DarkDLL float returnfloat()

{
float result = 12.3F;
return result;
}
But although the code will compile, the value returned bears no resemblance to 12.3.
To make the whole thing work as expected, we need to start by changing the return
type to DWORD
DarkDLL DWORD returnfloat()
and then the returned value must be modified to
return *(DWORD*)&result;
which means the whole routine is actually coded as:
DarkDLL DWORD returnfloat()

{
float result = 12.3F;
}
The string table caption for this routine could be:
RETURNREAL[%F%returnfloat
Notice on this occasion the term used in DarkBASIC Pro is RETURNREAL, while
the function is actually called returnfloat. Also, since there are no parameters, the
final part of the entry is missing.
Functions that Return Strings

When we need to return a string from a function, things get seriously weird! To
start with we need to include a file named globstruct.h using the line:
#include "globstruct.h"
You'll find this file in some examples embedded in DarkBASIC Pro in the folder
program files|dark basic professional|help|documents|files|Third Party Commands

Ä |TESTCOMMANDS2
The file must be copied into the folder containing your DLL source code.
Unfortunately, globstruct.h makes references to other header files from the DirectX
development kit. If you don't have this, you won't be able to compile your program.
However, assuming only simple functions are required in your DLL we can doctor

globstruct.h slightly to compile without reference to yet more header files.
Activity 53.9
Find globstruct.h and copy it into the DLL source code folder.
Make the following changes to the file's code:
Comment out the line #include "D3dx9tex.h"
Change the lines
// Bitmap and Surface Data (for drawing offscreen)

int iCurrentBitmapNumber;
LPDIRECT3DTEXTURE9 pCurrentBitmapTexture;
LPDIRECT3DSURFACE9 pCurrentBitmapSurface;
LPDIRECT3DSURFACE9 pHoldBackBufferPtr;
LPDIRECT3DSURFACE9 pHoldDepthBufferPtr ;
to
// Bitmap and Surface Data (for drawing offscreen)
int iCurrentBitmapNumber;
void* pCurrentBitmapTexture;
void* pCurrentBitmapSurface;
void* pHoldBackBufferPtr;
void* pHoldDepthBufferPtr;
Also change the line
char pEXEUnpackDirectory[_MAX_PATH];
to
char pEXEUnpackDirectory[260];
Back in the code for the DLL, we need to create a global variable immediately after
the new #include statement:
GlobStruct* g_pGlob = NULL;
Activity 53.10
In your DLL source file, add the lines
immediately after
#include "stdafx.h"
A function that returns a string has to be written as returning a DWORD type. But
it also needs to use a phantom parameter of the same type:
DarkDLL DWORD returnstring(DWORD olds)
Within the routine we need to treat the parameter as a pointer and delete any space
it is referencing with the lines:

if(olds)
g_pGlob->CreateDeleteString((DWORD*)&olds,0);
The string which is going to contain the result needs to be set up with lines such as:
LPSTR strresult=NULL;
g_pGlob->CreateDeleteString((DWORD*)&strresult,10);
The value at the end of the second line is the size of the string, taking into account
that strings in C++ must be assigned an extra character position for the terminating
null character.
Now, as long as we cast strresult to LPSTR we can treat it as a normal C++ string.
For example, we could store the character sequence ABCD in the string using the
line:
strcpy((LPSTR)strresult,"ABCD");
To return the string from the function we need to end with a line such as
return (DWORD)strresult;
casting the result to a DWORD type.
So, the code for the complete function would be:

{
if(olds)
g_pGlob->CreateDeleteString((DWORD*)&olds,0);
g_pGlob->CreateDeleteString((DWORD*)&strresult,5);
}
The Caption in the string table holds another surprise. We'll start by using a
DarkBASIC Pro type name for the new instruction
ABCD$[
and since it is designed to return a string, this expands to
ABCD$[%S
but the phantom parameter is ignored, so we continue directly to the function name:
ABCD$[%S%returnstring
Because there are no parameters, the final comment section of the string can be
omitted or the term void used.
Activity 53.11
Add returnstring() to your DLL file along with an entry in the string table.
Try out the new statement in a DarkBASIC Pro program.
After testing, remove the returnstring() function and its string table entry from
your C++ project.

More String Handling Functions
More practical string-handling functions are given in LISTING-53.1. Again these
are based on the string functions described in Chapter 8. The functions are:
INSERT$ (A$,B$,p) which returns a string created by embedding

string B$ within A$ starting at position p.
DELETE$(A$,p,n) which returns a string formed by deleting n

characters from A$ starting with the character
at position p.
REPLACE$(A$,B$,p) which returns a string formed by replacing the

pth character in A$ with the character B$.
LISTING-53.1 DarkDLL DWORD insertstr(DWORD olds, DWORD s1, DWORD s2, int post)
{
C++ Code for the New if(olds)
Statements g_pGlob->CreateDeleteString ( (DWORD*)&olds, 0 );
if (post < 1)
post = 1;
if (post > strlen((LPSTR)s1)+1)
post = strlen((LPSTR)s1)+1;
g_pGlob->CreateDeleteString((DWORD*)&strresult,
Ästrlen((LPSTR)s1)+strlen((LPSTR)s2)+1 );
strncpy(strresult,(LPSTR)s1,post-1);
strresult[post-1]='\0';
strcat(strresult,(LPSTR)s2);
strcat(strresult,&((LPSTR)s1)[post-1]);
}
DarkDLL DWORD deletestr(DWORD olds, DWORD s, int start, int num)

