From matti.picus at gmail.com  Fri Sep  2 09:24:19 2016
From: matti.picus at gmail.com (Matti Picus)
Date: Fri, 2 Sep 2016 16:24:19 +0300
Subject: [Matplotlib-devel] getting pypy to play nice with matplotlib
Message-ID: <29b654a6-7039-ce49-08c7-5a0075602156@gmail.com>

The pypy C-API progress, where we can basically run all of numpy (on a 
HEAD version numpy + nightly version of pypy).

That means I will start to look at passing the matplotlib test suite.

My first try segfaults, due to the tkagg backend blit function which 
calls into tk via id(data) and id(bbox_array).

That of course will not work on PyPy since the address of a random 
python object does not reflect anything usable in C.

So my question, as someone who is new to your codebase, is how to 
rewrite this using cffi or the standard C-API without impacting performance?

Matti


From tcaswell at gmail.com  Fri Sep  2 09:32:27 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Fri, 02 Sep 2016 13:32:27 +0000
Subject: [Matplotlib-devel] getting pypy to play nice with matplotlib
In-Reply-To: <29b654a6-7039-ce49-08c7-5a0075602156@gmail.com>
References: <29b654a6-7039-ce49-08c7-5a0075602156@gmail.com>
Message-ID: <CAA48SF_jV_pJa44u+66OGBMLYwAJE8uruh1B8QkZRHqXUfwBkA@mail.gmail.com>

Which branch are you working off of?  Matthew Brett recently did a whole
bunch of work over-hauling how that linking happens.

We currently have a blanket ban on cffi so that mpl will run in locked-down
environments, but I would be open to being convinced using it in a
localized way inside of the tk code.

How are you running the test suite?  It is odd that you are hitting tk at
all.

Tom

On Fri, Sep 2, 2016 at 9:24 AM Matti Picus <matti.picus at gmail.com> wrote:

> The pypy C-API progress, where we can basically run all of numpy (on a
> HEAD version numpy + nightly version of pypy).
>
> That means I will start to look at passing the matplotlib test suite.
>
> My first try segfaults, due to the tkagg backend blit function which
> calls into tk via id(data) and id(bbox_array).
>
> That of course will not work on PyPy since the address of a random
> python object does not reflect anything usable in C.
>
> So my question, as someone who is new to your codebase, is how to
> rewrite this using cffi or the standard C-API without impacting
> performance?
>
> Matti
>
> _______________________________________________
> 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/20160902/ba04fb21/attachment.html>

From matti.picus at gmail.com  Fri Sep  2 09:54:07 2016
From: matti.picus at gmail.com (Matti Picus)
Date: Fri, 2 Sep 2016 16:54:07 +0300
Subject: [Matplotlib-devel] getting pypy to play nice with matplotlib
In-Reply-To: <CAA48SF_jV_pJa44u+66OGBMLYwAJE8uruh1B8QkZRHqXUfwBkA@mail.gmail.com>
References: <29b654a6-7039-ce49-08c7-5a0075602156@gmail.com>
 <CAA48SF_jV_pJa44u+66OGBMLYwAJE8uruh1B8QkZRHqXUfwBkA@mail.gmail.com>
Message-ID: <825f7b62-0788-9329-d30a-3ccbabfb5f52@gmail.com>

I am working off of 1.5.1, should I be using HEAD? Would any of the data 
access API's like np.array.__array_interface__ be acceptable?
I did these two calls

<pypy> setup.py build

<pypy> setup.py nosetests --verbosity=10

and got a crash in matplotlib.tests.test_artist.test_remove

Is there a better way to run tests?

Thanks
Matti

On 02/09/16 16:32, Thomas Caswell wrote:
> Which branch are you working off of?  Matthew Brett recently did a 
> whole bunch of work over-hauling how that linking happens.
>
> We currently have a blanket ban on cffi so that mpl will run in 
> locked-down environments, but I would be open to being convinced using 
> it in a localized way inside of the tk code.
>
> How are you running the test suite?  It is odd that you are hitting tk 
> at all.
>
> Tom
>
> On Fri, Sep 2, 2016 at 9:24 AM Matti Picus <matti.picus at gmail.com 
> <mailto:matti.picus at gmail.com>> wrote:
>
>     The pypy C-API progress, where we can basically run all of numpy (on a
>     HEAD version numpy + nightly version of pypy).
>
>     That means I will start to look at passing the matplotlib test suite.
>
>     My first try segfaults, due to the tkagg backend blit function which
>     calls into tk via id(data) and id(bbox_array).
>
>     That of course will not work on PyPy since the address of a random
>     python object does not reflect anything usable in C.
>
>     So my question, as someone who is new to your codebase, is how to
>     rewrite this using cffi or the standard C-API without impacting
>     performance?
>
>     Matti
>
>     _______________________________________________
>     Matplotlib-devel mailing list
>     Matplotlib-devel at python.org <mailto:Matplotlib-devel at python.org>
>     https://mail.python.org/mailman/listinfo/matplotlib-devel
>


From tcaswell at gmail.com  Fri Sep  2 13:05:52 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Fri, 02 Sep 2016 17:05:52 +0000
Subject: [Matplotlib-devel] getting pypy to play nice with matplotlib
In-Reply-To: <825f7b62-0788-9329-d30a-3ccbabfb5f52@gmail.com>
References: <29b654a6-7039-ce49-08c7-5a0075602156@gmail.com>
 <CAA48SF_jV_pJa44u+66OGBMLYwAJE8uruh1B8QkZRHqXUfwBkA@mail.gmail.com>
 <825f7b62-0788-9329-d30a-3ccbabfb5f52@gmail.com>
Message-ID: <CAA48SF-soAsKZxsL0Oj2DtBZZnGEHY2WU+QR6vXGRe5uoEWR-g@mail.gmail.com>

On my phone so excuse the terse answer.

