Author: Albert Bertilsson
Date: 03:40:59 12/19/03
Go up one level in this thread
On December 19, 2003 at 05:24:35, Sune Fischer wrote: >On December 19, 2003 at 04:22:24, Joost Buijs wrote: > >>I agree, a collision is very unlikely, but not impossible. >>You will never be sure the calculated number of nodes are actually correct. >> >>Regards, >>Joost > >With 128 bit keys collisions won't be your main concern, even with 64 bit keys >it takes a week, give or take, to average 1 collision, so we're probably talking >a few million years with 128 bit keys. > >I consider it more likely that there is a bug in the move generator or that some >computer will overheat and produce wrong numbers, or there could be an error in >the transmission... > >You can never be completely sure of that either until its been verified >independently. > >-S. I completely agree, the concerns about hashing with zobrist is valid but marginal compared to everything else that can be/go wrong. /Regards Albert
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.