>Ok you're right. And so am I. It looks like the form will turn it on if it needs it, but will not turn it off if the table is buffered, and it's own table in the DE is not. IOW, try the test the other way around, and you'll find that the form does indeed change the buffering. Interesting.

I found this while investingating data integrity issues. Sometimes, one table was not updated while the others were. I had to get the testers to write down each form they visited and traced it backwards to one form opening and buffering the tables. The form didn't close the tables. Also the system was written with no subclassing, and all forms using default data session. Hopefully, we can use this to push for a rewrite! We've got two more MVPs joining the team shortly - I'll try and get them on the push!