Integrating agents and objects to develop distributed AI systems

The paper presents a language for the development of multi agent systems that allows a user to write programs with a small effort thanks to a predefined agent model and a set of specialized primitives for agent management. Moreover the user can develop agent parts at a low level, taking advantage of all the potentialities of C++, Common Lisp and Java object oriented languages, and reusing a large part of the C, C++, Lisp, Common Lisp and Java software available on the Internet.