Abstract

The way we have come to expect computer systems to behave is that we can simply add a new component to a running system and then this new component will begin to interwork with the running system without interuption of service. Describing and validating such systems of dynamic, reconfigurable components presents a challenge for contemporary methods of formal description. Milner's pi-calculus goes some way towards addressing this issue. In this short paper we show that the pi-calculus is particularly good at describing the behaviour of components of a distributed system. We give a pragmatic introduction to the picalculus and illustrate this conjecture, using an example of clients and servers collaborating on the Web. The formalisation gives us the capability to define distributable components and to formulate properties of systems built from such components. The formalisation is different from, and probably complementary to, object-oriented formulations of such components. We describe h...

Citations