Author: Dieter Buerssner
Date: 11:09:00 11/23/00
Go up one level in this thread
On November 23, 2000 at 13:54:23, John Merlino wrote: >On November 23, 2000 at 07:21:55, Dieter Buerssner wrote: > >>And for anybody interested in the problem. To me it seems, that CM does not >>like moves in the form Ng1f3 sent to the GUI. Coordinate notation (is this the >>correct term?) g1f3 seems to work fine. >> >>I think, this should be considered a bug of CM ... >> >>Regards, >>Dieter "every WB GUI has its specialities" Bürßner > >Here's the question: Is Yace incorrect (in terms of the Winboard spec) in >reporting moves in that format, or should Chessmaster "support" it? I need to >know whether or not I should inform the development team, because they are very >reluctant to support all of the oddities/inconsistencies/quirks in every engine. My reading of the WB protocoll description of Tim Mann indicates, that the GUI frontend should be able to read different formats for the move text. Almost all GUIs I tried until now, didn't have any problems with it. The two exceptions were CM (which I cannot try personally) and Chess Academy 5.1. When I spotted the problem with Chess Academy, I changed to coordinate notation. (which is not yet in the public available version of Yace). One question to you. Many WB GUI frontends suffer from buffer overflow problems, when sending long PVs. I recently made Yace "resolve" EGTB hits in the PV, so that users can see the actual mate. This can yield in very long PVs. I.e. WB crashes with more than about 450 characters, Fritz doesn't like to receive more than 30 half moves ... Can you give a limit for the length of the PV for CM (in moves or characters)? Regards, Dieter Tim Mann's text: For the actual move text from your chess engine (in place of MOVE above), xboard will accept any kind of unambiguous algebraic format, including coordinate notation, SAN, and some slight variants of SAN. You don't have to send the pure coordinate notation that xboard sends to your engine; xboard parses the output with its general-purpose move parser, which was built to extract human-typed game scores from netnews messages. For example, the following will all work: e2e4 e4 Nf3 ed exd exd5 Nxd5 Nfd3 e8q e8Q e8=q e8(Q) e7e8q o-o O-O 0-0 and many more.
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.