On the lattices of exact and weakly exact structures (Q2082619)

The paper under review introduces the notion of weakly exact structure on an additive category, by relaxing the definition of exact structure due to Quillen. Recall that if \(\mathcal A\) is an additive category, then an exact structure consists of a class of admissible exact sequences of the form \(X\stackrel{i}\rightarrowtail Y\stackrel{d}\twoheadrightarrow Z,\) such that \(i\) is the kernel of \(d\) and \(d\) is the cokernel of \(i\). We call admissible monic, respectively admissible epic morphisms \(i\), respectively \(d\) appearing in an admissible short exact sequence as above. Moreover the classes of admissible monic and epic morphisms have to satisfy some axioms, one of them being the closure under composition. The generalization to weakly exact sequences is obtained by removing this last axiom. Note that an exact structure comes together with an Ext-bifunctor, and that an additive category \(\mathcal A\) admits a unique maximal exact structure \({\mathcal E}_{\max}\). It is therefore natural to look at sub-bifunctors of the ext functor corresponding to the maximal exact structure on \( \mathcal A\). Perhaps one of the main motivation of the study of weakly exact structures is the following: It turns out that sub-bifunctors of the maximal Ext-functor correspond bijectively to weakly exact structures on \(\mathcal A\); actually it is shown that this correspondence is even a lattice isomorphism. Recall that extriangulated categories is a common generalization for categories endowed with an exact and a triangulated structure. In the same vain as above, the paper introduces and studies weakly extriangulated categories.