Gift from Pandora's box: the software crisis

Abstract

The thesis is an exploration in the history of software development. Its aim is to understand how and why software has developed the way that it has.

The thesis singles out major themes in the development of software. It begins by analysing the early separation of hardware and software and the development of the first high level languages, focusing particularly on the attempt to establish an international standard language and the conflict that arose between the proponents of ALGOL and the supporters of FORTRAN. The issues and themes which emerge at this stage are traced through an analysis of the development in the 1960s of third generation computers, and particularly the dramatic history of the IBM 360. The problems of software development during the 1960s created an awareness of how important and difficult programming was. Software was recognised not just as an appendage of hardware, but as a force in its own right. This was reflected in the commodification of software and In a growing sense of a "software crisis". This feeling of crisis gave rise to conscious attempts to provide solid theoretical foundations for the development of programming. Two main approaches emerged, structured programming and software engineering, and the tensions between these two approaches can be traced back to the early days of software development.

It is argued that the patterns and tensions to be found throughout the whole development of software are not accidental: they arise not simply from academic controversies but from the very nature of software and from the social circumstances of its production and use. It is concluded that there is no easy solution to the software crisis.