Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Open Source Chess Programs

Author: Tord Romstad

Date: 10:52:10 06/07/05

Go up one level in this thread


On June 07, 2005 at 12:27:03, Dann Corbit wrote:

>A much better idea is something like Scorpio.  It is the size of Faile, and yet
>quite strong.

A bit smaller, in fact:

Olithink 4.10:  1556 lines
TSCP 1.81:      2249 lines
Diablo 0.3:     5154 lines
Glaurung 0.2.3: 5255 lines
Scorpio 1.1:    5630 lines
Faile 1.4.4:    6236 lines

Glaurung is also quite small (and the next version will almost certainly
be smaller), and the code is designed to be as clear and straightforward
as possible.  I seem to have failed spectacularly; it appears that most
people find it very difficult to read.  I would appreciate if somebody
could try to explain me why, and help me to make future versions more
instructive.

>I think that TSCP is a very onerous start.  Back in the day, before SMP was the
>wave of the future, it would have been a good place to begin understanding
>things.  But I think that it is a poor choice now, given the alternatives.

I agree.  I would say the same about Gerbil.

>For Bitboard programs, Slowchess is a bit large, and Olithink is a bit cryptic.

I don't agree about Olithink.  It is the most beatiful and readable code I
have ever seen in a bitboard program.  No other program comes close.

Slowchess does not use bitboards.

>I'd recommend Beowulf (cough).  Beowulf looks large, but over half is the C
>version of the EGTB code.

Thanks for the suggestion -- I have never looked at Beowulf.

Tord



This page took 0.01 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.