[SciPy-Dev] [Scikit-learn-general] Renaming scikits.learn

[josef.pktd at gmail.com]

On Tue, Jul 19, 2011 at 12:04 PM, Olivier Grisel
<olivier.grisel at ensta.org> wrote:
> 2011/7/19 Wes McKinney <wesmckinn at gmail.com>:
>>
>> As Gaël mentioned, we ({scikits}.statsmodels) have more or less
>> decided to ditch the namespace. Of course we are a smaller group of
>> developers. I personally don't see the benefit of the namespace,
>> especially in larger and well-established libraries. I understand the
>> original intent of scikits was to create a scipy-like package but
>> allow everyone to have their own release cycle. For a library like
>> scikit-learn or statsmodels (which are already becoming quite large,
>> not as big as scipy but maybe someday!) it seems like mainly an
>> annoyance with few redeeming qualities.
>
> Do you plan to ditch the "scikits" name from the project branding as
> well or just from the python package?
>
> It would be great if other scikits.* developers could express their
> own opinion on the issue.

I'm pretty much indifferent either way.
I really like the name recognition and family relationship that the
scikits name includes, both across scikits and with scipy. On Windows
without any distribution, we had some problems early on with namespace
installation, but I didn't hear of any problems in a long time.
(Problems with EPD enstaller seems to be more a problem of EPD than of
the scikits.)

On the other hand we picked the name statsmodels from the beginning
for it's usefulness as a standalone name that is easily searchable,
and most times we only refer to scikits.statsmodels as statmodels.

So except for the grouping together of the scikits packages and scipy
in terms of association by name (e.g. we used it for GSOC
applications), I don't see any disadvantage for statsmodels, and not
having a namespace package still makes life a bit easier.

Josef

>
> --
> Olivier
> http://twitter.com/ogrisel - http://github.com/ogrisel
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-dev
>