Author: Joachim Rang
Date: 00:22:24 08/16/05
Go up one level in this thread
On August 16, 2005 at 02:49:42, Majd Al-Ansari wrote: >I think a lot of people are rooting for Fruit, but the truth is that it still >has a lot of improvement to make. It has to be able to use multiprocessors, >this is especially true since multi core processors will be the norm and not the >exception. Also it has to learn how to use EGTB. I was watching the game >against Zappa yesterday where Fruit had an evaluation of -2.0 while my Shredder >with EGTB was showing mate in 31. That is a HUGE disadvantage. I actually think EGTBs are not so important. They give you sometimes impressive mate scores but seldom change the outcome of the game. Against Zappe even with EGTBs Fruit would have lost. > One more thing >is that Fruit's evaluation is totally different from any other program, >therefore it needs a specially tuned book. In my experiments Fruit performs >really great with some books but is miserable in other books. > Any details about that could be _really_ helpful. Especially if you could quickly say which line play against Shredder in order to win. ;-) Joachim >Still these things can easily be resolved if its gifted author decides to go for >it. A version of Fruit with the mentioned improvements will be a truly >formidable program. > > >On August 15, 2005 at 18:24:36, Thomas Logan wrote: > >>On August 15, 2005 at 17:58:00, Jorge Pichard wrote: >> >>>Fruit still has to play against Crafty, Shredder, and Deep Junior, three >>>programs with Multy Processors and tough to beat even with single processor. >>> >>>Jorge >> >>That seems certain >> >>Too much , too soon >> >>Still great promise in fruit >> >>Tom
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.