Author: Vincent Diepeveen
Date: 18:35:24 03/26/01
Go up one level in this thread
On March 26, 2001 at 17:50:30, Thorsten Czub wrote: >On March 26, 2001 at 17:31:56, Robert Hyatt wrote: >>>>This is very unfair behaviour. > >>I read the issue as "hey, the interface is _not_ telling my program that >>the game has ended so I am not learning anything (ie book learning or >>whatever)". > >>This shows a terrible weakness in the protocol... where one side has to be >>the "master" and the other has to be the "slave". A plus for the winboard >>protocol, in fact... which doesn't make that kind of design decision. > >>I would basically agree with Vincent that if a end-of-game sentinel is going >>to be sent, it should be sent as a default, and should have to be turned off >>to eliminate it, rather than having to turn it on to enable it. I can see >>room for _great_ mischief in autoplayer games, simply by choosing defaults >>that favor one engine over another. > >right. >and the problem is that this all influences the results. >especially when doing long matches. > >you can imagine what i want to say with this. The biggest influence is not the objective results, but the way in which the user interprets them, as only the program saving the games is given the chance to interpret the results. Which means it looks better for that program as it is. You can even throw away lost games by doing this. There is NO way to double check it except if you carfeully checkout log files!
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.