Re: [gnugo-devel] A simple and tricky semeai case

[LEE SP]

> From: "SP LEE" <address@hidden>
> Reply-To: GNU Go development <address@hidden>
> To: address@hidden
> Subject: [gnugo-devel] A simple and tricky semeai case
> Date: Sun, 03 Aug 2003 18:07:07 -0700
>
> I have just made a simple semeai test case to test my modifications, 
> but I found neither tactical reading code nor semeai module was 
> called because one of the group (E6) was considered alive. I have to 
> post it here because it's beyond my current knowledge of gnugo.
>
>    A B C D E F G H J
> 9 O . . . . . . . . 9
> 8 . . . . . . . . . 8
> 7 . . O O X X X X . 7
> 6 . O O X O O O O X 6
> 5 . O . X + . . O X 5
> 4 . O . X . . . O X 4
> 3 . O + X O O O . X 3
> 2 . . O O X X X X . 2     WHITE (O) has captured 0 stones
> 1 . . . . . . . . . 1     BLACK (X) has captured 0 stones
>   A B C D E F G H J
>
> Black to move. Black H3 will kill white. However, black E5 only 
> makes a seki. I guess the owl code may be over-optimistic of its 
> finding.
>
> SP Lee

Sorry I was wrong. Both H3 and E5 make seki. However, if white moves first, 
it can kill black. In my last experiment, maybe something was wrong (I used 
many times loadsgf and white/black commands in gtp mode). Actually the 
semeai module is called after I restart gnugo and reload the sgf file.

SP Lee

_________________________________________________________________
与联机的朋友进行交流，请使用 MSN Messenger:  http://messenger.msn.com/cn