{
if(olds)
g_pGlob->CreateDeleteString ( (DWORD*)&olds, 0 );
if(start < 1 || start > strlen((LPSTR)s)||num<1)
{
Ä
strlen((LPSTR)s)+1);
strcpy(strresult,(LPSTR)s);
}
else
{
g_pGlob->CreateDeleteString((DWORD*)&strresult,num+1);
strncpy(strresult,(LPSTR)s,start-1);
strresult[start-1] = '\0';
if(start+num > strlen((LPSTR)s))
num = strlen((LPSTR)s)-start;
strcat(strresult,&((LPSTR)s)[start+num-1]);
}
}

LISTING-53.1 DarkDLL DWORD replacechr(DWORD olds, DWORD s, DWORD ch, int post)
(continued) {
if(olds)
C++ Code for the New g_pGlob->CreateDeleteString ( (DWORD*)&olds, 0 );
Statements LPSTR strresult=NULL;
Ä
strlen((LPSTR)s)+1);
if(post < 1 || post > strlen((LPSTR)s)||strlen((LPSTR)ch)>1)
strcpy(strresult,(LPSTR)s);
else
{
strncpy(strresult,(LPSTR)s,post-1);
strresult[post-1] = ((LPSTR)ch)[0];
strresult[post]='\0';
strcat(strresult,&((LPSTR)s)[post]);
}
}
Activity 53.12
Add the three functions given in LISTING-53.1 to your DLL source code.
Add appropriate entries in the string table.
Test the new statements in a DarkBASIC Pro program.
Summary
l A Dynamic Link Library (DLL) is a file containing a collection of functions.
l DLLs can be created using a variety of languages.
l In Visual Studio's C++, the core of a DLL can be created automatically.
l Functions within a DLL that are designed to be used elsewhere should be

preceded by the term __declspec(dllexport).
l By including all functions within an extern "C" block, function names used by
the operating system match those given in the source code.
l Functions not included in an extern block have new names allocated by the C++
compiler.
l C++ generated names are known as mangled or decorated function names.
l The mangled name of a function can be found within the DLL file generated by
the C++ compiler.
l Return values and parameters used when coding DLL functions must match
those required by DarkBASIC Pro.
l Real parameters should be declared as type DWORD.
l String parameters should be LPSTR or DWORD.
l A string table must be added to the DLL project.

l Each entry in a string table gives the name to be used when the function is called
from DarkBASIC Pro, the function's return type and parameters, the actual
function name, and a comment listing the parameter types.
l New commands can be highlighted in blue in the DarkBASIC Pro editor by

adding a .ini file to the editor's Keywords folder.
l Entries in a .ini file give the name of the command, the path to the corresponding
HTML help file and a parameter list.
l An HTML file should be added to the appropriate folder for each new command.
l If a function returns a string the DLL should include globstruct.h.
l Elements of globstruct.h must be edited or other .h files from DirectX SDK must
be made available.
l A function that returns a string must include a phantom DWORD parameter.
l The phantom parameter is omitted from the string table and ignored when calling
the function.

Using Standard DLLs
Introduction
DLLs don't have to be used just to create new statements for DarkBASIC Pro. It is
also possible to make use of other DLLs simply because they contain functions
which would be difficult or impossible to implement in DarkBASIC Pro.
However, any functions within a DLL using real or string values must still be written
using the techniques explained in the previous section of this chapter.
LISTING-53.2 shows the contents of a DLL file named functions.dll.

LISTING-53.2 // Functions.cpp : Defines the entry point for the DLL
Standard DLL Functions
Äapplication.
//
#include "stdafx.h"
BOOL APIENTRY DllMain( HANDLE hModule,
LPVOID lpReserved
)
{
return TRUE;
}
#define DarkDLL _declspec(dllexport)

extern "C"
{
DarkDLL LONGLONG factorial(int v)
{
LONGLONG result = 1;
for(int c = 2; c <= v; c++)
result *=c;
return result;
}
DarkDLL DWORD reciprocal(int v)

{
float result = 1/v;
}
}
The file contains two functions.
The first, factorial(), returns the factorial of a specified value. A factorial of a value
is the result obtained by multiplying the value by every integer from 2 up to itself.
For example, factorial 5 (written in mathematics as 5!) is 2*3*4*5, or, 120.
The second function, reciprocal(), returns the value of 1 divided by the number
given. For example, the reciprocal of 5 is 1/5, or 0.2.
Activity 53.13
Create a new C++ Win32 DLL project and type in the code given in
LISTING-53.2. Check that the code compiles correctly.