Use the numpy calls if likely OK

Work of of master or v2.x

Look at our travis.yml file for how to run the tests.

Python tests.py

Tom

On Fri, Sep 2, 2016, 09:54 Matti Picus <matti.picus at gmail.com> wrote:

> I am working off of 1.5.1, should I be using HEAD? Would any of the data
> access API's like np.array.__array_interface__ be acceptable?
> I did these two calls
>
> <pypy> setup.py build
>
> <pypy> setup.py nosetests --verbosity=10
>
> and got a crash in matplotlib.tests.test_artist.test_remove
>
> Is there a better way to run tests?
>
> Thanks
> Matti
>
> On 02/09/16 16:32, Thomas Caswell wrote:
> > Which branch are you working off of?  Matthew Brett recently did a
> > whole bunch of work over-hauling how that linking happens.
> >
> > We currently have a blanket ban on cffi so that mpl will run in
> > locked-down environments, but I would be open to being convinced using
> > it in a localized way inside of the tk code.
> >
> > How are you running the test suite?  It is odd that you are hitting tk
> > at all.
> >
> > Tom
> >
> > On Fri, Sep 2, 2016 at 9:24 AM Matti Picus <matti.picus at gmail.com
> > <mailto:matti.picus at gmail.com>> wrote:
> >
> >     The pypy C-API progress, where we can basically run all of numpy (on
> a
> >     HEAD version numpy + nightly version of pypy).
> >
> >     That means I will start to look at passing the matplotlib test suite.
> >
> >     My first try segfaults, due to the tkagg backend blit function which
> >     calls into tk via id(data) and id(bbox_array).
> >
> >     That of course will not work on PyPy since the address of a random
> >     python object does not reflect anything usable in C.
> >
> >     So my question, as someone who is new to your codebase, is how to
> >     rewrite this using cffi or the standard C-API without impacting
> >     performance?
> >
> >     Matti
> >
> >     _______________________________________________
> >     Matplotlib-devel mailing list
> >     Matplotlib-devel at python.org <mailto: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/20160902/2398d163/attachment.html>

From matthew.brett at gmail.com  Fri Sep  2 13:18:41 2016
From: matthew.brett at gmail.com (Matthew Brett)
Date: Fri, 2 Sep 2016 10:18:41 -0700
Subject: [Matplotlib-devel] getting pypy to play nice with matplotlib
In-Reply-To: <CAA48SF-soAsKZxsL0Oj2DtBZZnGEHY2WU+QR6vXGRe5uoEWR-g@mail.gmail.com>
References: <29b654a6-7039-ce49-08c7-5a0075602156@gmail.com>
 <CAA48SF_jV_pJa44u+66OGBMLYwAJE8uruh1B8QkZRHqXUfwBkA@mail.gmail.com>
 <825f7b62-0788-9329-d30a-3ccbabfb5f52@gmail.com>
 <CAA48SF-soAsKZxsL0Oj2DtBZZnGEHY2WU+QR6vXGRe5uoEWR-g@mail.gmail.com>
Message-ID: <CAH6Pt5qfkzX_7Gac3d0r+7i6MrsaC-_uH1j7_mJ-w=ryty8p8Q@mail.gmail.com>

On Fri, Sep 2, 2016 at 10:05 AM, Thomas Caswell <tcaswell at gmail.com> wrote:
> On my phone so excuse the terse answer.
>
> Use the numpy calls if likely OK
>
> Work of of master or v2.x
>
> Look at our travis.yml file for how to run the tests.
>
> Python tests.py

The newer tcl / tk linking finds the relevant functions from libraries
loaded by Python, at run time, rather than at compile time.  This is
already working for pypy over at Pillow :

https://github.com/python-pillow/Pillow/pull/1932

Cheers,

Matthew

From tcaswell at gmail.com  Sat Sep  3 22:02:06 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Sun, 04 Sep 2016 02:02:06 +0000
Subject: [Matplotlib-devel] pydata cookbook
Message-ID: <CAA48SF8=uQchvNPVoLVzqJ05vhHkkdOQ4fmnps1McYFZZfifmg@mail.gmail.com>

Folks,

Andy Terrel and Katy Huff are working on a pydata cookbook and have
solicited chapter proposals at https://github.com/pydata/pydata-cookbook
(see the readme there for more details).

The schedule for chapter proposals was Sept 1, but if anyone one on this
list is interested in writing a chapter (15-20 pgs) about matplotlib, I
suspect that there is still time to get it in.

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160904/f298b6c7/attachment.html>

From tcaswell at gmail.com  Mon Sep  5 14:28:38 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Mon, 05 Sep 2016 18:28:38 +0000
Subject: [Matplotlib-devel] Documentation help request for web frameworks
Message-ID: <CAA48SF9pekYGvNdpuEe+qpauaxLHh06P6jRVECOE6KzSx3Q-8w@mail.gmail.com>

Folks,

I know that there is a use of mpl as part of django / flask / bottle / zope
/ ... to write figures directly to buffers and serve them to users without
writing to disk  (an example of this is
http://scipy-cookbook.readthedocs.io/items/Matplotlib_Django.html , but I
think this is better done via `fig.savefig(bytes_io_buffer, format='png')`
rather than reaching in and touching the canvas directly).

This seems like the sort of thing that should be easy to find in our
documentation.  Is anyone interested in writing up 'canonical' examples?

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160905/d00bc7d6/attachment.html>

From Nicolas.Rougier at inria.fr  Wed Sep  7 01:37:41 2016
From: Nicolas.Rougier at inria.fr (Nicolas P. Rougier)
Date: Wed, 7 Sep 2016 06:37:41 +0100
Subject: [Matplotlib-devel] pydata cookbook
In-Reply-To: <CAA48SF8=uQchvNPVoLVzqJ05vhHkkdOQ4fmnps1McYFZZfifmg@mail.gmail.com>
References: <CAA48SF8=uQchvNPVoLVzqJ05vhHkkdOQ4fmnps1McYFZZfifmg@mail.gmail.com>
Message-ID: <ADAC2E1E-A297-4D7F-AC59-302B08386FE9@inria.fr>


I have some material that could be adapted for the chapter.
Anyone else interested in participating ?


