No label defined (Q3976059)

[For the entire collection see Zbl 0731.00025.] This paper presents an original approach to the treatment of fairness in the context of non-interleaving concurrency. This is on itself already a little surprising, since fairness is mostly held to be a problem concerning interleaving. (If all processes involved are executed truly concurrently there appears to be no reason for considering fairness.) Fairness is treated by considering notions of maximality of traces in the sense of Mazurkiewicz, giving a causality-based representation of finite and infinite concurrent behaviours. With respect to the notions of maximality concerned not all infinite traces are maximal. Maximal traces, however, correspond to fair executions. My personal view on this (which may be mistaken) is that the author uses a non-interleaving type of semantics in a technical sense to solve some problem of interleaving models of concurrency, viz. fairness, almost by coincidence. This view is corroborated by the fact that the author notes that there are still some mismatches with other approaches to fairness in the literature, particularly when more refined forms such as weak fairness are considered. Nevertheless, the idea expounded in the paper is interesting and it deserves to be investigated how far it can be stretched.