DOC: prepare for release

matthew-brett · matthew-brett · commit 6ee43743b21c · 2018-02-17T15:06:16.000Z
diff --git a/Changelog b/Changelog
@@ -32,6 +32,20 @@ Abbreviated authors are:
 * JBP - Jean-Baptiste Poline
 * JT - Jonathan Taylor
 
+* 0.4.2 (Saturday 17 February 2018)
+
+  Bugfix, refactoring and compatibility release.
+
+  * Fixes for compatibility with released versions of Sympy and Numpy,
+    including some incorrect results from the Euler calculations;
+  * Fixes for deprecated escape sequences in docstrings (thanks to Klaus
+    Sembritzki);
+  * Fixes for compatibility with Windows in various configurations, now
+    tested with Appveyor builds;
+  * Various continuous integration and doc build fixes;
+  * The advent of Windows wheels on release - most credit to the Scipy folks
+    for building Scipy on Windows.
+
 * 0.4.1 (Friday 10 February 2017)
 
   Bugfix, refactoring and compatibility release.
diff --git a/LICENSE b/LICENSE
@@ -1,4 +1,4 @@
-Copyright (c) 2006-2017, NIPY Developers
+Copyright (c) 2006-2018, NIPY Developers
 All rights reserved.
 
 Redistribution and use in source and binary forms, with or without
diff --git a/doc/conf.py b/doc/conf.py
@@ -62,7 +62,7 @@
 # The master toctree document.
 master_doc = 'index'
 
-# copyright = ':ref:`2005-2017, Neuroimaging in Python team.
+# copyright = ':ref:`2005-2018, Neuroimaging in Python team.
 # <nipy-software-license>`'
 copyright = '2005-2017, Neuroimaging in Python team'
 
diff --git a/doc/devel/guidelines/make_release.rst b/doc/devel/guidelines/make_release.rst
@@ -75,12 +75,12 @@ Release checklist
   for:
 
   * https://nipy.bic.berkeley.edu/builders/nipy-examples-2.7
-  * https://nipy.bic.berkeley.edu/builders/nipy-examples-3.4
+  * https://nipy.bic.berkeley.edu/builders/nipy-examples-3.5
 
   The matching outputs appear at:
 
   * https://nipy.bic.berkeley.edu/nipy-examples-2.7
-  * https://nipy.bic.berkeley.edu/nipy-examples-3.4
+  * https://nipy.bic.berkeley.edu/nipy-examples-3.5
 
   I use the following commands to get the output to my laptop and review with
   vim::
@@ -99,11 +99,11 @@ Release checklist
 
   Likewise for::
 
-    PY_VER=3.4
+    PY_VER=3.5
 
   I also did a by-eye comparison between the 2.7 and 3.4 files with::
 
-    diff -r nipy-examples-2.7 nipy-examples-3.4 | less
+    diff -r nipy-examples-2.7 nipy-examples-3.5 | less
 
 * Do a final check on the `nipy buildbot`_
 
@@ -156,6 +156,18 @@ Release checking - buildbots
 * Check the documentation doctests pass from
   http://nipy.bic.berkeley.edu/builders/nipy-doc-builder
 
+* Build and test the Nipy wheels.  See the `wheel builder README
+  <https://github.com/MacPython/nipy-wheels>`_ for instructions.  In summary,
+  clone the wheel-building repo, edit the ``.travis.yml`` and ``appveyor.yml``
+  text files (if present) with the branch or commit for the release, commit
+  and then push back up to github.  This will trigger a wheel build and test
+  on OSX, Linux and Windows. Check the build has passed on on the Travis-CI
+  interface at https://travis-ci.org/MacPython/nipy-wheels.  You'll need commit
+  privileges to the ``dipy-wheels`` repo; ask Matthew Brett or on the mailing
+  list if you do not have them.
+
+* The release should now be ready.
+
 Doing the release
 =================
 
@@ -166,6 +178,13 @@ Doing the release
 
     make source-release
 
+* For the wheel build / upload, follow the `wheel builder README`_
+  instructions again.  Edit the ``.travis.yml`` and ``appveyor.yml`` files (if
+  present) to give the release tag to build.  Check the build has passed on on
+  the Travis-CI interface at https://travis-ci.org/MacPython/nipy-wheels.  Now
+  follow the instructions in the page above to download the built wheels to
+  a local machine and upload to PyPI.
+
 * Once everything looks good, you are ready to upload the source release to
   PyPi.  See `setuptools intro`_.  Make sure you have a file ``\$HOME/.pypirc``,
   of form::
@@ -178,27 +197,12 @@ Doing the release
     username:your.pypi.username
     password:your-password
 
