Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rybka - not an endgame expert

Author: Vasik Rajlich

Date: 07:51:34 12/05/05

Go up one level in this thread


On December 05, 2005 at 10:45:30, Uri Blass wrote:

>On December 05, 2005 at 10:33:19, Vasik Rajlich wrote:
>
>>On December 05, 2005 at 08:34:26, Gabor Szots wrote:
>>
>>>Just two examples. Positions taken from from PET.
>>>
>>>
>>>8/5k2/4p3/4Pp1p/5P1P/3Rn2K/6p1/8 b - - bm g1=B;                id "PET026: rook
>>>vs knight";
>>>
>>>Here Rybka plays g1Q, although it does see it is only draw.
>>>
>>>
>>>8/8/1P1k2p1/8/5P1p/4Qb2/1q4PK/8 w - - bm Qe5+;                 id "PET046: Q+B
>>>vs Q";
>>>
>>>Rybka captures at f3.
>>>
>>>
>>>Both positions are easily solved by most of the engines. Position No26 usually
>>>in 0 seconds.
>>
>>Thanks for the positions, however my even worse answer is that I am aware of
>>this :( You will probably have to wait for Rybka 2.0 sometime next year.
>>
>>The only bug in this thread that will definitely be fixed for Rybka 1.0 is the
>>one given by GCP - a chess engine should be able to mate with KQQ v K :)
>>
>>Vas
>
>What is the reason for the problem to fix it?
>
>I thought that at least underpromotion is easy to fix.
>If the problem is some speed reduction you can have default version that does
>not consider underpromotions and slightly slower version that considers
>underpromotions.
>
>Uri

Uri,

the problem is programmer time. At some point I had some interesting things in
Rybka and decided that I was releasing in December no matter what. Don't worry,
the engine will soon be perfect :)

Vas



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.