From beffa at ieee.org  Wed Dec  2 03:31:38 2015
From: beffa at ieee.org (Federico Beffa)
Date: Wed, 2 Dec 2015 09:31:38 +0100
Subject: [Matplotlib-devel] TkAgg compilation
Message-ID: <CAKrPhPN9d=Tao2AK7mmTEvY+vxBPt9oKnzNJ+a_D4B+tXrSNMw@mail.gmail.com>

Hi,

I've packaged matplotlib for the Guix functional package manager
(http://www.gnu.org/software/guix/) which compiles software in
isolated environments. The 'setup.py' program fails to detect the
location and version of the Tlc/Tk libraries required to build the
'TkAgg' backend.

Looking into the problem I have the impression that the detection
method is a little bit out of date. Tcl/Tk ships with 'pkg-config' .pc
files and it would probably be more robust to use this approach (as
used to detect other libraries).

The attached patch does this and fixes the problem for us.

Regards,
Fede
-------------- next part --------------
A non-text attachment was scrubbed...
Name: matplotlib-setupext-tk.patch
Type: text/x-diff
Size: 1315 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20151202/bdc08855/attachment.patch>

From mdroettboom at continuum.io  Wed Dec  9 17:47:13 2015
From: mdroettboom at continuum.io (Michael Droettboom)
Date: Wed, 9 Dec 2015 17:47:13 -0500
Subject: [Matplotlib-devel] matplotlib.org DNS
Message-ID: <CADsn3C09MOFyYFb5KcgW2PUpU0b-qG-muswX3ATX9B2vKYBSDg@mail.gmail.com>

There was a glitch in our recent transfer of the matplotlib.org domain such
that our DNS went down for some portion of yesterday and today.  It should
be resolved now, but it may be some time until the records propagate back
out to everyone.

Apologies for the inconvenience.

Mike

-- 
Michael Droettboom
Continuum Analytics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20151209/be71bae3/attachment.html>

From tcaswell at gmail.com  Thu Dec 17 09:19:16 2015
From: tcaswell at gmail.com (Thomas Caswell)
Date: Thu, 17 Dec 2015 14:19:16 +0000
Subject: [Matplotlib-devel] mpl 1.5.1?
Message-ID: <CAA48SF8oW6ijFWbSahzVxdS+x=UcyvDQKiWWDpGMOxgQSkZxhw@mail.gmail.com>

Hey folks,

We have accumulated a number of important bug fixes on the 1.5.x branch,
(soon including https://github.com/matplotlib/matplotlib/pull/5690 which
fixes a segfault when using pyside) so we should strongly consider doing a
1.5.1 release.

I do not think we need to do a rc cycle given the low-risk and so we can
get this out inside of the 2.0 mike-finds-all-the-rabbit-holes development
cycle.

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20151217/ccdb9c58/attachment.html>

From ben.v.root at gmail.com  Thu Dec 17 11:24:17 2015
From: ben.v.root at gmail.com (Benjamin Root)
Date: Thu, 17 Dec 2015 11:24:17 -0500
Subject: [Matplotlib-devel] mpl 1.5.1?
In-Reply-To: <CAA48SF8oW6ijFWbSahzVxdS+x=UcyvDQKiWWDpGMOxgQSkZxhw@mail.gmail.com>
References: <CAA48SF8oW6ijFWbSahzVxdS+x=UcyvDQKiWWDpGMOxgQSkZxhw@mail.gmail.com>
Message-ID: <CANNq6F=eMLXeRDCnvJFQVwk-Yn=R1fML3WYn=QA3TWdGe_rNcw@mail.gmail.com>

If anything, the rc process is good for vetting the correctness of the
binary builds. We do conflate the beta and rc process to be the "rc
process", and I agree that we don't need a beta testing period for 1.5.1,
but an rc should still be done so that people can report issues with the
binary builds. Note, this can be very short (just a couple of days, really).

