Poliglot and List 512

[Igor Gorelikov]

Geschrieben von:/Posted by: Igor Gorelikov at 30 March 2004 15:52:38:

List 512 + Poliglot has problems in multi-game matches.
The first game Delfi 4.4 vs List 512 (plus Poliglot) is ended with a draw by repetition.
Apparently it was claimed by Poliglot:
584300 (second: 1/2-1/2 {Draw by repetition}
But after the second game started List continue to think on the first game and then exits.
From winboard.debug:
=====================
595780 (second: 19 +15 0 76
595780 (second: 20 +15 0 80
597040 (second: file "xboard.cpp", line 991, assertion "move_is_legal_debug(RootMove,Board)" failed
Fatal Error: Error: second chess program (polyglot.exe List_512.ini) exited unexpectedly
GameEnds(0, (null), 2)

Igor

[Fabien Letouzey]

Geschrieben von:/Posted by: Fabien Letouzey at 30 March 2004 16:00:56:
Als Antwort auf:/In reply to: Poliglot and List 512 geschrieben von:/posted by: Igor Gorelikov at 30 March 2004 15:52:38:

List 512 + Poliglot has problems in multi-game matches.
The first game Delfi 4.4 vs List 512 (plus Poliglot) is ended with a draw by repetition.
Apparently it was claimed by Poliglot:
584300 (second: 1/2-1/2 {Draw by repetition}
But after the second game started List continue to think on the first game and then exits.
From winboard.debug:
=====================
595780 (second: 19 +15 0 76
595780 (second: 20 +15 0 80
597040 (second: file "xboard.cpp", line 991, assertion "move_is_legal_debug(RootMove,Board)" failed
Fatal Error: Error: second chess program (polyglot.exe List_512.ini) exited unexpectedly
GameEnds(0, (null), 2)

Igor

Igor,
Can you post the last few lines of PolyGlot's log file? PolyGlot complains that a root move reported by the engine is illegal. I would like to see what this case looks like.
Apparently this is a minor issue (RootMove is the move currently beeing searched by the engine, not the best move so far), but it could also point to a bug in PolyGlot.
Fabien.

[Igor Gorelikov]

Geschrieben von:/Posted by: Igor Gorelikov at 30 March 2004 16:15:21:
Als Antwort auf:/In reply to: Re: Poliglot and List 512 geschrieben von:/posted by: Fabien Letouzey at 30 March 2004 16:00:56:

Igor,
Can you post the last few lines of PolyGlot's log file? PolyGlot complains that a root move reported by the engine is illegal. I would like to see what this case looks like.

Sorry, the Poliglot log was off. But since List continues to play the first game I assume that it sends the move from the first game.
Igor

[Fabien Letouzey]

Geschrieben von:/Posted by: Fabien Letouzey at 30 March 2004 16:20:24:
Als Antwort auf:/In reply to: Re: Poliglot and List 512 geschrieben von:/posted by: Igor Gorelikov at 30 March 2004 16:15:21:

Igor,
Can you post the last few lines of PolyGlot's log file? PolyGlot complains that a root move reported by the engine is illegal. I would like to see what this case looks like.

Sorry, the Poliglot log was off. But since List continues to play the first game I assume that it sends the move from the first game.
Igor

Logging is now much faster than before in PolyGlot.
I suggest you leave it on.
The problem is that I don't clear logfiles between sessions, so they can grow quite large.
In this case, it is vital that I have PolyGlot's log to see what is really going on. Having the engine log would be a plus.
If you described the problem accurately, it should be easy to reproduce.
Fabien.

