Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: UCI versus Winboard

Author: Bas Hamstra

Date: 05:16:54 08/21/02

Go up one level in this thread


On August 21, 2002 at 06:18:33, Georg v. Zimmermann wrote:

>
>>>There are things that are not a matter of opinion, In my opinion :-). For
>>>instance, In UCI, it is not straightforward to do some techniques for learning.
>>>Alos, creativity about different pondering methods is almost impossible.
>>>IMHO, UCI is a good protocol for anaylis, but for engine-engine matches it
>>>has a design that limits the engine.
>>
>>  Good point. UCI is good for analysis, probably much better than wb, but not
>>for matches.
>>
>>  José C.
>>
>
>I agree that UCI might be ok for analysis. But as you said, not for games. In
>UCI you dont even know when a game ended, you have to extract that information
>from the last few moves. Also in long games the ammount of redundant data
>exchanged between the engine and uci-interfaces gets unbearable for a purist.
>And how pondering is handled - well - "creativity about different pondering
>methods is almost impossible" just about says it all.

Ok, to some extent I agree, especially about the game-result.

>Winboard has the big advantage that you can implement part of the protocol, and
>the engine already works fine. Winboard was here first. Winboard Protover 2
>works great.

>Now we only need to copy some of the nice ideas from UCI to winboard protover
>3, like engine-setting dialog (I already posted a detailed proposel a while
>ago) or hash fill status.

IF this happened (and the engine settings dialog IS great of course):

- It still would be less "robust" than UCI
-
In UCI I can let the engine display "free format" search info on the fly, I get
detailed





















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.