Coordination applications of Linda

Abstract

While the limelight has been focussed on the use of Linda to build efficient parallel programs, we have been steadily building a portfolio of successes in other areas of coordinated computing. It's time to make more of these efforts. They will begin to crowd center stage as open Linda systems—Linda systems that support access to the same tuple space from different executables—become more widely available, spawning a creative frenzy of coordinated programming.

From the perspective of the Linda project, these efforts emphasize the general utility of the Linda model. Few other approaches having some promise as tools for parallel computation maintain their attractiveness as tools for other sorts of distributed ensembles.

From a larger perspective, it is important to emphasize the full panoply of coordination applications. Arguably, significant future developments in large-scale systems will more often than not be members of this group. Current (over) emphasis on parallel computation risks stunting the development of the tools necessary to meet the needs of these next generation systems.

This research is supported by National Science Foundation grant CCR-8657615, by the Air Force Office of Scientific Research under grant number AFOSR-91-0098, and by NASA under NASA Training Grant NGT-50719.

This article was processed using the II_EX macro package with LMAMULT style