The M-calculus: a Higher-Order Distributed Process Calculus
Abstract
This report presents a new distributed process calculus, called the -calculus. Key insights for the calculus are similar to those laid out by L. Cardelli for its calculus of ambients. Mobile Ambients and other recent distributed process calculi such as the Join calculus or the D-calculus introduce notions of distributed locations or localities, corresponding to a spatial partitioning of computations and embodying different features of distributed computations (e.g. failures, access control, process migration, etc). However these calculi remain unsatisfactory in that they account for a single predefined behavior associated with a locality: in a large distributed system, localities may be of different types and exhibit a wide range of behaviors. This report tries to remedy to this limitation in defining a distributed programming model that allows the explicit programming of locality behavior. More precisely, the -calculus can be understood as a generalization of the Join calculus and of G. Boudol's blue calculus that provides: distributed localities with programmable behavior (cells), higher-order processes, process mobility, and dynamic binding features.
Loading...