Author: Lance Perkins
Date: 15:40:37 03/09/05
Go up one level in this thread
I don't see what that "huge" advantage is in the case of chess engines. Chess engines are by nature stateful - PV, hash tables, killer tables, etc. In UCI, can anyone justify why the entire move history has to be sent for each move? In WB if a series of moves really has to be sent, then you just send the "force" command followed by the moves. Otherwise, its just the next single move. That's what I'd call 'elegant'. On March 09, 2005 at 15:55:20, Gian-Carlo Pascutto wrote: >On March 09, 2005 at 13:56:32, Lance Perkins wrote: > >>The UCI protocol pretends to be stateless, when in fact it is. So, what is it >>really? > >"More stateless" than xboard. And yes, that *is* a *huge* advantage. > >>Marrying the UI and the protocol? Now that's a hack. How would you localize >>that? It means you have to localize every engine, when you should only be >>localizing the UI. > >Since UCI solves this, I assume you are praising UCI? > >-- >GCP
This page took 0.01 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.