| Summary: | The Linda model provides an alternative paradigm of coordination for open, distributed systems, where agents may communicate anonymously via shared data spaces called tuple spaces. Although it promotes flexibility and openness with its time-, space- and name-uncoupling properties, it raises issues with regard to the access control and protection, which may be vital in certain applications. In this paper, we discuss two conventional access control
mechanisms-access control lists and capabilities-which
can be incorporated in the Linda·model, and their
suitability and implications on the flexibility of the model.
|
|---|