>>>>> "Stavros" == Stavros Macrakis <macrakis at alum.mit.edu> writes:
Stavros> cache) bandwidth is the limiting factor, not processor speed. I think there
Stavros> are old published papers with timings of calculations.
Wang's (or Moses?) thesis on integration has some timing information.
I think some integrals took on the order of 1-5 sec. On the machines
I've tried I think they take like 0.1 sec or less. Surely a large
part is from the faster processor. But some could also come from the
Lisp implementation itself.
Stavros> In the particular case of CRE exponents, I am not sure. There are unlikely
Stavros> to be many practical applications of exponents > 2^31, and certainly
Stavros> Maxima's algorithms haven't been tuned for those cases. On the other hand,
Stavros> *silent* failure is almost always unacceptable. *IF* we are going to use
Stavros> integer arithmetic, we have to err out in the case of overflow. I don't
Stavros> think there are mechanisms for doing that in all (any?) of the Lisps we run
Stavros> in without using generic arithmetic.
I agree that we need to do something. An explicit check might be
necessary, but perhaps only for GCL.
Ray