From: Woody
Sent: Thursday, July 26, 2001 1:55 PM

> I assumed that B would get a deadlock because the record changed while he
> was in a transaction. If that's not the case, then it's my mistake.

I think this is impossible. B get deadlock just because A locked the record
and B couldn't update it.

> Regardless, as long as A doesn't actually change anything directly on the
> server and his changes are only client side, there is no problem until he
> actually posts the changes, right?

Right.

Regards from Brazil,
Dufloth.