![[ee]](/mwf/v12/flags/ee.png "Estonia")
Maybe this is already fixed in 2.3.2, but still ...
Found another position, where most of Rybkas have erratic behaviour, when single best variation is analysed. The error is, that she does not show the best line (actually, no move at all) and calculations seem to stop very soon after start. The best line reveals only after stopping the engine or in some case when to analyze multiple lines together.
The position is:
Tested on E6600, 2GB of hash, Win XP 64-bit, Shredder 9 interface and following Rybkas:
2.1c mp 32-bit = OK
2.1c 32-bit = no move!
2.2 mp 64-bit = no move!
2.2 mp 32-bit OK
2.2 64-bit = no move!
2.2 32-bit = no move!
2.2n2 mp = no move!
2.2n2 mp 32-bit = OK
Winfinder 2.2 mp 64-bit = no move! (shows mate in 0???)
Winfinder 2.2 mp 32-bit = no move! (shows mate in 0???)
Winfinder 2.2 64-bit = no move! (shows mate in 0???)
Winfinder 2.2 32-bit = no move! (shows mate in 0???)
2.3 mp 64-bit = no move!
2.3 mp 32-bit = no move!
2.3 64-bit = no move!
2.3 32-bit = no move!
2.3 lk mp 64-bit = no move!
2.3 lk mp 32-bit = no move!
2.3 lk 64-bit = no move!
2.3 lk 32-bit = no move!
2.3.1 mp 64-bit = no move!
2.3.1 mp 32-bit = no move!
2.3.1 64-bit = no move!
2.3.1 32-bit = no move!
Actually, for all 2.3 versions (except 2.3.1 32-bit) the problem seem to start already some moves earlier, from position
There even multiline analyses does not always give a result.
Vasik, please note a systematic difference between results of 32-bit and 64-bit versions of 2.2mp and 2.2n2 mp (32-bit versions of those Rybkas and 32-bit 2.1c did solve the position correctly). That was the point I had in mind, when discussing in topic 64-bit computing. The difference is not random and it makes me believe, that 64-bit engine has slightly altered behaviour than 32-bit engine, although built of same source code.
Found another position, where most of Rybkas have erratic behaviour, when single best variation is analysed. The error is, that she does not show the best line (actually, no move at all) and calculations seem to stop very soon after start. The best line reveals only after stopping the engine or in some case when to analyze multiple lines together.
The position is:
6Q1/8/6P1/4Bk2/3K4/1p6/1P6/8 w - -
Tested on E6600, 2GB of hash, Win XP 64-bit, Shredder 9 interface and following Rybkas:
2.1c mp 32-bit = OK
2.1c 32-bit = no move!
2.2 mp 64-bit = no move!
2.2 mp 32-bit OK
2.2 64-bit = no move!
2.2 32-bit = no move!
2.2n2 mp = no move!
2.2n2 mp 32-bit = OK
Winfinder 2.2 mp 64-bit = no move! (shows mate in 0???)
Winfinder 2.2 mp 32-bit = no move! (shows mate in 0???)
Winfinder 2.2 64-bit = no move! (shows mate in 0???)
Winfinder 2.2 32-bit = no move! (shows mate in 0???)
2.3 mp 64-bit = no move!
2.3 mp 32-bit = no move!
2.3 64-bit = no move!
2.3 32-bit = no move!
2.3 lk mp 64-bit = no move!
2.3 lk mp 32-bit = no move!
2.3 lk 64-bit = no move!
2.3 lk 32-bit = no move!
2.3.1 mp 64-bit = no move!
2.3.1 mp 32-bit = no move!
2.3.1 64-bit = no move!
2.3.1 32-bit = no move!
Actually, for all 2.3 versions (except 2.3.1 32-bit) the problem seem to start already some moves earlier, from position
6b1/4k1P1/6P1/3BB3/3K4/1p6/1P6/8 b - -
There even multiline analyses does not always give a result.
