Re: [gnugo-devel] New owl attack pattern

[SP LEE]

> On Tue, 9 Sep 2003, SP LEE wrote:
>
> > > you might even make that
> > >
> > > ; (lib(A) == 3 || lib(B) == 2 || oplay_attack(*,d,B) &&
proper_eye(c)
> > >
> > > since it seems entirely possible that shortening the liberty could
> > change
> > > the tactical status if there is a tactical semeai involved.  I
don't
> > know
> > > if it would be worth the time to check or not though.
> > >
> > > Evan
> > >
> >
> > I don't quite understand this constraint. Can you give an example?
That
> > is, lib (A) > 3 AND lib (B) > 2 and * is still a good move. Thanks.
> >
> > SP Lee
>
> Sure:
>
> O to play.
>
> OOOOOO.
> O....O.
> OXXXXO.
> OX..XO.
> OXOOXO.
> OOXXOOO
> O.X.X.O
> OOX*XOO
> OOOX.O.
> OOOOOO.
>
> A bit contrived, but it works.  There is a tactical semeai above for
the
> life of the group, and winning it is insufficient to get two eyes.
> However, if O throws in at *, then it costs O a liberty to keep the
eye,
> and the group dies.  Like I said, it seems kinda rare.
>
> Evan Daniel
>

That's interesting. I think it's worth to let gg take more time to think
in this situation. Although it seems to be rare, but I hope gg can make
the situation if it knows the trick. I'll send the patch again after run
the regression tests.

SP Lee