Semantische benadering van component-gebaseerde software
01 / 2001 - onbekend
Over the past decade, software systems have become very large and often distributed among several computing platforms. Recent technological developments have given to software also the possibility to migrate between distributed locations. Complexity of software systems grows continuously, and the present situation requires a better understanding of the large grain of those systems based on a solid mathematical foundation. Recently, trends have emerged for software development through the planned integration of pre-existing software components, that is, reusable off-the-shelf basic software building blocks. The objective of this research proposal is to develop a formal framework for components necessary for moving the actual focus from syntactical issues to semantical issues. The hard task of the project is to prove that the field of semantics of programming languages is rich enough to model a software engineering problem by integrating several semantical concepts (like mobility, dynamicity, and security), and at the same time mature enough to organize the results in a form, usable and understandable without loss of formality. The work will benefit of the active research of the members of the ALP group in Leiden on semantics of concurrent and distributed languages, coordination languages, software architecture, and design methodologies for object-oriented systems. During the first year a mathematical model will be proposed for component-based systems, and it will be investigated whether it can be made optimal with respect to a notion of observation. Collaboration is foreseen with Dr. Arbab (Centrum voor Wiskunde en Informatica, CWI) and Dr. F. de Boer (Utrecht University). A visit to Bologna University is planned for a better understanding of security issues within component-based systems. In the second year, several description techniques will be proposed and formally justified against the formal model proposed. These techniques will include logically based descriptions of the components as well as more intuitive and graphical representations of the whole system. During this year, a visit is planned to the University of Munich in order to compare and contrast the candidate approach with the stream-based approach of Prof. Broy and his collaborators. Finally, in the third year a study will be started on how the existing methodologies for the design of object-oriented systems can be extended to component-based systems. The plan is to lay a foundation for a mathematically justifiable methodology based on stepwise refinement. Collaboration in this area is planned with members of the Åbo University in Finland.