Author: Russell Reagan
Date: 21:24:42 05/06/04
Go up one level in this thread
On May 06, 2004 at 20:05:03, Robert Hyatt wrote: >I don't follow. I currently use 6-bit indices _everywhere_. Because the two >endpoints are moot. We're not changing that. Just squeezing out the bit for >the square where the piece in question sits... it is meaningless and could be a >zero or 1 and produce the same attacks... He is saying you can't always get correct attacks from 5 bits. Another way of putting it is that you can't always squeeze out the square that the attacker sits on, because it may have already been squeezed out. Let's say you start with all 8 bits for the 1st rank. You squeeze out the end bits, squares a1 and h1. You want to get attacks for a rook on a1. You can't squeeze out the square the attacker is on (a1), because you already squeezed it out. Put a rook on a1 and try to get correct horizontal attacks from 5 bits in all cases for the 1st rank. Take a simple example, 5 bits = 00000. What is the attack bit pattern for the 1st rank? Here is the board. You get 5 bits worth of rank state info, plus the implied rook on a1. abcdefgh 8 -------- 7 -------- 6 -------- 5 -------- 4 -------- 3 -------- 2 -------- 1 R-----?? ^^^^^ 12345 <-- You only get 5 bits What is the correct attack bitboard for this position? - represents a 0 bit # represents a 1 bit Choice A: -------- -------- -------- -------- -------- -------- -------- -####### Choice B: -------- -------- -------- -------- -------- -------- -------- -######- Choice C: Impossible to know without more information.
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.