I probably ought to read threads more fully before jumping in ...

>Ken-
>
>>I denormalize in a limited amount - primarily for performance reasons on
>
>Your point is not opposition to mine, but is a different one altogether.