I wrote: > Actually it's quite easy to disable amalgamation over four-liberty > strings altogether, see the appended patch. It does solve this semeai > mistake as well as testcases semeai:48 and semeai:147. Probably it > does bad things elsewhere though.
Amazingly enough it doesn't seem to. Regression results are endgame:931 PASS R16 [R16] endgame:960 PASS J10 [J10] semeai:48 PASS 1 0 D8 [1 0 (D8|F8|G12|G10)] semeai:147 PASS 1 1 B15 [1 1 (B15|C15|D15|E15)] gifu05:1110 PASS L5 [L5] 5 PASS Total nodes: 1703011843 3335518 12414098 (+0.022% +0.051% +0.027%) Of these semeai:48 and semeai:147 are properly solved, endgame:931 is random, and endgame:960 and gifu05:1110 are both a bit lucky but good. Additionally it properly solves testcase 16 of STS-RV_0 and testcases 32 and 49 of STS-RV_1. This patch has by far the best impact to size ratio I've seen in a long time. :-) /Gunnar _______________________________________________ gnugo-devel mailing list gnugo-devel@gnu.org http://lists.gnu.org/mailman/listinfo/gnugo-devel
