Author: Helmut Conrady
Date: 11:11:30 04/22/04
Go up one level in this thread
On April 22, 2004 at 13:48:19, Ziad Haddad wrote: >Hi Stefan >It's great to have you in this forum. I think that all the top chess >programmers must give a particular interest to the computer chess forums because >they'll find there what we expect from their products, but also it's agreat >source to discover the eventual bugs of the product because i suppose that with >all the computer chess lovers community we got there the biggest beta testers >ever. > >I've got a special question for you. I posted about an eventual bug of the >Shredder's engines in the managment of the HashTables. In fact, when i test them >in simultaneous games against various engines i have found that as the number of >consecutive games is greater as the results of the shredder engine is becoming >bad (to avoid the influence of the openings choosen by the engine i always test >it from some positions that i picked from the Nunn Testsets or others ). But >after i switch my computer Off, open it again and reinitiate the matches the >results of Shredder are good. > >haven't you discovered such a bug? >Have you an idea concerning this problem? May be it's due to my computer. I've >tested mostly on a 2,4 MHz P4 with a Hash for each engine of 64 Mo, Learn Off >for Shredder. I've got a total of 128Mo of possible Hash for each engine. The >main time control used is 3mn per engine. This problem tends to be less >important with greater time controls!! May be Shredder have had this problem in >his last round game of the WCCC 2003 !? No that was indeed a threefold repetition bug. Helmut
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.