Nicolas


> On 04 Sep 2016, at 03:02, Thomas Caswell <tcaswell at gmail.com> wrote:
> 
> Folks,
> 
> Andy Terrel and Katy Huff are working on a pydata cookbook and have solicited chapter proposals at https://github.com/pydata/pydata-cookbook (see the readme there for more details).
> 
> The schedule for chapter proposals was Sept 1, but if anyone one on this list is interested in writing a chapter (15-20 pgs) about matplotlib, I suspect that there is still time to get it in.
> 
> Tom
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel


From tcaswell at gmail.com  Thu Sep  8 23:13:11 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Fri, 09 Sep 2016 03:13:11 +0000
Subject: [Matplotlib-devel] matplotlib v1.5.3
Message-ID: <CAA48SF9Qf3NHVTUDC+-P9n9RZ0H2jDAD5VmAUjmO7Tnx74kkjw@mail.gmail.com>

Folks,

I have backported a few critical bug-fixes tagged 1.5.3.

This release contains a few critical bug fixes:

eliminate fatal exceptions with Qt5.7
memory leak in the contour code
keyboard interaction bug with nbagg
automatic integration with the ipython event loop (if running) which fixes
'naive' integration for IPython 5+

Plan wait to collect the wheels before uploading to pypi and a wider
release.  I have kicked off conda-forge with 1.5.3.

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160909/58badac0/attachment.html>

From tcaswell at gmail.com  Fri Sep  9 00:05:36 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Fri, 09 Sep 2016 04:05:36 +0000
Subject: [Matplotlib-devel] mpl v2.0.0b4
Message-ID: <CAA48SF8E3ODykzLcybuYfbBgCCVa794F+6c38HVZxGhA1F-c=w@mail.gmail.com>

Folks,

I tagged v2.0.0b4 tonight.  As with 1.5.3 will make a bigger announcement
once the binaries are ready.

I think we are (finally) converging.

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160909/5ddf8a0a/attachment.html>

From thomas.robitaille at gmail.com  Fri Sep  9 07:07:07 2016
From: thomas.robitaille at gmail.com (Thomas Robitaille)
Date: Fri, 9 Sep 2016 12:07:07 +0100
Subject: [Matplotlib-devel] OpenGL backend efforts?
Message-ID: <CAGMHX_0wYKjTjhar8-b7=Fg8dgMpjKbyuVAYp+2E2PG-+Cf76w@mail.gmail.com>

Hi everyone,

I was curious as to whether there are currently any active efforts to write
a Matplotlib OpenGL backend? In the past I've seen several starts at a
backend, including for example in the VisPy package, but the latter has now
been removed:

https://github.com/vispy/vispy/pull/1224

Is anyone aware of more recent or maintained efforts?

Thanks!
Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160909/6f0d0ba8/attachment.html>

From jlong at stsci.edu  Fri Sep  9 11:43:38 2016
From: jlong at stsci.edu (Joseph Long)
Date: Fri, 9 Sep 2016 15:43:38 +0000
Subject: [Matplotlib-devel] matplotlib v1.5.3
In-Reply-To: <CAA48SF9Qf3NHVTUDC+-P9n9RZ0H2jDAD5VmAUjmO7Tnx74kkjw@mail.gmail.com>
References: <CAA48SF9Qf3NHVTUDC+-P9n9RZ0H2jDAD5VmAUjmO7Tnx74kkjw@mail.gmail.com>
Message-ID: <D3F8506D.59029%jlong@stsci.edu>

Hi Tom,

Thanks for backporting these! I'm attempting to get the folks at Continuum
Analytics to provide this version in 'conda', but they say the website at
http://matplotlib.org shows 1.5.1 as the latest stable release. Can you
clarify, and if necessary ping the appropriate web maintainer?

Thanks!
--
Joseph Long


From jenshnielsen at gmail.com  Fri Sep  9 11:49:13 2016
From: jenshnielsen at gmail.com (Jens Nielsen)
Date: Fri, 09 Sep 2016 15:49:13 +0000
Subject: [Matplotlib-devel] matplotlib v1.5.3
In-Reply-To: <D3F8506D.59029%jlong@stsci.edu>
References: <CAA48SF9Qf3NHVTUDC+-P9n9RZ0H2jDAD5VmAUjmO7Tnx74kkjw@mail.gmail.com>
 <D3F8506D.59029%jlong@stsci.edu>
Message-ID: <CAM-Pw01jV_c1O4cLJjEdmcpFp2k4KWUBQwjO=dKka9RSEWvUGw@mail.gmail.com>

That would be https://github.com/matplotlib/matplotlib.github.com/issues/18

The docs needs to be rebuild for 1.5.2



On Fri, 9 Sep 2016 at 16:44 Joseph Long <jlong at stsci.edu> wrote:

> Hi Tom,
>
> Thanks for backporting these! I'm attempting to get the folks at Continuum
> Analytics to provide this version in 'conda', but they say the website at
> http://matplotlib.org shows 1.5.1 as the latest stable release. Can you
> clarify, and if necessary ping the appropriate web maintainer?
>
> Thanks!
> --
> Joseph Long
>
> _______________________________________________
> 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/20160909/427c5265/attachment.html>

From tcaswell at gmail.com  Fri Sep  9 12:04:10 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Fri, 09 Sep 2016 16:04:10 +0000
Subject: [Matplotlib-devel] matplotlib v1.5.3
In-Reply-To: <CAM-Pw01jV_c1O4cLJjEdmcpFp2k4KWUBQwjO=dKka9RSEWvUGw@mail.gmail.com>
References: <CAA48SF9Qf3NHVTUDC+-P9n9RZ0H2jDAD5VmAUjmO7Tnx74kkjw@mail.gmail.com>
 <D3F8506D.59029%jlong@stsci.edu>
 <CAM-Pw01jV_c1O4cLJjEdmcpFp2k4KWUBQwjO=dKka9RSEWvUGw@mail.gmail.com>
Message-ID: <CAA48SF8FWj2yibWzGrokHbLg0_BAyqTaOpK7RTnhhku6D7nuVg@mail.gmail.com>

I have the 1.5.3 wheels and will upload to pypi after lunch.

