Prime numbers and mathematical discovery

Werner Koch
Wed Aug 28 18:07:01 2002

On Wed, 28 Aug 2002 12:35:01 -0300, Renato Martini said:

> I read the news just now, and I'm thinking how this will can affect the assimetric
> crypto programs...

No at all.  Probabilistic prime tests are sufficient and a proven
prime doesn't gain us any advantage.  It is a different case for
pre-defined primes used as parameters for many of keys; here one like
to have a proven prime (due to general crypto paranoia ;-). 

After all it won't help in factoring composite numbers which is what
you have to do to break RSA.