Hi Marc,
The last version of clisp that reliably built maxima and passed all
tests was 2.29.
I maintain maxima and clisp for darwinports and we've not been able to
track down
the problem yet. (Recent version of clisp build for some people on OS X
and
not others; the problems seem to occur after starting the bootstrap
compiler.)
My suspicion is now that realloc is being used incorrectly, which can
lead to
hard to reproduce memory corruption.
You might want to try to build clisp from source and run "make check".
That
will let you know if the problem is likely to be on the clisp side. You
may also
want to build maxima from source and have it run all of its tests as
well.
Sorry not to be of immediate help. But if you have the chance to try to
build
clisp and maxima from source, I'd be interested in what you find. This
problem
may be on the maxima side, if so, that would be helpful to know as
well. Eventually,
this bug will be tracked down.
Best Wishes,
Greg
Gregory Wright
Antiope Associates LLC
gwright@antiope.com
On Jun 6, 2004, at 3:12 PM, Marc Charpentier wrote:
> Hi all,
>
> after noticing that Bruno Haible had posted an updated version of
> clisp, I compiled and installed it.
> Then I recompiled Maxima.
>
> All went well.
> Or so I thought.
>
> After several hour's work with Maxima, I suddenly noticed that the
> differential operator diff(y, x) returns unevaluated.
>
> My Fink-installation of Maxima (I use OS X) works fine.
> (This installation has a clisp-maxima file in /sw/bin)
>
>
> Any ideas?
>
> Thanks in advance.
>
> Marc Charpentier
>
> _______________________________________________
> Maxima mailing list
> Maxima@www.math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima
>