Optimal approximation: an addendum