The LOAD DLL Statement
To use the functions in a DLL which have not been added as new statements in
DarkBASIC Pro requires the DLL to first be loaded into our program. This is done
using the LOAD DLL statement which has the format shown in FIG-53.12.
FIG-53.12
The LOAD DLL LOAD DLL filename , dllno
Statement
In the diagram:

the DLL. This file should be in the current
DarkBASIC Pro project's folder.
dllno is an integer value giving the ID to be assigned

to the DLL.
For example, we could load functions.dll into our current project using the line:
LOAD DLL "functions.dll",1
The DLL EXIST Statement

It is possible that a DLL will not be successfully loaded - normally because we've
forgotten to copy the DLL file into our current project's folder. To check that it has
been loaded successfully, we need to use the DLL EXIST statement which returns
1 if the DLL specified has been loaded, and zero, if there's been a problem. This
FIG-53.13
The DLL EXIST Statement
DLL EXIST ( dllno )
integer
In the diagram:
dllno is an integer value specifying the ID of the DLL

to be checked.
Our program should check for the DLL being loaded immediately after the LOAD
DLL statement:
LOAD DLL "functions.dll",1

IF DLL EXIST(1)=0
PRINT "DLL not found"
ENDIF
The CALL DLL Statement

With a DLL successfully loaded, we are free to execute any of the functions within
the DLL using the CALL DLL statement which has the format shown in FIG-53.14.

FIG-53.14
The CALL DLL
CALL DLL ( dllno , functionname [ , value )
Statement
In the diagram:
dllno is an integer value giving the ID of the DLL

containing the function to be executed.
functionname is a string giving the name of the function to be

executed.
value represents any parameter that is to be passed to

the function. There can be as many
comma-separated values as required by the
function.
For example, we could execute the factorial() function in our DLL using the line
CALL DLL(1,"factorial",5)
which would return the value of 5!
The DLL CALL EXIST Statement

If we're going to play extra safe - as we should - we should check that a function of
the name specified actually exists within the DLL before we attempt to execute that
function. This can be done using the DLL CALL EXIST statement whose format
is given in FIG-53.15.
FIG-53.15
The DLL CALL EXIST DLL CALL EXIST ( dllno , functionname )
Statement
integer
In the diagram:
dllno is an integer giving the ID of the DLL containing the

function being checked.
functionname is a string giving the name of the function to be

checked.
The statement returns 1 if the specified function exists within the DLL, otherwise
zero is returned.
The DELETE DLL Statement

When we've finished with a DLL, we can ask the operating system to remove the
DDL from memory. This is done using the DELETE DLL statement whose format

FIG-53.16
DELETE DLL dllno
The DELETE DLL
Statement
In the diagram:
dllno is an integer value giving the ID of the DLL

to be removed from memory.
In fact, the DLL is only removed if no other program is using it.
The program in LISTING-53.3 demonstrates the use of a standard DLL by

executing the factorial() and reciprocal() routines in the functions.dll.
LISTING-53.3 REM *** Load the DLL ***
LOAD DLL "Functions.dll",1
Using a Standard DLL REM *** IF it loaded THEN ***
IF DLL EXIST(1)
REM *** IF factorial exists, execute it ***
IF DLL CALL EXIST(1,"factorial")
result = CALL DLL(1,"factorial",5)
PRINT "5! = ",result
ENDIF
REM *** IF reciprocal exists, execute it ***
IF DLL CALL EXIST (1,"reciprocal")
result# = CALL DLL(1,"reciprocal",10)
PRINT " 1/10 = ",result#
ENDIF
REM *** Delete the DLL ***
DELETE DLL 1
ELSE
REM *** Display message if DLL not found ***
PRINT "DLL not found"
ENDIF
WAIT KEY
END
Activity 53.14
Type in the program in LISTING-53.3 (dll03.dbpro).
Copy the functions.dll file into the dll03 folder.
Run the program and check that the two functions are executed correctly.
Summary
l DLLs not designed to create new DarkBASIC Pro instructions can also be called
from a DarkBASIC Pro program.
l The functions in such a DLL must conform to DarkBASIC Pro's method of

dealing with real and string values.
l No string table is required for these DLLs.
l Use LOAD DLL to load a required DLL into a program.
l Use DLL EXIST to check that a DLL has been loaded.

l Use CALL DLL to call a function within a DLL.
l Use DLL CALL EXIST to check that a function of a specified name exists within
a DLL.
l Use DELETE DLL to delete a DLL from memory.
At last, we've reached the end of two massive volumes covering most of
DarkBASIC Pro's statements. I hope you have found the effort worthwhile and are
inspired to start work on a game of your own.
But remember, there's a lot more you can learn from keeping in touch on The Game
Creators forums where you will find people with an amazing amount of experience
who are more than willing to help with any problems you have.
If you find any mistakes in this text or have any ideas about extra topics that might
be covered, I would genuinely like to hear from you. You can contact me at
alistair@digital-skills.co.uk. You'll also find supplementary material on the
DarkBASIC Pro section of our website at www.digital-skills.co.uk.

