Author: Sune Fischer
Date: 15:26:01 02/11/03
Go up one level in this thread
On February 11, 2003 at 16:18:53, Daniel Clausen wrote:
>>window=pawn_unit:
>>
>>1 -1 0 1 1.a3
>>1 14 1 5 1.d3
>>1 23 1 14 1.d4
>>1 105 1 21 1.Nc3
>>2 -103 1 52 1.Nc3 Nf3
>>2 0 2 96 1.Nf3 Nf6
>
>
>>window=pawn_unit/10:
>>
>>1 -1 0 1 1.a3
>>1 14 1 5 1.d3
>>1 23 1 14 1.d4
>>1 105 1 21 1.Nc3
>>2 -103 2 52 1.Nc3 Nf3
>>2 0 2 96 1.Nf3 Nf6
>
>
>>window=pawn_unit/100: (pawn_unit=1000)
>>
>>
>>1 -1 0 1 1.a3
>>1 14 0 5 1.d3
>>1 23 0 14 1.d4
>>1 105 1 21 1.Nc3
>>2 -206 1 49 1.Nc3 Nf3
>>2 0 1 93 1.Nf3 Nf6
>
>
>It seems to me that you either have a flaw when displaying the current PV or
>some major bug in your search. (I think it's the first one only)
>
>Each time, the first PV for ply=2 contains an illegal last move. The errors only
>occurs on even plies though.
>
>Since the search has just started and I assume the HT was empty, it can't really
>be a collision-problem in the HT. But something's definitely wrong there. Since
>the error occurs already after a few searched nodes, it should be rather easy to
>find this one, no? (I assume it's reproducable?)
>
>There are errors on higher (even) plies too, but I'm sure they have the same
>source. Hope you find this pesky bug soon! :)
>
>Sargon
I have had this bug for ages.
I think it has to do with construction of the triangle array (I don't extract
from hash), happens somewhere in the qsearch I think.
Usually I just chop off the illegal moves.
I can't find anything wrong though, tried to debug it a few times but was never
successful. Any ideas?
This is my code:
inline void BOARD::UpdatePV(MOVE &move,int &d) {
ptree->pv[d].movenr=ptree->pv[d+1].movenr+1;
ptree->pv[d].movelist[0]=move;
memcpy(&ptree->pv[d].movelist[1],&ptree->pv[d+1].movelist[0],ptree->pv[d+1].movenr*sizeof(MOVE));
ptree->pv[d+1].movenr=0;
return;
}
More or less directly from Bruce's tutorial.
-S
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.