Re: [gnugo-devel] another semeai patch

[Paul Pogonyshev]

SP Lee wrote:
> I have revised this patch again. Now the modification goes inside the
> function estimate_lunch_eye_value. I think it's reasonable to assume the
> max_eye a lunch of size 2 can make is 1. I'm not sure what was behind the
> original idea of max_eye = 2.
>
> $ cvs diff -u owl.c
> Index: owl.c
> ===================================================================
> RCS file: /cvsroot/gnugo/gnugo/engine/owl.c,v
> retrieving revision 1.188
> diff -u -r1.188 owl.c
> --- owl.c       14 Jan 2004 22:21:27 -0000      1.188
> +++ owl.c       20 Jan 2004 04:10:55 -0000
> @@ -5344,7 +5344,7 @@
>      else {
>        *min = 0;
>        *probable = 1;
> -      *max = 2;
> +      *max = 1;
>      }
>    }
>    else if (size == 1) {
>
> The regression breakage is not much.
>
> nngs:490 FAIL; correct J18 got G18. Although J18 is better, now gnugo
> chooses G18 because it also can owl_attack H18 (before this patch G18 can't
> do that). There is no obvious error in the reasoning.
>
> trevorc:1080 PASS
> semeai:58 PASS
>
> strategy4:167 FAIL correct D4 got E5. The E5 is indeed a bad move, but it
> seems to be not directly linked to the patch above.

I tried this modification before, in a first version of paul_5_3.2
(http://mail.gnu.org/archive/html/gnugo-devel/2003-11/msg00012.html).
First version tried to handle lunches that provide escaping in a
completely different way that lunches which (may) give eyes, but
later Gunnar and me decided that there was nothing bad about the
current approach.  The modification above just turned up, because
i modified estimate_lunch_eyevalue().  Later, when the patch logic
changed, i simply didn't think of trying the change alone.

So, i'm trying to say that this change seems perfectly logical to me,
and i'm for it even though the breakage is not very impressive.  I
would actually be for accepting it even if it gave more fails than
passes.

Paul