🚀 go-pugleaf

Here is a game b/w sargon on a mac (white) and GNU chess
(black) . What is the story with blacks 23rd move?  Does it not see the
pawn on f4, or perhaps it thinks it has a check?  I can't reproduce the
problem by backing up a few moves, it always prefers 23 .. c8b7 if
given the chance, so it is something pretty subtle and no doubt nearly
impossible to find. There was definitely no mistake made on input and there
was no time problem for black so I would say it is a genuine bug.
The game continued as shown with no mistakes, so I guess both programs
knew where all the pieces were OK.


# xboard game file -- Mon Aug 31 17:59:56 1992
	mwr@koel vs. gnuchessx@localhost
	algebraic
1. b2b3 b8c6
2. b1c3 e7e6
3. g1f3 g8f6
4. e2e4 f8b4
5. c1b2 e8g8
6. e4e5 b4c3
7. b2c3 f6e4
8. c3b2 d7d6
9. d2d3 e4g5
10. f3d2 h7h6
11. b2a3 f7f6
12. e5d6 c7d6
13. f1e2 c6d4
14. e1g1 g8h8
15. f2f4 g5f7
16. e2h5 d4f5
17. d2c4 b7b5
18. d1f3 a8b8
19. c4e3 d8b6
20. a1e1 b6a5
21. e3f5 a5a3
22. f5e7 a3c5
23. f1f2 f7e5

Continues ...

24. f4e5 c8b7
25. e7g6 h8g8
26. f3h3 f8d8
27. e5f6 e6e5
28. f6f7 g8h7
29. e1f1 e5e4
30. g1h1 d8f8
31. h3f5 c5f5
32. f2f5 a7a6
33. h1g1 b7c6
34. f5f4 e4d3
35. c2d3 b8a8
36. f4g4 c6b7
37. g6f8 a8f8
38. h5g6 h7h8
39. f1e1 b7c6
40. g1h1 d6d5
41. g4f4 c6d7
42. e1e7 d7c6
43. e7c7 d5d4
44. c7c6 a6a5
45. c6c8 f8c8
46. f7f8q c8f8
47. f4f8 White wins