gnugo-devel
[Top][All Lists]
Advanced

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

Re: [gnugo-devel] Regression corrections/comments


From: Gunnar Farneback
Subject: Re: [gnugo-devel] Regression corrections/comments
Date: Thu, 01 Nov 2001 23:16:39 +0100
User-agent: EMH/1.14.1 SEMI/1.14.3 (Ushinoya) FLIM/1.14.2 (Yagi-Nishiguchi) APEL/10.3 Emacs/20.7 (sparc-sun-solaris2.7) (with unibyte mode)

Trevor wrote:
> A few random corrections/comments to the regression suite:
> 
> strategy4:208
> B7 seems OK to me.  Am I missing something?

Then I'm missing it too. B7 looks fine to me.

> >> +#useful in manyfaces:8
> >> +#new pattern 3.1.13 -trevor
> >> +Pattern D833
> >> +
> >> +?O?         vital corner point.
> >> +O.O
> >> +O??
> >> +O.*
> >> +??.
> >> +
> >> +:8,s,value(40)
> >
> >This doesn't look much like a corner and actually rather weird with
> >all the wildcards. Is this as intended?
> Intended to be a corner position - left off bars.
> I'd intended the wildcards, as extraneous B or W stones in those
> positions wouldn't necessarily invalidate the position.

I don't think it's a good idea to allow it to match in a position like

.XX.|
XXO.|
OO.O|
XOX.|
XO..|
.XX.|
----+

This move could be better needed elsewhere (assuming the leftmost O
stone is connected to some better eye potential) and would be more
useful one point to the left anyway. The problem is that at most three
(pattern matched) moves are tested by the owl code so this could cause
an effective move elsewhere never to be tested.

/Gunnar




reply via email to

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