Author: Ed Schröder
Date: 04:54:30 11/01/03
Go up one level in this thread
On November 01, 2003 at 05:10:01, Johan de Koning wrote: >On October 31, 2003 at 10:53:42, Christophe Theron wrote: > >>On October 31, 2003 at 01:53:45, Johan de Koning wrote: >> >>>On October 30, 2003 at 22:12:09, Christophe Theron wrote: >>> >>>>It is my understanding that neither Johan nor the ChessMaster manufacturer >>>>(don't even know who it is now, it has changed so many times) have helped the >>>>SSDF to get a decent setup for CM9000. >>> >>>1. Your understanding is irrelevant. >>>2. You fail to define "decent set-up". >>>E.g. Crafty can be tested in a convenient way under CM GUI. >>>Does that qualify as decent? >>> >>>And BTW, it seems that 2003 is a bad year for October in terms of not getting >>>pissed off. But with Graz in the pipeline I fear the worst for Nov and Dec. :-) >>> >>>... Johan >> >> >>OK Johan. Nice to see that you communicate on the subject. > >I'm glad you're using a wide definition of "communication". :-) > >>So, what exactly have you done to help the SSDF in solving the issues in >>question? > >What issues? >CM9000 plays perfectly legal chess. >It even runs almost any WB engine. Unfortunately ChessBase does not run as a >WB engine under CM. So some work-around is called for. I was happy to make >suggestions and debug the set-up. > >>Do you consider as fair the fact that they are using the opening book of a >>competitor? > >It would be unfair if the competitor would actually use that book and win >many games because of it. But I'm not too worried about both. > >>In this case, is the next Chess Tiger allowed to use the same opening book? Or >>to use a mix of all available commercial opening books? > >As long as it is publicly available, yes. >Then the set-up would be known, which is about the only thing that matters >when it comes to testing. > >... Johan Hi Johan, The problem as far as I see it is that TK is now tested with 2 ingredients that represent intelligence (book+learning=elo) that aren't the work of yours but the work of others. That's not okay. Of course you are not to blame, don't get me wrong. Needless to say the experiment is very interesting, but that is what it will remain: an experiment and no real testing, at least in my perception how to interprete the work of the SSDF people. My only concern on this issue is the image of the SSDF, it is good now, the step now taken is wrong. Just to make a suggestion: it was little work for me to add the opening book + learning stuff to the Winboard Rebel acting totally independent from the Interface. Maybe the work involved will be different for TK, I can't judge, but it would solve this problem instantly. Please save the SSDF :) My best, Ed
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.