Computer Chess Club Archives


Search

Terms

Messages

Subject: hashing in QS

Author: martin fierz

Date: 23:27:19 10/19/04


aloha!

i have this old hash code of mine and would like to update it to something
better, and am looking for some input. in the first versions of my program, my
move structure was about 100 byte large LOL and so i saved the index of the best
move as generated by the movelist into the hashtable, instead of the move
itself. since then, i reduced it to 12 byte and then finally read about those
who use only an int for move representation, so i'm finally there too now...
now.

i also have two hashtables, one for the main search and one for the qsearch. at
the moment these are incompatible because the qsearch and the main search have
different move generators, so the moveindex in one hashtable is useless in the
other. which is also the main reason i have two different tables.

so i want to store the move itself in the hashtable instead of the moveindex
now, as probably everybody is doing. the question i have is this: should i stick
with two separate tables for QS and main search? or should i just use a single
big table (seems more sensible to me). or should i do no hashing at all in QS?
any opinions on this?

cheers
  martin



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.