RE: [gnugo-devel] nando_3_9.4

[Portela Fernand]

Arend wrote:

> I have looked at it a bit more. I've put up an updated version at
> http://www.uni-bonn.de/~uzsxtn/nando_3_9.4a. (Mostly insignificant
> changes:
> * Merge with CVS, in particular in mark_changed_dragon() (addressed a
>   FIXME there)
> * A few coding-style-match-surrounding type cleanups.
> * Do worry even about loosing a single stone. (I think it's worth it
>   once the problem described below is addressed.)

Thank you. Unfortunately, I'm working in a Windows environment (unfortunate
indeed, but the fastest boxes here all run this OS), and weirdly, the
gnuwin32 port of patch 2.5.4 crashes on this patch. So I copied manually the
important changes from your revised patch in my working directory.

> One problem is in the position below.

I haven't found the time to look seriously at it, sorry, will do very soon.
In any case, I agree that such dragons should be declared alive.


Else, I'm thinking about refinements. I don't like the assupmtion that the
referenced worm for a OWL_ATTACK_MOVE_GAIN move reason must belong to the
actually attacked dragon. It's too restrictive. I first looked into droping
these new move reasons and replacing them by some EITHER_MOVE, but it
doesn't seem to be easy. So I'll probably implement a global array (just
like either_data[]). Then, I'll try to figure how to fill up the worm list
prepared by catalog_goal() with neighboring friendly worms.

/nando