After that i will rebuild and upload the docs.

Tom

On Fri, Sep 9, 2016, 11:49 Jens Nielsen <jenshnielsen at gmail.com> wrote:

> That would be
> https://github.com/matplotlib/matplotlib.github.com/issues/18
> The docs needs to be rebuild for 1.5.2
>
>
>
> On Fri, 9 Sep 2016 at 16:44 Joseph Long <jlong at stsci.edu> wrote:
>
>> Hi Tom,
>>
>> Thanks for backporting these! I'm attempting to get the folks at Continuum
>> Analytics to provide this version in 'conda', but they say the website at
>> http://matplotlib.org shows 1.5.1 as the latest stable release. Can you
>> clarify, and if necessary ping the appropriate web maintainer?
>>
>> Thanks!
>> --
>> Joseph Long
>>
>> _______________________________________________
>> 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/20160909/947ef2f1/attachment.html>

From tcaswell at gmail.com  Fri Sep  9 19:03:14 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Fri, 09 Sep 2016 23:03:14 +0000
Subject: [Matplotlib-devel] v1.5.3 and v2.0.0b4
Message-ID: <CAA48SF9s6xPK=9p43zbkq6DMH2PmvzJCRLw0kTO0RSjYdWPsaQ@mail.gmail.com>

Folks,

Source + wheels for both are up on pypi and conda forge for both is built.

I will get the docs updated + annoucenments out  tomorrow.

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160909/a696eac9/attachment.html>

From tcaswell at gmail.com  Sat Sep 10 22:40:53 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Sun, 11 Sep 2016 02:40:53 +0000
Subject: [Matplotlib-devel] force-pushed to matplotlib.github.io repo
Message-ID: <CAA48SF-zjTXSyoJx8KD1SUjZ-kkWdkPN5zmMzpjNjNu5z8SsiQ@mail.gmail.com>

Folks,

I made a mistake uploading the 1.5.3 docs, cleaned it up and then
force-pushed to the matplotlib.github.io repository.

If you happened to fetch or clone it in ~the last hour be aware

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160911/19df7eb6/attachment.html>

From tcaswell at gmail.com  Thu Sep 15 22:44:39 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Fri, 16 Sep 2016 02:44:39 +0000
Subject: [Matplotlib-devel] waffle-bot
Message-ID: <CAA48SF-+7PCas96Wj0E5GD6v1a6FO_P1zJ9XBpKK5KW-0r1kHw@mail.gmail.com>

Folks,

We have been half-heatedly signed up for waffle.io for a while now.  They
suggested we try turning on wafflebot (which I thought would just take care
of assigning 'needs review' to PRs under it's account instead of Mikes),
but it looks like it was doing a whole bunch more.

I have uninstalled it, we probably should have a discussion about what
parts of it we want to use before / if we turn it back on.

Sorry for the noise!

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160916/8224f32e/attachment.html>

From t at too.gy  Sat Sep 17 16:45:16 2016
From: t at too.gy (tgy)
Date: Sat, 17 Sep 2016 13:45:16 -0700 (MST)
Subject: [Matplotlib-devel] animation in SVG
In-Reply-To: <1425053384790-45021.post@n5.nabble.com>
References: <1425053384790-45021.post@n5.nabble.com>
Message-ID: <1474145116302-47524.post@n5.nabble.com>

Hi,

I am very interested in SVG animations and I would like to know how your
research turned out.

Thanks,

-- 
tgy



--
View this message in context: http://matplotlib.1069221.n5.nabble.com/animation-in-SVG-tp45021p47524.html
Sent from the matplotlib - devel mailing list archive at Nabble.com.

From tcaswell at gmail.com  Sun Sep 18 15:33:35 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Sun, 18 Sep 2016 19:33:35 +0000
Subject: [Matplotlib-devel] pdf docs
Message-ID: <CAA48SF9fpHB5nRLT3vmYUoVnUqE2So55ZgTo-Pu4Wu78VOp9KQ@mail.gmail.com>

Folks,

Last time I tried to build the pdf docs I ran into LaTeX exceptions that I
could not sort out (to be fair I did not try super hard).  Not clear if the
issue is a bad/missing latex setup on my machine, there is something wrong
with the docs, or it is a sphinx bug.

There are two issues (one on matplotlib/matplotlib one on
matplotlib/devdocs) noting that the pdf docs are missing from 1.5.3,
2.0.0b4, and devdocs.

Has anyone successfully built the pdf version of the docs recently?

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160918/de6cc5ce/attachment.html>

From tcaswell at gmail.com  Mon Sep 19 21:39:23 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Tue, 20 Sep 2016 01:39:23 +0000
Subject: [Matplotlib-devel] mpl weekly phone calls
Message-ID: <CAA48SF9ngsMCus0RWVWdnFGxB7FpyrthshvVG11WQv7Lp2EwBg@mail.gmail.com>

Folks,

We have been having weekly phone calls about matplotlib Mondays at 3pm EST
(which is a not awful compromise between Hawaii and England).

If you are interested in joining please let me know.

If people are interested, but can not make this time we will reschedule
(but do that off-list).

