Well, keep it up - I won't drift this any farther afield. I'm sticking with 7 for a while yet. I have taken 2 small apps into 8 for a feeler - both were broke. They were easily (?) fixed and continued to work then in 7 but it bothers me how many other undocumneted bahavior changes there are lurking in there. Hard to admit my code didn't work before - it did.

>I wasn't in the least upset and saw a good opportunity to elaborate a bit more for the originator's benefit (in case it's dismissed as not the same).
>
>To me this is the nastiest known data loss bug and I've made it my mission to work at getting to its root.
>
>An interesting thing about this condition and VFP8's SET TABLEVALIDATE... on my system, but not GarrettF's, no error is raised when buffered and without TRANSACTIONS and TABLEVALIDATE return is ignored. The TABLEVALIDATE has to be checked for not .T. in order to get the prescribed error.
>
>Yep, love users. They sure come up with the darndest things!