On 10/16/10 2:25 PM, Edwin Woollett wrote:
> On Oct. 16, 2010, Raymond Toy wrote :
> ------------------------------------------------
>> I don't think GCL can be configured to use a larger value for
>> call-arguments-limit.
> -----------------------------------
>
> Assuming the problem is GCL, how does that explain that
> the Windows Maxima ver. 5.21.1 has no problem with my
> qdraw example, and Windows Maxima ver. 5.22.2 chokes?
Maybe because draw changed from 5.21.1 to 5.22.2? At the very least,
it's a hint that 5.21.1 isn't the same as 5.22.2.
> -------------------------------------------------------
>> But I also ran this with cmucl, and as best as I can tell the call to
>> apply in make-scene-2d is never given anything more than 6 elements long
>> (for the example with qdraw that was posted).
> --------------------------------------------------------------
>
> The list I am feeding draw2d in my example has 76 elements,
> as shown here:
Interesting. When I compile draw.lisp, and run your test, I get an
error from append saying it can't handle a 76-element list.
> -------------------------------------
>
> If you know the error is coming from an apply
> in make-scene-2d, and know that at most six
> list elements are involved, then the fact that
> my list drlist has 76 elements may be irrelevant.
Difficult to say. I don't know gcl very well, and I don't know how to
use gcl's debugger at all, and I've never looked at draw.lisp before, so
debugging this will take some time.
But having your examples helps tremendously.
>
>
> A smaller qdraw job, again designed to produce an
> eps file using 5.22.1, with a 44 element list,
> (restarting maxima) goes through with no
> apply arg complaint, but produces an **empty**
> eps file in c:/work5/test4.eps with 0 bytes.
Bummer. Yet another bug to track down.
Ray