Author: Miguel A. Ballicora
Date: 06:08:52 04/19/02
Go up one level in this thread
On April 19, 2002 at 01:54:32, Gian-Carlo Pascutto wrote: >On April 18, 2002 at 19:41:40, Gian-Carlo Pascutto wrote: > >>[D]5N2/2n5/7k/5Pp1/ppr5/4B3/Pb4PP/3R2K1 w - g6 0 2 >> >>White can play fxg6 e.p and *check*. >> >>My program has been broken on this kind of position, for, eh, >>forever probably. > >To clarify: > >the issue isn't about finding the right move (I have no idea >which one that would be), but to see if the program understands >that after gxf6 black is in check. > >This move uncovers a check along a different diagonal than the >one where the pawn moves along, and it's also not one going >through the from or to squares. My program doesn't check this; >I never thought it would be possible. Others may have the smae >bug too, especially if you have an optimized incheck function. Thanks for posting this. Gaviota does not have that problem but thanks to your position I found a stupid bug on setboard(). I have a code that checks if the ep flag is correct, i.e. if there is a pawn to capture etc. Well, I had confused a column and a row in one line. Now it is fixed. So, gaviota had problems with this position as loaded in setboard but it did not have problems in the tree. Watch out, can you solve this other case? (it is mate in one) [D]8/8/8/kpP4R/p7/P1p5/2q5/K4BB1 w - b6 Other positions to test (inspired in your example) Mate in 5 and Mate in 6 The first test the problem is in the root and the second in the tree. [D]3qrN2/4n3/4B2k/5Ppp/8/8/5RPP/2B4K w - g6 [D]3qrN2/4n1p1/4B2k/5P1p/8/B7/5RPP/7K w - - Regards, Miguel > >-- >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.