posix source for GambitFruit4bxSE ?

[govert]

I had it, but lost it.

Does anyone know where I can find it again?

[govert]

Don't bother. I found it

http://www.computerchess.info/tdbb/phpB ... &sk=t&sd=a

[Denis P. Mendoza]

-minor code fixes from material.cpp

if (bq && bn) {
opening -= Queen_Knight_combo;
opening -= Queen_Knight_combo;

to

if (bq && bn) {
opening -= Queen_Knight_combo;
endgame -= Queen_Knight_combo;

[govert]

OK. I am guessing that this bugfix addresses the endgame material calculations?

[Chris Formula]

it deserves a recompile with updated version number to differentiate it from the previous builds.

chris

[Eelco de Groot]

I hope it will be tested for longer time controls, I am very curious if the Rebel reductions combined with Toga's selective search will work well. Thorsten's results at slightly longer timecontrols than Blitz, 40 moves in ten minutes were good so I hope that with the strong Blitz qualities it will also still do well in 40 moves in 40 minutes. It should be possible to port some of Toga's SMP code too to Gambit Fruit.

I hope I will find out where I lost a good deal of my branching factor compared with Toga, Fruit and Gambit Fruit. Some of it is lost in the Singular Extensions, but that can't be the whole problem, some of it is certainly lost in my Root Search now, some goes in doing two checks in quiescence search, Denis build is faster but adding it all up I still am losing a lot in the iterative search and I think it hurts too much in endgames especially. Sorry if this is not really about Gambit Fruit but I just tested Walter Eigenmann's position #3 again and while I am happy that I now do find the solution and the solution is more stable, at the end of the search I have great trouble of finding both a winning score for 1. h4! and there is a gigantic difference in attained plydepth. Some tinkering to do...

[D]8/1p3p2/p7/8/2k5/4P1pP/2P1K1P1/8 w - -

Engine: Blueberry Beta 4 DM70 Build 298 (256 MB)
by F. Letouzey, T. Gaksch, E. de Groot

1/02 0:00 -1.39 1.c3 Kxc3 (5)

1/02 0:00 -0.96 1.Kd1 (7)

1/02 0:00 -0.72 1.Kd2 (13)

1/03 0:00 +0.67 1.h4 Kc3 (51)

2/09 0:00 +2.55 1.h4 Kc3 2.h5 b5 (520)

3/24 0:00 +2.44 1.h4 Kd5 2.h5 Ke6 3.e4 a5 4.h6 Kf6 (13.107)

4/24 0:00 +2.44 1.h4 Kd5 2.h5 Ke6 3.e4 a5 4.h6 Kf6 (14.000)

5/24 0:00 +1.97 1.h4 Kd5 2.h5 Ke6 3.e4 a5 4.h6 Kf6
5.Kf3 (15.574)

6/30 0:00 +0.25 1.h4 Kd5 2.h5 Ke6 3.Kd2 Kf6 4.Kd3 Kg7 (138.608)

7/30 0:00 +0.19 1.h4 Kd5 2.h5 Ke6 3.Kd2 Kf6 4.Kd3 Kg7
5.Ke4 (158.694)

8/39 0:02 -0.49 1.h4 Kd5 2.Kd2 Ke6 3.c4 Kf6 4.e4 Kg6
5.Kc2 a5 6.Kb3 (1.963.239) 965

9/40 0:02 -0.49 1.h4 Kd5 2.Kd3 Ke6 3.Kd4 Kf5 4.Kc5 Kg4
5.Kb6 Kxh4 6.c4 a5 7.Kxa5 Kg5 8.Kb4 (2.395.883) 942

10/40 0:02 -0.49 1.h4 Kd5 2.Kd3 Ke6 3.Kd4 a5 4.Kc3 a4 (2.586.887) 925

11/44 0:04 -0.41 1.h4 Kd5 2.Kd3 Ke5 3.c4 a5 4.h5 Kf6
5.Kd4 a4 6.Kc3 Kg5 7.Kb4 Kxh5 8.Kxa4 Kg5
9.Kb5 Kf5 (4.614.877) 937

12/44 0:06 -0.49 1.h4 Kd5 2.Kd3 Ke5 3.c4 Ke6 4.Kd4 a5
5.Kc5 a4 6.Kb4 Kf5 7.Kxa4 (6.315.669) 965

