Article View: gnu.chess
Article #1638Re: Activate Analysis mode
From: Simon Waters
Date: Mon, 13 May 2002 16:59
Date: Mon, 13 May 2002 16:59
54 lines
3047 bytes
3047 bytes
Lukas Geyer wrote: > > "Bohr, Mike" <Mike_Bohr@gmx.net> writes: > > > how can I activate the analysis mode? I've xBoard GNU chess v5.02 > > GNU chess is not yet capable of analysis, we are working on it. The command line 'post' output is very similar to the analysis mode of most engines, which is why I have a minimal working analysis change somewhere, it is five minutes work. i.e. ./gnuchess st 10000 post book off e4 (or "load position.fen", "go") ctrl-C to stop thinking. Will print out depth information, nodes searched, and principal variation, plus some weird symbols which do mean something honest. What we have remaining is "merely" to tell gnuchess how to understand when xboard wants it to stop thinking and do something (like tell it the analysis so far), ironically this is harder than the analysis code itself. I think also that analysis has to print out some other number, which we don't currently track in the code, but I don't think that is a big issue. I'm keen to get analysis in as it makes the program work better with SCID, as well as helping us spot problems with it's game. Meantime if you want to know more about GNU Chess thinking than "Show Thinking" gives you I'm afraid it is the command line. Lukas has done much of the work for handling the communication aspects, I have to check it, and we will probably add pondering at the same time. Simon PS: Anyone want to think about sorting moves. Looking at the code we spend a long time calculating the "SwapOff" value of moves, and use this to sort moves. I'm sure there are probably computationally cheaper ways of deciding which moves to consider first. Even just "capture piece value - capturing piece value" is pretty close, and a LOT faster. PPS: Similarly preserving information on HashMv/Killer1/Killer2 and similar between moves could lead to a significant gain in strength.
Technical Details
Message-ID:
<3CDFF100.481FD91A@wretched.demon.co.uk>Path:
arc1-ca.pugleaf.net!archiver1.google.com!news1.google.com!newsfeed.stanford.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!kibo.news.demon.net!news.demon.co.uk!demon!wretched.demon.co.uk!not-for-mailReferences:
<abo759$j263u$2@ID-143325.news.dfncis.de> <87bsbkgnyz.fsf@fullp.ath.cx>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 10d71d,f0e784582fcc7ab1 X-Google-Attributes: gid10d71d,public X-Google-ArrivalTime: 2002-05-13 10:00:00 PST Path: arc1-ca.pugleaf.net!archiver1.google.com!news1.google.com!newsfeed.stanford.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!kibo.news.demon.net!news.demon.co.uk!demon!wretched.demon.co.uk!not-for-mail From: Simon Waters <Simon@wretched.demon.co.uk> Newsgroups: gnu.chess Subject: Re: Activate Analysis mode Date: Mon, 13 May 2002 17:59:44 +0100 Organization: Eighth Layer Limited Message-ID: <3CDFF100.481FD91A@wretched.demon.co.uk> References: <abo759$j263u$2@ID-143325.news.dfncis.de> <87bsbkgnyz.fsf@fullp.ath.cx> NNTP-Posting-Host: wretched.demon.co.uk X-NNTP-Posting-Host: wretched.demon.co.uk:193.237.109.203 X-Trace: news.demon.co.uk 1021309187 nnrp-14:9786 NO-IDENT wretched.demon.co.uk:193.237.109.203 X-Complaints-To: abuse@demon.net X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.19 i686) X-Accept-Language: en-GB, en-US, en MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 54