Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Has Christophe re-written Tiger?

Author: Steve Maughan

Date: 07:38:55 10/08/03

Go up one level in this thread


Christophe,

Thanks for the reply.

>I needed more flexibility in order to be able to add knowledge without slowing
>the program down. In other words in the new version the NPS and the amount of
>knowledge have nothing to do with each other. The new version will be at the
>same time faster and have a lot more knowledge than previous versions.

[...]

>The current version retains all the knowledge of CT15. But this knowledge is
>now significantly out of sync with the new concepts, so it needs some work,
>and I plan to improve it drastically - that's what the rewrite is all about.
>
>Anyway, from now on it does not make any sense anymore to look at the NPS and
>conclude anything about the amount of knowledge in the program.

Interesting... the only other place I've heard this type of statement is in the
Brutus article.  Are you moving away from your array based representation of the
board?  Also, an interesting comment about being "out of sync" since with most
chess evaluation routines this isn't a problem, which I guess implies that
you're doing something quite different.  Do you regard this re-write as a
potential long-term breakthrough?

Regards,

Steve



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.