Is software engineering really engineering?
It is argued that many of the problems that beset software engineering are in fact problems with modelling and a need to provide appropriate science and technology for speeding up the craft to engineering transition.
Abstract
It is argued that many of the problems that beset software engineering are in fact problems with modelling. For example: both reuse and codification require shared and understandable models if they are ever to be effcctivelv utilised. Codification is a vital ingredient of any engineering culture and a lack of suitable, widely understood, tool-based modelling frameworks mitigates strongly against effective codification and the effective use of already codified experience. Effectively capturing context is a crucial issue in a world of abstract artefacts, especially in areas such as estimation and process improvement. The author provides a a personal view of some of what still needs to be undertaken Fundamentally, there is a need to provide appropriate science and technology for speeding up the craft to engineering transition. In addition, a proper understanding of the role of models and modelling in software engineering is crucial. Two further related and important areas of research are the use of quantitative data [software measures] for the building of more powerful predictive models and empirical sofware engineering studies.