[Pkg-exppsy-maintainers] Licensing

Michael Hanke michael.hanke at gmail.com
Sun Feb 24 19:16:11 UTC 2008


On Sat, Feb 23, 2008 at 09:20:59PM -0500, Christoph T. Weidemann wrote:
> Hi Michael!
> 
> Thanks for the info! I was not aware of the scipy license
> compatibility page. I was a bit surprised by their reasoning and
> especially by the fact that they won't even consider LGPL. I would
> definitely be interested to hear what the NIPY people think of this
> when you bring it up at the code sprint.
> 
> In any case, here's my take on it:
> 
> There doesn't seem to be a great chance that pymvpa will be part of
> scipy or NIPY anytime soon, and especially in the later case I'm not
> even sure if it should be (the pattern analyses should be general
> enough to be useful for all kinds of data, not just those from
> neuroimaging). Likewise, as far as I know there's no company who's
> currently interested in helping to improve pymvpa and who wouldn't do
> it if we switched to GPL. (The situation might be different for
> pynifty.)
> 
> So it seems to me that this compliance with their licensing policy,
> which (it seems) neither of us is particularly happy with, is
> premature at best. If there is serious interest in having pymvpa
> become part of one of these packages in the future, or if a company
> wants to help out, but doesn't want to be bound by the GPL we could
> always reconsider then (or ask them to reconsider). In short, at the
> present point I don't think we should cast away the GPL, just because
> some other popular packages do.
> 
> Also, I found the following note at
> http://projects.scipy.org/neuroimaging/ni/wiki/ReadMe :
> "This NIPY distribution contains no GNU General Public Licensed
> (GPLed) code so it may be used in proprietary projects. There are
> interfaces to some GNU code but these are entirely optional."
> So, if integration with NIPY is a goal, why not have that achieved
> with an interface to the GPLed code. Those who don't want to be bound
> by the GPL can't use pymvpa, those who want to use it, have to share
> alike.
As I said, I did not do it for any single 'money'-reason given on the SciPy
licensing page. I do not agree with any of the
we-need-to-attract-companies-points given there.
Furthermore I do not expect that any company would take pymvpa for
anything or even contribute to it ;-)

I do not even speculate about having code merged into NiPy and/or SciPy.
The _only_ argument which is valid for me (and my only reason) is that
that having very special interest projects like pymvpa and using some
license for one project and some other for another, slows down the overall
development process. You can easily get into licensing problems when you
want to merge code licensed under something else. Having related
projects (wrt language, field, ...) share a single license makes life
easier. This is maybe also the reason why Perl licensing statements
mostly look like: 'Same as Perl itself.'

SciPy/NumPy simply is _the_ big player in the field - using its license
seem reasonable to me. And in the end it is all free software
(simplified view of the reality -- I know ;-).


Cheers,

Michael

-- 
GPG key:  1024D/3144BE0F Michael Hanke
http://apsy.gse.uni-magdeburg.de/hanke
ICQ: 48230050



More information about the Pkg-exppsy-maintainers mailing list