A number of the "deficiencies" of Maxima in that article
seem to revolve around misconceptions of what it means to
declare something. This may have changed in Macsyma, but at
least in the early design, declarations had no effect whatsoever
on rat, ratsimp, radcan.
I think it actually shows maxima to pretty good advantage, even though
Mupad seems to score higher according to the author's criterion.
1. Maxima gets lots of stuff right. The stuff it gets wrong is
understandable, for the most part.
2. Maxima is much faster (15X) than Mupad, to run the whole test.
3. Maxima is free and open source.
As I recall, in Wester's original test, Macsyma was high scorer,
outscoring the competition of Maple, Mathematica, ....
RJF
Barton Willis wrote:
> If you have not read "Maxima Vs MuPad," (
> http://www.tex-sales.se/Artiklar/MaximaMuPAD.pdf )
> I think it's worth a look. While I think it would be wrongheaded to target
> specific bugs revealed
> from the tests in this comparison for (ad-hoc) fixing, the document
> might give us guidance
> on what we need to fix.
>
> Barton
>
> _______________________________________________
> Maxima mailing list
> Maxima@www.math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima