I hope you don't mind if I ask you another question
conscerning writing a kernel in OCAML.  I was writing
a few programs to find primes in C and in OCAML so
that I could compare the performance.  It appears that
OCAML offers a very nice compromise of code length and
code execution speed.  However, I also noticed
occasional pauses in the primes generation during
which (I assume) OCAML was doing its (two levels) of
garbage collection.  I fear that reference counting
might be a more appropriate algorithm for *GC in a
kernel* than what I guess is two levels of mark and
sweep. 

Can you comment on this or perhaps offer any
workarounds?  I did play with the Gc module's control
structure, but didn't really gain any huge
improvements.

Thanks for your time,
-jj

__________________________________________________
Do You Yahoo!?
Send your FREE holiday greetings online!
http://greetings.yahoo.com
-------------------
Bug reports: http://caml.inria.fr/bin/caml-bugs  FAQ: http://caml.inria.fr/FAQ/
To unsubscribe, mail caml-list-request@inria.fr  Archives: http://caml.inria.fr