> And what happen if somebody concurrently set (fastupdate = on)? > Can we miss conflicts because of that?
I think it'd be better to have that option require AccessExclusive lock, so that it can never be changed concurrently with readers. Seems to me that penalizing every single read to cope with this case would be a bad trade-off.
As Andrey Borodin mentioned, we already do. Sorry for buzz :)