Solutions
No solution required. The new statements, POS and OCCURS, should now be
highlighted in blue.
Activity 53.2
Activity 53.6
The code in StringFunctions.cpp should be:
// StringFunctions.cpp : Defines the entry
Äpoint for the DLL application.
// Activity 53.7
#include "stdafx.h" No solution required.
BOOL APIENTRY DllMain
( Activity 53.8
HANDLE hModule,
DWORD ul_reason_for_call, No solution required.
LPVOID lpReserved
)
{ Activity 53.9
return TRUE;
} No solution required.
#define DarkDLL __declspec(dllexport)
Activity 53.10
extern "C"
{ No solution required.
DarkDLL int pos(LPSTR s1,LPSTR ch)
{
int result = -1;
Activity 53.11
if (s1[c] == ch[0]) The code in StringFunctions.cpp should now be:
{
result = c+1;
// StringFunctions.cpp : Defines the entry
}
break;
Äpoint for the DLL application.
//
return result;
}
#include "stdafx.h"
DarkDLL int occurs(LPSTR s1,LPSTR ch)
{
int result = 0;
BOOL APIENTRY DllMain
if (s1[c] == ch[0])
(
result++;
HANDLE hModule,
return result;
}
LPVOID lpReserved
}
)
{
return TRUE;
Activity 53.3 }
The string table caption for occurs() should be: #define DarkDLL __declspec(dllexport)
OCCURS[%LSS%occurs%string,string extern "C"
{
DarkDLL int pos(LPSTR s1,LPSTR ch)
Activity 53.4 {
int result = -1;
A very simple DarkBASIC Pro program such as for(int c = 0; c < strlen(s1);c++)
if (s1[c] == ch[0])
PRINT POS("ABCD","C") {
PRINT OCCURS("ABBBCDBBDB","B") result = c+1;
WAIT KEY break;
END }
return result;
should act as a basic test of the new statements. }
DarkDLL int occurs(LPSTR s1,LPSTR ch)

{
int result = 0;

if (s1[c] == ch[0])
result++;
return result;
}

{
if(olds)
g_pGlob->CreateDeleteString
Ä
((DWORD*)&olds,0);
g_pGlob->CreateDeleteString
Ä((DWORD*)&strresult,5);
}
}
A DarkBASIC Pro program such as
PRINT ABCD$()
WAIT KEY
END
will call the new routine and display ABCD.
Activity 53.12
String table entries for the new functions should be:
INSERT$[%SSSL%insert%string,string,int
DELETE$[%SSLL%deletestr%string,int,int
REPLACE$[%SSSL%replacechr%string,character,int
A DarkBASIC Pro program such as
PRINT INSERT$("ABCD","XYZ",2)
PRINT DELETE$("ABCD",2,2)
PRINT REPLACE$("ABCD","X",3)
WAIT KEY
END
should give an indication that the new statements are

working.
Activity 53.13
Activity 53.14

INDEX
3D primitives 749 Constructive Solid Geometry (CSG) 788
3D Vectors 1350 Contact joint 1340
column vectors 1350 CONVERT OBJECT FVF 1386
direction 1350 CONTROL CAMERA USING ARROWKEYS 856
elements 1350 Coordinates system 744
magnitude 1350 COPY MATRIX4 1367
row vectors 1350 COPY MEMBLOCK 1287
4 by 4 Matrices 1365 COPY VECTOR3 1353
4D Vectors 1363 CREATE NET GAME 1394
homogeneous coordinates 1363 CREATE NET PLAYER 1412
w 1363 CREATE VERTEX SHADER FROM FILE 1384
__declspec(dllexport) 1448 Creating New DBPro Statements 1446
CROSS PRODUCT VECTOR3 1360
A Culling 784
Accessing Memory 1280

D
ADD LIMB 917
ADD MATRIX 1368 DBO file format 948
ADD MESH TO VERTEXDATA 1265 DELETE BSP 1173
ADD VECTOR3 1356 DELETE CAMERA 866
Advanced Camera Techniques 871 DELETE DLL 1464
Advanced Terrain Statements 1186 DELETE LIGHT 890
Alpha channel 750, 820 DELETE MATRIX4 1372
Ambient Lighting 886 DELETE MEMBLOCK 1286
APPEND OBJECT 970 DELETE MESH 915
AUTOCAM 851 DELETE MESH FROM VERTEXDATA 1271
AUTOMATIC CAMERA COLLISION 858 DELETE OBJECT 777
AUTOMATIC OBJECT COLLISION 1082 DELETE OBJECTS 777
Axes 744 DELETE OBJECT COLLISION BOX 1082
DELETE PARTICLES 1104
B DELETE PIXEL SHADER

