>Hi Y'all,
>
>Every now and then I get this quirky behaviour and have not been able to figure out why. Here's the simple version:
>
>Work area one: "real" table, unbuffered
>Work area two: one record view on above parameterized by primary key, buffering set to Table, Optimistic
>
>- most times, updating the view does not move the record pointer; view uses "key field only" to do the update
>
>- every now and then (some forms), after a table update is performed, the correct record is indeed updated but the record pointer is left on the last physical record in the table (the GO BOTTOM record, not even EOF())
>
>* looking at forms that work, there seems to be no difference in buffering and all have "Key fields only" selected etc. Also, the update is not changing an index such that the movement is due to an index change on the "real table".
>
>It's sort of like the SQL statement being generated is scanning the whole table to find more than one record but then stops on the last record (even though that type of behaviour usually results in an EOF() condition).
>
>Anyone seen this behaviour?

I remember having that problem early this year. It was the first time it happened to me. Unfortunately, I don't remember if I ever found the reason and/or how to fix it.