On Tue, 27 Jul 2010 22:06:18 +0000, john gilmore wrote: > >Yes, 61, which is prime, is better than 64 = 2^6, which is composite. > > ... > > >If division-method hashing is used a prime divisor/modulus is highly >desirable. Clustering at the prime divisors of a composite modulus does >occur. I dislike arguments from authority, but 1) this is not the place for a >bibliography and 2) RKFATWTF. > For pure division-method, surely. Modulo 64 merely extracts the rightmost 6 bits of the original string. But you had suggested CKSUM, then modulo. And if CKSUM is of high quality, I'd expect any modulus to give good results.
RKFATWTF!? Not in my lexicon. Nor in Google's, apparently. But the last three characters are familiar and hauntingly apt. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