DELETE TERRAIN
1386
1181
DELETE VECTOR3 1353
BACKDROP 864 Dereference 1282
Binary Space Partitioning 1164 Directional lighting 886
Bit 1280 DIVIDE MATRIX4 1368
Bounding box 1071 DIVIDE VECTOR3 1355
BSP COLLISION 1172 DLL 1446
BSP COLLISION HIT 1172 DLL CALL EXIST 1464
BUILD TERRAIN 1188 DLL EXIST 1463
Bump map 1038 DOT PRODUCT VECTOR3 1357
Bytes 1280 Doors 935
Dungeon Torch 1122
C E
CALL DLL 1463
Cameras 751, 836 Edge 749
CAMERA ANGLE 844 EFFECT EXIST 1378
CAMERA POSITION 844 Elevator model 948
Castle model 822 emitter 1102
CHANGE MESH 931 extern "C" 1448
CHANGE MESH FROM MEMBLOCK 1310
CHECK LIMB LINK
CLEAR CAMERA VIEW
947
869
F
Client/server 1390
CLONE OBJECT 778 FADE OBJECT 817
Collision box 1071 File formats 963
Collision sphere 1071 FILL MATRIX 1222
COLOR AMBIENT LIGHT 889 FIRE PARTICLES 1119
COLOR BACKDROP 864 FIX OBJECT PIVOT 773
COLOR LIGHT 890 FOG 901
COLOR LIMB 922 FOG COLOR 902
COLOR OBJECT 815 FOG DISTANCE 902
COLOR PARTICLES 1107 Frame rate 750
Column vectors 1350 FREE NET GAME 1411
FREE NET PLAYER 1412 Internet File Transfers 1436
FTP 1436 INSTANCE OBJECT 779
FTP CONNECT 1436 INTERSECT OBJECT 1083
FTP DELETE FILE 1442 INVERSE MATRIX4 1370
FTP DISCONNECT 1440 IPX 1391
FTP FIND FIRST 1438 IS EQUAL MATRIX4 1367
FTP FIND NEXT 1439 IS EQUAL VECTOR3 1359
FTP GET FILE 1440 IS IDENTITY MATRIX4 1366
FTP PROCEED 1441
FTP PUT FILE
FTP SET DIR
1443
1438 J
FTP TERMINATE 1442
FX Files 1376 JOIN NET GAME 1396
G L
GET FTP DIR$ 1438 LAN 1390
GET FTP ERROR$ 1437 LENGTH VECTOR3 1355
GET FTP FAILURE 1436 LIGHT DIRECTION 900
GET FTP FILE NAME$ 1439 LIGHT EXIST 898
GET FTP FILE SIZE 1439 Light map 1035
GET FTP FILE TYPE 1439 LIGHT POSITION 900
GET FTP PROGRESS 1442 LIGHT RANGE 899
GET FTP STATUS 1437 LIGHT TYPE 899
GET GROUND HEIGHT 1218 LIGHT VISIBLE 899
GET INDEXDATA 1268 Lighting 752, 886
GET MATRIX HEIGHT 1217 Limbs 917
GET MAXIMUM PIXEL SHADER VERSION 1377 LIMB ANGLE 939
GET MAXIMUM VERTEX SHADER VERSION 1377 LIMB DIRECTION 940
GET MEMBLOCK PTR 1281 LIMB EXIST 936
GET MEMBLOCK SIZE 1286 LIMB NAME$ 945
GET NET MESSAGE 1401 LIMB OFFSET 937
GET OBJECT COLLISION 1080 LIMB POSITION 939
GET PICK DISTANCE 984 LIMB SCALE 938
GET STATIC COLLISION 1089 LIMB TEXTURE 946
GET STATIC COLLISION HIT 1087 LIMB TEXTURE NAME 946
GET TERRAIN GROUND HEIGHT 1191 LIMB VISIBLE 937
GET TERRAIN HEIGHT 1184 LINK LIMB 928
GET TERRAIN SIZE 1193 LOAD BSP 1165
GET TOTAL TERRAIN HEIGHT 1186 LOAD DLL 1463
GET VERTEXDATA 1255 LOAD EFFECT 1378
GET VERTEXDATA DIFFUSE 1258 LOAD IMAGE 800
GET VERTEXDATA INDEX COUNT 1267 LOAD MESH 914
GET VERTEXDATA NORMALS 1253 LOAD OBJECT 950, 963
GET VERTEXDATA POSITION 1248 LOAD TERRAIN 1194
GET VERTEXDATA VERTEX COUNT 1247 Local axes 747
GHOST MATRIX OFF 1232 LOCK OBJECT 874
GHOST MATRIX ON 1231 LOCK VERTEXDATA FOR LIMB 1251
GHOST OBJECT OFF 817 LOCK VERTEXDATA FOR MESH 1246
GHOST OBJECT ON 816 LOOP OBJECT 966
GHOST PARTICLES OFF 1114
M
GHOST PARTICLES ON 1113
GLUE OBJECT TO LIMB 931
Gomuku 1415
MAKE BITMAP FROM MEMBLOCK 1299
H MAKE CAMERA
MAKE FILE FROM MEMBLOCK
863
1292
MAKE FIRE PARTICLES 1119
HIDE LIGHT 887 MAKE IMAGE FROM MEMBLOCK 1302
HIDE LIMB 927 MAKE LIGHT 889
HIDE OBJECT 776 MAKE MATRIX 1213
HIDE OBJECT BOUNDS 1072 MAKE MATRIX4 1365
HIDE PARTICLES 1103 MAKE MEMBLOCK 1281
Host 1392 MAKE MEMBLOCK FROM BITMAP 1297
MAKE MEMBLOCK FROM FILE 1294
I MAKE MEMBLOCK FROM IMAGE
MAKE MEMBLOCK FROM MESH
1302
1306
MAKE MEMBLOCK FROM SOUND 1303
Importing 3D Objects 962 MAKE MESH FROM MEMBLOCK 1309
MAKE MESH FROM OBJECT 912 OBJECT COLLISION CENTER 1075
MAKE OBJECT 914 OBJECT COLLISION RADIUS 1075
MAKE OBJECT BOX 757 OBJECT EXIST 779
MAKE OBJECT COLLISION BOX 1077 OBJECT FRAME 972
MAKE OBJECT CONE 760 OBJECT HIT 1069
MAKE OBJECT CYLINDER 759 OBJECT IN SCREEN 987
MAKE OBJECT CUBE 756 OBJECT INTERPOLATION 972
