Author: Dann Corbit
Date: 12:37:08 10/10/98
Go up one level in this thread
The biggest opportunity I see for chess software is in the database arena. The database systems used by current chess systems are pitiful toys compared to the technology that is available. Let's take opening books [for instance]. We might take a bunch of GM games and decide to create opening books from them. But even in a case like that, there can be [and indeed are] blunders. Yet the computer will almost blindly trust the move, since it is in the opening book [or at least spend a tiny fraction of time analyzing it]. What happens once you wander out of book? Then the computer tends to be on its own. But perhaps there is an excellent path back in the book openings that you can arrive back at. Again, some computer systems do not hunt for these opportunities. The endgame functions for tablebase systems are well defined, but the query systems could be much more transparent. Why not make a database server with one terabyte of storage space and huge compressed tablebase files that a computer can query via tcp/ip? That way, instead of millions of machines chewing up gigabytes each, you have a few machines. Sort of like fics servers, but they are knowlege base systems. Would not work too well with a phone connection, but with a full speed internet tap, I bet it would be worth while.
This page took 0.01 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.