Re: [gnugo-devel] Owl persistent cache

[Daniel Bump]

> > answer at once. Unfortunately, in LOOSE situations, we would have to go
> > through the move generation to produce the moves tested, which is the
> > most expensive part and which requires all the other expensive work of
> > updating the owl data.
> 
> You have a good point here, but remember that most sitations, at least
> open ones, are WINs.  This is because GNU Go doesn't yet try chancy
> invasions, but extends from safe positions.  The losses appear mostly
> in closed positions.  Maybe we should only try this if we have a win
> to start with.

After Gunnar's patch, the active areas look pretty reasonable,
so  I don't think it's a good idea to run the owl code twice.

> Is the autolevel feature tuneable?  It would be pity if GNU Go uses
> settings for one machine when playing on another.  Good input to such
> a tuning function would be reading nodes per second and owl nodes per
> second (if such a measure would mean anything).

There are actually two autolevels, one in GNU Go and the other
in gnugoclient. In the games on NNGS, the levels are set by
gnugoclient. The autolevel in GNU Go definitely needs work but
isn't relevant here.

Dan