Indirect naming in distributed programming languages

  • Authors:
  • D. D. Jung;E. Sibert

  • Affiliations:
  • IBM, 1701 North St., Endicott, NY;Syracuse University, School of Computer & Information Science, Syracuse, NY

  • Venue:
  • ACM SIGPLAN Notices
  • Year:
  • 1989

Quantified Score

Hi-index 0.01

Visualization

Abstract

Direct naming of processes in message communication is not desirable in distributed programming, because it requires a priori knowledge of sending and receiving process names. This not only complicates the construction of library or utility programs, but also is not suitable for a dynamic process creation environment. Indirect naming of processes in distributed message communication can be achieved either by introducing channels (ports) or by using process variables. This paper discusses how a new distributed programming language, COPL, achieves indirect naming effectively by using process variables without introducing shared objects, such as channels, in a distributed environment where dynamic process acitivation and selection of communicating partners are important.