13/55 0:10 -0.37 1.h4 Kd5 2.Kd3 Ke5 3.c4 Kf5 4.Kd4 b6
5.c5 bxc5+ 6.Kxc5 Kg4 7.Kb6 Kxh4
8.Kxa6 Kg4 9.Kb6 Kg5 10.Kc5 (9.773.209) 949

14/55 0:36 -0.37 1.h4 Kd5 2.Kd3 Ke5 3.c4 Kf5 4.Kd4 b6
5.c5 bxc5+ 6.Kxc5 Kg4 7.Kb6 Kxh4
8.Kxa6 Kg4 9.Kb6 Kg5 10.Kc5 (34.030.081) 931

15/69 21:35 -0.29 1.h4 Kd5 2.Kd3 Ke5 3.c4 b6 4.c5 bxc5
5.Kc4 Kf5 6.Kxc5 a5 7.Kb5 Kg4 8.Kxa5 Kxh4
9.Kb5 f5 10.Ka5 Kh5 (1.142.844.751) 881

16/69 25:57 0.00 1.h4 Kd5 2.c4+ Kd6 3.h5 Ke7 4.c5 f5
5.h6 Kf6 6.Kd2 Kg6 7.Kc3 Kxh6 8.Kb4 Kg7
9.Ka5 Kf8 10.Kb4 Kg7 (1.392.247.522) 894

17/69 108:42 +2.66 1.h4 Kd5 2.c4+ Kd6 3.c5+ Ke6 4.Kd3 f5
5.Kc3 Ke7 6.Kb3 Ke8 7.h5 Kf7 8.Ka4 Kf6
9.Ka5 Kg5 10.Kb6 a5 11.Kxa5 Kxh5
12.Kb6 Kg6 13.Kxb7 Kf6 (5.284.624.243) 810

best move: h3-h4 time: 148:38.156 min n/s: 820.063 CPU 100.0% n/s(1CPU): 820.063 nodes: 7.313.310.000


8/1p3p2/p7/8/2k5/4P1pP/2P1K1P1/8 w - -

Engine: Gambit Fruit 1.0 Beta 4bxSE (256 MB)
by Ryan Benitez, Thomas Gaksch and Fabien

1/02 0:00 -1.21 1.c3 Kxc3 (3)

1/02 0:00 -0.87 1.Kd1 (4)

1/02 0:00 -0.63 1.Kd2 (7)

1/02 0:00 -0.36 1.h4 (10)

2/02 0:00 -0.30 1.h4 Kd5 (23)

3/04 0:00 -0.03 1.h4 Kd5 2.h5 (100)

4/07 0:00 +0.11 1.h4 Kd5 2.h5 Ke5 (180)

5/07 0:00 +0.54++ 1.h4 Kd5 2.h5 Ke5 3.h6 (502)

6/08 0:00 +0.86 1.h4 Kd5 2.h5 Ke5 3.h6 Kf6 (875)

7/08 0:00 +1.09 1.h4 Kd5 2.h5 Ke5 3.h6 Kf6 4.h7 Kg7 (1.642)

8/11 0:00 +0.65-- 1.h4 Kd5 2.h5 Ke5 3.h6 Kf6 4.Kd3 Kg6 (3.089)

9/13 0:00 -0.27-- 1.h4 Kd5 2.h5 Ke5 3.h6 Kf6 4.Kf3 Kg6
5.Kxg3 Kxh6 (7.466)

10/14 0:00 -0.27 1.h4 Kd5 2.h5 Ke5 3.h6 Kf6 4.Kf3 Kg6
5.Kxg3 Kxh6 (11.836)

11/15 0:00 -0.43 1.h4 Kd5 2.Kd3 Ke5 3.Kc4 Kf5 4.Kc5 Kg4
5.Kb6 Kxh4 6.Kxb7 (29.034)

12/16 0:00 -0.42 1.h4 Kd5 2.Kd3 Ke5 3.Kc4 Kf5 4.Kc5 Kg4
5.Kb6 Kxh4 6.c4 a5 7.Kxa5 (47.596)

13/18 0:00 -0.77 1.h4 Kd5 2.Kd3 Ke5 3.h5 b6 4.c4 Kf6
5.c5 bxc5 6.Kc4 Kg5 7.Kxc5 Kxh5 (98.361)

14/22 0:00 -0.66 1.h4 Kd5 2.Kd3 Ke5 3.c4 Kf5 4.Kd4 a5
5.c5 Kg4 6.Kc4 Kxh4 7.Kb5 Kg5 8.Kxa5 (320.749)