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160920/e88d54ec/attachment.html>

From tcaswell at gmail.com  Mon Sep 19 22:24:24 2016
From: tcaswell at gmail.com (Thomas Caswell)
Date: Tue, 20 Sep 2016 02:24:24 +0000
Subject: [Matplotlib-devel] new github features
Message-ID: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>

Folks,

Have people had a chance to look at the new review tools on github?  Do we
want to work these into our workflow?  It looks like we can set it so that
PRs can only be merged via the web UI with at least one approve review and
no 'needs work' reviews if we want.   My initial reaction is we should try
to use the review tools, but hold on on the engineering control on merges
until we have a better sense of what that would look like.

Has anyone looked at the 'projects' feature yet?

A suggestion from Nelle Varoquaux is to follow skimage/sklearn and on
review add a [mrg+N] to the title to help other devs find PRs that are
almost ready to merge, but need another set of eyes.

Tom
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160920/7dc4884c/attachment.html>

From bussonniermatthias at gmail.com  Mon Sep 19 22:32:48 2016
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Mon, 19 Sep 2016 19:32:48 -0700
Subject: [Matplotlib-devel] new github features
In-Reply-To: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
References: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
Message-ID: <CANJQusX8VxaAG4FTpQoy_=MHihtmh5thrkJtdGDJrQ=R5jajJQ@mail.gmail.com>

Hi all,

One data point, it seem that some users/dev (at least on Jupyter side)
are complaining that they don't get the "grouped-review-comments" by
emails. So are waiting for feedback they never get until they check
the PR.

Still haven't figured out when this happens.

I'm curious if the "Approved" review get reset after a new commit is
pushed, and If so and the N reviewer is enforced, it can be annoying
if someone approved with a small condition, like typo fixed or
something similar.
-- 
M

On Mon, Sep 19, 2016 at 7:24 PM, Thomas Caswell <tcaswell at gmail.com> wrote:
> Folks,
>
> Have people had a chance to look at the new review tools on github?  Do we
> want to work these into our workflow?  It looks like we can set it so that
> PRs can only be merged via the web UI with at least one approve review and
> no 'needs work' reviews if we want.   My initial reaction is we should try
> to use the review tools, but hold on on the engineering control on merges
> until we have a better sense of what that would look like.
>
> Has anyone looked at the 'projects' feature yet?
>
> A suggestion from Nelle Varoquaux is to follow skimage/sklearn and on review
> add a [mrg+N] to the title to help other devs find PRs that are almost ready
> to merge, but need another set of eyes.
>
> Tom
>
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel
>

From efiring at hawaii.edu  Mon Sep 19 23:00:18 2016
From: efiring at hawaii.edu (Eric Firing)
Date: Mon, 19 Sep 2016 17:00:18 -1000
Subject: [Matplotlib-devel] new github features
In-Reply-To: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
References: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
Message-ID: <5e9d29d9-9e95-9dc4-0d23-4b02686fc87f@hawaii.edu>

On 2016/09/19 4:24 PM, Thomas Caswell wrote:
> Folks,
>
> Have people had a chance to look at the new review tools on github?  Do
> we want to work these into our workflow?  It looks like we can set it so
> that PRs can only be merged via the web UI with at least one approve
> review and no 'needs work' reviews if we want.   My initial reaction is
> we should try to use the review tools, but hold on on the engineering
> control on merges until we have a better sense of what that would look like.

I'm not a fan of this new mechanism.  One of the problems is that 
sometimes one doesn't know at the outset whether one will be doing a 
thorough review, or just starting to look and make a few comments.  I 
don't much like the automated comments it generates based on the review 
checkbutton, either.  Making this mechanism mandatory would be quite an 
irritant for little or no benefit, I think.

Some automated operations are good.  For example, if a "closes #1234" in 
an obvious location in a PR causes #1234 to be closed when the PR is 
merged, that's good.  Unfortunately, I haven't figured out how this 
works--it seems like sometimes it does, sometimes it doesn't. 
Presumably I just don't know the rule github is using.  So here, a 
better design would be a box in the PR specifically for linking in 
"closes" issue numbers.

Thumbs-up emoji counts are not helpful; I need to look to see whose 
thumb it is, and then I'm still left wondering exactly what was being 
approved.  And was it a thoughtful approval, or just a twitch?

So overall, with my usual curmudgeon's hat on, covering my small brain, 
I favor keeping the automation simple, obvious, and fairly minimal.  For 
everything else, let's use normal human communications, not bots.

>
> Has anyone looked at the 'projects' feature yet?
Haven't looked.  Should I?
>
> A suggestion from Nelle Varoquaux is to follow skimage/sklearn and on
> review add a [mrg+N] to the title to help other devs find PRs that are
> almost ready to merge, but need another set of eyes.

Aha! So that's what those are!  The "N" is the number of eye-sets needed?
I don't object to having a few conventions like that.  If it ends up 
helping us cut down the open PR and issue counts, that would be great.
Conventions are better than bots.

Eric



From story645 at gmail.com  Tue Sep 20 12:40:03 2016
From: story645 at gmail.com (Hannah)
Date: Tue, 20 Sep 2016 12:40:03 -0400
Subject: [Matplotlib-devel] new github features
In-Reply-To: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
References: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
Message-ID: <CAOuQ7j27cGL-vAjktGzs-XobxnBu-KmojWQk7gHsWPgLRe+9EQ@mail.gmail.com>

I think the projects board (it looks like a trello type thing) could be a
good way to organize MEPs and/or new feature requests with a lot of moving
pieces (like the categorical work...)


