Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF and the programmers............

Author: Robert Hyatt

Date: 13:55:40 03/22/98

Go up one level in this thread


On March 21, 1998 at 06:24:13, Ed Schröder wrote:

>>Posted by Robert Hyatt on March 19, 1998 at 16:22:25:
>
>>this cuts two ways.  IE I have gotten some games played via the
>>autoplayer
>>with rebel vs crafty, with the rebel "anti-doubles" option enabled, and
>>Rebel is quite capable of aborting a game (somehow) without Crafty
>>having
>>any say-so in it at all.  So this has been there from day one I
>>suppose??
>
>Yes, since Rebel7 this "anti-doubles" software is added. When
>AUTO232 is in charge there are two options starting Rebel:
>
>REBEL.EXE  This will activate the "anti-doubles" software and also
>will terminate a game when the Rebel score is +5.00 or -5.00. Also
>boring (drawn) endgames such as KRKR, KQKQ etc. will be terminated.
>
>This whole process speed up the testing with a factor 2-3. Not bad
>and exclusively programmed for myself in the first place. Later I
>decided to release it also for the public, why not.
>
>Now to your question, the Rebel driver has nothing to say. It doesn't
>rule over Crafty. When Rebel terminates a game (for example by
>reaching a +5.00 score) Rebel just will not accept any keyboard
>input anymore.
>
>Then, since Rebel refuses to play any longer both Pc's are doing just
>nothing and are waiting till the "/t" parameter time value is reached.
>Then the Rebel driver saves his own game into the database and a new
>game is started.
>
>When I sent Rebel7 to SSDF they after some time contacted me saying
>they couldn't test Rebel7 without AUTO232 because of the:
>
>- anti-doubles software.
>- the +5.00 / -5.00 decision to abort a game.
>
>All understandable. Since Rebel7 needed a bugfix I promised the
>SSDF to add a parameter to Rebel7 which bypasses all the special
>software. So with the patch we introduced the "A" option. So if
>you don't like my special AUTO232 software just start Rebel as
>REBEL.EXE A
>
>- Ed -


OK... I misunderstood what I was seeing in the games you sent me a
long while back.  I only saw games that suddenly ended with no
explanation
of any kind...  which the timeout would explain.

However, as I mentioned, the "position learning" idea somewhat
invalidates
this approach, because even if Crafty plays the same book line twice,
there's
no guarantee it will play the same after it leaves book because of
this...



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.