The rezzing is a good point of interest..that's something I can test with
I'll keep looking at this..I'm just afraid it will point to a difference in OS or something to do with VMs...
(Someone mentioned that it is 'bad form' to make calls like that in the destructor..and I agree... I may add a handler to flush the out buffer and see
if that helps. If I can find a reproducible trigger, I won't have to make a guess at what might be happening
)