Successful open source software projects often require a steady supply of self motivated software developers. However, little work has been done from a relational/network perspective to study the factors that drive the developers to participate in OSS projects. In this paper, we investigate the participation dynamics in a social network, particularly in an online open source community called Ohloh. Through a REST-based API, we collected information about 11,530 open source software projects involving 94,330 developers. Using social network analysis and statistical analysis methods, we examine a set of social and technical factors in the Ohloh dataset, which we define as the determinants that significantly influence the developers' participation choices. We found that the determinants include (1)homophily in programming language, (2)project mutual acquaintance, and (3)project age. In addition, our research findings provide the possibility of predicting developers' participation choices based on the discovered determinants, and therefore can have important implications for OSS project management and in designing social network enabled recommendation systems.