Moses Glasner wrote:
> James Amundson writes:
> >
> > We aren't dropping GCL, but we will no longer rely on it. That's all. In
> > order for GCL to remain a viable option beyond the near future, someone
> > will have to take over its maintenance.
>
> With the latest CVS compiled under clisp-2.27 when doing load
> on several packages, eg, fourie, ineq, vect, kach, functs, desol,
> I get the following:
>
> (C3) load(fourie);
>
> *** - FUNCALL: argument (LAMBDA (U) (MEMQ U '(T NIL))) is not a
> function.
> To get a function in the current environment, write (FUNCTION ...).
> To get a function in the global environment, write (COERCE
> '... 'FUNCTION).
> 1. Break [17]>
>
> But with maxima-pre59 compiled under gcl-2.4.0 all these packages load.
> Is this a problem with Clisp or am I doing something wrong?
>
> --Moses
>
It is Maxima problem with Clisp. Clisp is more restrictive (and more
standard compliant as far as I understand) and treats, say,
(apply '(lambda (u) (print u)) '((1 2 3)))
as an error. Clisp requires explicit object of function type in funcall like
(apply #'(lambda (u) (print u)) '((1 2 3)))
In GCL both constructions are OK.
So we have to patch Maxima.
Vadim
--
[ Vadim V. Zhytnikov <vvzhy@mail.ru> <vvzhy@td.lpi.ac.ru> ]