Can Component/Service-Based Systems Be Proved Correct?

Abstract

Component-oriented and service-oriented approaches have gained a strong enthusiasm in industries and academia with a particular interest for service-oriented approaches. A component is a software entity with given functionalities, made available by a provider, and used to build other application within which it is integrated. The service concept and its use in web-based application development have a huge impact on reuse practices. Accordingly a considerable part of software architectures is influenced; these architectures are moving towards service-oriented architectures. Therefore applications (re)use services that are available elsewhere and many applications interact, without knowing each other, using services available via service servers and their published interfaces and functionalities. Industries propose, through various consortium, languages, technologies and standards. More academic works are also undertaken concerning semantics and formalisation of components and service-based systems. We consider here both streams of works in order to raise research concerns that will help in building quality software. Are there new challenging problems with respect to service-based software construction? Besides, what are the links and the advances compared to distributed systems?Comment: 16 page