[Ecls-list] Maxima + ECL status

[Robert Dodier]

Carl Witty wrote:

> I'd like to give it a try.  (I can't commit for a "long haul" myself,
> but I could spend some time on it in the near future.)

Super -- any time you can invest is greatly appreciated.

> Could you explain what the current status is, and how I could
> replicate your current setup?  Maybe we could set up a wiki page
> somewhere?  (We could put it on wiki.sagemath.org, if you don't have a
> better place.)

Well, there's not much to say. The ECL-specific stuff is on the
branch patches-for-ecl-branch in Maxima CVS.
I think this incantation or something like it should fetch the branch:

cvs -d :pserver:anonymous at maxima.cvs.sourceforge.net:/cvsroot/maxima
get -rpatches-for-ecl-branch maxima

You'll need to build ECL from current CVS (the 0.9j release won't
work).
The --enable-smallcons flag for configure is recommended (by the
ECL maintainer) and --enable-unicode is discouraged.

Following the instructions in INSTALL.lisp (in the top-level maxima
directory) up to and including step 4.
Then cd src, launch ECL, and try to build Maxima via defsystem
(further steps in INSTALL.lisp) or asdf. Sorry if this is too vague;
I haven't yet figured out what works.

The resulting Maxima program (launched by (cl-user::run)) can do
a lot of stuff, but it has nontrivial strangeness. It gets part way
through the Maxima test suite (run_testsuite()) before halting with
an error. Dunno what's going on there.

I think Maxima + ECL works better if the code is interpreted
instead of compiled. Perhaps you can verify this.

I've cc'd this message to the Maxima and ECL mailing lists.
Follow-ups should probably be directed to the ECL list.
I'll copy this message, such as it is, to the Maxima wiki for
reference. http://maxima.sourceforge.net/wiki

Thanks for your interest, & let us know how it turns out.

Robert Dodier