Chain rule for conic derivatives (Q351650)

The classical chain rule is abstracted in the standard form to some one-sided but ``smooth'' situation. Unfortunately, the article lies far beyond the mainstream in generalized differentiation. This is corroborated by the list of obsolete references not reflecting any modern research in differentiation of arbitrary mappings and new uncustomary forms of the chain rule (as presented, e.g., in the recent two volume set by \textit{B. S. Mordukhovich} [Variational analysis and generalized differentiation. I: Basic theory. II: Applications. Grundlehren der Mathematischen Wissenschaften 331/332. Berlin: Springer (2006; Zbl 1100.49002)], not even mentioned in the paper). The term ``conic'' is unhappily used, for it wrongly implies some connection with ordered vector spaces and other modern problems of one-sided approximation.