MAKE OBJECT FROM LIMB 919 OBJECT LOOPING 971
MAKE OBJECT PLAIN 760 OBJECT PLAYING 971
MAKE OBJECT SPHERE 758 OBJECT POSITION 780
MAKE OBJECT TERRAIN 1186 OBJECT SCREEN 982
MAKE OBJECT TRIANGLE 761 OBJECT SIZE 781, 973
MAKE PARTICLES 1102 OBJECT SPEED 972
MAKE SNOW PARTICLES 1118 OBJECT VISIBLE 780
MAKE SOUND FROM MEMBLOCK 1305 ODE 1318
MAKE STATIC COLLISION BOX 1087 ODE ADD FORCE 1338
MAKE TERRAIN 1180 ODE COLLISION GET MESSAGE 1336
MAKE VECTOR3 1351 ODE COLLISION MESSAGE EXISTS 1336
Manipulating Vetices 1246 ODE CREATE DYNAMIC BOX 1318
Map 1164 ODE CREATE DYNAMIC CYLINDER 1322
Matrices 1212 ODE CREATE DYNAMIC TRIANGLE MESH 1324
MATRIX EXIST 1237 ODE CREATE STATIC BOX 1321
MATRIX POSITION 1230 ODE CREATE STATIC TRIANGLE MESH 1325
MATRIX TILE COUNT 1228 ODE DESTROY OBJECT 1334
MATRIX TILES EXIST 1228 ODE END 1319
MATRIX WIREFRAME STATE 1220 ODE GET BODY HEIGHT 1335
MAXIMIZE VECTOR3 1359 ODE GET BODY LINEAR VELOCITY 1335
MEMBLOCK 1285 ODE GET OBJECT 1336
MEMBLOCK EXIST 1286 ODE GET OBJECT ANGULAR VELOCITY 1338
Memory Blocks ODE GET OBJECT VELOCITY 1337
Media 1297 ODE MAKE DYNAMIC SPHERE 1322
Strings 1288 ODE SET ANGULAR VELOCITY 1331
Merging 3D primitives 788 ODE SET BODY MASS 1332
MESH EXIST 915 ODE SET BODY ROTATION 1332
Meshes 912 ODE SET CONTACT FDIR1 1328
MINIMIZE VECTOR3 1360 ODE SET LINEAR VELOCITY 1328
Mipmaps 799 ODE SET WORLD CFM 1327
MOVE CAMERA 837 ODE SET WORLD ERP 1326
MOVE OBJECT 764 ODE SET WORLD GRAVITY 1320
MOVE OBJECT distance 772 ODE SET WORLD STEP 1325
Multiple cameras 863 ODE START 1319
MULTIPLY MATRIX4 1369 ODE UPDATE 1319
MULTIPLY VECTOR3 1354 OFFSET LIMB 920
N P
NET BUFFER SIZE 1408 Packets 1392
NET GAME EXISTS 1413 Particles 1102
NET GAME LOST 1413 PARTICLES EXIST 1114
NET GAME NOW HOSTING 1411 PARTICLES POSITION 1115
NET MESSAGE 1402 Peer-to-peer 1390
NET MESSAGE EXISTS 1402 PERFORM CHECKLIST FOR EFFECT 1379
NET MESSAGE PLAYER FROM 1403 PERFORM CSG DIFFERENCE 790
NET MESSAGE PLAYER TO 1403 PERFORM CSG INTERSECTION 790
NET MESSAGE TYPE 1406 PERFORM CSG UNION 788
NET PLAYER CREATED 1409 PERFORM CHECKLIST FOR NET CONNECTIONS 1391
NET PLAYER DESTROYED 1409 PERFORM CHECKLIST FOR NET PLAYERS 1397
Network 1390 PERFORM CHECKLIST FOR NET SESSIONS 1395
Networked games 1390 PERFORM CHECKLIST FOR OBJECT LIMBS 944
NORMALIZE VECTOR3 1358 PICK SCREEN 986
Normals 752 PICK OBJECT 983
surface 752 PICK VECTOR 985
vertex 753 PITCH CAMERA 842
PITCH OBJECT 770
O
Pixel shader 1376
Planes 745
PLAY ANIMATION TO IMAGE 805
PLAY OBJECT 965
OBJECT ANGLE 781
POINT CAMERA 838
OBJECT COLLISION 1070
POINT LIGHT 893
POINT OBJECT 771 SET CARTOON SHADING ON 1054
Points 746 SET CUBE MAPPING ON 1042
Pointers 1280 SET CURRENT CAMERA 865
Polygon 749 SET DETAIL MAPPING ON 811
POSITION CAMERA 836 SET DIRECTIONAL LIGHT 892
POSITION LIGHT 891 SET EFFECT CONSTANT 1383
POSITION MATRIX 1229 SET EFFECT ON 1380
POSITION OBJECT 762 SET EFFECT TECHNIQUE 1383
POSITION PARTICLE EMISSIONS 1105 SET EFFECT TRANSPOSE 1384
POSITION PARTICLES 1104 SET GLOBAL COLLISION 1071
POSITION TERRAIN 1182 SET GLOBAL OBJECT CREATION 785
PREPARE MATRIX TEXTURE 1220 SET GLOBAL SHADOW COLOR 1050
PROCESS BSP COLLISION 1171 SET GLOBAL SHADOW SHADES 1050
SET GLOBAL SHADOWS 1048
R
SET IDENTITY MATRIX4 1366
SET INDEXDATA 1270
SET LIGHT MAPPING ON 1035
SET LIGHT RANGE 891
RANDOMIZE MATRIX 1214
SET LIGHT TO OBJECT ORIENTATION 897
READ MEMBLOCK 1292
SET LIGHT TO OBJECT POSITION 895
REMOVE LIMB 928
SET LIMB EFFECT 1381
ROLL CAMERA 843
SET LIMB SMOOTHING 934
ROLL OBJECT 771
SET MATRIX 1235
Roman candle 1121
SET MATRIX HEIGHT 1215
ROTATE CAMERA 838
SET MATRIX NORMAL 1234
ROTATE LIGHT 895
SET MATRIX PRIORITY 1232
ROTATE LIMB 920
SET MATRIX TILE 1223
ROTATE MATRIX4 1371
SET MATRIX WIREFRAME 1219
ROTATE OBJECT 768
