Author: Robert Hyatt
Date: 18:51:28 07/16/99
Go up one level in this thread
On July 16, 1999 at 14:58:58, José de Jesús García Ruvalcaba wrote: >On July 16, 1999 at 14:06:09, Robert Hyatt wrote: > >>Ok... >> >> For all you whiners, naggers, complainers, grumblers, bellyachers, and other >>lazy bums that won't pony up and buy a 6gig disk, the endgame database probe >>idea of your dreams is now ready for testing. :) >> >>A crafty user (I won't mention names, but it is in the comments in main.c if >>you are really interested) submitted some code that looked ok. It is now in >>the just released version 16.13... >> >>Here is what it does. >> >>If the initial position is a mate in N (tablebase hit), then for each move, >>a probe is done. If the move is not mate in N-1, then that move is excluded >>from the set of moves to search. This leaves non-tablebase moves, one of which >>_must_ lead to a mate in N-1. We now hope Crafty can find it via a search. >> >>I had to further modify iterate.c, as with the submitted code, crafty would >>still think it was in a tablebase position, and stop the search after 4 plies. >> >>Now you can download the files of your choice with pawns, without having to >>grab the promotion cases. It is _still_ probably better to have them, but >>at least it will now avoid most of the ugly draws that could happen without >>them using the old approach. >> >>Reports of success or failure always welcome... >> >>Bob > > Does it mean that egtb.cpp has been modified? >José. No. egtb.cpp has nothing to do with this. This is an internal Crafty issue, and doesn't bother Eugene's code at all...
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.