Author: Anthony Cozzie
Date: 15:33:01 09/04/03
Go up one level in this thread
On September 04, 2003 at 18:13:45, Gian-Carlo Pascutto wrote: >On September 04, 2003 at 17:13:42, Anthony Cozzie wrote: > >>>Uh, that's a tricky comparison. You don't know how I or Vincent order >>>our moves >> >>Crafty is GPL, ergo you should be at least as good or you need to read the >>crafty source again :) > >Crafty isn't GPL. OK, but the source is available. You get the point. >Your statement is wrong, because the programs are simply different. >If you have a program with attacktables, you'll use em in moveordering. >If you don't, you won't be able to, and maybe move ordering will hence be >worse. You could add attacktables and get better move ordering, but then >you go slower of course. > >An example in point is that Yace uses(used?) MVV/LVA instead of SEE. > >Technology from one program isn't necessarily transferable to another. You are right in principle, but I've seen crafty's move ordering, and there is nothing there that requires "crafty specific" data. Killer moves and history tables are relatively easy to do, and crafty uses mvv/lva IIRC. >> and I fail to see why you 'know' that we are searching slower >>>(hint: NPS is not a good indicator). >> >>why not? if it takes 1ms to evaluate a node, then split overhead is relatively >>smaller. admittedly I've never written a parallel search, but this seems >>logical to me. > >The way the nodes are counted is not necessarily the same. I'm not quite sure what to make of this . . . >-- >GCP
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.