Vasik, please note a systematic difference between results of 32-bit and 64-bit versions of 2.2mp and 2.2n2 mp (32-bit versions of those Rybkas and 32-bit 2.1c did solve the position correctly). That was the point I had in mind, when discussing in topic 64-bit computing. The difference is not random and it makes me believe, that 64-bit engine has slightly altered behaviour than 32-bit engine, although built of same source code.
![[hu]](/mwf/v12/flags/hu.png "Hungary")
Both work for me (see below). Could something strange be happening on your machine, with priorities, etc?
Re. 32 bit vs 64 bit, this is purely a matter of speed. The non-determinism of the MP version may mask this, but it's true.
Vas
New game,
Analysis by Rybka 2.3.1 32-bit :
1.Qe8
+- (30.21) Depth: 2 00:00:00
1.Qe8
+- (27.71) Depth: 3 00:00:00
1.Qe8
+- (25.35) Depth: 3 00:00:00
1.Qe8 Kg4 2.g7 Kf3 3.Qh5+ Kf2 4.g8Q Ke1 5.Qg3+ Kf1 6.Qd1#
+- (#6) Depth: 4 00:00:00
1.Qe8 Kg4 2.g7 Kf3 3.Qh5+ Kf2 4.g8Q Ke1 5.Qg3+ Kf1 6.Qd1#
+- (#6) Depth: 5 00:00:00
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 5 00:00:00
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 6 00:00:00 1kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 7 00:00:00 1kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 8 00:00:00 1kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 9 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 10 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 11 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 12 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 13 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 14 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 15 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 16 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 17 00:00:00 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 18 00:00:01 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 19 00:00:01 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 20 00:00:01 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 21 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 22 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 23 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 24 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 25 00:00:01 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 26 00:00:01 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 27 00:00:02 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 28 00:00:02 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 29 00:00:02 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 30 00:00:02 7kN
(, 01.04.2007)
New game,
Analysis by Rybka 2.3.1 32-bit :
1...Bxd5
+- (19.04) Depth: 2 00:00:00
1...Bxd5
+- (22.82) Depth: 3 00:00:00
1...Kd7
+- (21.54) Depth: 3 00:00:00
1...Bxd5
+- (21.89) Depth: 3 00:00:00
1...Bxd5
+- (24.39) Depth: 4 00:00:00
1...Bxd5
+- (23.47) Depth: 4 00:00:00
1...Bxd5 2.Kxd5 Kd8
+- (25.97) Depth: 5 00:00:00
1...Kd8 2.Bxg8
+- (20.80) Depth: 5 00:00:00
1...Kd8 2.Bxg8 Kd7
+- (23.30) Depth: 6 00:00:00 1kN
1...Kd8 2.Bxg8 Ke7
+- (27.33) Depth: 6 00:00:00 3kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7
+- (29.83) Depth: 7 00:00:00 4kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7
+- (30.12) Depth: 7 00:00:00 6kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7 4.Bf7
+- (32.62) Depth: 8 00:00:00 8kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7 4.Bf7
+- (29.53) Depth: 8 00:00:00 9kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7 4.Bf7 Kc6
+- (32.03) Depth: 9 00:00:00 9kN
1...Kd8 2.Bxg8 Ke7 3.Bd5 Kd7 4.g8Q Ke7 5.Bf7
+- (#5) Depth: 9 00:00:00 11kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 9 00:00:00 12kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 10 00:00:00 13kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 11 00:00:00 13kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 12 00:00:00 14kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 13 00:00:00 15kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 14 00:00:00 16kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 15 00:00:00 18kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 16 00:00:00 19kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 17 00:00:00 21kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 18 00:00:00 22kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 19 00:00:00 24kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 20 00:00:00 26kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 21 00:00:00 28kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 22 00:00:00 30kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 23 00:00:00 32kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 24 00:00:01 35kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 25 00:00:01 37kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 26 00:00:01 39kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 27 00:00:01 42kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 28 00:00:01 44kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 29 00:00:01 47kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 30 00:00:01 49kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 31 00:00:01 51kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 32 00:00:01 54kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 33 00:00:01 56kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 34 00:00:01 59kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 35 00:00:01 61kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 36 00:00:01 63kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 37 00:00:01 66kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 38 00:00:01 68kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 39 00:00:01 71kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 40 00:00:01 73kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 41 00:00:01 75kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 42 00:00:01 78kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 43 00:00:02 80kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 44 00:00:02 83kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 45 00:00:02 85kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 46 00:00:02 87kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 47 00:00:02 90kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 48 00:00:02 92kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 49 00:00:02 95kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 50 00:00:02 97kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 51 00:00:02 99kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 52 00:00:02 102kN
(, 01.04.2007)
Re. 32 bit vs 64 bit, this is purely a matter of speed. The non-determinism of the MP version may mask this, but it's true.
