>Hi Mark.
>
>I meant to ask about row/table buffering's impact on updating many records (instead I wrote optimistic/pessimistic). I guess I'm really trying to understand why one should use remote views at all. I'd prefer to use one solution.

Since a remote view is just a SQL stored in the DBC along with all the other properties set for the view, I guess it really depends on the preference of the developer. Instead of having all the code needed to create the view in the DBC, it is just as easy to have all this code in a class.

Most of my views (local and remote) are parameterized, and I just find this easier than SPT. Open the view at the beginning of form instanciation with NoData, define the parameter when needed and issue the requery function.

My primary use of SPT is when I need to retrieve data for an ad-hoc lookup table or something else that I will be using on-the-fly as read-only.

As for buffering, I primarily use table buffering on views even if there is only 1 record in the table. That assures me that no updating will occur until my save method is fired even if the record pointer is moved for any reason.