Release 2.7.0 on January 1, 2015

[aclark4life]

OK I'm dropping off again, please use @aclark4life if you need anything or email aclark@aclark.net. Thank you 🍻

[aclark4life]

I'm back! How are we looking for Jan 1 @hugovk @wiredfool ?

[hugovk]

We have eight open PRs, two aren't ready: #897 and the experimental #1013. #1054 doesn't have tests but it's not a showstopper if we want it in the release.

[wiredfool]

I'm digging out from under visitors and getting my wrist working well again. hoping to have some good time to work through stuff tonight.

[wiredfool]

I've merged #1054, I think it's ready enough. We're down to only PRs that I don't think should be merged for the release.

At this point, I think we should draw a line and only fix regressions or documentation.

[aclark4life]

Great!

[wiredfool]

I've run the tests on my PPC/bigendian machine and we're only getting the usual known issues with Jpeg2k.

[cgohlke]

I get two test failures/errors on Windows. The error is trivial. The Scipy is version 0.14.1:

======================================================================
ERROR: TestImageFont.test_find_linux_font
----------------------------------------------------------------------
Traceback (most recent call last):
  File "D:\Build\Pillow\Pillow-git\Tests\test_imagefont.py", line 246, in test_find_linux_font
    self._test_fake_loading_font('/usr/local/share/fonts/Arial.ttf')
  File "D:\Build\Pillow\Pillow-git\Tests\test_imagefont.py", line 229, in _test_fake_loading_font
    font = ImageFont.truetype('Arial')
  File "D:\Build\Pillow\Pillow-git\PIL\ImageFont.py", line 266, in truetype
    return FreeTypeFont(filepath, size, index, encoding)
  File "D:\Build\Pillow\Pillow-git\Tests\test_imagefont.py", line 227, in loadable_font
    return ImageFont._FreeTypeFont(filepath, size, index, encoding)
  File "D:\Build\Pillow\Pillow-git\PIL\ImageFont.py", line 137, in __init__
    self.font = core.getfont(font, size, index, encoding)
IOError: cannot open resource

======================================================================
FAIL: Test_scipy_resize.test_imresize4
----------------------------------------------------------------------
Traceback (most recent call last):
  File "D:\Build\Pillow\Pillow-git\Tests\test_scipy.py", line 38, in test_imresize4
    assert_equal(im2, res)
  File "X:\Python27-x64\lib\site-packages\numpy\testing\utils.py", line 260, in assert_equal
    return assert_array_equal(actual, desired, err_msg, verbose)
  File "X:\Python27-x64\lib\site-packages\numpy\testing\utils.py", line 718, in assert_array_equal
    verbose=verbose, header='Arrays are not equal')
  File "X:\Python27-x64\lib\site-packages\numpy\testing\utils.py", line 644, in assert_array_compare
    raise AssertionError(msg)
AssertionError:
Arrays are not equal

(mismatch 75.0%)
 x: array([[ 1.  ,  1.25,  1.75,  2.  ],
       [ 1.5 ,  1.75,  2.25,  2.5 ],
       [ 2.5 ,  2.75,  3.25,  3.5 ],
       [ 3.  ,  3.25,  3.75,  4.  ]], dtype=float32)
 y: array([[ 1. ,  1. ,  1.5,  2. ],
       [ 1. ,  1. ,  1.5,  2. ],
       [ 2. ,  2. ,  2.5,  3. ],
       [ 3. ,  3. ,  3.5,  4. ]], dtype=float32)

----------------------------------------------------------------------

[wiredfool]

The font one should probably only run on the appropriate platform, or at least, not windows. I'll look at scipy more closely.

[wiredfool]

I've got to say, the existing one doesn't exactly look right. But it's not like it's getting tested on travis, since Ubuntu 12.04 ships with scipy 0.9, which doesn't even have the imresize method. Not that we're installing it to find that out.

I'm assuming that that has passed in the past?

[wiredfool]

OK, I found these -- they're from the scipy tests, and were added after the last release. https://github.com/scipy/scipy/blob/master/scipy/misc/tests/test_pilutil.py

[wiredfool]

The scipy change appears to be because of a change in the way that image resizing works. The new version looks more reasonable to me, which fits with the conclusions that we've had from the changes in the resizing code this iteration.

[wiredfool]

@cgohlke Check master now.

Also as a heads up, I'm planning on a 2.6.2 release tomorrow with the PNG dos fix. When the tests pass on my branch, I'll push the last changes in. (it's 2.6.1 + png dos fix + pypy2.4 regression fix + reverting the test changes on for pypy2.4)

[wiredfool]

Ok, versions bumped. 12 hours or so till I tag & push.