[SciPy-dev] Re: [OFF-TOPIC]: organizations and licenses Re: [SciPy-user] plotting enhancements and copyright

eric eric at scipy.org
Thu Nov 1 16:09:42 CST 2001


Hey Tony,

I haven't heard any major complaints about this, but it does seem somewhat
relevant to the current conversation, so I'll toss this out.  There are
reasons why SciPy is not GPL'd beyond the fact that Enthought is a
commercial entity.  I would have made the same choice for SciPy had I
remained in academics.

By its very nature, SciPy (and Python itself to some extent) is most
valuable as a building block integrated into larger systems -- not as a
stand-alone package.  Pretty much anything that computes and needs
signal/image processing, optimization algorithms, etc. can benefit from
SciPy.  This includes both Open Source packages as well as commercial
software and hardware.  Commercial entities cannot use GPL'd code (at least
how I read it) in their products unless the product is free, so a GPL'd
SciPy would be useless to them.  However, a BSD style license opens SciPy to
use and contributions from these companies.  If SciPy can serve and be
developed by both communities, it will grow more quickly.  I believe SciPy
is good for science (both academic and commercial), and the faster it grows
the better.

Packages like Emacs don't have the same issues because, while they are used
by companies to develop products, they do not become a part of products.  As
a result, companies that are happy using Emacs would pass over SciPy and
choose to either roll-their-own or purchase a licensed alternative when it
came to scientific tools.

As you said, the licensing decision has already been made.  Still, I thought
it worthwhile to explain, at least in part, the reasoning behind the choice.

eric





More information about the Scipy-dev mailing list