15/22 0:00 -0.66 1.h4 Kd5 2.Kd3 Ke5 3.c4 Kf5 4.Kd4 a5
5.c5 Kg4 6.Kc4 Kxh4 7.Kb5 Kg5 8.Kxa5 (558.346)

16/23 0:00 -0.78 1.h4 Kd5 2.Kd3 Ke5 3.c4 Kf5 4.Kc3 Kg4
5.c5 Kxh4 6.Kb4 Kg5 7.Ka5 Kf5 8.Kb6 a5
9.Kxa5 (963.901)

16/26 0:01 -0.54 1.Kd2 b6 2.h4 Kd5 3.c4+ Ke5 4.Kc3 f5
5.c5 bxc5 6.Kc4 f4 7.exf4+ Kxf4 8.h5 Kg5
9.h6 Kxh6 10.Kxc5 (1.441.293) 1040

17/30 0:01 -0.70 1.Kd2 b6 2.h4 Kd5 3.c4+ Ke5 4.Kc3 f5
5.c5 bxc5 6.Kc4 f4 7.exf4+ Kxf4 8.h5 Kg5
9.h6 Kxh6 10.Kxc5 Kg5 (2.105.175) 1069

17/30 0:02 -0.64 1.c3 a5 2.h4 Kd5 3.c4+ Ke4 4.c5 f6
5.h5 Kf5 6.Kd3 Kg5 7.Kc4 Kxh5 8.Kb5 Kg5
9.Kxa5 Kf5 (3.224.332) 1085

18/30 0:04 -0.60 1.c3 a5 2.h4 Kd5 3.c4+ Ke4 4.c5 f6
5.h5 Kf5 6.Kd3 Kg5 7.Kc4 Kxh5 8.Kb5 Kg5
9.Kxa5 Kf5 10.Kb4 (4.553.476) 1115

19/36 0:07 -0.46 1.c3 f5 2.h4 Kd5 3.c4+ Ke5 4.c5 Kf6
5.Kd3 Kg6 6.Kc3 Kh5 7.Kb4 Kxh4 8.Ka5 Kg5
9.Kb6 a5 10.Kxa5 Kf6 (8.583.258) 1147

20/39 0:10 -0.42 1.c3 f5 2.h4 Kd5 3.c4+ Ke5 4.c5 Kf6
5.Kd3 Kg6 6.Kc3 Kh5 7.Kb4 Kxh4 8.Ka5 Kg5
9.Kb6 a5 10.Kxa5 Kf6 11.Kb4 (12.654.263) 1172

20/39 0:12 -0.06++ 1.h4 Kd5 2.c4+ Ke4 3.c5 Kf5 4.Kd2 Kg4
5.Kc3 a5 6.Kc4 a4 7.Kb4 a3 8.Kxa3 Kh5
9.Kb4 f5 10.Kc3 Kxh4 11.Kd3 (15.062.754) 1185

21/39 0:14 +0.74++ 1.h4 Kd5 2.c4+ Ke4 3.c5 Kf5 4.Kd2 Kg4
5.Kc3 Kxh4 6.Kb4 Kg5 7.Ka5 f6 8.Kb6 a5
9.Kxa5 Kf5 10.Kb6 Ke5 11.Kxb7 Kd5 (17.691.037) 1198

22/43 0:20 +2.41++ 1.h4 Kd5 2.c4+ Ke4 3.c5 Kf5 4.Kd2 Kg4
5.Kc3 Kxh4 6.Kb4 Kg5 7.Ka5 f5 8.Kb6 a5
9.Kxa5 Kf6 10.Kb6 Ke6 11.Kxb7 f4
12.exf4 (24.321.941) 1210

23/43 0:23 +2.51 1.h4 Kd5 2.c4+ Ke4 3.c5 Kf5 4.Kd2 Kg4
5.Kc3 Kxh4 6.Kb4 Kg5 7.Ka5 f5 8.Kb6 a5
9.Kxa5 Kf6 10.Kb6 Ke6 11.Kxb7 f4
12.exf4 Kd5 (28.845.283) 1216

24/44 0:37 +3.74++ 1.h4 Kd5 2.c4+ Ke4 3.c5 Kf5 4.Kd2 Kg4
5.Kc3 f6 6.Kb4 f5 7.Ka5 f4 8.exf4 Kxf4
9.Kb6 a5 10.Kxa5 Ke5 11.Kb6 Kd4
12.Kxb7 Kxc5 13.h5 Kd5 (45.173.081) 1204

