Hello,
On Tue, 20 Dec 2005, Yigal Weinstein wrote:
> But Mathematica which has the nastiest
> license possibly known to man is by many Debian users the only CAS that is
> considered worthy of pratical use when it comes to many problems.
I beg to disagree. As far as I know: number theorists use g-pari/kant,
group theorists use gap and commutative algebra/algebraic geometry
people use macaulay/singularity. While none of these are packaged for
Debian most of them do not have a license as restrictive as Mathematica.
I would also add that programs written in them tend to run significantly
faster than corresponding mathematica ones.
However, these are packages written by mathematicians for
mathematicians. What you want is perhaps a package written for those who
do not want to worry about the mathematical details :)
> I am rather nieve but trying to explore Maxima, Yacas, Axiom etc. the one
> thing I get from all of these is the lack of a user friendly environment.
My reaction to requests for UFE is:
"How about an environment friendly user :)"
In the *nix environment there are always multiple ways to do things
and different programs to do each thing and do it well. Programs like
Mathematica which try to be one-package-that-fits-them all will never
satisfy a specialist in Commutative Algebra or Number Theory or Group
Theory (not taking into account that the company behind Mathematica
has alienated a number of mathematicians that went to work with it).
> There is a breaking point and I think I am moving closer and closer to it.
I do not see what you want to break out of. If you really need to use
Mathematica you can certainly buy a "Linux" license for it.
> 1. Is there no GNU CAS that even comes close to Mathematica?
In the sense that you mean perhaps not. There is similarly no user
environment in GNU that comes even close to Mac OS X --- perhaps the
aims are different. (See Linus' recent rant to the effect that if you
make an environment which assumes that people using it are stupid
then only stupid people will use it).
> 2. Why are programmers not working together on projects? I mean there is
> gnuplot, pplot, xmgrace etc. and they are worse at graphing than MM or
> Maple. We have so many brilliant people working on these programs and yet
> the outcome is almost never the same quality as the propriatary stuff.
Again I must disagree.
I don't know enough about the other projects but g-pari (which is
packaged for Debian) is a collaborative project which even had a
"hand-over" of leadership. The original leader (I think) H. Cohen
even won a big award in 2004 for his work on g-pari. I would imagine
that similar things can be said about all the other packages
mentioned above.
I think you are trying to fit a round peg in a square hole.
(a) Figure out what you want.
(b) Cut out what you don't need.
(c) Go back to (a) if (b) made a change.
(d) Install/build/buy-if-necessary the things you need
to get your job done.
Hope this helps.
Regards,
Kapil.
--
Attachment:
signature.asc
Description: Digital signature