CCP was founded in 1987 as a natural combination of concurrent logic programming and constraint logic programming.

Since that time, various successes have been recorded. CCP is the definitive ac count of ,accumulative" concurrency. Together with actors, CCP is one of the major formulations of fine-grained, asynchronous concurrency integrating such phenome na as dynamic, networks and ,scope extrusion" in a natural way. The semantic the ory for determinate CCP is simple and elegant and led to the development of new practically useful combinators for search. Work on indeterminate CCP also led to simple and elegant models, and spurred further work in models for asynchronous languages. Timed CCP calculii with a rich semantic theory have been developed, successfully integrating ideas from synchronous programming. Work has begun on developing proof theories and static analysis techniques for CCP. Concrete ,failure- free" languages have been designed for distributed computing, and advanced com pilation techniques investigated. On the practical side, the powerful and well-devel oped Oz system, combining several language ideas and representing many work- years of development effort, is now available. Toontalk is a powerful visual programming system based on CCP ideas.

Together with these successes are some important questions still to be answered definitively. What are the ,natural" application areas for CCP? What are the kinds of computational systems that are substantially easier to develop or maintain in CCP languages -- the ,path to the sea" for CCP? Development of these ,natural" domains of applicability will be crucial in sustaining work for the next ten years. Should CCP ideas be implemented in the context of separate programming languages, or in the context of modern-day programming languages such as Java? Is the Internet gener ally, and network communities specifically, a context for ,computing for the masses" in which declarative programming ideas will find striking applicability? Can CCP ideas server as the foundation for a software development methodology for real-time and hybrid systems that yields provably correct software? On the language side, what is a coherent notion of types for CCP languages that takes into account the par tial nature of information in CCP? At a conceptual level is there a coherent program ming framework which captures the ideas of CCP, higher-order programming and object-oriented programming?

We invite your participation in a seminar at Dagstuhl to review the state of the field today, and to discuss research plans for the next ten years.