Author: Uri Blass
Date: 07:44:56 01/27/06
Go up one level in this thread
On January 27, 2006 at 10:35:34, Uri Blass wrote: >On January 27, 2006 at 06:42:16, Vasik Rajlich wrote: > >>On January 26, 2006 at 21:13:25, Uri Blass wrote: >> >>>On January 26, 2006 at 19:50:56, Pablo Ignacio Restrepo wrote: >>> >>>> >>>>Hello to everyone into WWW with all respect. >>>> >>>>I have played many games in private against the great Rybka in orthodox lines, >>>>just a few antichess. >>>> >>>>I have made many matches between engines, and Rybka have been the absolute >>>>engine champion. >>>> >>>>Rybka is doing an incredible space job playing against other engines >>>>.......Rybka is playing a superior chess in 2006, its look as a Kasparov or >>>>Karpov in more strong and perfect level that them .......... >>>> >>>>Today I have started my job into Playchess in 2006. I hope will be doing new >>>>things playing engines and I would like to will punting good games for everyone >>>>into chess communities........ >>>> >>>>.................... First game, first new discover ...... Rybka have gone far >>>>away of the 256 chessbase interface move, and have got a draw some strange >>>>............ here the game ........... >>> >>>The 256 move limit is no real limit. >>> >>>The problem is simply that engines get wrong information and it can cause them >>>to lose on time. >>> >>>I believe that rybka allow you repetition because of the wrong information. >>> >>>The wrong information that engines get from the interface is wrong time(3 >>>minutes for the rest of the game in case of blitz 3 minutes per game) and not >>>getting move but getting fen. >>> >> >>Uri -- >> >>this is news to me. Are you sure that some GUIs will send the fen rather than >>the move list? Working around this on the engine side would require remembering >>previously analyzed positions, which defeats the UCI approach. >> >>Vas > >I am not sure about latest GUI but I am sure that after more than 256 moves >fritz sent movei only fen and not a move based on the log file. > >It usually does not happen when games have more than 256 moves so the best way >to test it is simply to do nunn match from a game that has already 250 moves and >is not going to be finished in a few moves. > >I found that in the first moves everything was ok but after getting the limit of >256 moves the engine simply did not get moves but only fen after every move. > >Uri The problem is also wrong times that movei got from the interface of fritz8(I am not sure if it is the last interface of fritz8 and I think that I downloaded some update later and did not check it again after updating the interface). I can add that I will probably not try to fix the problem because if my memory is correct I need to write a function that get the last move of the opponent based on the fen and I do not like to do it for rare cases under chessbase. I also do not remember at this point exactly what commands movei got from the chessbase interface and I do not think that it is my job to write a code for these rare cases. Note that code for these rare cases can only help for games of less than 300 moves under chessbase because after 300 moves the interface simply refuse to make moves and if you give it a game with more than 300 moves it refuse to read it(again I am not sure about the last interface of fritz8 and I do not have fritz9). Uri
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.