[Igor Gorelikov]

Geschrieben von:/Posted by: Igor Gorelikov at 30 March 2004 17:24:18:
Als Antwort auf:/In reply to: Re: Poliglot and List 512 geschrieben von:/posted by: Fabien Letouzey at 30 March 2004 16:20:24:

I set the next game between two UCI engines (ie two Poliglots involved).
It's ended again with exiting List.
[Event "Computer chess game"]
[Site "GORELIKOV"]
[Date "2004.03.30"]
[Round "1"]
[White "SOS 4 for Arena"]
[Black "List 512"]
[Result "*"]
[TimeControl "60+3"]
1. e4 Nc6 2. Nf3 e5 3. Be2 Nf6 4. d3 Be7 5. c4 Bc5 6. O-O d6 7. Nc3 O-O 8.
Be3 Bxe3 9. fxe3 Bd7 10. Nd5 Nxd5 11. cxd5 Ne7 12. Qb3 c6 13. d4 exd4 14.
dxc6 Bxc6 15. Nxd4 Ng6 16. Bc4 Ne5 17. Bd5 Rc8 18. Rad1 Qc7 19. Rf4 a5 20.
Rdf1 Bxd5 21. exd5 b6 22. Qa3 Qd7 23. e4 Rfe8 24. Nf5 Rc5 25. Qb3 b5 26.
Qa3 b4 27. Qg3 f6 28. R4f2 Kh8 29. h3 Rc4 30. Qe3 Qc7 31. Kh1 Rc8 32. a3 b3
33. Qxb3 Rxe4 34. Qd1 Qc5 35. Qb1 Ra4 36. Re2 Rc4 37. Kh2 Re8 38. Kh1 a4
39. b4 axb3 40. Qxb3 Rc1 41. Ree1 Rxe1 42. Rxe1 g6 43. Ng3 Qa5 44. Rc1 Qd2
45. Rd1 Qf4 46. Rf1 Qh4 47. Rc1 Nd7 48. Rc7 Re1+ 49. Kh2 Qd4 50. Rc8+ Nf8
51. Rxf8+ Kg7 52. Nf5+ gxf5 53. Qg3+ Kxf8 54. Qxe1 Qxd5 55. a4 Kf7 56. a5
Qb5 57. Kg3 d5 58. Qd2 Qb3+ 59. Kf2 Qc4 60. g3 d4 61. Qe2 d3 62. Qh5+ Kg7
63. Qd1 Qd4+ 64. Kf3 Qe4+ 65. Kf2 h5 66. Qf3 Qc4 67. Qe3 Qa2+ 68. Kf1 Qd5
69. Qe7+ Kg6 70. Qe8+ Kg5 71. Qe3+ f4 72. Qxf4+ Kg6 73. g4 Qxa5 74. gxh5+
Qxh5 75. Qg3+ Qg5 76. Qxd3+ f5 77. Kf2 Qf4+ 78. Qf3 Qe4 79. Qxe4
*
From winboard.debug:
====================
538000 (second: 31 +15 0 4930 fxe4
538000 (second: engine_step(): not a move "f5e4 ponder"
Fatal Error: Error: second chess program (polyglot.exe List_512.ini) exited unexpectedly
From poliglot's log (for List 512)
==================================
( ENGINE info score cp 15 depth 31 nodes 4930 time 0 pv f5e4
) XBOARD 31 +15 0 4930 fxe4
( ENGINE bestmove f5e4 ponder
engine_step(): not a move "f5e4 ponder"

Igor

[Fabien Letouzey]

Geschrieben von:/Posted by: Fabien Letouzey at 30 March 2004 17:35:13:
Als Antwort auf:/In reply to: Poliglot and List 512 (next game) geschrieben von:/posted by: Igor Gorelikov at 30 March 2004 17:24:18:

I set the next game between two UCI engines (ie two Poliglots involved).
It's ended again with exiting List.
From winboard.debug:
====================
538000 (second: 31 +15 0 4930 fxe4
538000 (second: engine_step(): not a move "f5e4 ponder"
Fatal Error: Error: second chess program (polyglot.exe List_512.ini) exited unexpectedly
From poliglot's log (for List 512)
==================================
( ENGINE info score cp 15 depth 31 nodes 4930 time 0 pv f5e4
) XBOARD 31 +15 0 4930 fxe4
( ENGINE bestmove f5e4 ponder
engine_step(): not a move "f5e4 ponder"

Thanks!
It looks like a bug in List, but I can be wrong.
Engines can send "bestmove (m)" or "bestmove (m1) ponder (m2)".
It seems like List sent "bestmove (m) ponder", which is neither.
Can you confirm with List's log file?
Fabien.