-    [server-login]
-    username:your.pypi.username
-    password:your-password
-
 * Once everything looks good, upload the source release to PyPi.  See
   `setuptools intro`_::
 
     python setup.py register
-    python setup.py sdist --formats=gztar,zip upload
-
-* Trigger binary builds for Windows from the buildbots. See builders
-  ``nipy-bdist32-26``, ``nipy-bdist32-27``, ``nipy-bdist32-32``.  The ``exe``
-  builds will appear in http://nipy.bic.berkeley.edu/nipy-dist . Download the
-  builds and upload to pypi.
-
-* Trigger binary builds for OSX from travis-ci:
-
-    * https://travis-ci.org/MacPython/nipy-wheels
-    * https://github.com/MacPython/nipy-wheels
-
-  Upload the resulting wheels to pypi from http://wheels.scipy.org;
+    python setup.py sdist
+    twine upload sdist/nipy*.tar.gz
 
 * Tag the release with tag of form ``0.5.0``::
 
diff --git a/doc/users/installation.rst b/doc/users/installation.rst
@@ -128,47 +128,47 @@ Now :ref:`install-numpy-scipy`.
 
 .. _install-numpy-scipy:
 
-Install Numpy and Scipy on Linux or macOS
------------------------------------------
-
-Now you've followed the instructions above, install Numpy and Scipy with:
-
-.. code-block:: bash
-
-    pip3 install --user numpy scipy
-
 Install Python 3, Pip, NumPy and Scipy on Windows
 -------------------------------------------------
 
 It's worth saying here that very few scientific Python developers use Windows,
 so if you're thinking of making the switch to Linux or macOS, now you have
 another reason to do that.
 
-That said, if you are installing on Windows, we recommend the Python 3 version
-of `Anaconda`_.  This is a large installer that will install many scientific
-Python packages, as well as Python itself, and Pip, the package manager.
+Option 1: Anaconda
+^^^^^^^^^^^^^^^^^^
+
+If you are installing on Windows, you might want to use the Python 3 version of
+`Anaconda`_.  This is a large installer that will install many scientific
+Python packages, including NumPy and Scipy, as well as Python itself, and Pip,
+the package manager.
 
 The machinery for the Anaconda bundle is not completely open-source, and is
 owned by a company, Continuum Analytics. If you would prefer to avoid using
-the Anaconda installer:
+the Anaconda installer, you can also use the Python standard Pip installer.
+
+Option 2: Standard install
+^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+If you don't have Python / Pip, we recommend the instructions `here
+<http://docs.python-guide.org/en/latest/starting/install3/win>`_ to install
+them.  You can also install Python / Pip via the Python 3 installer from the
+https://python.org website.
 
-1. Download and install the Python 3 installer from the https://python.org website;
-2. Download and install Pip following the instructions at `install pip with
-   get-pip.py`_;
-3. Download and install the `Visual C++ redistributable packages for VC++ 2015
-   <https://www.microsoft.com/en-us/download/details.aspx?id=53587>`_;
-4. Download the following packages from Christoph Gohlke's `unofficial Windows
-   binaries`_:
+If you already have an old Python installation, you don't have Pip, and you
+don't want to upgrade, you will need to download and install Pip following the
+instructions at `install pip with get-pip.py`_.
 
-   * numpy (MKL version);
-   * scipy (MKL version);
+Now open a Cmd or Powershell terminal and run:
 
-   and install these packages with Pip.
+.. code-block:: bash
+
+    pip3 install --user numpy scipy
 
 Install Nipy
 ============
 
-Now you have Pip:
+Now you have Python and Pip:
 
 .. code-block:: bash
 

Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-Copyright (c) 2006-2017, NIPY Developers`
	`1`	`+Copyright (c) 2006-2018, NIPY Developers`
`2`	`2`	`All rights reserved.`
`3`	`3`
`4`	`4`	`Redistribution and use in source and binary forms, with or without`