Author: Reinhard Scharnagl
Date: 13:59:39 03/14/05
Go up one level in this thread
On March 14, 2005 at 15:56:46, Tim Foden wrote: >On March 14, 2005 at 12:45:22, Michael Yee wrote: > >>(2) New commands from GUI to engine : getstartpos, getmovelist, getresult >> >>This enables the protocol to be truly extensible since variants that aren't >>officially supported by the GUI can still be supported by letting one engine be >>the "referee" or "arbiter". > >Maybe it would be a good idea to: > >1) allow an extra third engine as a referee. This could be useful in the case >of two engines where you trust neither to be a referee. A tried and trusted >third could be used in this job. This seems to be covered. But because of an initial lack of engines and for testing purposes any engine should be usable as a referee. That is what I have proposed. After some time special engines would qualify as good referees. But more important is, that this approach would simplyfy the writing of a GUI. Introducing a new game variant also would need to write only one engine instead of having simultaneously an additional referee engine at hands. >2) formalise what is required of a "referee" only. This would allow a simple >"referee engine" to be written, to perform arbritration only. Because a so called referee engine alone would not make sense without playing engines to be refereed, there is no need to write a reduced referee engine. May be an early engine prototype would already be sufficent to test and support a GUI. But it is of course better to replace it later by the completed engine, because this would cover all needs. I think it is time to make convergent builds of GUI and engine to refine that XCI proposal by real practice. Reinhard.
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.