Empirical Software Engineering Research on Free/Libre/Open Source Software

Public available data sources are an important knowledge generator from which researchers can obtain, mostly in a non-intrusive way, data and facts from software projects. We present a methodological approach to the data sources commonly found in libre (free, open source) software projects over the Internet, explain how to extract these data and enhance them and offer some ways of analyzing it from various perspectives. The whole process has been implemented with tools that automatize the process so that an ample amount of analysis from various angles (that range from software maintenance and software evolution to the social structure of the underlying organization in charge of the development) of a huge amount of software projects has been used as case studies. This paper demonstrates that it is possible to build research methodologies that can be applied to a large quantity of software projects and that empirical software engineering studies have not to refer to a limited number of software projects. Although specifically targeted to libre software development, many of the techniques and lessons learned can be generally applied to other types of software environments