Author: Dieter Buerssner
Date: 14:28:42 09/22/02
Go up one level in this thread
On September 21, 2002 at 06:20:39, Mogens Larsen wrote: >On September 21, 2002 at 06:13:13, pavel wrote: > >>According to Tony, SSDF is most likely going to test YACE is the very near >>future. > >Yes, I know. That was almost a month ago. With Deep Fritz 7 arriving and soon >the new Tiger (and others), I wouldn't hold my breath. That the testing of Yace didn't happen yet, is only my fault. Peter Berger volunteered to do some testing with an autoplayer setup. Initially I suggested Remi's autoplayer. But this seems to be not as reliable, as one wants (and perhaps there is even no way at all, to support the time control of SSDF). But we approached a very well working version now (I mean functionality, not playing strength). I feeled, that an engine without book learning might have a rather significant disadvantage. Also, I want, that Yace uses a freely available (- your -) book. So, I had to write some code for UCI, an interface, that does not make it really easy to write book learning code (IMHO). At the moment, it seems, that the engine works correctly under the Shredder 6 autoplayer (I neither have S6, nor do I have 2 computers to try autoplayer). I hope, after fixing one last little bug, Peter has found, I can send a version to the SSDF. I will send the almost same version to you for the Yace web page, when the SSDF starts testing. The dfferences will be just a few small things in the default setup. At least some versions of Fritz 7 autoplayer can win on time (because of a few seconds), which the special SSDF-version should avoid. Anything else, will be the same. I guess, you can imagine, that all this could not be tested within hours. Cheers, Dieter
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.