Vas
New game,
6Q1/8/6P1/4Bk2/3K4/1p6/1P6/8 w - - 0 1
Analysis by Rybka 2.3.1 32-bit :
1.Qe8
+- (30.21) Depth: 2 00:00:00
1.Qe8
+- (27.71) Depth: 3 00:00:00
1.Qe8
+- (25.35) Depth: 3 00:00:00
1.Qe8 Kg4 2.g7 Kf3 3.Qh5+ Kf2 4.g8Q Ke1 5.Qg3+ Kf1 6.Qd1#
+- (#6) Depth: 4 00:00:00
1.Qe8 Kg4 2.g7 Kf3 3.Qh5+ Kf2 4.g8Q Ke1 5.Qg3+ Kf1 6.Qd1#
+- (#6) Depth: 5 00:00:00
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 5 00:00:00
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 6 00:00:00 1kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 7 00:00:00 1kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 8 00:00:00 1kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 9 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 10 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 11 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 12 00:00:00 2kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 13 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 14 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 15 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 16 00:00:00 3kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 17 00:00:00 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 18 00:00:01 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 19 00:00:01 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 20 00:00:01 4kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 21 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 22 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 23 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 24 00:00:01 5kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 25 00:00:01 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 26 00:00:01 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 27 00:00:02 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 28 00:00:02 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 29 00:00:02 6kN
1.Qf7+ Kg5 2.g7 Kh4 3.g8Q
+- (#4) Depth: 30 00:00:02 7kN
(, 01.04.2007)
New game,
6b1/4k1P1/6P1/3BB3/3K4/1p6/1P6/8 b - - 0 1
Analysis by Rybka 2.3.1 32-bit :
1...Bxd5
+- (19.04) Depth: 2 00:00:00
1...Bxd5
+- (22.82) Depth: 3 00:00:00
1...Kd7
+- (21.54) Depth: 3 00:00:00
1...Bxd5
+- (21.89) Depth: 3 00:00:00
1...Bxd5
+- (24.39) Depth: 4 00:00:00
1...Bxd5
+- (23.47) Depth: 4 00:00:00
1...Bxd5 2.Kxd5 Kd8
+- (25.97) Depth: 5 00:00:00
1...Kd8 2.Bxg8
+- (20.80) Depth: 5 00:00:00
1...Kd8 2.Bxg8 Kd7
+- (23.30) Depth: 6 00:00:00 1kN
1...Kd8 2.Bxg8 Ke7
+- (27.33) Depth: 6 00:00:00 3kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7
+- (29.83) Depth: 7 00:00:00 4kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7
+- (30.12) Depth: 7 00:00:00 6kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7 4.Bf7
+- (32.62) Depth: 8 00:00:00 8kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7 4.Bf7
+- (29.53) Depth: 8 00:00:00 9kN
1...Kd8 2.Bxg8 Ke7 3.Bxb3 Kd7 4.Bf7 Kc6
+- (32.03) Depth: 9 00:00:00 9kN
1...Kd8 2.Bxg8 Ke7 3.Bd5 Kd7 4.g8Q Ke7 5.Bf7
+- (#5) Depth: 9 00:00:00 11kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 9 00:00:00 12kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 10 00:00:00 13kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 11 00:00:00 13kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 12 00:00:00 14kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 13 00:00:00 15kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 14 00:00:00 16kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 15 00:00:00 18kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 16 00:00:00 19kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 17 00:00:00 21kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 18 00:00:00 22kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 19 00:00:00 24kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 20 00:00:00 26kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 21 00:00:00 28kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 22 00:00:00 30kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 23 00:00:00 32kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 24 00:00:01 35kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 25 00:00:01 37kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 26 00:00:01 39kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 27 00:00:01 42kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 28 00:00:01 44kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 29 00:00:01 47kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 30 00:00:01 49kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 31 00:00:01 51kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 32 00:00:01 54kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 33 00:00:01 56kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 34 00:00:01 59kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 35 00:00:01 61kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 36 00:00:01 63kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 37 00:00:01 66kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 38 00:00:01 68kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 39 00:00:01 71kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 40 00:00:01 73kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 41 00:00:01 75kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 42 00:00:01 78kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 43 00:00:02 80kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 44 00:00:02 83kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 45 00:00:02 85kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 46 00:00:02 87kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 47 00:00:02 90kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 48 00:00:02 92kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 49 00:00:02 95kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 50 00:00:02 97kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 51 00:00:02 99kN
1...Kd7 2.Bxg8 Kc6 3.Bd5+ Kb5 4.g8Q Kb6 5.Qb8+ Ka5 6.Bc7+ Ka4 7.Bxb3#
+- (#6) Depth: 52 00:00:02 102kN
(, 01.04.2007)
I made additional tests on another computer: P4 HT, 512MB hash, WinXP 32bit, Shredder 9 interface
The results were pretty similar:
2.1c mp 32-bit = OK
2.1c 32-bit = no move!
2.2 mp 32-bit = OK
2.2 32-bit = no move!
2.2n2 mp 32-bit = OK
Winfinder 2.2 mp 32-bit = no move! (shows mate in 0???)
Winfinder 2.2 32-bit = no move! (shows mate in 0???)
2.3 mp 32-bit = no move!
2.3 32-bit = no move!
2.3 lk mp 32-bit = no move!
2.3 lk 32-bit = no move!
2.3.1 mp 32-bit = no move!
2.3.1 32-bit = no move!
The only exeption was 2.2n2 mp 32-bit, that gives result under win32bit. I was not able to test 64 bit versions on this computer.
I tried the same position under Arena 1.1 too. The problem is there!
The results were pretty similar:
2.1c mp 32-bit = OK
2.1c 32-bit = no move!
2.2 mp 32-bit = OK
2.2 32-bit = no move!
2.2n2 mp 32-bit = OK
Winfinder 2.2 mp 32-bit = no move! (shows mate in 0???)
Winfinder 2.2 32-bit = no move! (shows mate in 0???)
2.3 mp 32-bit = no move!
2.3 32-bit = no move!
2.3 lk mp 32-bit = no move!
2.3 lk 32-bit = no move!
2.3.1 mp 32-bit = no move!
2.3.1 32-bit = no move!
The only exeption was 2.2n2 mp 32-bit, that gives result under win32bit. I was not able to test 64 bit versions on this computer.
I tried the same position under Arena 1.1 too. The problem is there!
Attachment: Rybkapromotion.wmv (1172k)
Vas, as I mentioned in the first post, the Rybka 2.3.1 32-bit was an exeption for the second FEN, so your second sample is irrelevant.
Hi, Vas!
I have already tested the first position on 4 different computers, with different CPUs (C2D, P4HT, P740M and PII), hash sizes (2GB, 512MB, 128MB, 16MB), operating systems (WinXP 64bit, WinXP32bit, Win98), chess interfaces (Shredder9, Arena 1.1), but the error occured on each of those. Only 1 of those computers is overclocked, the others have always run at default clock speed. I run Rybka with default settings; only hash size is manually set. As 2 of my computers had uptime more than 2 month, I tried rebooting, but this did not help at all. The problem appeared to be smallest on good old PII+Win98+Shredder9, where only WinFinder 2.2 failed (again indicating 0 moves to mate!). On P740+WinXP32 all tested Rybkas (all 32bit, non MP versions) failed.
Maybe debugging version might help to reveal what's wrong? Or should I really deliver my comps to some junkyard?
I uploaded a short video, where is possible to see the bug. If I didn't stop the calculation, the line never appears.
