Multiprocessing via intercommunicating LISP systems

Multiprocess environments comprising several intercommunicating LISP systems are straightforward to implement due to certain fundamental characteristics of the LISP language. Experiences with four methods of establishing the necessary communications linkages are described. The features of LISP which support experimentation with interprocess communication are identified. Two key characteristics of the language are important in this regard: 1. LISP programs can construct and interpret new code as they run; 2. Structures within LISP systems are accessible by name. The most flexible of the communication methods use only the ordinary LISP input and output functions, supplemented by a small amount of system-dependent code to create communication linkages that can be treated by LISP as file structures.