On Mon, Sep 19, 2016 at 10:24 PM, Thomas Caswell <tcaswell at gmail.com> wrote:

> Folks,
>
> Have people had a chance to look at the new review tools on github?  Do we
> want to work these into our workflow?  It looks like we can set it so that
> PRs can only be merged via the web UI with at least one approve review and
> no 'needs work' reviews if we want.   My initial reaction is we should try
> to use the review tools, but hold on on the engineering control on merges
> until we have a better sense of what that would look like.
>
> Has anyone looked at the 'projects' feature yet?
>
> A suggestion from Nelle Varoquaux is to follow skimage/sklearn and on
> review add a [mrg+N] to the title to help other devs find PRs that are
> almost ready to merge, but need another set of eyes.
>
> 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/20160920/ac92e40d/attachment.html>

From bussonniermatthias at gmail.com  Tue Sep 20 13:10:15 2016
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Tue, 20 Sep 2016 10:10:15 -0700
Subject: [Matplotlib-devel] new github features
In-Reply-To: <5e9d29d9-9e95-9dc4-0d23-4b02686fc87f@hawaii.edu>
References: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
 <5e9d29d9-9e95-9dc4-0d23-4b02686fc87f@hawaii.edu>
Message-ID: <CANJQusUZfwS+n5KW3baQ=vNJXo6Y2PrO5sMw4G=2dHr14ALT8w@mail.gmail.com>

Hi Eric,

> Some automated operations are good.  For example, if a "closes #1234" in an obvious location in a PR causes #1234 to be closed when the PR is merged, that's good.  Unfortunately, I haven't figured out how this works--it seems like sometimes it does, sometimes it doesn't. Presumably I just don't know the rule github is using.  So here, a better design would be a box in the PR specifically for linking in "closes" issue numbers.

I can inform you on that.

It has to be in one of the _commit message_, once the commit is in
master the PR get closed.

It is confusing as when there is a single commit, GitHub make the the
default for the pull-request description,
but being in the commit allows you to close issues by pushing directly
on master (for project that do it).

See https://help.github.com/articles/closing-issues-via-commit-messages/

Cheers,
-- 
M

On Mon, Sep 19, 2016 at 8:00 PM, Eric Firing <efiring at hawaii.edu> wrote:
> On 2016/09/19 4:24 PM, Thomas Caswell wrote:
>>
>> Folks,
>>
>> Have people had a chance to look at the new review tools on github?  Do
>> we want to work these into our workflow?  It looks like we can set it so
>> that PRs can only be merged via the web UI with at least one approve
>> review and no 'needs work' reviews if we want.   My initial reaction is
>> we should try to use the review tools, but hold on on the engineering
>> control on merges until we have a better sense of what that would look
>> like.
>
>
> I'm not a fan of this new mechanism.  One of the problems is that sometimes
> one doesn't know at the outset whether one will be doing a thorough review,
> or just starting to look and make a few comments.  I don't much like the
> automated comments it generates based on the review checkbutton, either.
> Making this mechanism mandatory would be quite an irritant for little or no
> benefit, I think.
>
> Some automated operations are good.  For example, if a "closes #1234" in an
> obvious location in a PR causes #1234 to be closed when the PR is merged,
> that's good.  Unfortunately, I haven't figured out how this works--it seems
> like sometimes it does, sometimes it doesn't. Presumably I just don't know
> the rule github is using.  So here, a better design would be a box in the PR
> specifically for linking in "closes" issue numbers.
>
> Thumbs-up emoji counts are not helpful; I need to look to see whose thumb it
> is, and then I'm still left wondering exactly what was being approved.  And
> was it a thoughtful approval, or just a twitch?
>
> So overall, with my usual curmudgeon's hat on, covering my small brain, I
> favor keeping the automation simple, obvious, and fairly minimal.  For
> everything else, let's use normal human communications, not bots.
>
>>
>> Has anyone looked at the 'projects' feature yet?
>
> Haven't looked.  Should I?
>>
>>
>> A suggestion from Nelle Varoquaux is to follow skimage/sklearn and on
>> review add a [mrg+N] to the title to help other devs find PRs that are
>> almost ready to merge, but need another set of eyes.
>
>
> Aha! So that's what those are!  The "N" is the number of eye-sets needed?
> I don't object to having a few conventions like that.  If it ends up helping
> us cut down the open PR and issue counts, that would be great.
> Conventions are better than bots.
>
> Eric
>
>
>
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel

From rmay31 at gmail.com  Wed Sep 21 17:14:33 2016
From: rmay31 at gmail.com (Ryan May)
Date: Wed, 21 Sep 2016 15:14:33 -0600
Subject: [Matplotlib-devel] new github features
In-Reply-To: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
References: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
Message-ID: <CAKH0P+U7a5K43ybHfPYtT3Rerb9dDc0zU4_yAvegnzEWSAAw4w@mail.gmail.com>

Hi,

In general, given the griping I've seen over time for how hard it is to do
code review on GitHub Pull Requests, I think we (the open source community)
owe it to GitHub to try out the new feature they rolled out to address and
give them feedback--that's the only way the feature can ever meet our needs.

For our (matplotlib) case, I think we get a few good gains:
- line comments are batched so we can cut down on the number of
notifications
- Your work is saved so you can come back later and finish--and adjust when
you see something later that invalidates your comment *before* commenting
for all to see
- Reviews are atomic items, so you can get an idea of what a specific
reviewer thinks (and not lose pieces)
- By having the state of Approved/Needs Changes actually tracked, there's
no need to read through the entire PR to see what objections are left.
(Granted, it does mean people have to come back and approve it once the
changes are made)

Seems like a lot to gain for little to no extra effort--having used it, I
certainly fail to see any problem. The matter of turning on required
reviews for protected branches is a whole other matter, though I will say
you can always let admins override.

