>Is there any reason that I would want to buffer my local (updatable) views? I can't think of any reason that it would be necessary.

If you want to be able to revert to the user's starting point then you have to buffer either the views or the underlying table. It makes sense to me to buffer the actual data entry point (the view) and issue a Tableupdate() when you are ready to commit the changes to the table. If the user decides to cancel the changes mid-stream then you just issue a Tablerevert() on the view.

Just my $0.02 at 0200. :-)