>Yes, but only once in a big way and that was last century.
>There is no reason for optimizers not to pick the best way for both ways to formulate,
>but OTOH I've been bitten by assuming too much intelligence baked into the products I use ;-)

Ok, thanks