Author: Robert Hyatt
Date: 18:17:35 05/26/02
Go up one level in this thread
On May 26, 2002 at 18:48:11, Vincent Diepeveen wrote: >On May 26, 2002 at 18:15:41, Robert Hyatt wrote: > >>On May 26, 2002 at 16:07:41, Vincent Diepeveen wrote: >> >>>On May 25, 2002 at 18:02:45, Gian-Carlo Pascutto wrote: >>> >>>>On May 25, 2002 at 17:29:03, Vincent Diepeveen wrote: >>>> >>>>>you know very little from programming GCP. >>>> >>>>I know Sjeng can do it :) >>>> >>>>-- >>>>GCP >>> >>>then you can do more than Bob and all of his advisors >>>could imagine about a year ago in extensive discussions >>>where i showed what bitboards CANNOT easily do. >>> >>>This was no problem for bob as he doesn't believe in >>>knowledge. He believes in a bit of crude/rude/crafty >>>knowledge which then has to search zillions of nodes >>>a second. >>> >>>Most likely you do something simplistic which is rude >>>and not accurate. >>> >>>Be happy with it. >>> >>>For DIEP that accuracy is not enough. >>> >>>That's why i can't use bitboards. Not at 64 bits not at 32 bits. >>> >>>Best regards, >>>Vincent >> >> >>I have said this before, I will say it again. Bitboards are mathematically >>provable to be just as good as any other chess board representation. Think >>about it for a minute and you will understand why. There is nothing you can >>do with your offset (mailbox) board representation that I can't do with >>bitboards, and vice-versa... > >As proven a year ago it is over 3 to 4 times slower to do certain >things using pure bitboards. The word impossible is not good then, >but the word 'not a smart datastructure to use' is more valid, though >for me it comes down to the same. > >Best regards, >Vincent Bitboards are also _better_ in many cases also. I _still_ believe is it a "wash" until 64 bit hardware is prevalent. Then bitmaps will have a performance advantage overall.
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.