Re: [matplotlib-devel] matplotlib 1.3.0rc4 (was: matplotlib 1.3.0rc3 tomorrow)

2013-07-02 Thread Paul Hobson
On Fri, Jun 21, 2013 at 10:42 AM, Michael Droettboom md...@stsci.eduwrote:

  I have made a new release candidate (1.3.0rc4) that fixes the following
 vs. 1.3.0rc3:

 - It doesn't add a setup.cfg file to the tarball

 - It doesn't install the KnownFailure nose plugin as a pkg_resources
 entry_point (this conflicted with IPython's plugin of the same name)

 - We get a known failure from the pep8 test if pep8 isn't installed

 Hopefully that's enough to get to the point of giving this release
 candidate some wider exposure before putting out a final release.  We can
 use some sprint time at Scipy to get this release polished if desired and
 necessary, too.

 Mike


Hey folks,

I just downloaded the RC4 Gohlke binary for 64-bit Python 3.3, and the test
suite is blowing up. Mostly related to the baseline images not being
included. Is there a trick to get the test suite to meaningfully run on
Windows? The test suite for Python 2.7.5 (via Anaconda) went relevatively
well(1 failure + 1 error, I think).

Thanks,
-paul
--
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel


[matplotlib-devel] matplotlib 1.3.0rc4 (was: matplotlib 1.3.0rc3 tomorrow)

2013-06-21 Thread Michael Droettboom
I have made a new release candidate (1.3.0rc4) that fixes the following 
vs. 1.3.0rc3:


- It doesn't add a setup.cfg file to the tarball

- It doesn't install the KnownFailure nose plugin as a pkg_resources 
entry_point (this conflicted with IPython's plugin of the same name)


- We get a known failure from the pep8 test if pep8 isn't installed

Hopefully that's enough to get to the point of giving this release 
candidate some wider exposure before putting out a final release. We can 
use some sprint time at Scipy to get this release polished if desired 
and necessary, too.


Mike

On 06/19/2013 06:10 PM, Michael Droettboom wrote:

On 06/19/2013 05:45 PM, Michael Droettboom wrote:

On 06/19/2013 05:03 PM, Russell E. Owen wrote:

In article51c0b2f6.5070...@stsci.edu,
  Michael Droettboommd...@stsci.edu  
  wrote:



I have tagged a 1.3.0rc3 and uploaded a tarball to Sourceforge.

We may not get all the binaries up in the next little while, so I'll
wait for those and then make an announcement on matplotlib-users.  After
a couple of weeks, assuming no serious problems, we'll be ready for
1.3.0 final.

Thanks again to everyone for their help with this release!

I have uploaded the MacOS 10.6 64-bit binary. However, there were a few
unit test failures, including the font complaint that I first saw in an
earlier 1.3.0 prerelease, plus one about pep8 that I've never seen
before.


The font issue was reportedly fixed by 9ccacff 
https://github.com/pwuertz/matplotlib/commit/9ccacffe81e0d4a48e4a4b19064460b94c6f77c9, 
but apparently not.  I'll ping Peter Wuertz about that.
My bad.  It was determined that this was too difficult to fix at the 
moment.  See issue https://github.com/matplotlib/matplotlib/pull/2117.


Let's just ignore this for now -- for finesse we can trap this case 
and mark it as known fail, but I don't think that needs to hold 
things up for now.


Mike


--
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev


___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel


--
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel