Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: DanChess 1.07 is an amazing newcomer ! ( Long Post)

Author: David Dahlem

Date: 07:59:38 10/27/04

Go up one level in this thread


On October 27, 2004 at 10:52:30, Uri Blass wrote:

>On October 27, 2004 at 10:35:35, Daniel Clausen wrote:
>
>>On October 27, 2004 at 10:09:48, Uri Blass wrote:
>>
>>[snip]
>>
>>>Baron got only depth 5 in 41 seconds and it is not the only case(it happened
>>>also at move 19 with 21 seconds).
>>>
>>>I think that maybe it is better if programmers tell their program to refuse to
>>>play if they do not get enough nodes per second when the user may choose the
>>>minimal number of nodes per second based on the speed of his computer.
>>>
>>>I think to make minimal number of nodes per second as a parameter and probably
>>>suggest to set it as 1/2 of the number of nodes per second that the program get
>>>in the opening position.
>>>
>>>In case that the program does not get enough nodes per second it will refuse to
>>>play and report game aborted by insufficient number of nodes.
>>>
>>
>>What are typical reasons for such misbehaviours? Based on the posts here in CCC,
>>these things happen quite frequently.
>
>I think that the reason is that the computer does not give time to the process
>of the engine.
>
>I want the engine to refuse to play if it does not get at least 90% in the cpu
>time and I want also the opponents to do it because I am not interested in
>unfair games.
>
>I have no idea how to detect cpu time by the program(I can do it only by looking
>at the window task manager) and the only idea that I thought about is to use the
>number of nodes to detect situations when it is clear that the engine does not
>get at least 90% cpu time.

I believe Mike Byrne, in Crafty SE, uses limitation of cpu time to weaken
certain personalities. So he must know how to detect cpu time. :-)

Regards
Dave

>
>>
>>The suggestion above though sounds like "rebooting Windows if something's wrong"
>>or "the doctor giving you cough syrup if she doesn't have a clue what's wrong".
>>Hardly something I'd call a solution.
>>
>>Sargon
>
>I agree but what is the solution to know that my engine get enough cpu time.
>I do not like to release an engine and after it get results of tournaments when
>it is clear that it does not get enough cpu time from the output.
>
>I want to be sure that games are not played when something is wrong and it is
>not the fault of the engine.
>
>Uri



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.