Browse by author
Lookup NU author(s): Emeritus Professor John Dobson
Full text is not currently available for this publication.
It is becoming increasingly widely recognised that the role and importance of non-functional requirements in the development of complex computer- based systems has hitherto been inadequately appreciated. Specifically, it has been common practice for system developers to concern themselves primarily with functionality and only to consider non-functional requirements, such as timeliness and reliability, when they have got the functionality right. Unsatisfactory experience with many system developments has led to the realisation that non-functional requirements need to be taken into account both early and throughout the system development life cycle. However, in order to achieve this goal, we require a clearer understanding of the nature and role of non-functional requirements. This report presents a set of concepts, models and categorisations intended to facilitate the development of such understanding. The report builds on other work, primarily in the area of the design of distributed systems, to present these models. It addresses aspects of process, product and evaluation. (1) It studies the process because some non-functional requirements relate to or constrain the process and because we need to show how non-functional requirements are transformed and resolved during development. (2) Most non-functional requirements relate to the product, so this is the prime focus of the concepts and models presented. (3) Since it is also important that we can assess the extent to which systems satisfy requirements, the report presents a model of evaluation. We use the models to categorise non-functional requirements, specifically a list of such requirements supplied by ARE for our study, and to articulate reasons why we believe that the functional/non-functional dichotomy is over-simplistic and misleading. We suggest an improved approach to the categorisation of all types of requirements which is based on our models. Finally, we draw conclusions on the status of the models and goals for possible future work.
Author(s): Dobson JE, McDermid JA
Publication type: Report
Publication status: Published
Series Title: Computing Laboratory Technical Report Series
Year: 1991
Report Number: 320
Institution: Computing Laboratory, University of Newcastle upon Tyne
Place Published: Newcastle upon Tyne
