Description of the Consortium

The INRIA partner intends to continue the work around FunLoft, SchemeBigloo, and non-interference. The initial work on which reactive programming is based has been conducted in the team corresponding to the INRIA partner. The first version of SugarCubes (J-F. Susini thesis) is also issued from this team. The CNAM partner is presently developing SugarCubes, a set of Java classes for reactive programming in Java. The aspects related to dynamicity are particularly important in SugarCubes, as well as those dedicated to distributed programming. The LRI partner develops ReactiveML, built on top of Ocaml. The safety of Ocaml is preserved by ReactiveML. Recent work has proposed means for the introduction of dynamic aspects in ReactiveML. A deep unity exists between the partners INRIA, CNAM, and LRI, coming from the use of the same model of parallelism (the reactive-synchronous parallelism) and of the same implementation techniques (« reactive engine », implemented in SugarCubes, SchemeBigloo, ReactiveML, and FunLoft). However, the researches of the three partners are based on three different programming frameworks: basically C and Scheme for INRIA, Java for CNAM, and ML for LRI. Moreover, LRI develops LucidSynchrone which is a dataflow synchronous language with dynamic aspects that would certainly be worth considering in the project. Experience of CNAM on embedded applications (active RFID) gives a possibility of defining new application domains for the approach proposed by the project.