Author: Robert Hyatt
Date: 14:49:50 12/01/99
Go up one level in this thread
On December 01, 1999 at 16:51:06, Jeremiah Penery wrote: >On December 01, 1999 at 14:41:26, Robert Hyatt wrote: > >>On December 01, 1999 at 14:26:34, Ricardo Gibert wrote: >> >>>The blitz rating for crafty has plummeted to under 2800 in a fairly short period >>>of time. Has the latest version acquired a bug or we witnessing a normal rating >>>oscillation at its low ebb? For awhile there crafty sported a 3250+ rating that >>>must have sent frightened GMs diving under their desks for cover. What happened? >> >> >>It is a bug. I just found one serious thing, but I have simply not had time >>to do further analysis. I am currently in a hospital bed (fortunately with a >>phone and my trusty notebook) recovering from surgery. Once I get home later >>this week, I hope I can fix everything. > >Oh, my! Add my hopes that it was nothing serious, and that you make a speedy >recovery. :) Not particularly serious, and all is well now, just need to get back to the house and out of this poke/test/prod zoo... > >>The current bug dealt with having a "char passed_pawn_value[8]" array... which >>I scaled up so that a passer on the 7th starts at a score of 132. Unfortunately >>+132 (in a char) is a negative number which was not what I intended. :) > >What version was this bug introduced in? I've been having problems with passed >pawns on the 7th rank lately with 17.2. It always seems to allow them, and then >often doesn't try to capture them when it easily could. Eventually, it chokes >on them and loses. :( look at the source files, module data.c, and look at the passed_pawn_value[] initialization. If it has a 132 in it, change it to 127 and that problem will go away. I will try to release a new version tonight once I watch this on ICC to be sure it is reasonable now...
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.