Subject: gcd problem / really bignums for exponents
From: Stavros Macrakis
Date: Wed, 14 Feb 2007 11:17:02 -0500
On 2/14/07, Robert Dodier <robert.dodier at gmail.com> wrote:
>
> On 2/14/07, Richard Fateman <fateman at cs.berkeley.edu> wrote:
>
> > OK, here's the choice. Making factorization (and other computations
> using
> > canonical polynomials) n% slower, or when some exponent larger than
> about
> > 2^29 occurs, giving a warning: please reformulate the problem so that
> such
> > large exponents do not occur.
>
> Arguments to f+ are not necessarily exponents. Intermediate results
> can overflow a fixnum even if the exponents in question do not.
> There may be no exponents in sight anyway (i.e. f+ applied to some
> problem not involving polynomials).
>
In the CRE package, f+ is used only for exponents, if I'm not mistaken.
Elsewhere, it may be used for other things.
The two cases need to be kept separate.
-s