I found and tried the alpertron applet. It was fun to play with but started getting really slow as I upped the lenght of the numbers. I think it has too many extra functions in the package that I'm not looking for, like knowing precisely how many factors a given number has.
At a point I tried to play with it and the PSP Sieve's factors and it said something like the number was too big.
Definition of "very large"... hmm... something around 10,000,000 digits for the eff awards would be great.
They state on the rules that
Quote:
The primality proof must be a deterministic proof for a distinct integer
The claim must be for the primality for a distinct integer. The proof of primality of that distinct integer must be constructive, definitive, reproductable, verifiable and deterministic.
Probableprimality proofs are not acceptable. Claims involving probable primes will not be accepted.
Your claim must explicitly identify a distinct prime number. For example, claims involving Mills' Theorem real number or Matijasevic polynomial without providing a specific solution will not be accepted.

So I guess going around the problem is a nono for them.
And yea, I know I'm dreaming way up in the clouds, and totally off my league, but isn't dreaming what always takes us further?