[CFT] math/py-pandas: Update to 0.11.0rc1

[John W. O'Brien]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 04/20/2013 07:21 PM, Geoffrey Mainland wrote:
> On 4/20/13 8:46 PM, John W. O'Brien wrote:
>> freebsd-python@: Looping in math/py-matplotlib maintainer See the
>> earlier part of this thread for background.
>> 
>> A pandas dev and I just discovered that one of the pandas test 
>> failures is induced because the value sys.getdefaultencoding()
>> returns changes upon importing matplotlib.pyplot.
>> 
>> https://github.com/pydata/pandas/issues/3360#issuecomment-16709420
>>
>>
>> 
Could you help me assess whether this is expected behavior, and, if
>> so, how we could expect it and accommodate it in pandas. If you
>> think this is a bug in matplotlib, I will open a PR.
>> 
>> Regards, John
> 
> Hi John,
> 
> I'm sorry, I don't think I understand. The github link you sent
> above identifies PyGTK as the culprit---it sets the default
> encoding. This doesn't seem to be a problem with the FreeBSD port,
> or even a problem with matplotlib, but is a problem with PyGTK. Is
> this not accurate?

You're right, of course. I discovered that after sending my earlier
message. In fact, I was just composing a reply to retract my question
and let matplotlib off the hook.

I'm surprised that this PyGTK problem doesn't show up more often. Have
you run into it before with MPL?

Cheers,
John
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iQEcBAEBAgAGBQJRcyQnAAoJEBRzAKlhyP/Ff0gH/1WP1WHW3AWmmxZ9rqqQqGZQ
8A6EPlplsvJXdvlXksCArPf7F6sGIbNeL+heV+FEu9oYp+Qmkth1CPPvl250jw+W
CVhym944iRP9uqQiCIWUxpPwT5sWkxGcmnlpaHTi9O/wVLg0aQDe9lu6iFQYVQ7h
zbWwckGuU42EWUHsJExWPUYx4AWTTk3DJNXKCg+7YhVGvgj0ClmP6vJPF7zsXsHI
xSMX6EsgQINggo6q/WwDT87sHTgPhNWZTQzinZADJ+x68Q0SD5zoojW+QNRLIinW
fBu6KtaPpry1GWKGPXzLgPaAWVtAHF0MQ3OZ1ZxtRE+2suWElj0ZtlH91UF177I=
=LHtf
-----END PGP SIGNATURE-----