Author: Matthias Gemuh
Date: 05:01:31 03/19/03
Go up one level in this thread
On March 19, 2003 at 07:05:12, Uri Blass wrote: >On March 19, 2003 at 06:51:15, Uri Blass wrote: > >>On March 19, 2003 at 06:33:01, Matthias Gemuh wrote: >> >>>I decided to stop chess programming but even the latest version of my program >>>sucks. How can I quit in peace? >>>It calculates this attack information (bitboards of attackers to 64 squares) >>> BITBOARD AttacksTo[64] >>>from scratch at each node. I tried to do this incremementally and it quickly got >>>messy and buggy because of sliding pieces, castle, en passant. >>>How do you attack attack boards (even the conventional type)? >>> >>>/Matthias. >> >>I do not know what is the conventional type >> >>I update my attack information incrementally but I use no bitboards. >>I also think that my attack information may be improved and I have not tables >>that tell me the information in Ed's site. >> >>The information that I have is for every square the directions that it is >>attacked by white and by black and for every direction the attacking square. >> >>I started by updating them not incrementally and later I started to update them >>incrementally only in part of the cases. >> >>I think that trying to do too much is a mistake and you may start by updating >>them incrementally only in part of the cases(for example you can update them >>incrementally only for rook moves that are not captures when the rook moves in >>the same file) >> >>Only when you see that there are no bugs you can continue to do it for rook >>moves in the same rank and only when you see that there is no bug you can try >>another piece. >> >>I worked some monthes about updating my attack information faster before >>I decided to develop my chess program and in the first months I had only a >>program to calculate perft and I worked on doing it faster and fixing bugs. >> >>Uri > >I can add that I decided not to have function for moves but functions for >deleting and adding pieces. > >It took me time to think about the idea to update the attack tables >incrementally after deleting a piece and update it incrementally after adding a >piece. > >I believe that function for moves may be slightly faster but I decided that I >have enough complexity and the advantage in speed is not big so practically >every move is a sequence of delete piece and add piece in my move generator. > >Uri Hi Uri, I will see how your ideas fit my bitboard attackers. Conventional attack boards are like Ed's. Some version of GnuChess may have them also. Maybe I shall look to see how they are updated. Thanks, Matthias.
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.