Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Crafty will never win a title

Author: Dann Corbit

Date: 12:55:29 08/23/01

Go up one level in this thread


On August 23, 2001 at 15:48:54, Olaf Jenkner wrote:

>Crafty is open source. Each programmer can look at it and take the ideas.
>Reading the table after a competition we should think of it. Crafty would be
>some positions higher if nobody can see the source.

I think it highly probable that crafty will win a title at some point.  It's
pretty much a matter of probability.  You don't have to be the strongest program
to win, for that matter.  If another program is twice as strong as you are, and
you have enough contests, eventually you will win one of them.

Open source does not instantly translate to another system.  For instance,
Christophe Theron would be hard put to translate many of the ideas, because he
uses 0x88 style and not bitboards.

The entire system that comprises the crafty source code will take an expert
chess programmer several years to understand.  I think probably they have better
things to do with their time.

That's on the one hand.  On the other hand, you would be a blooming idiot if you
were a professional chess programmer and did not scan through the excellent open
source programs like:
* Crafty
* Phalanx
* Amy
* ExChess
etc. to get good ideas.

My guess is that some really good ideas (or tricky details like how to implement
SMP smoothly) will be absorbed by the pros.

But just because you have open source does not mean that you can never exceed
the closed black boxes at some point.  You could not stay permanently in the
lead, because if you found some huge advantage [e.g. a revolutionary new
algorithm] eventually it would be understood and used by everyone.



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.