Re: [matplotlib-devel] backporting the boxplot fix

2015-04-26 Thread Thomas Caswell
The branch 1.4.x was the branch for bug fixes that would eventually turn into the _next_ 1.4 series release. We do not plan on doing a 1.4.4 so the 1.4.x branch no longer is needed. The commit used for the releases are tagged (ex v1.4.3). The v1.4.3-doc branch is so that we can correct any docum

Re: [matplotlib-devel] backporting the boxplot fix

2015-04-26 Thread Mike Kaufman
So it seems I was using the old branch (now deleted?) v1.4.x rather than v1.4.3-doc, which does have the commit for the fix in question. I was naively assuming that v1.4.x was always the latest release in the v1.4 series. False alarm. Sorry. M On 4/25/15 4:26 PM, Thomas Caswell wrote: > The c

Re: [matplotlib-devel] backporting the boxplot fix

2015-04-25 Thread Thomas Caswell
The commit that fixes that https://github.com/matplotlib/matplotlib/commit/40720ef9fb5de75d908d0ce433d5c3bb8902884f should be in 1.4.1 an onward. Exactly which version are you using? There will be no 1.4.4. On Tue, Apr 21, 2015 at 11:00 AM Michael Kaufman wrote: > Is there any possibility of b

[matplotlib-devel] backporting the boxplot fix

2015-04-21 Thread Michael Kaufman
Is there any possibility of back-porting the fix to the boxplot positions to v1.4.x? This would be ticket #3563. I had thought that this was fixed in 1.4, but it seems to be there again. v1.5-devel (where the boxplot works fine) is not-very-usable for me due to the GTK idle bug. Thanks, M ---