Author: Robert Hyatt
Date: 07:09:49 09/07/02
Go up one level in this thread
On September 07, 2002 at 04:40:44, Uri Blass wrote: >On September 06, 2002 at 21:23:44, Robert Hyatt wrote: > >>On September 06, 2002 at 18:57:50, Uri Blass wrote: >> >>>On September 06, 2002 at 18:48:01, Eugene Nalimov wrote: >>> >>>>Actually, I generated all 3+3 and 4+2 pawnless tables but two; twose last two >>>>are being generated right now. >>> >>>I think that some of these are not interesting. >>>For example it seems clear that KQQQ vs KN is always a win for the stronger side >>>except cases when the position is a stalemate. >>> >>>Uri >> >> >>Right, but you gotta have them to generate KPXX vs KX where X=anything... > >Only if you care about the exact number of moves. >You can have tablebases of distance to mate or conversion for KQQQ vs KN that is >not a stalemate. > >I agree that it is simpler to generate KQQQ vs KN than not generating it and >thinking how to change the source code of the move generator later. > >Uri Let me rephrase then. To use Eugene's generator, you have to have all promotion tables before generating the corresponding pawn table. You could write anything you want, of course, but I interpreted the question as using an existing TBGEN type program, not writing his own, which is a _very_ non-trivial task.
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.