Nalimov 5 pieces Tablebases (OS X)

Get your specific HIARCS/Junior support questions answered here as well as up-to-the-minute news!

Moderators: Watchman, Mark Uniacke, mrudolf

Post Reply
Rene Pfammatter
Member
Posts: 3
Joined: Fri Aug 03, 2007 8:47 am

Nalimov 5 pieces Tablebases (OS X)

Post by Rene Pfammatter »

I believe there is a small bug in SigmaChess/Hiarcs with the 5-piece Tablebases.
Examples:
3KQ3/8/8/8/8/2n5/3b4/3k4 w - - 0 1
Hiarcs: mate in 25
Shredder Online: mate in 39

3k4/2rp4/8/8/8/8/3Q4/3K4 w - - 0 1
Hiarcs: mate in 25
SO: mate in 51

Hiarcs plays the right moves, but the mate announcements are incorrect (until it reaches mate in 25 in the above examples). With Scid, The Tablebase Files were correctly used, so i think they are fine.
User avatar
Harvey Williamson
Site Admin
Posts: 6079
Joined: Sun Jul 29, 2007 6:57 am
Location: Media City, UK
Contact:

Re: Nalimov 5 pieces Tablebases (OS X)

Post by Harvey Williamson »

Rene Pfammatter wrote:I believe there is a small bug in SigmaChess/Hiarcs with the 5-piece Tablebases.
Examples:
3KQ3/8/8/8/8/2n5/3b4/3k4 w - - 0 1
Hiarcs: mate in 25
Shredder Online: mate in 39

3k4/2rp4/8/8/8/8/3Q4/3K4 w - - 0 1
Hiarcs: mate in 25
SO: mate in 51

Hiarcs plays the right moves, but the mate announcements are incorrect (until it reaches mate in 25 in the above examples). With Scid, The Tablebase Files were correctly used, so i think they are fine.
I dont have a mac version with me to test it at the moment. Can you try it again but switch off swindle in the engine parameters?

If you still have the problem please email mark at Hiarcs support.

Best Wishes,

Harvey
User avatar
Harvey Williamson
Site Admin
Posts: 6079
Joined: Sun Jul 29, 2007 6:57 am
Location: Media City, UK
Contact:

Re: Nalimov 5 pieces Tablebases (OS X)

Post by Harvey Williamson »

Rene Pfammatter wrote:I believe there is a small bug in SigmaChess/Hiarcs with the 5-piece Tablebases.
Examples:
3KQ3/8/8/8/8/2n5/3b4/3k4 w - - 0 1
Hiarcs: mate in 25
Shredder Online: mate in 39

3k4/2rp4/8/8/8/8/3Q4/3K4 w - - 0 1
Hiarcs: mate in 25
SO: mate in 51

Hiarcs plays the right moves, but the mate announcements are incorrect (until it reaches mate in 25 in the above examples). With Scid, The Tablebase Files were correctly used, so i think they are fine.
I dont have a mac version with me to test it at the moment. Can you try it again but switch off swindle in the engine parameters?

If you still have the problem please email mark at Hiarcs support.

Best Wishes,

Harvey
User avatar
Andreas Guettinger
Member
Posts: 27
Joined: Tue Jul 31, 2007 5:07 pm
Location: Bern, Switzerland

Post by Andreas Guettinger »

Seems to be a problem of the GUI, so swindle has nothing to do with it.

Here from Terminal:

