Topology and evolution of the open source software community

Abstract

The Open Source Software (OSS) development movement is a classic example of a social network. It is also a prototype of a complex evolving network. After two years of collecting developer and project information from SourceForge, we have sucient data to determine the dynamic and structural mechanisms that govern the evolution and topology of this complex system. First, we analyzed the empirical data we received from SourceForge in order to ascertain the statistics of the topolog- ical information of the OSS developer social network. By inspecting the network's evolution over time, we were able to gain knowledge about the appearance of the complex network and characteristics of the system. Secondly, we generated a model to depict the evolution of this network. Finally, using Java/Swarm, we simulated the evolution of the OSS developer network using our model to compare against the real data. In simulation, we used several di erent models to replicate the empirical data: random network theory, scale-free network, scale-free network with constant tness and scale-free networks with dynamic tness. From these simulation itera- tions, we were able to verify and validate our model for this network and to increase our understanding of the OSS developer collaboration network.