Author: Tom Likens
Date: 04:43:27 11/01/03
Go up one level in this thread
On October 31, 2003 at 09:24:22, Robert Hyatt wrote: >On October 30, 2003 at 17:24:02, Randall Shane wrote: > >>On October 30, 2003 at 10:59:59, Robert Hyatt wrote: >><snippage> >>> You only need a >>>total of 768 RNG's. >> >>Here's what is probably a bad idea, but... >> >>If one only needs a total of 768 random numbers, why have a random number >>generator in the program at all? Generate them once. Use them as an >>appropriately defined array of constants in the program. > >You can do that. In fact, you probably should do that for simplicity. > >However, I need random numbers for other reasons. IE I have some randomness >built into my opening move book selection code. But doing it for the Zobrist >numbers, and avoiding random generation there would make perfect sense. You >could spend some time tweaking the numbers also, if you find a few that are >too close in hamming distance. > > >> >>I can see many programs sharing the same set of numbers, for instance. >>Could help toward being able to share both book and learning files (if hash >>signatures are used to store the data) between different programs. > >I agree... Yes, Robert perhaps you and Vincent could work out the set of numbers!! :-) :-) ;-) Sorry, couldn't resist. --tom
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.