Uncovering the relationship between OSS user support networks and OSS popularity

a r t i c l e i n f o Keywords: OSS user support Structural embeddedness Junctional embeddedness In-degree centralization Betweenness centralization OSS popularity The open source model of software development has become an attractive alternative to the traditional proprietary approach. However, the incomplete understanding of the phenomenon has continued to prompt researchers to investigate factors that could increase the use and popularity of open source software (OSS). While a key antecedent for OSS use highlighted in the previous literature is the software quality, we propose that effective online user support is also necessary to increase its popularity. As an understudied area, this paper seeks to understand the role of online user support networks in facilitating OSS use. Based on the network embeddedness theory, it suggests that properties of the user support network i.e., variation in structural and junctional embeddedness, measured as the in-degree and betweenness centralizations respectively, would affect OSS popularity in terms of the number of active users and downloads of the software. Testing on a sample of 176 OSS projects from Sourceforge.net, we showed that a negative quadratic relationship exists between the variation in structural embeddedness of the OSS user support network and the software popularity. Further, as hypothesized, the variation in junctional embeddedness was found to positively impact the OSS popularity. Theoretical and practical implications of the study are discussed. Open source software (OSS) is having a substantial impact on software and its development processes with a large number of software products now containing open source components. OSS is unique in its production, modification, and distribution. A key characteristic of OSS is that it is often developed in a public and collaborative manner. Also, owners typically make their source code available and impose special licenses that allow users to access, change, and sometimes even distribute the software [33,54]. However, the high failure rate of OSS projects [12] has motivated substantial previous research that focused on understanding how to improve the quality of OSS to increase the possibility of success. For instance, previous studies have examined the effects of the technical aspects of the OSS (e.g. [19,32]), the role of learning tools (e.g. [25]), the empowerment (e.g. [27]) and work effectiveness of OSS developers (e.g. [43,48]) on OSS development and quality. Nevertheless, even if an OSS project is completed by generating a stable working version, it may not be able to acquire a critical mass of …