Computer Chess Club Archives


Search

Terms

Messages

Subject: Suggestion

Author: Ricardo Gibert

Date: 07:34:09 06/11/01

Go up one level in this thread


On June 10, 2001 at 15:19:44, Bruce Moreland wrote:

>On June 10, 2001 at 12:54:40, Mogens Larsen wrote:
>
>>On June 10, 2001 at 12:27:28, Bruce Moreland wrote:
>>
>>>My basic approach will be to call the eval function, which thanks to a change
>>>made two days ago will probably recognize that the game is an endgame, and
>>>reward the program if its king is in the center.  That's certainly good enough
>>>to win KQ vs K, and is probably good enough to win KR vs K.  What more do you
>>>need than that?
>>
>>That sounds good enough to me. Just curious about your general approach, ie.
>>encouraging people en masse to add Nalimov EGTB support.
>>
>>Mogens.
>
>I don't know what your last sentence means.
>
>I think endgame databases are cool, and if I were going to distribute something
>that was designed for strength, I would include support for them, assuming I got
>permission from Eugene.
>
>This program is going to be fairly strong just because it will have most of the
>important pieces present, but I don't think it would gain much from having
>endgame tables, and it would get bigger and more complicated.
>
>I'm already having to restrain myself from adding tricky stuff.
>
>bruce

Here's a suggestion, which I think could make Gerbil more interesting and
requires only a small amount of additional effort:

Instead of releasing Gerbil as a single program, why not release it as a series
of programs, where Gerbil_1 is the 1st version you got to work, Gerbil_2 adds
hash tables and Gerbil_3 adds an opening book, etc.

In this way, you won't have to "restrain" yourself and people can select the
version most suitable to them to learn from.

The usefulness of Gerbil will be greatly enlarged and the amount of additional
work by you will be relatively small. All you would be doing, essentially, is
releasing the intermediate versions in addition to the "final" version. A small
amount of additional work will be needed to make the intermediate versions
"presentable".

Does this sound like a good idea? I think it would greatly add to the
educational value of Gerbil and really set it apart from similar efforts like
SCP, TSCP, etc.



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.