===
Please, somebody, test those positions and let me know if you can also see the problem! I think the bug might be unknown, because Vas is denying it or at least he did not say at once that this is the same old promotion bug and is already fixed.
I have already tested the first position on 4 different computers, with different CPUs (C2D, P4HT, P740M and PII), hash sizes (2GB, 512MB, 128MB, 16MB), operating systems (WinXP 64bit, WinXP32bit, Win98), chess interfaces (Shredder9, Arena 1.1), but the error occured on each of those. Only 1 of those computers is overclocked, the others have always run at default clock speed. I run Rybka with default settings; only hash size is manually set. As 2 of my computers had uptime more than 2 month, I tried rebooting, but this did not help at all. The problem appeared to be smallest on good old PII+Win98+Shredder9, where only WinFinder 2.2 failed (again indicating 0 moves to mate!). On P740+WinXP32 all tested Rybkas (all 32bit, non MP versions) failed.
Maybe debugging version might help to reveal what's wrong? Or should I really deliver my comps to some junkyard?
I uploaded a short video, where is possible to see the bug. If I didn't stop the calculation, the line never appears.
===
Please, somebody, test those positions and let me know if you can also see the problem! I think the bug might be unknown, because Vas is denying it or at least he did not say at once that this is the same old promotion bug and is already fixed.
![[de]](/mwf/v12/flags/de.png "Germany")
Hello,
I tested your first position on my computer (P4, xp32) under Arena 1.1, Shredder 10 and Fritz 9 and with various versions of Rybka. I only get the error you described under Shredder 10. Under Arena and Fritz everything works fine, Rybka shows analysis lines and says mate in 4 (only Winfinder 2.2 says mate in 0). I looked into this further and I think I found the reason. It seems that Shredder does not display info lines during the first second of analysis, but Rybka reaches a maximum depth of 59 in about half a second and then stops to send info lines to the GUI. Therefor Shredder shows no analysis output.
I hope this can be of help.
I tested your first position on my computer (P4, xp32) under Arena 1.1, Shredder 10 and Fritz 9 and with various versions of Rybka. I only get the error you described under Shredder 10. Under Arena and Fritz everything works fine, Rybka shows analysis lines and says mate in 4 (only Winfinder 2.2 says mate in 0). I looked into this further and I think I found the reason. It seems that Shredder does not display info lines during the first second of analysis, but Rybka reaches a maximum depth of 59 in about half a second and then stops to send info lines to the GUI. Therefor Shredder shows no analysis output.
I hope this can be of help.
Vielen Dank, Günther!
You have a very good point! For Shredder interface the mystery is solved!
I did under Arena only one test on each computer and it happened to be always the WinFinder (that has certain "mate in 0 moves bug" here!), so I thought that the problem automatically applies also to Arena. And therefore did not expect that the GUI might be the reason.
You have a very good point! For Shredder interface the mystery is solved!
I did under Arena only one test on each computer and it happened to be always the WinFinder (that has certain "mate in 0 moves bug" here!), so I thought that the problem automatically applies also to Arena. And therefore did not expect that the GUI might be the reason.
Indeed, this looks like a GUI problem.
In the Shredder classic GUI, please go to "Extras -> Options" and make sure that "Show All Principal Variations" is checked.
This would be consistent with your observation that your weaker machines have fewer problems. These are the machines where Rybka might need more than 1 second.
If this does not fix your problem, please either post here again or send me an email (vrajlich at yahoo dot com).
Vas
In the Shredder classic GUI, please go to "Extras -> Options" and make sure that "Show All Principal Variations" is checked.
This would be consistent with your observation that your weaker machines have fewer problems. These are the machines where Rybka might need more than 1 second.
If this does not fix your problem, please either post here again or send me an email (vrajlich at yahoo dot com).
Vas
Thanks, Vas!
It seems, that problem is mostly solved (as far as it concerns Shredder).
For WinFinder there still seems to be something odd in this position. The "Mate in 0" should not be normal message there.
