On Sat, Nov 23, 2002 at 06:28:41PM -0500, Stavros Macrakis wrote:
> ...
> As for Zeilberger, I have written to the RISC Combinatorics group in
> Austria to request a clarification of the license terms of their
> implementation of the Zeilberger algorithm. One interpretation of their
> conditions prohibits the commercial *use* of the implementation (e.g. an
> engineer at General Electric using Maxima with Zeilberger.mc) -- this
> restriction is incompatible with the GPL; another interpretation of
> their conditions prohibits the *incorporation* of the implementation in
> a commercial product (e.g. Maple or Mathematica), which of course Maxima
> is not and cannot be (because of the GPL).
As you describe, neither of these interpretations is compatible with
the GPL. GPL products can certainly be commercial. For instance, Red
Hat is a commercial product, with a large percentage made of GPL code.
> Even if they don't wish to license its use within Maxima, presumably
> someone could reimplement the algorithm, unless it's patented (which I
> doubt).... Of course, that may be non-trivial.
I've read the description of the algorithm, and it did not sound that
difficult. But I might be missing some corner cases. It's described
in the book 'A=B'. (But I'm very surprised Maxima doesn't already
implement it.)
Best,
Dylan Thurston