Author: Dieter Buerssner
Date: 09:39:55 02/01/05
Go up one level in this thread
On January 31, 2005 at 08:11:18, Marc Bourzutschky wrote: >Unlike Thompson, I did not use a "fully-retro" algorithm, i.e., instead of >obtaining lost in N by backing up (with reverse moves) from wins in N and then >verifying that the backed up positions are indeed lost in N, I simply loop over >the still unresolved positions and determine lost in N by checking whether all >(forward) moves end on the other side winning in <= N. Unless there are a lot >of draws there is probably not much benefit, if any, in doing a fully retro >algorithm. I assume, your generator is rather general. How long does it take to generate KBBKN or KQPKQ? (From memory those belong to the toughest ones of the 5-men). I have a generator that is very similar to Wu Beal and fully retro. The mentioned TBs take about 10 minutes on my P4 2.53; the old Nalimov generator, that was available at Hyatt's ftp took over 8 hours on the same computor). >For the 1 ply consistency check I actually transformed the four output files >(KNNNNKQ and KQKNNNN with both white and black to move) to the Nalimov format, >including the non-blocking check trick. I then added code to Eugene's program >for KNNNNK and KNNNNKQ (pretty ugly) and ran his verification option. What exactly is the verifaction option doing? And another question from your original post: "Originally I had planned to implement a general 7-man tablebase generator using Johan de Koning's clever algorithm which requires very little memory." Is the algorithm published? Or your private knowledge (that you want/have to keep for yourself)? Regards, Dieter
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.