Author: Dann Corbit
Date: 17:23:29 10/11/99
Go up one level in this thread
On October 11, 1999 at 19:37:12, Paul Svelmoe wrote: >Hi, > I've heard some of the talk about how hard it would be to generate 6 man >tablebases such as the one that's in demand now (kqpkqp). Is there any way it >could be done within a week or 2, modifying the generator, etc., if the white >pawn was at g7 and the black pawn at d4, thereby eliminating all other pawn >position possibilities? Also, I might add, it would be acceptable to disallow >underpromotion. Also, I might add, we could probably get 20 or 30 programmers >working on the task 2 or 3 hours a day for modifying the generator. (Would it >really be that difficult??). Also, I might add, one of us probably has access >to 64 bit hardware. To use kqpkqp, you would have to have all the children tablebase files done. Consider how many permutations you can get for each of those 'p' guys and you will see that the match will be over long before you are. > As for crafty, how does one go about modifying wcrafty? When you download it, >all you get is the executable with no source code as in regular crafty. Get the source files instead. E.g.: ftp://ftp.cis.uab.edu/pub/hyatt/v16/crafty-16.19.zip >If it's >not possible, what command-line options would be best to optimize it for the >position white: Kh6, Pg6, Qd4 and black: Kb2, Pd5, Qf5? No idea. Probably rewrite it from scratch. As far as crafty is concerned, all positions are considered with no speciality. It just throws its host of algorithms at it and spits back a pv and a ce.
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.