Re: [Bug-gnubg] rollout settings

[Jim Segrave]

On Tue 29 Oct 2002 (11:20 -0000), Ian Shaw wrote:

> Thanks for the clarifications, everyone. The rotation behaviour is
>  as logical as once could expect, at least until the randomisation of
>  the sequence is implemented. I always roll out in multiples of 36
>  anyway, but I'll now be reassured when looking at others results.

If there are no major problems with the rollout patches I just
submitted, I think I'll look at adding randomisation of the order of
the rolls within a rotation so stopping a 1296 trial rollout is less
likely to be way out.

I also want to look at being able to rollout the top n moves of a
match/game/session automatically. If anyone else is working on that,
let me know. No promises on timescales.

-- 
Jim Segrave           address@hidden