Author: Bas Hamstra
Date: 01:46:42 06/21/00
Go up one level in this thread
On June 20, 2000 at 14:46:59, James Robertson wrote:
>On June 20, 2000 at 05:19:36, Bas Hamstra wrote:
>
>>On June 20, 2000 at 04:48:05, James Robertson wrote:
>>
>>>On June 20, 2000 at 04:44:57, Tony Werten wrote:
>>>
>>>>On June 20, 2000 at 03:50:36, James Robertson wrote:
>>>>
>>>>>On June 20, 2000 at 03:09:49, Tony Werten wrote:
>>>>>
>>>>>>On June 19, 2000 at 19:54:39, James Robertson wrote:
>>>>>>
>>>>>>>On June 19, 2000 at 19:48:36, Larry Griffiths wrote:
>>>>>>>
>>>>>>>>I have found bitboards to be an even trade-off on my Pentium system. I have to
>>>>>>>>update about 6 bitboards when a piece moves and this generates a lot of
>>>>>>>>instructions. I get it back in my IsKingInCheck code so it evens out.
>>>>>>>
>>>>>>>While detecting check is faster with bitboards if you have many pieces on the
>>>>>>>board, I think it is actually slower in endgame positions. :(
>>>>>>
>>>>>>I don't understand this. I have written it before but here it is again. You only
>>>>>>have to look at the to and from square of the last move.
>>>>>
>>>>>I do not understand.... could you please elaborate?
>>>>
>>>>Sure,
>>>>
>>>>If you exclude a rochade, a check can only occur in 2 ways.
>>>>
>>>>1) The piece last moved is giving check ( on the to-square )
>>>>2) The piece just moved revealed a slider which is giving check. ( behind the
>>>>from square )
>>>>
>>>>The code should look something likes this. ( there might be some mistakes)
>>>>
>>>>if (last_move==rochade)
>>>>{
>>>> square=king_square;
>>>> if (possible to attack square from rook_square)
>>>> {
>>>> repeat
>>>> square+=dir_of_rook;
>>>> until (square!=empty)
>>>> }
>>>> return (square==rook_square)
>>>>}
>>>>
>>>>if (possible for moved_piece to attack king_square from to_square)
>>>>{
>>>> if moved_piece==KING,KNIGHT or PAWN then return (true)
>>>> square=king_square;
>>>> repeat
>>>> square+=dir_of_to_square;
>>>> until (square!=empty)
>>>> if (square==to_square return) (true)
>>>>}
>>>>
>>>>if (possible for BISHOP or ROOK to attack king_square from from_square)
>>>>{
>>>> square=king_square;
>>>> repeat
>>>> square+=dir_of_from_square;
>>>> if (square!=empty) then return (piece_on_square can attack king_square)
>>>> until (square is not on board)
>>>>}
>>>>
>>>>return false;
>>>>
>>>>cheers,
>>
>>I doubt if this is faster than a simple call to SquareAttacked when you are
>>using BB's?
>
>Possibly. I intend to test that. My program generates attacks a rate of about
>2.6mil per second on a P233. I'll see what the above code is capable of.
Please let me know what the outcome is.
I do around 5M SquareAttacks/sec on a Celeron 466 (so about the same if you
factor out hardware).
Regards,
Bas Hamstra.
>
>James
>
>>
>>Regards,
>>Bas Hamstra.
>>
>>
>>
>>
>>>
>>>This idea looks nice. I think I can use it to improve my existing code....
>>>Thanks!
>>>
>>>James
>>>
>>>>
>>>>Tony
>>>>
>>>>PS If you are looking for a selfcheck, you only have to do the from_square. ( If
>>>>you didn't start of in check )
>>>>
>>>>>
>>>>>Thanks,
>>>>>James
>>>>>
>>>>>>
>>>>>>It has nothing to do with the number of pieces.
>>>>>>
>>>>>>Tony
>>>>>>
>>>>>>>
>>>>>>>James
>>>>>>>
>>>>>>>> I like
>>>>>>>>to have fast move generation code, but most of my gains have been through
>>>>>>>>alpha-beta, hash-table, killer-move and movelist ordering etc.
>>>>>>>>
>>>>>>>>Larry.
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.