Author: milix
Date: 07:08:14 03/11/05
Go up one level in this thread
On March 11, 2005 at 09:46:45, milix wrote: >On March 10, 2005 at 12:41:10, Ingo Bauer wrote: > >Are you talking as a user or as a chess programmer? Because I see this in my debug file: 7375>1:uci 67563<1:id name AICE 0.91 67625<1:id author Anastasios Milikas 67750<1:uciok 67953>1:isready 68610<1:readyok 69360>1:ucinewgame 69360>1:isready 69391<1:readyok 69516>1:position startpos moves e2e4 69516>1:go wtime 302000 btime 302000 winc 2000 binc 2000 69547<1:bestmove c7c5 72563>1:position startpos moves e2e4 c7c5 a2a3 72563>1:go wtime 301172 btime 303974 winc 2000 binc 2000 81485<1:bestmove b8c6 and now i take back my program's last move (here wboard enters the 'force' mode) and play Na6 instead, so I'm analyzing from now on. The communication goes: 88000>1:position startpos moves e2e4 c7c5 a2a3 b8a6 88000>1:go wtime 301172 btime 297521 winc 2000 binc 2000 and we have a problem. We are analyzing but we think that we are playing. > >>Hello >> >>>Pondering is one problem, learning is another one. How can someone be sure in >>>UCI if he analyzes a game or playing a game? UCI has many good features when the >>>desired function is game/position analysis, but for actual games wboard is >>>better. >> >>Pls reread what I have written? I deleted everything except the relevant part >>for that matter. In short: If there is a "ucinewgame" it should learn out of a >>game if there is nothing it is analysis or position learning. It is possible to >>decide that even without a ucinewgame, but because some critisised that the new >>command would be implemented in 2004. >> >>Btw: Shredder engine and its learning abbilities itself prove that learning is >>more than just "possible". If you do not like the fact that Shredder may not be >>unbiased, take an independent engine that is quite good in learning positional >>and out of a game: Gandalf 6. >> >>Bye >>Ingo >> >>>>>4. In UCI it is on you to implement a book learning in your engine! >>>>> >>>>>* ucinewgame >>>>>this is sent to the engine when the next search (started with "position" and >>>>>"go") will be from a different game. This can be a new game the engine should >>>>>play or a new game it should analyse but also the next position from a testsuite >>>>>with positions only. If the GUI hasn't sent a "ucinewgame" before the first >>>>>"position" command, the engine shouldn't expect any further ucinewgame commands >>>>>as the GUI is probably not supporting the ucinewgame command. So the engine >>>>>should not rely on this command even though all new GUIs should support it. As >>>>>the engine's reaction to "ucinewgame" can take some time the GUI should always >>>>>send "isready" after "ucinewgame" to wait for the engine to finish its >>>>>operation.
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.