Author: Robert Hyatt
Date: 09:55:32 06/20/02
Go up one level in this thread
On June 20, 2002 at 04:41:44, Russell Reagan wrote: >On June 20, 2002 at 00:59:42, Robert Hyatt wrote: > >>That is why I _still_ maintain that 90% of the aggregate computer chess >>advancements were academia-based. Because all academia-developed ideas have >>been published for all to see, increasing the body of computer-chess knowledge >>slowly but surely. > >So, for example, alpha-beta and null-move were published, so they are in common >usage today and have advanced computer chess overall. What Fritz, Junior, Tiger, >Shredder, Rebel, etc. do is unpublished, so their achievements have done nothing >to further the computer chess community as a whole. Correct... Not to mention other ideas. null-move search. aspiration window. PVS. Iterated search. mtd(f). Internal iterative deepening. "pondering". Parallel search. time usage. hash tables in general. Hash replacement strategies. Evaluation terms such as square of the pawn. Endgame tables. non-generic search methods like proof number, conspiracy number, SSS*, etc... move ordering. bitmaps. 0x88. singular extensions. fractional extensions. threat extensions. The list goes on and on. And _everybody_ has access to this body of knowledge so that they don't have to start from scratch. > >That makes sense, but it doesn't necessarily mean that their achievements are >any less significant. In other words, if all of the top commercial programs were >suddenly open source, and each author wrote a book about the inner workings of >their program, don't you think there would be another revolutionary idea born >within the near future, or at least that overall the entire community would >discover new ideas to push the envelope with? I think so, but you have a few >decades of experience over me, so you'd know better than I :) > >Russell I didn't say they were any less significant. I just find the one-way information pathway distasteful. Would _you_ feel comfortable taking ideas from someone without offering him some ideas in return? During every ACM event we held, all the authors (and others that were interested) used to hold impromptu discussions explaining what we had done new for this year. Everybody took all those ideas home, used the ones they liked, and also tried to come up with _new_ ideas to surprise everyone the next year and win the event. Then we "revealed all" again and repeated the cycle, year-in and year-out. And progress was quite remarkable. That is done to an extent today, but there are a few "black holes" if you know what I mean. Information only flows _in_ to a black hole, nothing can escape the event horizon and get back out.
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.