Re: [gnugo-devel] Bug in 3.0.0

[Daniel Bump]

Gunnar wrote:

> > I'm replaying the game with 3.0.0 in case it is a persistent caching
> > problem.
> 
> Only a single genmove is executed in the sequence of gtp commands
> leading to the crash. The persistent cache just doesn't come into
> play. (Gnugoclient starts a brand new GNU Go process each time it
> plays a new game.)

The fact that a new process is started with each game seems to me
irrelevant.

Unless gnugoclient starts a new process or clears the cache every move I
don't see any reason the persistent cache might not be involved
in the original crash. 

Am I missing something? The replay is up to move 70.

Dan