Routes to Sustainable Software: Transitioning to Peer Production

Abstract

A key challenge of science policy is to achieve sustained benefit from scientific grant making. As software has become more important to the practice of science, policy makers and scientists have become more concerned about a perceived lack of sustained benefit from software produced by grant funded projects. Policy makers and scientists have sought to learn from the successes of open source software development and have encouraged scientific software projects to draw on “peer production” as a route to sustainable software projects. In this paper we examine grant funded projects to build two taxonomies through qualitative analysis. The first is a taxonomy of organizational configurations: ways in which projects organize to build software. The second is a taxonomy of organizational transitions: modes of organizational change observed around the software produced by these projects. We use these taxonomies to identify three routes to peer production and explain both why peer production is rare and why some routes are more traveled than others. We did not observe grant-funded projects changing themselves to become peer production communities during or after the grant period. Instead we describe alternative routes to peer production via already existing projects. We conclude with implications for science policy.