[Fwd: Java versus Lisp. Comparison needed.]

Robert Dodier wrote:
.....
>
> I'm going to guess that the program creates and destroys a lot of
> objects. Maybe Java and GCL are spending a lot of time in garbage
> collection, and SBCL handles that more efficiently? Just a guess.
>   

I think that's a good hypothesis. In fact I would be totally delighted 
if that were the case, since crappy garbage collection routines are just 
as good as fantastic garbage collection routines if you don't need to 
use them (much),  and I suspect that one has stressed Java GCs.  Maybe 
comp.lang.lisp people will know.

It has been my suspicion that (compared to Java) GC routines in (some) 
Lisps are really great {SBCL? Allegro, Lispworks} , but don't get much 
respect from the usual benchmarking goobers.
Of course some GC routines in some Lisps are also crappy.  Perhaps that 
is the case with GCL.  Maybe CLISP?

It may also be the case that totally other factors are in play, like 
interpreted GCL vs compiled SBCL.

RJF