Hi all,
I have a different question concerning P-1 and ECM.
Some time ago I asked which power to put small primes
into when multiplying them into E ( factor = gcd(a^E-1,N) ).
Paul Leyland, I believe, replied that the power for prime p should
be trunc( ln(B1) / ln(p) ) ( log(B1) with base p ),
where B1 is the bound up to which we put primes into E.
But what if there is a stage 2 with a higher bound B2?
Should it be trunc( ln(B2) / ln(p) ) then? Or still the stage 1 bound?
In his Diplomarbeit about ECM
( see ftp://ftp.informatik.tu-darmstadt.de/pub/TI/reports/berger.diplom.ps.gz ),
Franz-Dieter Berger mentiones on page 40f that his experience
shows that it is better to use the stage 2 bound.
Any opinion from the factoring gurus here on the list?
Ciao,
Alex.
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
