gnugo-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[gnugo-devel] nando_3_17.2


From: Portela Fernand
Subject: [gnugo-devel] nando_3_17.2
Date: Fri, 31 Jan 2003 00:14:34 +0100

A small piece of code in check_self_atari() verifies that an apparently
bad candidate move (a self-atari) could possibly be the first sacrificial
move of a combination (either a snapback or oiotoshi). This patch just
reactivates this check, since it corresponds exactly to the position in
regression test nando:31.

I haven't run the regressions (box busy running a twogtp match). I will
do asap (I would be surprised if it has any negative effects).

/nando

- check for snapback/oiotoshi reactivated in check_self_atari()

Index: engine/readconnect.c
===================================================================
RCS file: /cvsroot/gnugo/gnugo/engine/readconnect.c,v
retrieving revision 1.43
diff -u -r1.43 readconnect.c
--- engine/readconnect.c        26 Jan 2003 10:30:08 -0000      1.43
+++ engine/readconnect.c        30 Jan 2003 23:04:06 -0000
@@ -3326,7 +3326,7 @@
 static int
 check_self_atari(int pos, int color_to_move)
 {
-#if 0
+#if 1
   int lib;
 #endif
   
@@ -3336,10 +3336,12 @@
   if (is_ko(pos, color_to_move, NULL))
     return 1;
 
-#if 0
+#if 1
   /* FIXME: At some time I added this exceptional case but I can no
    * longer see how it would be useful. It might still be, however, so
    * I leave the code in for a while. /gf
+   *
+   * Code reactivated, see nando:31. /nn
    */
   if (approxlib(pos, color_to_move, 1, &lib) >= 1
       && approxlib(lib, OTHER_COLOR(color_to_move), 3, NULL) <= 2




reply via email to

[Prev in Thread] Current Thread [Next in Thread]