From help-octave-request at bevo dot che dot wisc dot edu  Thu Jun  4 10:07:52 1998
Subject: Re: symbolic math in octave
From: "John W. Eaton" <jwe at bevo dot che dot wisc dot edu>
To: George White <gwhite at bodnext dot bio dot dfo dot ca>
Cc: daniel dot heiserer at bmw dot de, help-octave@bevo.che.wisc.edu
Date: Thu, 4 Jun 1998 10:07:43 -0500 (CDT)

On 29-May-1998, George White <gwhite at bodnext dot bio dot dfo dot ca> wrote:

| In order to have symbolic math in octave, a GNU licensed symbolic math
| engine is needed.

If the two programs are linked together to form a single executable,
this is correct, because the combination forms a derivative work, and
the GPL says that it must be possible to distribute the resulting work
under the terms of the GPL.

But, as I understand it, if Octave communicates with another program
through a pipe, the result is not considered a derivative work, so the
auxiliary program need not be free software.

In any case, I believe it would be better to use freely
redistributable software for this purpose.

| Such an engine does exist (maxima, a fork of Macsyma),
| but it seems to need a BSD-hosted clisp to run.

Are you sure?  I thought that it could now be used with GNU Common
Lisp, and that it should work on just about any Unixy system.

| The problem remains that running two interpreters (octave and the lisp
| host for maxima) requires lots of memory (maxima is a big package). 
| Since many octave users are financially constrained, symbolic math may be
| a stretch (it certainly isn't feasible to run maxima and octave on our
| m68k NeXT systems with 16M Ram!). 

As it is now possible to buy 64MB for around $100, it seems to me that
this is becoming less of a problem.

If anyone is interested in working on an interface that would allow
maxima to be used with Octave, please contact me.

Thanks,

jwe