> > The probability at the top of the range is presumably such that the cost > > in time of an erroneous LL test balances the extra time needed for the > > next FFT size, for which the chance of error is small. > > I think (and hope) the default setup is a bit more conservative than that.
I guess it is too. The "break even" approach suggests tolerating up to ~30% error rates. This would be an incentive to double check more promptly. > There's the unfortunate possibility of someone running a first LL test on an > exponent and incorrectly finding it composite due to an avoidable error - we > want this to be "reasonably improbable" even though double-checking would > eventually find the mistake! I'm not sure how "unfortunate" such an occurence would be. (1.8% I believe) The down side is the long straggly tail of exponents which have not been double checked. Colquittand Welsh's out of sequence discovery of M110503 in 1988 gave me a nice target: I proved it was the 29th Mersenne using high school multiplication on a K6 in about 6 months. Quite a coup for Moore's Law I think! I think the discovery of another prime with less tha 10,000,000 digits would be more fun than the award of the $100,000. David Eddy > > Regards > Brian Beesley _________________________________________________________________ Be one of the first to try Windows Live Mail. http://ideas.live.com/programpage.aspx?versionId=5d21c51a-b161-4314-9b0e-4911fb2b2e6d _______________________________________________ Prime mailing list [email protected] http://hogranch.com/mailman/listinfo/prime