SET NET CONNECTION 1392
ROTATE PARTICLES 1106
SET OBJECT AMBIENCE 1029
Rotation 748
SET OBJECT COLLISION 1070
Absolute 769
SET OBJECT COLLISION TO BOXES 1076
Relative 769
SET OBJECT COLLISION TO POLYGON 1076
SET OBJECT COLLISION TO SPHERE 1074
S SET OBJECT CULL
SET OBJECT DIFFUSE
784
1030
SET OBJECT EMISSIVE 1031
SAVE MESH 913 SET OBJECT EFFECT 1379
SAVE OBJECT 949 SET OBJECT FILTER 813
SAVE TERRAIN 1193 SET OBJECT FOG 903
SCALE LIMB 922 SET OBJECT FRAME 968
SCALE LIMB TEXTURE 925 SET OBJECT INTERPOLATION 969
SCALE MATRIX4 1370 SET OBJECT LIGHT 1034
SCALE OBJECT 775 SET OBJECT ROTATION 769
SCALE OBJECT TEXTURE 802 SET OBJECT RADIUS 1074
SCALE VECTOR3 1354 SET OBJECT SPECULAR 1030
SCROLL LIMB TEXTURE 927 SET OBJECT SPECULAR POWER 1031
SCROLL OBJECT TEXTURE 808 SET OBJECT SPEED 967
Seamless tiling 804 SET OBJECT TEXTURE 807
SEND NET MESSAGE 1401, 1404 SET OBJECT TO CAMERA ORIENTATION 873
SET ALPHA MAPPING ON 1044 SET OBJECT TRANSPARENCY 810
SET AMBIENT LIGHT 888 SET OBJECT WIREFRAME 783
SET BLEND MAPPING ON 1041 SET PARTICLE CHAOS 1110
SET BUMP MAPPING ON 1038 SET PARTICLE EMISSIONS 1108
SET BSP CAMERA 1173 SET PARTICLE FLOOR 1112
SET BSP CAMERA COLLISION 1167 SET PARTICLE GRAVITY 1110
SET BSP CAMERA COLLISION RADIUS 1169 SET PARTICLE LIFE 1113
SET BSP COLLISION HEIGHT ADJUSTMENT 1170 SET PARTICLE SPEED 1111
SET BSP COLLISION OFF 1171 SET PARTICLE VELOCITY 1109
SET BSP COLLISION THRESHOLD 1171 SET PIXEL SHADER OFF 1386
SET BSP MULTITEXTURING 1173 SET PIXEL SHADER ON 1386
SET BSP OBJECT COLLISION 1167 SET PIXEL SHADER TEXTURE 1386
SET BSP OBJECT COLLISION RADIUS 1169 SET POINT LIGHT 893
SET CAMERA ASPECT 846 SET RAINBOW SHADING ON 1056
SET CAMERA FOV 847 SET REFLECTION SHADING ON 1057
SET CAMERA RANGE 848 SET SHADING OFF 1058
SET CAMERA ROTATION 840 SET SHADOW POSITION 1051
SET CAMERA TO FOLLOW 853 SET SHADOW SHADING ON 1045
SET CAMERA TO IMAGE 871 SET SHADOW SHADING OFF 1048
SET CAMERA TO OBJECT ORIENTATION 873 SET SPHERE MAPPING ON 1039
SET CAMERA VIEW 845 SET SPOT LIGHT 892
SET TERRAIN HEIGHTMAP
SET TERRAIN LIGHT
1187
1190
V
SET TERRAIN SCALE 1187
SET TERRAIN SPLIT 1191 Vertex 749
SET TERRAIN TEXTURE 1188 Vertex buffer 785
SET TERRAIN TILING 1189 Vertex data buffer's structure 1266
SET TEXTURE TRIM 1226 Vertex shader 1376
SET VECTOR3 1352 Video texture 805
SET VECTOR3 TO CAMERA POSITION 875
SET VECTOR3 TO CAMERA ROTATION
SET VECTOR3 TO PARTICLES POSITION
876
1116 W
SET VECTOR3 TO PARTICLES ROTATION 1116
SET VERTEX SHADER STREAM 1386 WAN 1390
SET VERTEX SHADER STREAMCOUNT 1386 Wireframe 749
SET VERTEX SHADER VECTOR 1386 World 1164
SET VERTEXDATA DIFFUSE 1257 World axes 744
SET VERTEXDATA NORMALS 1254 World units 747
Sky spheres 828 WRITE MEMBLOCK 1284
SET VERTEXDATA POSITION 1250 WRITE MEMBLOCK 1290
SET VERTEXDATA UV 1256
Shaders 1376
Shadows
SHIFT MATRIX
1045
1227
X
SHOW LIGHT 888
SHOW LIMB 928 XROTATE CAMERA 840
SHOW OBJECT 776 XROTATE OBJECT 766
SHOW PARTICLES
Skybox
1104
1197 Y
SNOW PARTICLES 1118
Solitaire 994 YROTATE CAMERA 841
Spaceship 1122 YROTATE OBJECT 767
Spot lighting 886
Z
SQUARED LENGTH VECTOR3 1356
Standard DLLs 1462
Static collision 1068
STATIC LINE OF SIGHT 1093
ZROTATE CAMERA 841
STATIC LINE OF SIGHT coordinates 1095
STOP OBJECT 968 ZROTATE OBJECT 767
String table 1449
Subdivisions 1212
SUBTRACT MATRIX4 1368
SUBTRACT VECTOR3 1357
Surface reflection 1028
Switching between cameras 866
T
TCP/IP 1391
Terrain 1180
TERRAIN POSITION 1183
Textures 750
TEXTURE LIMB 923
TEXTURE OBJECT 798
TEXTURE TERRAIN 1183
Tiling 801
TOTAL OBJECT FRAMES 966
TURN CAMERA 842
TURN OBJECT 770
TRANSLATE MATRIX4 1371
TRANSPOSE MATRIX4 1372
U
UNGLUE OBJECT 934
UNLOCK VERTEXDATA 1250
UPDATE MATRIX 1214
Also Available
from our website www.digital-skills.co.uk
Hands On DarkBASIC Pro Volume 1 742 pages

