Author: Eric Oldre
Date: 10:37:17 02/15/05
Go up one level in this thread
On February 15, 2005 at 08:46:30, Andrew Wagner wrote: >Well, after CCT7 I've got the chess programming bug in my blood again. But I >have no desire to go back to working on Trueno, mostly because I think I proved >my point with it. A competitive chess engine CAN be written in a VB-type >language. > >Anyway, my idea now is to create an engine with parts that are as generic as >possible. The idea being that I want to be able to easily swap out, say, an >alphabeta search, and replace it with a mini-max search. Or switch from >bitboards to 0x88 easily. The minimum goal of the project would be to be able to >write all my own parts and have them function interchangeably. Ideally, I'd like >to really make things generic and re-usable to the extent that I could swap in >parts from other engines, eventually. So if I want to compare the speeds of >Fruit's eval() with that of Crafty, I can do it with minimal code changing, >regardless of whether or not they use the same board representation or type of >search, or whatever. I think I'd like to write it in Perl, too, just for grins >and giggles. Not something you hear much of. > >Anyway, thoughts? Suggestions? Accusations of sheer dementia? Bring 'em on! Although I think that creating this program would be both interesting and challenging from a academic standpoint. I don't think we'd be able to draw very useful results from it. I think what we would quickly find is that an eval like crafty's works wonderfully when coupled with a crafty like board representation. But without the underlying bitboards, it's efficiency drops dramatically, things that used to be worth calculating are no longer worth the cost, and vice-versa. Conversely, I also believe that we will find a Fruit like eval works best when used with a Fruit like board representation. Because (i assume) that eval has been specifically designed to make use of the advantages inherent in it's board representation. I also agree that trying to use perl to code such a project would be challenging to say the least. Designing it would be a challenge and a good excersize in creating a modular program. But I wouldn't expect to be able to draw any conclusions about the effectiveness of different chess programming strategies from it. Good luck, Eric
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.