Author: Terry McCracken
Date: 05:03:20 07/28/03
Go up one level in this thread
On July 28, 2003 at 07:36:58, Uri Blass wrote: >On July 28, 2003 at 07:17:08, Terry McCracken wrote: > >>On July 28, 2003 at 02:26:49, margolies,marc wrote: >> >>>Ok terry, eduard-- I agree with both of you. Whenever you enter yourselfs in >>>computer chess competitions from now on, you have may permission to bring all >>>five volumes of eco with you into the computer competition from now on, >>>similarly all five volumes of ece-- i'll right you a note for FIDE. >>>Also, in fairness with fide rules, we should drug test the computer operators >>>with serum and urine so that you boys wont be unfairly hampered as humans, which >>>I still presume that you are. >>>best regards, marc >> >> >>No we're ALIENS.... >> >>Now what the hell are you talking about? I said programmes should have _all_ >>basic Endgame Knowledge....IE Fundamental Mates with K+Q vs K, K+R vs K , >>K+BB vs King, K+BN vs K, and how to win Basic Endgames, such as K+Q vs K+R etc. >> >>I don't agree that EGTBS should be needed to cover this type of Elementry Chess >>Knowledge. They also should understand a B of the Wrong Colour can't force a >>pawn in on the Rook File, also they should see other Elementry Draws with pawns >>advanced to the seventh rank, IE Bishop File and Rook File can Force Draw >>against K+Q vs K+P on these files if the opposing King is too far away and the >>other King is there supporting these pawns. >> >>Terry > >I do not plan to implement tablebases and I plan to implement this knowledge in >the future in movei(I already have part of it in my latest version and latest >movei knows to win KQ vs KR) > >I do not like the fact that people criticize programs like Junior for not having >the knowledge because the purpose of the programmers was to teach the program to >play better and not to solve the problem without tablebases. > >I even read that some deleted knowledge after implementing tablebases because >the knowledge is not needed. > >I do not think that people should be surprised when a program that was developed >for task A fails in task B. > >A lot of top programs are developed to play better and the programmers do not >care about what happens when you do not use tablebases. > >By the same logic you can decide that the program should not use hash >tables(humans do not have memory to remember million of lines). > >They should not search more than 10 nodes per second(humans are unable to search >more than it and if you give programs to search more than 10 nodes per second >then it is the same as giving humans more time). > >Uri You can't be serious? Terry
This page took 0.01 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.