Author: Mike S.
Date: 16:25:48 02/09/06
Go up one level in this thread
On February 09, 2006 at 18:27:05, Chessfun wrote: >(...) >My own thoughts on without tablebases is since there are a few engines very week >on endings without tablebase support, much better to play with since most >engines not only support them but even those without seek to use them in newer >versions. I was making a suggestion towards a selected subset of tbs., as a kind of performance compromise (to reduce slowdown effects but at the same time, get most of the practical 5-piece benefit = R vs. R): http://www.talkchess.com/forums/1/message.html?484714 I wouldn't mind if you would test all types of engines in the endgame tests (those which access tbs. and those which do not). If you test in Fritz, the GUI will provide "some" tbs. support if the engine in itself cannot, but only in tablebase positions if they are reached in the game. This is in contrast to other GUIs like WinBoard which don't do this. It is a philosophy question, how to handle this. I would think, it is more typical for a normal practise to let the software make use of such a function, than to unneccessarily restrict engines when the GUI could help them to perfectly finish a game. OTOH, that may lead to (some) results which an non-tbs. engine could not have achieved in another GUI. Regards, Mike Scheidl P.S. My opening book PB12moves is involved in the CCRL tests! :-) I'm glad that I could offer something useful that way, as it seems. http://www.computerschach.de/index.php?option=com_remository&Itemid=52&func=fileinfo&id=51 http://members.aon.at/computerschach/links.htm#downloads (links to infos)
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.