Developing a Compositional Calculus of Component Connectors

Project code: C-QuattroResearch group: Coalgebraic models of computation (SEN3.3)Start: 2004-03-01End: 2008-03-01Project coordinator: Farhad ArbabThe C-Quattro (C4) project aims to develop the compositional calculus of component connectors called Reo. Reo is intended as a glue language for orchestrating components in a component-based system. The C4 project has both formal and practical aspects.

Project code: C-Quattro
Research group: Coalgebraic models of computation (SEN3.3)

Start: 2004-03-01
End: 2008-03-01

Project coordinator: Farhad Arbab

The C-Quattro (C4) project aims to develop the compositional calculus of component connectors called Reo. Reo is intended as a glue language for orchestrating components in a component-based system. The C4 project has both formal and practical aspects.

The formal, foundational research is carried out within the C4 project by the PhD student Helle Hansen. This research focuses on coalgebraic descriptions of automata and other models of infinite behaviour, logic-based methods of synthesis and verification of automata, and the relationship between logic and coalgebra. Together with David Costa, Helle Hansen has implemented a tool in Haskell which given a specification of a stream function f in Rutten's bitstream calculus constructs a Mealy machine implementing f. This construction is based on the coalgebraic notion of derivatives of stream functions, a concept similar to the (Brzozowski) derivative of regular languages and expressions.

The practical goals of C4 are to create a working distributed implementation of Reo. Post Doctoral Researcher Dave Clarke is working towards this goal. In the past year, Dave developed (with David Costa) an implementation scheme for Reo, implemented these as the Reolite engine, and applied the engine to examples involving both reconfiguration and transactional processing for web services (with John Sloan), and formulated logical tools for reasoning about Reo.

For the coming year, Helle plans to look at the theory of logic and automata on infinite objects from a coalgebraic perspective. This includes investigating whether coalgebra can provide new insights into certain existing (but complicated) automata constructions.

For the coming year, Dave plans to eliminate problems in the implementation scheme (with David Costa), to continue the implementation, and to develop further techniques for reasoning about connector reconfiguration. In addition, he plans to investigate linguistic approaches to expressing component connectors which promise to be more lightweight than the 100% channel-based approach.

Members

Prof.dr. F. Arbab, Dr. D.G. Clarke, D.F. de Oliveira Costa, M.Sc. H.H. Hansen,
Prof.dr. J.J.M.M. Rutten

Key publications

  • F. Arbab, Reo: A Channel-based Coordination Model for Component Composition. In Mathematical Structures in Computer Science, February 2004.
  • F. Arbab and J. J. M. M. Rutten, A Coinductive Calculus of Component Connectors. In LNCS post-proceedings of WADT'02. Also Technical Report SEN-R0216, CWI, Amsterdam 2002.
  • F. Arbab, Abstract Behavior Types: A Foundation Model for Components and Their Composition. InProc. First International Symposium on Formal Methods for Components and Objects (FMCO '02), LNCS vol. 2852, pp. 33-70. Also Technical Report SEN-R0305, CWI, Amsterdam 2003.
  • F. Arbab, C. Baier, J. J. M. M. Rutten, and M. Sirjani, Modeling component connectors in Reo by Constraint Automata. In Electronic Notes in Theoretical Computer Science, 2004. Also Technical Report SEN-R0304, CWI, Amsterdam 2003.