>The alternative would be to select into an empty field and then run the function afterward via a SCAN. Would that be preferable?
>
I would guess that derived table solution would be quicker than select + update using the original function, but you may want to test it to be sure.