6. Distributed Constraint Solving for Functional Logic Programming

　The purpose of this research is to improve the usability and to widen the scope of the prototype of a software system for distributed constraint solving in a functional-logic programming language environment. The system is based on an implementation of "deterministic lazy conditional narrowing" in the computer algebra system Mathematica augmented by facilities for expressing systems of (linear, polynomial, differential, and partial differential) equations over the complex numbers. In the course of the execution of the program, the interpreter collects these constraints and forwards them to appropriate external solvers that can handle them. Parallelization annotations in the program guide the distribution and scheduling of the constraint sets.

　We will brushup the system by developing "electronic tutorials" that describe its various aspects in the form of application examples that are embedded in Mathematica notebooks (browsable hypermedia documents with executable contents). In the same way, the internal structure of the system will be documented by "literal programs". In order to improve the applicability of the system, we will integrate more constraint solvers, improve the efficiency of the system with respect to concurrent execution, and extend the calculus to handle more general kinds of constraints and to incorporate some higher-order features.