Foundations for Reasoning About (Practical) Implicit Invocation Systems

点击翻译按钮获取中文摘要

A critical issue for large-scale systems design and evolution is the choice of an architectural style that permits the integration of separately-developed components into larger systems. Familiar styles include those based on remote procedure call, shared variables, asynchronous message passing, etc. One increasingly important architectural style for system composition is implicit invocation (II). However, there is no established basis for reasoning about II systems. In particular it is difficult to answer questions like: What will be the effect of announcing a given event? Have enough event bindings been declared? Does a given component announce the sufficient events to permit effective integration? If I add a new component to an existing system, will it break the existing system? The goal of this project is to provide such a basis for reasoning about systems designed using the II architectural style. Two central benefits are expected. The first benefit is a step towards being able to reason fully about systems constructed in this style. The second and probably more important benefit is to replace current ad hoc reasoning approaches used by practitioners who use the II style with a collection of sound ideas that allow better informal reasoning about such systems. ***