Author: Christophe Theron
Date: 15:12:25 11/25/01
Go up one level in this thread
On November 25, 2001 at 13:10:28, Victor Fernandez wrote:
>On November 25, 2001 at 12:43:59, Steve Maughan wrote:
>
>>Absurdly not true. I doubt any commercial programmer releases an upgrade
>>without doing some development of the evaluation function. Most programmers
>>spend their time either looking at the search algorithms, the evaluation
>>function and I assume less than 10% of the time trying to make the existing code
>>faster.
>
>That could say it if you were a commercial chess programmer (or a chess
>programmer),
>but I believe that it is not the case. The main interest of the comercial chess
>programmers (except for honorable exceptions ) it is to defeat to the other
>comercial chess programs, and for they cut it code.
I'm a commercial chess programmer and I can confirm that what you say above is
WRONG.
Take Chess Tiger, Rebel, Fritz, Junior, Shredder, Gandalf, The King, and so on:
what you say above is simply wrong for these programs.
All the chess programmers that are making progress are going in the opposite
direction: they add knowledge, both in the evaluation and in the search
algorithms.
We spend a little bit of our time on speed optimization, just to make sure that
things do not go out of control in this area, but MOST of our time is spend on
"adding intelligence" to the program.
You have a poor view of what chess programming is in reality. Please reconsider.
Christophe
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.