[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[gnugo-devel] Small break-in code tuning
From: |
nando |
Subject: |
[gnugo-devel] Small break-in code tuning |
Date: |
Sat, 15 May 2004 15:08:34 +0200 |
Hi,
I wanted to study the break-in code (that portion of the code is very new
to me), so I tried to solve the problem case I've submitted recently. The
solution appeared to be very simple. The regression breakage is just one
PASS (nando:32) and no FAILs. And apparently no noticeable impact on the
engine performance.
nando
- small break-in tuning
Index: readconnect.c
===================================================================
RCS file: /cvsroot/gnugo/gnugo/engine/readconnect.c,v
retrieving revision 1.77
diff -u -r1.77 readconnect.c
--- readconnect.c 29 Apr 2004 23:59:21 -0000 1.77
+++ readconnect.c 15 May 2004 12:50:11 -0000
@@ -2998,7 +2998,11 @@
}
}
- if (tried_moves == 0 && distance < 1.0) {
+ /* Because of a couple differences between the break-in and the
+ * connection reading code, we can't afford to be as optimistic
+ * as in recursive_connect2() here. See nando:32
+ */
+ if (tried_moves == 0 && distance < 0.89) {
SGFTRACE(NO_MOVE, WIN, "no move, probably connected");
#if USE_HASHTABLE_NG
READ_RETURN_HASH_NG(BREAK_IN, str, depth - stackp,
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [gnugo-devel] Small break-in code tuning,
nando <=