The author compares calculus of communicating systems (CCS) with communicating sequential processes (CSP), noting the ways in which they are similar and how they differ. He describes operational semantics, bisimulation, traces, and when processes are and are not equivalent. Features such as nondeterministic choice, recursion, and (of course) communication are looked at. All in all, if someone wants a short introduction to CCS, CSP, or both, or if they know one and want to learn the other quickly, this would be a reasonable paper to read.

However, the purpose of the paper is not obvious. The author seems to want to build a broad framework within which these and other models of concurrency can be placed; algebra of communicating processes (ACP) and Petri nets are mentioned a few times. The nature of such a framework is not clearly outlined, though, and anyone already familiar with CCS and CSP will not find much that is new. Above all, there is the question of why we should care about this in the late 1990s. Both Milner and Hoare have gone on to other things (the pi calculus and the duration calculus), because they found their earlier algebras difficult to extend to handle such things as dynamic process creation and real time constraints. This paper needs a conclusion to indicate what is new and interesting about it.