Author: Thomas Gaksch
Date: 13:51:38 03/17/05
Go up one level in this thread
On March 17, 2005 at 15:43:27, Frank Schneider wrote:
>Hi Dan, Fabien, Thomas!
>
>First, something is missing from your diff:
>
>
><big snip>
>>unable to open %s
>>search_full.c
>
>
>Second: in search_full.cpp, Toga's code is:
>
>history_pruning = false;
>save_old_depth = new_depth;
>
> if (!in_check && played_nb >= HistoryMoveNb && new_depth < depth &&
>threat_extension == 0) {
> value = sort->value; // history score
> if (value < HistoryValue) {
> if (node_type != NodePV){
> // always true
> new_depth -= Fractional;
> history_pruning = true;
> }
> if (value < HistoryBound && new_depth > 1*Fractional){
> new_depth -= Fractional;
> history_pruning = true;
> }
> }
> }
>}
>
>
>The inner "if (node_type != NodePV)" is always true.
>
>Best,
>Frank
hi frank,
that is true what you mentioned. for tests i have often commented out the and
condition like
if (UseHistory && depth >= HistoryDepth /* && node_type != NodePV */) {
if it is commented out the other condition makes sense.
but anyway. ther could be bugs in my few lines. last time i programmed c is 12
years ago.
thomas
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.