Contents include:
Background Concepts Video Cards and the Screen

Starting DarkBASIC Pro File Handling
Data Handling Music Files
Selection Displaying Video Files
Iteration Accessing the Keyboard
Drawing Statements Mathematical Functions
Modular Programming Images
String Functions Sprites
The Game of Hangman Sound
Arrays 2D Vectors
The Game of Bull and Touch Two-Player Space Duel Game
Advanced Data Types and Operators Using the Mouse
Bitmaps Using a Joystick
Hands On Milkshape 338 pages (also available as an ebook)

Contents include:
Background Concepts
Milkshape Basic Controls
Principles of 3D Construction
Vertices, Edges and Faces
3D Primitives
Manipulating Vertices
Reshaping Meshes
Extrusion
Using Milkshapes Additional Tools
Groups
Creating Models of Real World Objects
Texturing Your Model
Animation
Exporting Your Model to DBPro

Dark Basic Pro Vol 2

Uploaded by

Copyright:

Available Formats

Dark Basic Pro Vol 2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dark Basic Pro Vol 2

Uploaded by

Copyright:

Available Formats

Hands On DarkBASIC Pro

All rights reserved.

No part of this work may be reproduced or used in any form

Although every effort has been made to ensure accuracy, the

DarkBASIC Professional is produced by The Game Creators Ltd

Printed September 2006

Title : Hands On DarkBASIC Pro Volume 2

Other Titles Available:

Hands On DarkBASIC Pro Volume 1

Chapter 36 Importing 3D Objects

Chapter 37 Screen Control

Chapter 39 Advanced Lighting and Texturing

Chapter 42 The Elevators Game

Chapter 43 Handling BSP Models

Chapter 44 Creating Terrain

Chapter 45 Using Matrices

Chapter 46 Manipulating Vertices

Chapter 47 Accessing Memory

Chapter 48 Open Dynamics Engine

The SET OBJECT EFFECT Statement . . . . . . . . . . . . . 1379

Chapter 51 Network Programming

Chapter 52 Using File Transfer Protocol

Chapter 53 Dynamic Link Libraries

What You Need

How to Get the Most out of this Text

1. Read a chapter or section through without taking notes or worrying

3. Embedded in the material are a series of activities. Do each task as you

Braces represent choice

Raised tiles Sunken tiles

LIMB OFFSET Y ( objno , limbno )

Line Continuation Symbol

POSITION OBJECT 2,OBJECT POSITION X(2),OBJECT POSITION Y(2)-0.1,OBJECT POSITION Z(2)+0.1

POSITION OBJECT 2,OBJECT POSITION X(2),

Vertex and Surface Normals

The Major Planes in 3D

Vertices, Edges and Polygons

DarkBASIC Pro: 3D - Concepts and Terminology 743

The Coordinate System

744 DarkBASIC Pro: 3D - Concepts and Terminology

3D Axes and the Viewer

DarkBASIC Pro: 3D - Concepts and Terminology 745

746 DarkBASIC Pro: 3D - Concepts and Terminology

Determining the Position of a

DarkBASIC Pro: 3D - Concepts and Terminology 747

X-axis Z-axis X-axis

Axi when viewed from

By specifying a negative angle of rotation, an object will rotate anti-clockwise.

vectno is an integer value giving the ID to be assigned

The statement returns 1 if the vector is created successfully; otherwise zero is

result = MAKE VECTOR3(1)

748 DarkBASIC Pro: 3D - Concepts and Terminology

A 3D Vector 12.3 4.6 1.7

The line between two vertices is known as an edge (see FIG-30.12).

Every 3D shape in a game is constructed from polygons (normally triangles), as

DarkBASIC Pro: 3D - Concepts and Terminology 749

A Sphere with Few Polygons A Sphere with Many Polygons

Images with an Alpha Channel

750 DarkBASIC Pro: 3D - Concepts and Terminology

The image seen at

DarkBASIC Pro: 3D - Concepts and Terminology 751

LISTING-31.1 REM * Set display resolution and backdrop *