Author: Normand M. Blais
Date: 13:57:07 12/07/99
Go up one level in this thread
On December 07, 1999 at 16:07:15, Tom Kerrigan wrote: I've benefited from Tom's efforts and I would like to say a few words. As far as I'm concerned TSCP is perfect. I needed some simple code so I can understand how a "real" chess program is built. A move generator and a board representation are relatively easy to create; but what is more difficult IMO, are the search engine and the evaluation function. A specific example is the use of a PV array with an alphabeta search function. Most algorithms that I saw don't show the relation between the PV and the AB function. I've learned a lot from TSCP. Thanks Tom >>Basic search algorithm optimized for speed (TSCP or modified TSCP) >>Basic evaluation optimized for speed (modified TSCP) > >I can certainly optimize TSCP for speed by copying most of Stobor. It would be >about 10 times faster and the code would be an order of magnitude harder to >understand. So what's the point? I'd have two copies of Stobor and my Simple >Chess Program will be a far, far cry from simple. > >I suggest looking elsewhere for a program that's fast. > >>Hash table support >>Thinking on opponents time > >These features would also make the code significantly more complicated. > >>Interface with winboard > >Already done. > >>Makefile for MS C++ > >Don't feel it's necessary. It takes all of 30 seconds to make your own. And >besides, I'm including the executable already. > >-Tom
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.