Ryan


On Mon, Sep 19, 2016 at 8:24 PM, Thomas Caswell <tcaswell at gmail.com> wrote:

> Folks,
>
> Have people had a chance to look at the new review tools on github?  Do we
> want to work these into our workflow?  It looks like we can set it so that
> PRs can only be merged via the web UI with at least one approve review and
> no 'needs work' reviews if we want.   My initial reaction is we should try
> to use the review tools, but hold on on the engineering control on merges
> until we have a better sense of what that would look like.
>
> Has anyone looked at the 'projects' feature yet?
>
> A suggestion from Nelle Varoquaux is to follow skimage/sklearn and on
> review add a [mrg+N] to the title to help other devs find PRs that are
> almost ready to merge, but need another set of eyes.
>
> Tom
>
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel
>
>


-- 
Ryan May
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160921/265347a3/attachment.html>

From rmay31 at gmail.com  Wed Sep 21 17:26:57 2016
From: rmay31 at gmail.com (Ryan May)
Date: Wed, 21 Sep 2016 15:26:57 -0600
Subject: [Matplotlib-devel] new github features
In-Reply-To: <CANJQusX8VxaAG4FTpQoy_=MHihtmh5thrkJtdGDJrQ=R5jajJQ@mail.gmail.com>
References: <CAA48SF-1Pzqf2CaoZWC_cxWs37NAO263PmGy05=2HpDJA0X+nA@mail.gmail.com>
 <CANJQusX8VxaAG4FTpQoy_=MHihtmh5thrkJtdGDJrQ=R5jajJQ@mail.gmail.com>
Message-ID: <CAKH0P+VEsoAHh=Yum_udweTj6swKhLKa3Tn8Lh+GKh+ZTCed2A@mail.gmail.com>

Matthias,

Regarding the emails, could this be due to notification settings? I notice
that PR reviews are a separate item you have to check in the notification
center.

Ryan

On Mon, Sep 19, 2016 at 8:32 PM, Matthias Bussonnier <
bussonniermatthias at gmail.com> wrote:

> Hi all,
>
> One data point, it seem that some users/dev (at least on Jupyter side)
> are complaining that they don't get the "grouped-review-comments" by
> emails. So are waiting for feedback they never get until they check
> the PR.
>
> Still haven't figured out when this happens.
>
> I'm curious if the "Approved" review get reset after a new commit is
> pushed, and If so and the N reviewer is enforced, it can be annoying
> if someone approved with a small condition, like typo fixed or
> something similar.
> --
> M
>
> On Mon, Sep 19, 2016 at 7:24 PM, Thomas Caswell <tcaswell at gmail.com>
> wrote:
> > Folks,
> >
> > Have people had a chance to look at the new review tools on github?  Do
> we
> > want to work these into our workflow?  It looks like we can set it so
> that
> > PRs can only be merged via the web UI with at least one approve review
> and
> > no 'needs work' reviews if we want.   My initial reaction is we should
> try
> > to use the review tools, but hold on on the engineering control on merges
> > until we have a better sense of what that would look like.
> >
> > Has anyone looked at the 'projects' feature yet?
> >
> > A suggestion from Nelle Varoquaux is to follow skimage/sklearn and on
> review
> > add a [mrg+N] to the title to help other devs find PRs that are almost
> ready
> > to merge, but need another set of eyes.
> >
> > Tom
> >
> > _______________________________________________
> > Matplotlib-devel mailing list
> > Matplotlib-devel at python.org
> > https://mail.python.org/mailman/listinfo/matplotlib-devel
> >
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel
>



-- 
Ryan May
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20160921/9bf3b89c/attachment.html>

From nelle.varoquaux at gmail.com  Fri Sep 30 12:32:28 2016
From: nelle.varoquaux at gmail.com (Nelle Varoquaux)
Date: Fri, 30 Sep 2016 09:32:28 -0700
Subject: [Matplotlib-devel] BIDS machine shop & matplotlib
Message-ID: <CAE-UAvSwNbUC8jGNqCX-OiJ8qvPA6XLyJF46o2=rtG4Pm11pJA@mail.gmail.com>

Hello matplotlib-devs !

As some of you may know, I am involved with St?fan van der Walt in an
initiative around Berkeley campus known as the "BIDS Machine Shop". It
consists of trying to bridge the gap in software development and best
practises around campus by introducing undergraduates  to research
software development and open source communities.
St?fan and I will manage students on different research projects and
open source packages, including matplotlib and scikit-image. I'll be
managing the matplotlib side of the machine shop, and you should see
(I hope) a bunch of contributions from our students popping up on
github.

I am going to create a label "Machine Shop" and a project "Machine
Shop" for the duration of the semester (maybe longer if this
experiment is a success). I'll assign "easy fix" tickets to our
students. I expect them to do peer-review before opening the PRs for
broader reviews. Please, wait for the tag "need review" to be set on
those PR before reviewing theme. Once they fixed these issues, I'll
assign them to more complex tasks.

If you have questions/suggestions, don't hesitate to chime in.

Thanks,
Nelle

PS: we should have a website up and running soon with more details.

From pmhobson at gmail.com  Fri Sep 30 14:15:48 2016
From: pmhobson at gmail.com (Paul Hobson)
Date: Fri, 30 Sep 2016 11:15:48 -0700
Subject: [Matplotlib-devel] BIDS machine shop & matplotlib
In-Reply-To: <CAE-UAvSwNbUC8jGNqCX-OiJ8qvPA6XLyJF46o2=rtG4Pm11pJA@mail.gmail.com>
References: <CAE-UAvSwNbUC8jGNqCX-OiJ8qvPA6XLyJF46o2=rtG4Pm11pJA@mail.gmail.com>
Message-ID: <CADT3MED4awtMu+gGtmoAeEE1QhEbs5sHW8rEdS8YLVhiYJ54fA@mail.gmail.com>

Nelle,

I think the "needs review" tag is automatically added. I might be wrong
though.
-p

On Fri, Sep 30, 2016 at 9:32 AM, Nelle Varoquaux <nelle.varoquaux at gmail.com>
wrote:

