Re: [gnugo-devel] worms from try_move

[Gunnar Farneback]

Arend wrote:
> in effect, you are talking about systematically reevaluating worm
> attributes during owl (life and death) reading (certainly not for all
> trymove()s, as Gunnar has pointed out). This might actually be worth
> trying out, it could gain accuracy in reading, at the cost of a severe
> speed penalty.

But I don't think we should do that by starting to modify the worm
array. The appropriate place to put such information would be in the
local_owl_data struct.

/Gunnar