Comments on “model uncertainty in structural reliability” by Ove Ditlevsen