[Pkg-exppsy-maintainers] Licensing

[Michael Hanke]

Hi Christoph,

On Sat, Feb 23, 2008 at 01:28:55PM -0500, Christoph T. Weidemann wrote:
> Hi all!
> 
> First of all congrats to the first release of PyMVPA and thanks for
> all the hard work! I'm looking forward to using and contributing to
> it!
> I have a question / suggestion regarding the license of PyMVPA: Is
> there any good reason for PyMVPA to use the MIT license rather than
> the GPL? If there's no compelling reason to use the MIT license, I
> would like to suggest abandoning the MIT license and switching to the
> GPL. I'll try to lay out my reasoning behind this suggestion below.
> 
> I think the requirement to pass on the freedom that comes with GPL
> licensed software to anybody it is re-distributed do (modified or not)
> has done a lot for the spread of free software. In short, I think it's
> pretty safe to assume that without this requirement there would be
> less free software today. Likewise, I think it's reasonable to assume
> that new software that is released with such a requirement, will
> increase the chances that new software that builds on it will also be
> free. Thus, using the GPL in favor of more permissive licenses is
> benefiting the free software community. The following text makes a
> strong case why the GPL is a better way to promote the freedom of the
> users than more permissive licenses such as the one currently adopted
> for pymvpa:
> http://www.fsf.org/licensing/essays/why-copyleft.html
> 
> There's also a long (but interesting) essay by RMS on the GNU project
> that (among other things) discusses this issue:
> http://www.gnu.org/gnu/thegnuproject.html
> (alternatively you can access this essay by hitting C-h C-p in Emacs)
> The sections most relevant to the issue of licensing is the one titled
> "Is a program free for every user?", "Copyleft and the GNU GPL", and
> "The GNU Library GPL". As pointed out in that last section, it does
> make sense (from the point of view of promoting free software) in some
> cases to use a more permissive license than the GPL, but to me it's
> pretty clear that PyMVPA is not such a case.
> 
> I would also like to point out that a dual licensing scheme in which
> users could pick between the MIT license and the GPL would not ensure
> user's freedoms as effectively as using the GPL as the only license
> (indeed from the point of promoting users' freedoms such a dual
> licensing scheme would provide no improvement over just using the MIT
> license alone).
Thanks for showing that you care for the philosophical/political aspects
of free software development. I pretty much share your point of view.
Let me explain the current situation:

When I started with the devlopment of pynifti and pymvpa both were GPL
licensed. I got contacted by SciPy people, because they were interested
in merging pynifti into scipy's IO filter section. They asked whether I
could change the license to MIT or BSD, because the GPL is incompatible
with the SciPy license -- they even have a webpage with information
about it.

http://www.scipy.org/License_Compatibility

The reason why I choose the change the license is, that SciPy is a very
powerful framework and moving neuroimaging software closer to it will
make it more likely the SciPy is developed with the other tools in mind
-- and therefore hopefully make life easier, as the necessary effort
for maintainance and development is shared among the (few) people
working in this field. Please note that I did not do it for the reasons
given on the page cited above!

A number of people in the SciPy dust-circle also work on NiPy

	http://neuroimaging.scipy.org/

which will eventually become a great set of tools. PyMVPA is a bit
targeted to it, therefore its MIT license.

Yarik and I are going to see those guys during a coding sprint in Paris
in March -- I guess we can talk about licensing at that time, but I'm
not confident at all that they even consider changing the license to
something else.

However, if you manage to convince them to go GPL -- I will be with you.
but I hope you also understand my reasons for not doing it now and
alone. Anyway, you can be sure that I did not have any commercial
interests in mind when choosing the license ;-)


Cheers,

Michael



-- 
GPG key:  1024D/3144BE0F Michael Hanke
http://apsy.gse.uni-magdeburg.de/hanke
ICQ: 48230050
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.alioth.debian.org/pipermail/pkg-exppsy-maintainers/attachments/20080223/6613f5f3/attachment.pgp