A direct path to dependable software

Abstract

What would it take to make software more dependable? Until now, most approaches have been indirect: some practices – processes, tools or techniques – are used that are believed to yield dependable software, and the argument for dependability rests on the extent to which the developers have adhered to them. This article argues instead that developers should produce direct evidence that the software satisfies its dependability claims. The potential advantages of this approach are greater credibility (since the argument is not contingent on the effectiveness of the practices) and reduced cost (since development resources can be focused where they have the most impact).