[SciPy-dev] MIT license

[Emanuele Olivetti]

Matthieu Brucher wrote:
>
>
> 2008/4/27 Dominique Orban <dominique.orban at gmail.com
> <mailto:dominique.orban at gmail.com>>:
>
>     On Sun, Apr 27, 2008 at 2:57 PM, Nathan Bell <wnbell at gmail.com
>     <mailto:wnbell at gmail.com>> wrote:
>     > We've received some MIT-licensed code to read matrices in the
>     >  Harwell-Boeing file format[1].
>     >
>     >  My understanding is that the MIT license is essentially the same as
>     >  the BSD, so is there any need to relicense the code?
>      Generally, can
>     >  MIT licensed code be included in SciPy without further
>     consideration?
>     >
>     >  [1] http://projects.scipy.org/scipy/scipy/ticket/354
>
>     Nathan,
>
>     I chose the MIT license because I had to pick one. I am not partial to
>     it and am open to picking another. The HB and RB readers depend on
>     Konrad Hinsen's FortranFormat module, which I think is part of his
>     ScientificPython package---that's released under the "CeCILL" license
>     (www.cecill.info/index.en.html
>     <http://www.cecill.info/index.en.html>) so I'm not sure how that
>     works with
>     SciPy licensing.
>
>
>
> CeCILL is the French version of teh GPL, so if those headers are
> needed to build the package, I don't think they can be included into
> Scipy.
>

CeCILL (v2) is a free software licence _similar_ to GPL and compatible
to GPL.
Here is a brief comment by Free Software Foundation:

http://www.gnu.org/philosophy/license-list.html

To Dominique: as most of the minor licenses be careful to adopt it; its
robustness is not well tested and you could have unexpected troubles
later.


Emanuele