5& -8 10 44306 c8g4 g1f3 c6d4 e2f2 g4f3 g2f3
5& -1 25 114646 e7b4 e1c1 c8g4 g1f3 b4c3 b2c3 g4f3 g2f3
5. -59 33 158853 e7b4 e1c1 c8g4 g1f3 b4c3 b2c3 g4f3 g2f3
The first column shows the ply that Gnuchess 4.0 is searching. The
second column shows the score. The third is time taken and the fourth
is the no. of positions seen. The rest are the best moves found at this
ply.
Notice that although the second line and the last line shows the same
move order, their score is different! I strongly suspect that the bug
is in the printing of the score, rather than a bug in the calculation of
the score, but I haven't understood the source code enough to justify my
guess.
Of course Gnuchess 3.1 does not exhibit this discrepancy.
PS. Is this considered a bug report?
Chua.