Abstract

This thesis conducts an exploratory study of Open Source Software (OSS) from various perspectives in order to discover and demonstrate fertile research areas in OSS that can benefit from the public archives of OSS projects. It follows a horizontal research method which combines theoretical model building with empirical data analysis. On the theoretical side, it classifies existing quantitative OSS studies and categorizes the public archives. It defines the concept of an OSS project by delineating its four critical components--community, methodology, products, and services. It specifies the roles in OSS communities, examines the speed, cost, and quality of OSS development, and reveals the impacts of programming languages on software projects. Most importantly, it originates a new approach to OSS adoption research which comprises strategic level study of OSS adoption and an assessment framework for OSS projects. A rich set of propositions are formulated for future study. On the empirical side, it analyzes summary statistics of 48,331 OSS projects and more detailed attributes of 1,907 projects which use Python as one of their programming languages. It depicts the portraits of OSS projects in general, and Python projects in specific. Caveats, for pitfalls and weaknesses of OSS projects discovered from the analysis, such as small development teams and competing projects, are announced with suggestions for improvements. Inspiring original ideas into a burgeoning research domain, this thesis contributes to both comprehension and practice of OSS.