Author: Peter Fendrich
Date: 13:04:44 06/14/03
Go up one level in this thread
On June 14, 2003 at 15:53:51, John Merlino wrote: >On June 14, 2003 at 15:07:31, Peter Fendrich wrote: > >>On June 14, 2003 at 04:56:03, Ross Boyd wrote: >> >>>Hi all, >>> >>>My engine (TRACE) apparently hangs under CM9000. I don't own CM9000 and >>>therefore can't test it. So, I'm wondering if there is anything unique about >>>CM9000's implementation that I need to handle. >>> >>>Or is there a way to get CM9000 to print a debug file (like WB)? >>> >>>I output pv and bestmove in algebraic ie. e2e4 a8Q etc.... should it be in SAN >>>format? PLEASE say No... :-) >>> >>> >>>The best idea would be to implement a log file for TRACE, I know. But life is so >>>short... and besides, I'm feeling so incredibly lazy. :-) >>> >>>Thanks for any advice/suggestions... >>> >>>Ross >> >>There is one thing that isn't following the WB protocol exactly (at least in >>CM8000). It sends the command "." to the engine regualary in order to get some >>information about the search. Within the WB protocol it is only sent during >>analyze but CM8000 (and probably CM9000) sends it during all search types. >>Maybe your program goes banana when it receives a "." in the middle of the >>search? >> >>/Peter > >A clarification on this. CM8000 (and CM9000) will only send the "." command if >the "Thinking Lines" window is open, so that it can get a regular update on the >progress of the search. If that window is closed, the engine will never receive >the "." command. > >jm ok, it still isn't 100% WB protocol. OTOH I think it would be a perfectly logical change of that protcol. It's quite easy to take care of it in an engine of your own but you have to be aware of the difference. /Peter
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.