Code: Select all

HIARCS by Mark John Uniacke
Copyright (c) 1992-2007 Mark Uniacke, Applied Computer Concepts Ltd. All rights reserved.
uci
id name HIARCS 11.2 SP
id author Mark John Uniacke
option name Ponder type check default true
option name Licensed To type combo default Andreas Guettinger var Andreas Guettinger
option name OwnBook type check default true
option name Book File type string default H11Book2.hcs
option name Book Status type string default Loaded OK
option name Book Positions type combo default 691458 var 691458
option name Book Style type combo default Tournament var Wild var Wide var Varied var Tournament
option name Book Learning type check default true
option name Position Learning type check default true
option name Hash type spin min 1 max 2048 default 16
option name Clear Hash type button
option name Retain Hash type check default true
option name Use Tablebases type check default true
option name Tablebase Depth type spin min 0 max 7 default 2
option name NalimovCache type spin min 1 max 256 default 1
option name NalimovPath type string default /tbs
option name Hiarcs Draw Value type spin min -100 max 100 default 0
option name Playing Style type combo default Normal var Normal var Aggressive var Solid
option name Selectivity type spin min 0 max 7 default 7
option name Forward Pruning type spin min 0 max 15 default 10
option name Threat Depth type spin min 0 max 7 default 5
option name Smart Search type check default true
option name Futility type check default true
option name Steady Search type check default false
option name Combinations type check default false
option name Hyper-Modern Play type check default true
option name Swindle type check default true
option name GUI Time Lag type check default false
option name Verbose UCI type check default false
option name AutoSave PGN type check default false
option name UCI_Elo type spin min 1250 max 2858 default 2270
option name UCI_LimitStrength type check default false
option name UCI_Chess960 type check default false
option name MultiPV type spin min 1 max 127 default 1
option name UCI_EngineAbout type string default HIARCS 11.2 SP is activated. Copyright (c) 2007 Applied Computer Concepts Ltd.  All rights reserved. Visit www.hiarcs.com for updates.
registration checking
registration ok
uciok
setoption name NalimovPath value /TB
position fen 3KQ3/8/8/8/8/2n5/3b4/3k4 w - - 0 1 
go ponder
info depth 1 seldepth 0 time 2 nodes 21 nps 304759 score mate 32 tbhits 22 pv d8d7 


HIARCS by Mark John Uniacke
Copyright (c) 1992-2007 Mark Uniacke, Applied Computer Concepts Ltd. All rights reserved.
uci
id name HIARCS 11.2 SP
id author Mark John Uniacke
option name Ponder type check default true
option name Licensed To type combo default Andreas Guettinger var Andreas Guettinger
option name OwnBook type check default true
option name Book File type string default H11Book2.hcs
option name Book Status type string default Loaded OK
option name Book Positions type combo default 691458 var 691458
option name Book Style type combo default Tournament var Wild var Wide var Varied var Tournament
option name Book Learning type check default true
option name Position Learning type check default true
option name Hash type spin min 1 max 2048 default 16
option name Clear Hash type button
option name Retain Hash type check default true
option name Use Tablebases type check default true
option name Tablebase Depth type spin min 0 max 7 default 2
option name NalimovCache type spin min 1 max 256 default 1
option name NalimovPath type string default /tbs
option name Hiarcs Draw Value type spin min -100 max 100 default 0
option name Playing Style type combo default Normal var Normal var Aggressive var Solid
option name Selectivity type spin min 0 max 7 default 7
option name Forward Pruning type spin min 0 max 15 default 10
option name Threat Depth type spin min 0 max 7 default 5
option name Smart Search type check default true
option name Futility type check default true
option name Steady Search type check default false
option name Combinations type check default false
option name Hyper-Modern Play type check default true
option name Swindle type check default true
option name GUI Time Lag type check default false
option name Verbose UCI type check default false
option name AutoSave PGN type check default false
option name UCI_Elo type spin min 1250 max 2858 default 2270
option name UCI_LimitStrength type check default false
option name UCI_Chess960 type check default false
option name MultiPV type spin min 1 max 127 default 1
option name UCI_EngineAbout type string default HIARCS 11.2 SP is activated. Copyright (c) 2007 Applied Computer Concepts Ltd.  All rights reserved. Visit www.hiarcs.com for updates.
registration checking
registration ok
uciok
setoption name NalimovPath value /TB
position fen 3k4/2rp4/8/8/8/8/3Q4/3K4 w - - 0 1 
go ponder
info depth 1 seldepth 0 time 1 nodes 23 nps 304996 score mate 51 tbhits 24 pv d2a5 
first mate in 32
second mate in 51
regards,
Andy
User avatar
Harvey Williamson
Site Admin
Posts: 6079
Joined: Sun Jul 29, 2007 6:57 am
Location: Media City, UK
Contact:

Post by Harvey Williamson »

Thanks Andy,

I will ask Mark to pass this on to Ole at Sigma Chess.

