On Sun, Jul 25, 2010 at 9:15 AM, mok-kong shen
<mok-kong.shen at t-online.de> wrote:
> ?array (u,integer,5);
> ?hh():=block(u[2]:333, print(u[2])); compile(hh); hh();
>
> leads to an error message.
Well, that's a bug in the Maxima-to-Lisp translator.
At the moment I'm not sure what is the correct behavior;
maybe the translator should emit (MEVAL '$U).
I'll have to look at it more carefully.
Can you please open a bug report.
See: http://sourceforge.net/projects/maxima/bugs
Be advised that the translator code probably has other bugs,
and in any event Maxima's treatment of subscripted doodads
(lists/matrices vs declared arrays vs Lisp arrays vs hash tables)
is pretty messy.
best
Robert Dodier