Community Structure in OSS Projects

The structure and organization of people who work on any complex software project can in large part affect the quality and design. Unlike traditional software, Open Source Software (OSS) Projects have no mandated organizational structure. Rather the community is a self organizing and very dynamic entity. Due to the wealth of recorded information regarding open source projects, the organization of the community and changes to it can be observed and analyzed. We believe that subcommunities form within large and complex OSS projects such as the Apache1 webserver due to the number of volunteers and the various subsystems within the codebase. Evaluating whether these subcommunities exist and examining their behavior can help us better understand how self organizing software communities work. In addition, this information may shed light on how OSS projects are able to produce high quality, complex software. We have mined the activities in the Apache development mailing list and software repository. From this data we have been able to reconstruct ∗Some of the data gathering and processing work was performed prior to this course for other research. We would therefore like to acknowledge the help of Prem Devanbu and Alex Gourley. http://httpd.apache.org the social network of volunteers who have contributed to the Apache project. Using recent advances in complex network theory, we have extracted and evaluated subgroups from within the social network and observed their changes over time. Our results show that there are indeed a number of subgroups within the Apache social network and that there is a correlation between the age of the project and “strength” of these subcommunities. We present our findings in this paper.