2.00 0:00 +319.97 1.Qf8+
2.00 0:00 +M2 1.Qf8+ Kxg6 2.Bf4 Kh5 3.Qf5+ (733) 44
3.00 0:00 +M1 1.Qf8+ Kg5 2.Qf6+ Kg4 3.g7 Kh5 4.Qf3+ Kg5 5.g8Q+ Kh4 6.Qh1+ (803) 48
4.00 0:00 +M0 1.Qf8+ Kg5 2.g7 Kh4 3.g8Q Kh5 4.Qf3+ Kh4 5.Qh1+ (913) 54
best move: Qg8-f8 time: 0:55.859 min n/s: 16 nodes: 913
It seems, that problem is mostly solved (as far as it concerns Shredder).
For WinFinder there still seems to be something odd in this position. The "Mate in 0" should not be normal message there.
2.00 0:00 +319.97 1.Qf8+
2.00 0:00 +M2 1.Qf8+ Kxg6 2.Bf4 Kh5 3.Qf5+ (733) 44
3.00 0:00 +M1 1.Qf8+ Kg5 2.Qf6+ Kg4 3.g7 Kh5 4.Qf3+ Kg5 5.g8Q+ Kh4 6.Qh1+ (803) 48
4.00 0:00 +M0 1.Qf8+ Kg5 2.g7 Kh4 3.g8Q Kh5 4.Qf3+ Kh4 5.Qh1+ (913) 54
best move: Qg8-f8 time: 0:55.859 min n/s: 16 nodes: 913
This is a known bug in the winfinder. The mate depths can't be trusted, although when a mate is reported, you can be sure that there is some sort of a mate.
Vas
Vas
![[it]](/mwf/v12/flags/it.png "Italy")
I notice, in game 2, a contradictory response by Winfinder 2.2. Don't know if it's an already known bug
New game - Rybka 2.2 32-bit
Analysis by Rybka WinFinder 2.2 32-bit:
1. +- (#21): 1...Be6 2.Bxe6 Kxe6 3.g8Q+ Kf5 4.g7 Kg4 5.Qxb3 Kf5 6.Qh3+ Kg5 7.g8Q#
2. +- (#20): 1...Kd7 2.Bxg8 Kc8 3.Bd5 Kd7 4.g8Q Ke7 5.Bf7 Kd7 6.Qe8#
3. +- (#20): 1...Kd8 2.Bxg8 Kc8 3.Bd5 Kd7 4.g8Q Ke7 5.Bf7 Kd7 6.Qe8#
4. +- (#20): 1...Ke8 2.Bxg8 Kd7 3.Bd5 Kd8 4.g8Q+ Ke7 5.Bf7 Kd7 6.Qe8#
5. +- (#19): 1...Bh7 2.gxh7 Kd7 3.g8Q Ke7 4.Qe6+ Kd8 5.h8Q#
6. +- (#18): 1...Bf7 2.gxf7 Kd7 3.g8Q Ke7 4.Qe8#
Funny.
New game - Rybka 2.2 32-bit
6b1/4k1P1/6P1/3BB3/3K4/1p6/1P6/8 b - - 0 1
Analysis by Rybka WinFinder 2.2 32-bit:
1. +- (#21): 1...Be6 2.Bxe6 Kxe6 3.g8Q+ Kf5 4.g7 Kg4 5.Qxb3 Kf5 6.Qh3+ Kg5 7.g8Q#
2. +- (#20): 1...Kd7 2.Bxg8 Kc8 3.Bd5 Kd7 4.g8Q Ke7 5.Bf7 Kd7 6.Qe8#
3. +- (#20): 1...Kd8 2.Bxg8 Kc8 3.Bd5 Kd7 4.g8Q Ke7 5.Bf7 Kd7 6.Qe8#
4. +- (#20): 1...Ke8 2.Bxg8 Kd7 3.Bd5 Kd8 4.g8Q+ Ke7 5.Bf7 Kd7 6.Qe8#
5. +- (#19): 1...Bh7 2.gxh7 Kd7 3.g8Q Ke7 4.Qe6+ Kd8 5.h8Q#
6. +- (#18): 1...Bf7 2.gxf7 Kd7 3.g8Q Ke7 4.Qe8#
Funny.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill