I think I know the range of work needed to get some non-lisp library linked
into a specific lisp on a specific platform. It can range from almost no
work, to "apparently unlimited/impossible". To get the library working for
ALL lisps is harder.
To a certain extent the mess of maxima comes from the diversity of
platforms, and this inhibits advanced development if we insist that anything
included must be reduced to the level of the least capable lisp.
As an example of one-lisp implementation... to have maxima speak math
expressions, see the code in tts.cl and math7speak.lisp in
http://www.cs.berkeley.edu/~fateman/speech-lisp/
The core of the program is 12 (non-comment) lines in tts.cl
I don't know how much work it would be to get running in other lisps. For
some, I presume it would not be much different. For others it might be
nearly impossible. Why do I mention this?---
Is speech as interesting as LAPACK? Well, if you are interested in math
education, or use of math by sight-disabled people, it might be important.
I am not against making accommodations for LAPACK, if someone wants to do
that. I'm just pointing out that if we spend our time doing
non-symbolic-math things, we are not building on Maxima's strength. People
who are primarily interested in high-speed linear algebra are relatively
numerous, have many resources available e.g. supercomputer centers, and they
have access to substantial funding.
Regards
RJF
> -----Original Message-----
> From: Robert Dodier [mailto:robert.dodier at gmail.com]
> Sent: Saturday, December 23, 2006 5:04 PM
> To: Richard Fateman
> Cc: maxima at math.utexas.edu
> Subject: Re: [Maxima] LAPACK with maxima
>
> Richard,
>
> > Instead of inserting LAPACK source into the Maxima source, how about
> > pointing to a few binary forms, one for each plausible host environment,
> > with a suitable interface. Perhaps we could even point to matlisp, which
> > apparently solves the problem of the foreign function interface for
> cmucl
> > and allegro.
>
> That would make Maxima an even bigger mess.
> We have a pretty good way to get Fortran code running, namely f2cl.
> I don't see why we would want to retreat to platform and Lisp specific
> solutions, given that we have a pretty good chance of getting stuff to
> work on multiple platform/Lisp combinations.
>
> I think you've seriously understated the work involved in getting
> some library linked into Maxima (be it Lisp, .so, .dll, whatever).
> It is certainly not something we can expect typical users to carry out.
> It is a significant effort even for the Maxima developers.
>
> > There are many other programs in netlib that could potentially be of
> > interest to maxima users, and for someone on this list to speculate on
> what
> > is most useful and spend time converting or interfacing such a program -
> -
> > without an application in mind -- seems to have the potential for a
> major
> > distraction.
>
> We can safely say that LAPACK is useful to many people.
> It is hardly a distracting speculation.
>
> For what it's worth,
>
> Robert