Abstract:
Openness is one qualify which modern CBSs strive to
possess. This work on the evaluation of CBSs was
developed in the context of research work to measure
open systems, and that work forms the basis of this paper.
Openness in large computer based systems is a quality
sought by many, claimed by many, but proven by few.
Openness in many systems with which we have daily
contact is a qualify which makes those systems operate in
a fair and effective manner. For example, we have a
standard (open) electrical wiring system, giving all
customers great flexibility in the choice of appliances.
Openness is seen as a modern panacea for the ills and
shortfalls of large, complex computer based systems, by
offering cost effective upgrade paths, or evolvability.
This paper explores ways in which the relative
openness of large computer based systems may be
measured reliably, repeatably and fairly. The aim of the
project upon which this paper is reporting was to
determine the relative openness between various
architectures, and the (long-term) consequences of
utilising various architectures. Along the way, it has
provided significant evidence of the usefulness of the open
system approach for large computer based systems. This
paper certainly provides a basis upon which to judge
whether or not a system which claims to be open lives up
to the claims of being a cost effective, evolvable complex
computer based system.
The outcome is a measurement system which should be
able to deal with all practical propositions, is based on
solid evidence and provides solid analytical measures
which are expressed in plain language (outlining
consequences). A most important outcome of this paper is
beginning to establish whether or not Openness is of any
use.