> Hello matplotlib-devs !
>
> As some of you may know, I am involved with St?fan van der Walt in an
> initiative around Berkeley campus known as the "BIDS Machine Shop". It
> consists of trying to bridge the gap in software development and best
> practises around campus by introducing undergraduates  to research
> software development and open source communities.
> St?fan and I will manage students on different research projects and
> open source packages, including matplotlib and scikit-image. I'll be
> managing the matplotlib side of the machine shop, and you should see
> (I hope) a bunch of contributions from our students popping up on
> github.
>
> I am going to create a label "Machine Shop" and a project "Machine
> Shop" for the duration of the semester (maybe longer if this
> experiment is a success). I'll assign "easy fix" tickets to our
> students. I expect them to do peer-review before opening the PRs for
> broader reviews. Please, wait for the tag "need review" to be set on
> those PR before reviewing theme. Once they fixed these issues, I'll
> assign them to more complex tasks.
>
> If you have questions/suggestions, don't hesitate to chime in.
>
> Thanks,
> Nelle
>
> PS: we should have a website up and running soon with more details.
> _______________________________________________
> 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/20160930/a7f35de9/attachment.html>

From nelle.varoquaux at gmail.com  Fri Sep 30 14:20:00 2016
From: nelle.varoquaux at gmail.com (Nelle Varoquaux)
Date: Fri, 30 Sep 2016 11:20:00 -0700
Subject: [Matplotlib-devel] BIDS machine shop & matplotlib
In-Reply-To: <CADT3MED4awtMu+gGtmoAeEE1QhEbs5sHW8rEdS8YLVhiYJ54fA@mail.gmail.com>
References: <CAE-UAvSwNbUC8jGNqCX-OiJ8qvPA6XLyJF46o2=rtG4Pm11pJA@mail.gmail.com>
 <CADT3MED4awtMu+gGtmoAeEE1QhEbs5sHW8rEdS8YLVhiYJ54fA@mail.gmail.com>
Message-ID: <CAE-UAvQC-_bihmTNYaCoVdLwoNWDBrVTWZJSFRM_fbr1T3AgsQ@mail.gmail.com>

On 30 September 2016 at 11:15, Paul Hobson <pmhobson at gmail.com> wrote:
> Nelle,
>
> I think the "needs review" tag is automatically added. I might be wrong
> though.

Yes. I plan on removing it by hand everytime one of my students open a
PR not ready to be merged. It is going to be painful, but I don't see
any other way?

I actually think it is slightly problematic to have it added
automatically. A bunch of PRs are not ready to be merged or review,
and not many people remove the tag after reviewing if changes need to
be applied. That means most of our PRs have this tag, so it is hard to
identify the one that actually need to be reviewed.

Chers,
N

> -p
>
> On Fri, Sep 30, 2016 at 9:32 AM, Nelle Varoquaux <nelle.varoquaux at gmail.com>
> wrote:
>>
>> Hello matplotlib-devs !
>>
>> As some of you may know, I am involved with St?fan van der Walt in an
>> initiative around Berkeley campus known as the "BIDS Machine Shop". It
>> consists of trying to bridge the gap in software development and best
>> practises around campus by introducing undergraduates  to research
>> software development and open source communities.
>> St?fan and I will manage students on different research projects and
>> open source packages, including matplotlib and scikit-image. I'll be
>> managing the matplotlib side of the machine shop, and you should see
>> (I hope) a bunch of contributions from our students popping up on
>> github.
>>
>> I am going to create a label "Machine Shop" and a project "Machine
>> Shop" for the duration of the semester (maybe longer if this
>> experiment is a success). I'll assign "easy fix" tickets to our
>> students. I expect them to do peer-review before opening the PRs for
>> broader reviews. Please, wait for the tag "need review" to be set on
>> those PR before reviewing theme. Once they fixed these issues, I'll
>> assign them to more complex tasks.
>>
>> If you have questions/suggestions, don't hesitate to chime in.
>>
>> Thanks,
>> Nelle
>>
>> PS: we should have a website up and running soon with more details.
>> _______________________________________________
>> Matplotlib-devel mailing list
>> Matplotlib-devel at python.org
>> https://mail.python.org/mailman/listinfo/matplotlib-devel
>
>

From efiring at hawaii.edu  Fri Sep 30 14:41:29 2016
From: efiring at hawaii.edu (Eric Firing)
Date: Fri, 30 Sep 2016 08:41:29 -1000
Subject: [Matplotlib-devel] BIDS machine shop & matplotlib
In-Reply-To: <CAE-UAvQC-_bihmTNYaCoVdLwoNWDBrVTWZJSFRM_fbr1T3AgsQ@mail.gmail.com>
References: <CAE-UAvSwNbUC8jGNqCX-OiJ8qvPA6XLyJF46o2=rtG4Pm11pJA@mail.gmail.com>
 <CADT3MED4awtMu+gGtmoAeEE1QhEbs5sHW8rEdS8YLVhiYJ54fA@mail.gmail.com>
 <CAE-UAvQC-_bihmTNYaCoVdLwoNWDBrVTWZJSFRM_fbr1T3AgsQ@mail.gmail.com>
Message-ID: <efaead86-48bd-2997-b994-97b201a209f1@hawaii.edu>

On 2016/09/30 8:20 AM, Nelle Varoquaux wrote:
> I actually think it is slightly problematic to have it added
> automatically. A bunch of PRs are not ready to be merged or review,
> and not many people remove the tag after reviewing if changes need to
> be applied. That means most of our PRs have this tag, so it is hard to
> identify the one that actually need to be reviewed.

I agree.  Automatic adding of this tag is pointless, not helpful, as far 
as I can see.

Eric

From efiring at hawaii.edu  Fri Sep 30 22:51:18 2016
From: efiring at hawaii.edu (Eric Firing)
Date: Fri, 30 Sep 2016 16:51:18 -1000
Subject: [Matplotlib-devel] encoders for animations
Message-ID: <94e0e688-44a8-6132-c662-6bada1c842d6@hawaii.edu>

For making video files the animation module presently supports ffmpeg 
(the default), avconv (a fork of ffmpeg), and mencoder.

Does anyone use mencoder? If so, why?  And would you be seriously 
inconvenienced or otherwise harmed if we ended our only partially 
successful support of mencoder as an option in the animation module?

Thank you.

Eric