Erratum to “Régularité et interpolation”

Abstract

In “Regularite et interpolation” [J. Algebraic Geometry 8 (1999) 471–481] the proof of the equivalence in the definition of (m, b)-regularity is not correct: there is a gap in the proof of (∗∗)′ ⇒ (∗)′ ; and this implication is indeed false (unless b = A) as mentioned in Remark B below. Nevertheless, it does not affect the other results of the article since we only use the definition (∗)′, except at one point where it can be easily repaired as explained in Remark A. The other implications stated are correct (namely (∗)′ ⇒ (∗) ⇒ (∗∗) ⇔ (∗∗)′), they are used in the proof of Corollary 2. Therefore one has to replace Definition 1 with the following,