Ben Root

On Thu, Dec 17, 2015 at 9:19 AM, Thomas Caswell <tcaswell at gmail.com> wrote:

> Hey folks,
>
> We have accumulated a number of important bug fixes on the 1.5.x branch,
> (soon including https://github.com/matplotlib/matplotlib/pull/5690 which
> fixes a segfault when using pyside) so we should strongly consider doing a
> 1.5.1 release.
>
> I do not think we need to do a rc cycle given the low-risk and so we can
> get this out inside of the 2.0 mike-finds-all-the-rabbit-holes development
> cycle.
>
> Tom
>
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20151217/b9058c3b/attachment.html>

From tcaswell at gmail.com  Fri Dec 18 23:10:37 2015
From: tcaswell at gmail.com (Thomas Caswell)
Date: Sat, 19 Dec 2015 04:10:37 +0000
Subject: [Matplotlib-devel] v2.x branch
Message-ID: <CAA48SF_bnSmUJwvSR-b1Ak5jgVMQXCyHRdySAsQwWFJNK8zg1Q@mail.gmail.com>

Hey folks,

Due to a mistake in merging the master branch accidentally was merged into
v2.0.x. The v2.0.x branch has been deleted on matplotlib/matplotlib (but
there is reference to it on tacaswell/matplotlib at it's last state).

The new branch for v2.0 development is v2.x.

A new branch name is being used to remove the master branch commits from
the v2.0 release branch without a rebase on upstream.

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20151219/a8d362b9/attachment.html>

From tcaswell at gmail.com  Mon Dec 28 14:17:13 2015
From: tcaswell at gmail.com (Thomas Caswell)
Date: Mon, 28 Dec 2015 19:17:13 +0000
Subject: [Matplotlib-devel] article on scales
Message-ID: <CAA48SF9k2efrEXudhsibNPTQq1KB7YOcGHn79Cx1oSFvGSJHCg@mail.gmail.com>

There has been some discussion about how to generalize/unify the mixture of
scaling functionality in mpl (axes scales, color.norm, and the units
framework).  Mike Bostock has pulled the scale functionality out of d3 [1],
we might want to take a look at that for API inspiration (also scroll to
the bottom of the post and note what color maps he uses as demo :) ).

Tom

[1] https://medium.com/@mbostock/introducing-d3-scale-61980c51545f
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20151228/c7ba2003/attachment.html>

From ellisonbg at gmail.com  Mon Dec 28 18:10:24 2015
From: ellisonbg at gmail.com (Brian Granger)
Date: Mon, 28 Dec 2015 15:10:24 -0800
Subject: [Matplotlib-devel] article on scales
In-Reply-To: <CAA48SF9k2efrEXudhsibNPTQq1KB7YOcGHn79Cx1oSFvGSJHCg@mail.gmail.com>
References: <CAA48SF9k2efrEXudhsibNPTQq1KB7YOcGHn79Cx1oSFvGSJHCg@mail.gmail.com>
Message-ID: <CAH4pYpTPaq2w_d62iWyn=6k9rUrkoCADz-VHNAMA_H9gaaGmZA@mail.gmail.com>

Nice, I would really like to see a nice python scale library that was
independent of any particular viz package that everyone could start to
use...

On Mon, Dec 28, 2015 at 11:17 AM, Thomas Caswell <tcaswell at gmail.com> wrote:
> There has been some discussion about how to generalize/unify the mixture of
> scaling functionality in mpl (axes scales, color.norm, and the units
> framework).  Mike Bostock has pulled the scale functionality out of d3 [1],
> we might want to take a look at that for API inspiration (also scroll to the
> bottom of the post and note what color maps he uses as demo :) ).
>
> Tom
>
> [1] https://medium.com/@mbostock/introducing-d3-scale-61980c51545f
>
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel
>



-- 
Brian E. Granger
Associate Professor of Physics and Data Science
Cal Poly State University, San Luis Obispo
@ellisonbg on Twitter and GitHub
bgranger at calpoly.edu and ellisonbg at gmail.com