Understanding the Role of Core Developers in Open Source Development

Abstract

Most Open Source Software projects fail, while only very few succeed. In this study, we examine the factors that may influence the success or failure of OSS projects. We particularly focus on OSS core developer roles. We separate core developers from other developers in the community and empirically examine their different roles. Based on our analysis of a data set comprising 300 open source projects, we demonstrate that both core-developer leadership and project-advocating activities are crucial in the development of OSS projects. This research also offers insight into the sustainability of OSS development and provides advice to practitioners in the OSS community on how to make their projects successful.

The attraction of contributors in free and open source software projects

“…Each group of users and developers contributes a unique set of complementary resources to FOSP: users provide inputs including bug reports, suggestions of new features, and translation of documentation; and developers implement new features, fix bugs, and deal with sponsors. These roles are reflected in the ways the success of FOSP has been measured (Crowston et al, 2005;Long, 2006), including developers' contribution to source code modularity (Shaikh and Cornford, 2003), number of lines of code generated (Mockus et al, 2000), velocity of closing bugs (Herbsleb and Mockus, 2003), and the number of downloads (Crowston et al, 2004;Krishnamurthy, 2002;Grewal et al, 2006). Raja and Tretter (2006), Crowston and Scozzi (2002), and Comino et al (2007) viewed success as the ability of a project to advance through development phases (e.g., from alpha to beta, and from beta to stable).…”

A Multi-repository Approach to Study the Topology of Open Source Bugs Communities: Implications for Software and Code Quality

“…Many researchers contributed to FOSS body of knowledge by investigating the relative time it takes to fix bugs [11], [12], characterizing bugs according to types of errors [13], classifying defect-prone files [14], helping us understand the role of developers in the bug fixing process [15], proposing ways of coping cope with problems associated with opening up bug repositories [16], and investigating the structure and the coordination practices adopted by development teams during the bug-fixing process. However, while each project is unique, data from one or two projects will most likely not give us a comprehensive picture of the complex bug reporting and fixing process.…”

“…However, while each project is unique, data from one or two projects will most likely not give us a comprehensive picture of the complex bug reporting and fixing process. A review of the research literature on FOSS bugs (Table I) shows that researchers, with the exception of [15], study one or two, at most, nine projects [17]. Thus, a relatively large sample of projects may yield an added dimension that would have been difficult to observe from a small number of projects.…”

“…Eclipse [9] Bugzilla XML Predict the time to fix a bug. 300 projects [15] SourceForge bug and issue tracker MySQL queries paths performed to obtain the metadata required for our analysis. The primary data comes from the FLOSSMetrics project database.…”

A Survey on Mining Software Repositories

“…Google code is useful for studying patterns of code, because it is possible to search codes using several conditions like package, language, class, function, and licenses and to identify codes of files for the various projects. These deployment logs or source codes from multiple source code repositories could be analyzed together with other data sources such as bug tracking systems at the same time [135].…”