I'm still getting a couple of errors, these may have been mentioned in #779, this is on Ubuntu 12.04

======================================================================
FAIL: matplotlib.tests.test_delaunay.test_gauss.test
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/home/sofia/.local/lib/python2.7/site-packages/matplotlib/testing/decorators.py", line 36, in failer
    result = f(*args, **kwargs)
  File "/home/sofia/.local/lib/python2.7/site-packages/matplotlib/testing/decorators.py", line 140, in do_test
    '(RMS %(rms).3f)'%err)
ImageComparisonFailure: images not close: /home/sofia/code/matplotlib/result_images/test_delaunay/gauss-lin-con.png vs. /home/sofia/code/matplotlib/result_images/test_delaunay/expected-gauss-lin-con.png (RMS 10.677)

======================================================================
FAIL: matplotlib.tests.test_delaunay.test_trig.test
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/home/sofia/.local/lib/python2.7/site-packages/matplotlib/testing/decorators.py", line 36, in failer
    result = f(*args, **kwargs)
  File "/home/sofia/.local/lib/python2.7/site-packages/matplotlib/testing/decorators.py", line 140, in do_test
    '(RMS %(rms).3f)'%err)
ImageComparisonFailure: images not close: /home/sofia/code/matplotlib/result_images/test_delaunay/trig-lin-con.png vs. /home/sofia/code/matplotlib/result_images/test_delaunay/expected-trig-lin-con.png (RMS 19.105)

----------------------------------------------------------------------
Ran 1061 tests in 1013.936s

FAILED (KNOWNFAIL=266, failures=2)

(The known fails are due to not having inkscape)

the diff images can be found here

Looks like #576 wasn't backported in full. Sliders are still rendered as triangles when valmin == valinit. See also #581 (comment)

I made a small mistake in the new mono font rendering (used by the tests). See #793.

On Windows matplotlib.test() still fails with hundreds of IOError: [Errno 24] Too many open files. Any chance 4cb2aaa could be backported? Probably not that easy since the with statement was introduced in Python 2.5.

@cgohlke: It would be nice to backport that commit, but as you say it won't be very direct as we'd have to replace the with statements. I'll attempt this -- would you be willing to test any PR on Windows/Python2.5?

@cgohlke: See #795. I'm not sure it will resolve your issue, but I'm curious to know if it does! I think it should only be merged if we plan to have another release candidate, as it's a fairly significant change.

@mdboom: Thank you. I tested #795 and it does indeed not solve the problem. Good to know. I don't think #795 (as is) can be applied to v1.1.1 because it is not compatible with Python 2.4 and Python 2.5 would require from __future__ import with_statement in a couple of files.

As a workaround for the Too many open files errors, the max number of allowed open files can be increased with import ctypes;ctypes.cdll.msvcr90._setmaxstdio(2048). All tests pass on win-amd64-py2.7.

@cgohlke thanks for the catch on the slider valmin glitch. I finished the backport in #797

@cgohlke re the too many open file handles bug, does #798 help?

#798 does help: matplotlib.test() passes on win-amd64-py2.7 without increasing maxstdio. Thank you.

I've updated #795 to remove the few remaining with statements. Can you test on Python 2.5? (Though it doesn't actually solve the underlying issue -- I think #798 is better for that -- we perhaps should just leave well enough alone and not merge #795, though it should help with the number of open file handles, just not on the same scale as #798).

@mdboom given the thread safetly issue I mentioned on #798, we should continue testing #795 too. I will test it on python 2.4 later today. Anyone who is a thread expert, please look at the discussion on #798 about thread safety using class level versus instance level caching of the FT2Font instances, and pitch in with any suggestions.

@mdboom: Where do I find the updated #795? Also, not critical, the PIL images in testing.compare.py are not getting closed, e.g. https://github.com/matplotlib/matplotlib/blob/v1.1.x/lib/matplotlib/testing/compare.py#L219. This seems fixed in master too.

platform:
Windows 7, 64 bit

python:
Enthought Python Distribution -- www.enthought.com
Version: 7.1-1 (32-bit)

matplotlib:
http://www.lfd.uci.edu/~gohlke/pythonlibs/#matplotlib
matplotlib-1.1.1rc.win32-py2.7.‌exe

With or without the command 'import ctypes;ctypes.cdll.msvcr90._setmaxstdio(2048)',
I get hundreds of errors like the following when running 'matplotlib.test()':

======================================================================
ERROR: matplotlib.tests.test_legend.test_legend_auto2.test
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\nose\case.py", line 187, in runTest
    self.test(*self.arg)
  File "C:\Python27\lib\site-packages\matplotlib\testing\decorators.py", line 36, in failer
    result = f(*args, **kwargs)
  File "C:\Python27\lib\site-packages\matplotlib\testing\decorators.py", line 130, in do_test
    err = compare_images(expected_fname, actual_fname, self._tol, in_decorator=True)
  File "C:\Python27\lib\site-packages\matplotlib\testing\compare.py", line 219, in compare_images
    expectedImage = Image.open( expected ).convert("RGB")
  File "C:\Python27\lib\site-packages\PIL\Image.py", line 1952, in open
    fp = __builtin__.open(fp, "rb")
IOError: [Errno 2] No such file or directory: 'C:\\Temp\\result_images\\test_legend\\expected-legend_auto2.png'

----------------------------------------------------------------------
Ran 1068 tests in 337.868s

FAILED (KNOWNFAIL=534, errors=508)

@breisfeld: The Windows installer does not include the test data. To get the test data, you'll need to checkout from git.

If you don't have git and don't want to go through the trouble of getting it installed, you can also get a zip file of the current 1.1.x branch here:

https://github.com/matplotlib/matplotlib/zipball/v1.1.x

Extract that somewhere and copy the lib/matplotlib/tests/baseline_images folder to C:\Python27\lib\site-packages\matplotlib\tests.

The tests should then have their data. (Not tested, as I don't have my Windows VM handy here).

@breisfeld: The official release candidate binaries at https://sourceforge.net/projects/matplotlib/files/matplotlib/matplotlib-1.1.1/ include the test data.

Closing; the release has been made.