Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SEE for forward pruning in Q. Search and nullmove cooking problems

Author: Heiner Marxen

Date: 08:08:52 08/14/99

Go up one level in this thread


On August 13, 1999 at 23:03:12, Robert Hyatt wrote:

>On August 13, 1999 at 06:35:13, Vincent Diepeveen wrote:
>
>>
>>At Paderborn there were never more than 2 terminals showing DB junior.
>>It was with BIG Capitals. It's main problem and weakness was that it
>>played without book obviously. Most of the times replying d5.

[snip]
>As I mentioned before... it is stateless.  As a result, they didn't deal with
>book moves at all. They didn't catch repetitions.  They didn't catch 50 move
>rule draws.

Also important (IMHO): they did not ponder, and they did not save any
hash table entries across moves.

Basically, for a web server every request (click) is a completely
independant computation.  A connection with earlier request is possible
via hidden form fields and cookies, but that is not be much more
information as just to identify the requestor, much less than a TT.

Vincent, perhaps you should try to design a web interface (CGI) for
your program in order to see the problems/limitations involved.
I did so (for some other task), and learned it that way.

Cheers, Heiner



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.