Tuning against kgs:130 and kgs:120. Solves the former. The latter cannot be solved with eye pattern tuning (getting deep, looks lively problem.) I didn't run full regressions, but I doubt anything else changes.
BTW, kgs:270 looks very suspicious to me... --- eyes.db 24 Jan 2004 06:04:56 +0200 1.48 +++ eyes.db 06 Apr 2005 20:51:51 +0300 @@ -4563,6 +4563,26 @@ Pattern 70001 :1122 +Pattern 70002 +# Special corner shape. + +|X*. +|. ..$ ++----- + +:1122 + + +Pattern 70003 +# Special corner shape. + +|XX. +|. ..$ ++----- + +:1112 + + Pattern 70010 @..>X.( @@ -5976,6 +5996,18 @@ Pattern 80000 :1111 + +Pattern 80010 +# Special corner shape. + +|) +|*.. +|. ..$ ++----- + +:1122 + + # .... # Topology: .... # @@ -6360,6 +6392,30 @@ Pattern 90800 ############################### +Pattern 100001 + +|! +|. +|.*. +|.*..$ ++----- + +:1122 + + +Pattern 100002 +# `analyze_eyegraph' believes it is 0012, but I doubt. Must be some +# liberty shortage artifact or something. /pp + +|! +|. +|.X. +|....$ ++----- + +:0112 + + Pattern 100100 @ _______________________________________________ gnugo-devel mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnugo-devel