25/48 0:56 +6.08++ 1.h4 Kd5 2.c4+ Ke4 3.c5 a5 4.Kd2 a4
5.Kc3 a3 6.Kb3 Kxe3 7.h5 Kf2 8.h6 Kxg2
9.h7 Kf1 10.h8Q g2 11.Qa1+ Kf2
12.Qd4+ Kf3 13.Kxa3 f5 14.Kb4 (67.239.825) 1179

26/52 1:18 +6.26 1.h4 Kd5 2.c4+ Ke4 3.c5 a5 4.Kd2 a4
5.Kc3 a3 6.h5 Kf5 7.Kb3 Kg5 8.Kxa3 Kxh5
9.Kb4 Kg5 10.Ka5 f5 11.Kb6 f4
12.exf4+ Kf5 13.Kxb7 Kxf4 14.c6 (93.043.524) 1182

27/52 1:44 +5.45-- 1.h4 Kd5 2.c4+ Ke4 3.c5 a5 4.Kd2 a4
5.Kc3 a3 6.Kb3 f5 7.h5 Ke5 8.Ka2 Kf6
9.Kxa3 f4 10.exf4 Kg7 11.Ka4 Kf8
12.Kb5 Kg7 13.Kb6 Kh6 14.Kxb7 (123.676.401) 1185

28/52 2:16 +5.45 1.h4 Kd5 2.c4+ Ke4 3.c5 a5 4.Kd2 a4
5.Kc3 a3 6.Kb3 a2 7.Kxa2 f5 8.h5 Ke5
9.Kb3 Kf6 10.Ka3 f4 11.exf4 Kg7
12.Kb4 Kh6 13.Kb5 Kh7 14.Kb6 (162.047.780) 1186

29/55 3:35 +6.04++ 1.h4 Kd5 2.c4+ Ke4 3.c5 a5 4.Kd2 a4
5.Kc3 a3 6.Kb3 a2 7.Kxa2 f5 8.h5 Ke5
9.Kb3 Kf6 10.Ka3 Kg7 11.Kb4 Kf6
12.Kb5 Ke6 13.Kb6 Ke5 14.h6 (248.024.144) 1149

30/56 5:58 +6.51++ 1.h4 Kd5 2.c4+ Ke4 3.c5 a5 4.Kd2 a4
5.Kc3 a3 6.Kb3 a2 7.Kxa2 f5 8.h5 Ke5
9.Kb3 Kf6 10.Kb4 Kg5 11.Kb5 Kxh5
12.Kb6 f4 13.exf4 Kg4 14.f5 (399.482.022) 1113

31/78 9:24 +6.51 1.h4 Kd5 2.c4+ Ke4 3.c5 Kf5 4.Kd2 Kg4
5.Kc3 Kxh4 6.Kb4 Kg5 7.Ka5 Kf5 8.Kb6 Ke4
9.Kxb7 Kxe3 10.c6 Kf2 11.c7 Kxg2
12.c8Q a5 13.Qg4 Kh2 14.Qh4+ (612.219.450) 1083

32/78 66:32 +8.74++ 1.h4 Kd5 2.c4+ Kd6 3.c5+ Ke5 4.Kd2 Kf5
5.Kc3 Kg4 6.Kb4 Kxh4 7.Ka5 Kg5 8.Kb6 a5
9.Kxa5 Kf5 10.Kb6 Ke4 11.Kxb7 Kd3
12.c6 Ke2 13.c7 Kf2 14.e4 (3.945.717.508) 988

33/78 178:27 +8.58 1.h4 Kd5 2.c4+ Ke4 3.c5 a5 4.h5 Kf5
5.Kd3 a4 6.Kc3 a3 7.Kb3 a2 8.Kxa2 Kg5
9.Ka3 Kxh5 10.Kb4 Kg5 11.Ka5 Kf5
12.Kb6 Ke4 13.Kxb7 Kd3 14.c6 (10.224.576.130) 954

best move: h3-h4 time: 200:11.422 min n/s: 961.747 CPU 100.0% n/s(1CPU): 961.747 nodes: 11.551.750.000

Gambit Fruit reaches 33 plies in about three hours on my Athlon with 2009 MHz, and Blueberry does about half of that if I ignore the logarithmic scaling a bit, comparing the time to reach 17 plies Gambit Fruit 4bxSE has a roughly 6000 times faster time to depth ...

Eelco