evand comments on Defining the normal computer control problem