>> Furthermore, it could be started and stopped, whereas the RV-based version tried to bring down the entire result set when trying to do an update (even if opened NODATA).
>
>Really? I have never heard of this.
>
Like I said, something may not have been set right, but frankly I was very surprised myself.

Meanwhile back at the ranch, you mentioned in another post about modifying the RV when the table structure changes on the back end. This turned out to be a major problem for us in that we were designing, developing and fielding the app pretty much simultaneously. Obviously not the best way to do things but we didn't have much choice. Table maintenance became a hot issue, as we could not rely on the users to get the latest version on a regular basis. By eliminating RV's on the client side, and making changes only on the server side, we minimized the impact of a table change as much as possible.

This is probably not an issue in three+ tier development (or, perhaps not as much an issue) but it was one more reason for us to be wary of RV's. Eventually we also added version checking and update on startup which helped as well, but in the beginning it was a real nightmare.