Abstract. In this paper we present a short, straightforward and self-contained derivation of the Boggs-Tolle-Wang characterization of those quasi-Newton methods for equality-constrained optimization which produce iterates which are g-superlinearly convergent.