Author: Thomas Mayer
Date: 08:55:49 07/11/03
Go up one level in this thread
Hi Grzegorz,
>//Transposition table entry
>unsigned long *trans_b;//Lock
>bool *val_side; //side on move
>int *val_gatunek; //flag
>int *val_hash; //score
>int *val_depth;//height
>int *val_morder;//move index from move generator
>BYTE *val_pt; //idx for another good move (for move ordering)
>BYTE *val_age;// age of this entry
putting that all in a struct might be way faster... just by the way...
> if (height>=level-2 && wrpx<beta && val_gatunek[spr_poz]==UBOUND)
> {
> do_null=false;
> }
outch, whats that ? New to me... :)
> else
> if (val_age[spr_poz]==m_cur_age)
> {
I don't understand this - for solving position this does not play any role, but
I do instead the following:
if (height>=level)
{
val_age[spr_poz]==m_cur_age)
-> just to prevent that entry's from older search which stay important are not
overwritten. (I have two hashtables - one with some overwriting rules and the
other one always replace)
>if (level>=height || m_cur_age!=val_age[spr_poz])
>{
I think this is the problem ! (On the other hand, I wonder why you get anyway so
deep) -> with this replacementscheme you can't solve the position. Try it out
with always replace ?! (The reason why I use two hashtables)
Greets, Thomas
P.S.: If that does not help: try out without nullmove
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.