Re: [gnugo-devel] new and revised regressions

[Arend Bayer]

Gunnar wrote:

> Some new test cases and a few revisions of old ones. I'm open for
> suggestions of changes to the nngs3:1090 revision.

> +# White must defend against a combination attack. M8 works but is only
> +# a temporary measure until the M6 liberty is filled. Otherwise there
> +# are a number of effective moves which only differ with respect to
> +# the number and size of ko threats left behind.
> +# R7 looks like the most solid alternative, with N9 as second choice.
> +# We may want to add more moves as acceptable.
>  loadsgf games/nngs/gnugo-3.2-merlin-200205071828.sgf 224
>  1090 reg_genmove white
> -#? [!S6]*
> +#? [R7|N9|M8]*

I don't see the point of this test at all. All defense moves GNU Go
comes up do work, and only differ in the ko threats they leave back.
(All of them require one more defense move when the liberty at M6 gets
filled.)

Arend