Re: [gnugo-devel] float's in surround.c

[bump]

> I wasn't referring to surround_3_12.1, sorry to make that unclear.
> What I am referring to is the computation of the convex hull:

Yes, it might be good to try to avoid the functions ceil and floor.

Let me explain what I think the next step is. Right now what I'm doing is
adding three patterns which look like:

Pattern SU2

....             run with an ikken tobi
.*..
....
?Oo?

:8,d,shape(3)

....
.*..
....
?Ao?

; is_surrounded(A) && surround_map(A,*)!=1
> add_surround_escape_move(*,A)

It's important that the definition of surrounded is such
that after this move is played, the definition is no longer
satisfied. This is because the ikken tobi is not a connection
and the stone played may live in a different dragon. After 
yesterday's patch, this is OK.

SURROUND_ESCAPE_MOVE is a new move reason that is triggered
by these patterns. The problem will then be how much
strategic effect to allow for such a move reason. There
will be three data: the effective size of the dragon,
its weakness, and the size of the surround area. 

I think the strategic effect should be inversely
proportional to the surround size since a dragon that is
tightly surrounded is under stronger attack than one
that sits in a large open area.

Right now I'm trying to construct a suitable strategic effect.

Dan