On Sat, Jul 26, 2008 at 8:04 PM, James Youngman <address@hidden> wrote:
> I have a working GMP-based version (essentially, it's the example
> taken from the GMP docs that you mentioned earlier). For the moment,
> the code always uses GMP if it's available. The GMP version is
> dramatically faster for the pathalogical case mentioned in the docs,
> and about 2x slower for "easy" cases with small integers.
It turns out that it needs a little more work, since it does not
currently produce the factors in ascending order. I guess we
probably can't spuriously change that behaviour.
Options:
1. Sort the factors before printing them
2. Bluster that we can break the rules for N>2^64, since that never
worked before anyway
3. Stare hard at the code to find a way to efficiently do the
factorisation and emit the factors in numerical order
James.