Keynote speech: Industrializing software development

Abstract

The objective of the talk is to discuss economic, organizational, and technological aspects of software industrialization. While it is impossible to predict exactly when the industrial revolution in software will occur, it is clear that when it happens it will cause a dramatic redistribution of wealth and a decline of the software monopolies. There is the economic reason why software components as an industry (predicted in the late sixties by Doug McIlroy) never materialized: it is the emergence of the software industry, whose very existence is based on unspecified, irregular and extremely complex interfaces. Organizationally, there is no division of labor, a very low level of professionalism, and a reward system that is based on number of features, rather than on the level of reliability, correctness, and security. Finally, technologically we still have to learn to produce comprehensive, well-organized catalogs of highly generic, reliable components with precise time and space performance characteristics. © Springer-Verlag Berlin Heidelberg 2005.