Best Wishes,

Harvey
Rene Pfammatter
Member
Posts: 3
Joined: Fri Aug 03, 2007 8:47 am

Post by Rene Pfammatter »

Swindle on/off doesn't change it.
Oh, i forgot about the Terminal, shame on me, that would have been a logical test ;-) So the Problem seems to be with SigmaChess, not Hiarcs.
And Andy is right, the first one is a mate in 32, my mistake.

regards
Rene
User avatar
Andreas Guettinger
Member
Posts: 27
Joined: Tue Jul 31, 2007 5:07 pm
Location: Bern, Switzerland

Re: Nalimov 5 pieces Tablebases (OS X)

Post by Andreas Guettinger »

Rene Pfammatter wrote:I believe there is a small bug in SigmaChess/Hiarcs with the 5-piece Tablebases.
Examples:
3KQ3/8/8/8/8/2n5/3b4/3k4 w - - 0 1
Hiarcs: mate in 25
Shredder Online: mate in 39

3k4/2rp4/8/8/8/8/3Q4/3K4 w - - 0 1
Hiarcs: mate in 25
SO: mate in 51

Hiarcs plays the right moves, but the mate announcements are incorrect (until it reaches mate in 25 in the above examples). With Scid, The Tablebase Files were correctly used, so i think they are fine.
Strange, crafty (and Hiarcs started from terminal) say it's mate in 32 moves in the first position.

Crafty v20.14 (1 cpus)

White(1): 3KQ3/8/8/8/8/2n5/3b4/3k4 w - - 0 1
1. Kd7! Nb1! 2. Qg6 Kc1! 3. Qd3 Bc3! 4. Ke6! Bb2 5. Kf5! Nc3 6. Ke5
Nb1+ 7. Kf4 Ba1! 8. Qb3 Bb2! 9. Ke3! Na3! 10. Ke4! Kb1! 11. Kd5!
Nc2 12. Qd3! Ba1 13. Kc4! Kc1! 14. Qd7! Nd4 15. Qb7! Nc2 16. Kd3!
Ne1+! 17. Ke3! Bc3 18. Ke2! Kc2! 19. Qb6 Ng2 20. Qg6+ Kb3! 21. Qxg2!
Kc4! 22. Ke3! Bd4+ 23. Ke4! Bc3! 24. Qc2! Kb4! 25. Kd5 Ba1 26. Qc4+
Ka3! 27. Qa6+ Kb2 28. Qd3! Ka2! 29. Qc2+! Bb2! 30. Kc4! Ka1! 31.
Kb3! Bc1 32. Qxc1#

Crafty v20.14 (1 cpus)

White(1): 3k4/2rp4/8/8/8/8/3Q4/3K4 w - - 0 1
1. Qa5! d6! 2. Qb6 Kd7 3. Ke2! Rc2+ 4. Kf3 Rc3+! 5. Kg4 Rc5! 6. Qa7+
Kd8! 7. Qb7! Re5 8. Qc6! Ke7! 9. Qc7+! Ke6! 10. Qd8! Rc5! 11. Qe8+!
Kf6! 12. Qd7! Rc4+! 13. Kf3! Ke5! 14. Qe8+ Kd5! 15. Qf7+ Kd4 16.
Qf5! Rc3+ 17. Kf4! Rd3! 18. Qc8! Kd5! 19. Qc2! Rd4+ 20. Kf5! Rh4!
21. Qd3+! Kc6! 22. Ke6! Rh6+! 23. Ke7! Rh4! 24. Qxd6+! Kb5! 25. Qd2!
Rc4 26. Qe3! Rc6 27. Kd7! Rb6! 28. Qd4 Ka5! 29. Qc5+! Ka6! 30. Qd5!
Rb7+! 31. Kd6 Rb6+! 32. Kc7! Rb4! 33. Qd3+! Ka5! 34. Kd6! Rh4 35.
Qf5+ Kb4! 36. Kd5! Ka3 37. Qf3+! Kb4! 38. Qe2! Ka3 39. Qe3+! Ka4!
40. Kc5! Rh5+! 41. Kc4! Rh4+! 42. Kc3! Rb4 43. Qe8+ Ka5! 44. Qa8+!
Kb5! 45. Qd5+! Ka4! 46. Qc6+! Ka5! 47. Qc5+! Ka6! 48. Kxb4! Kb7 49.
Kb5 Ka8 50. Kb6! Kb8 51. Qf8#

I think Shredder uses it's own system of tablebases, maybe that explains the difference in the first position.
regards,
Andy
User avatar
Mark Uniacke
Hiarcs Author
Posts: 1458
Joined: Sun Jul 29, 2007 1:32 pm
Location: United Kingdom
Contact:

Post by Mark Uniacke »

Thank you all for the help tracking this problem down!

I have reported it to Ole and invited him to join us on the forum.
Best wishes,
Mark

https://www.hiarcs.com
Post Reply