Open Source Software Development
Sherae Daniel, University of Pittsburgh (formerly University of Maryland)
Ritu Agarwal, University of Maryland
Katherine Stewart, University of Maryland
David Darcy, Florida International
Ilana Diamant, University of Pittsburgh
Rachel Chung, Carlow
Pratyush Sharma, University of Pittsburgh
Laura Dabbish, Carnegie Mellon University
Ben Collier, Carnegie Mellon University
Jim Herbsleb, Carnegie Mellon University
Marcelo Cataldo, Carnegie Mellon University
Likoebe Maruping, University of Louisville
(click here for latest abstracts) 
Diversity and OSS Projects
with Ritu Agarwal and Katherine Stewart
Diversity is a defining characteristic of global collectives on the Internet. Although substantial evidence suggests that diversity can have profound implications for a variety of outcomes including performance, member engagement, and withdrawal behavior, the effects of diversity have predominantly been investigated in the context of organizational workgroups or virtual teams. In this paper we use a diversity lens to study the success of non-traditional virtual work groups exemplified by open source software (OSS) projects. Building on the diversity literature, we propose that three types of diversity (separation, variety and disparity) influence two critical outcomes for OSS projects: community engagement and market success. We draw on the OSS literature to further suggest that the effects of diversity on market success are moderated by the development stage of the project. We instantiate the operational definitions of the three forms of diversity to the specific and unique context of open source projects. Using archival data from 337 projects hosted on SourceForge, we find that disparity diversity, reflecting variation in participants’ activity-based reputation, is positively associated with community engagement. The impact of separation diversity, measured as diversity in the languages spoken by participants, negatively impacts market success but has an unexpected positive effect on community engagement. Variety diversity, reflected in dispersion in project participant roles, positively influences community engagement and market success. The impact of variety diversity on market success is conditional on the development stage of the project: projects in earlier stages of development benefit more from variety diversity than those in more advanced stages. We discuss how the study’s findings on the effects of OSS project participant diversity advance the literature on antecedents of OSS success and expand our theoretical understanding of diversity, and present the practical implications of the results for managers of distributed collectives.
Role Conflict and OSS Participants
with Likoebe Maruping, Marcelo Cataldo and Jim Herbsleb
Software applications developed within the OSS community have enjoyed tremendous success and for-profit organizations are keen to tap into this significant pool of software development talent (Chesbrough 2003; Feller et al. 2008). Examples companies include Netscape, IBM and Google (Maccormack et al. 2006). These companies seek to benefit from the talent of a global and sometimes voluntary workforce by paying some employees to contribute to OSS projects (Raymond 1999, Shah 2006). This merging of open and traditional software development may cause developer stress based on conflicting OSS community and traditional software development norms. Specifically, developers must balance company intellectual property concerns with the reciprocal and community-based norms that drive OSS development (Stewart and Gosain 2006). When these values are not in sync, contributors that aim to abide by conflicting values may exhibit dysfunctional attitudes. Employee stress with respect to their role can be destructive to organizational outcomes (Maas and Matejka 2009; Aranya and Ferris 1984). This study develops an OSS context specific model that describes the relationship between clashing software development cultures and employee organizational commitment. We leverage the rich OSS literature and the research that focuses on organizational-professional conflict (OPC) to develop hypotheses linking clashing cultures and organizational commitment. These hypotheses are tested using a combination of archival data and a survey of 127 GNOME developers. The findings presented in this paper contribute to OSS literature and offer findings that will enable organizations to more successfully engage OSS communities.
Absorptive Capacity and OSS Projects
with Ritu Agarwal and Katherine Stewart
This study explores antecedents of OSS project performance from a knowledge-focused perspective. In particular, it examines the development and effects of absorptive capacity for an OSS project. In describing how OSS absorptive capacity is developed, this study identifies characteristics and behaviors of project participants that indicate an OSS project’s absorptive capacity. We underscore the importance of the characteristics and behaviors of two different sets of project participants: those in the Internet-based user community (IBUC) and those in the development group. Archival data about 74 OSS projects that use the SourceForge platform are examined to empirically test the hypotheses developed. This study makes several contributions to theory and practice. In exploring the effect of absorptive capacity in an OSS project, this study identifies the nuances of the construct when it is used to understand non-traditional organizations. In particular we suggest some cases when increased knowledge acquisition is associated with decreased performance; for example, increased IBUC relationships with other projects decreases development activity. This study also extends the OSS literature by specifically linking characteristics and behaviors of the IBUC to development activity. For instance, a dynamic IBUC that has strong communication with the development group is associated with improved development activity. These findings inform project managers regarding the participants to target and behaviors to encourage that will lead to superior performance for an OSS project.
OSS Projects and Complexity Evolution
with Katherine Stewart and David Darcy
OSS is a growing phenomenon, representing an increasing amount of the software engineering work that is being done. OSS projects represent new means of collaborating to build, distribute, and support software. Prior work on OSS has tended to focus on either the “big” projects such as Linux, Apache, etc, or on “the rest,” such as the population (or a sample) of SF. Very few have made distinctions among “the rest” (i.e., beyond dead and not dead, or better performing and worse performing). This study provides a finer-grained view of the landscape of OSS development projects as viewed through the lens of software evolution. In doing so, we make three main contributions to the literature. First, we combine statistical analysis of objective longitudinal data on software complexity with qualitative analysis of information from projects’ public websites to elaborate a categorization scheme that differentiates among 6 types of OSS projects. Second, in developing this categorization scheme we add to the work on software evolution by identifying a set of projects in which aspects of complexity (size and structural complexity) move in different directions over time; in particular a set of projects for which structural complexity is decreased as size is increased. The identification of this category of projects is unique in that most prior work has shown positive correlations between size and structural complexity. The third contribution of this study is methodological; we demonstrate the application of a new statistical technique, FDA, which is useful in overcoming challenges in studying software evolution.
This paper explores the application of functional data analysis (FDA) as a means to study the dynamics of software evolution in the open source context. Several challenges in analyzing the data from software projects are discussed, an approach to overcoming those challenges is described, and preliminary results from the analysis of a sample of open source software (OSS) projects are provided. The results demonstrate the utility of FDA for uncovering and categorizing multiple distinct patterns of evolution in the complexity of OSS projects. These results are promising in that they demonstrate some patterns in which the complexity of software decreased as the software grew in size, a particularly novel result. The paper reports preliminary explorations of factors that may be associated with decreasing complexity patterns in these projects. The paper concludes by describing several next steps for this research project as well as some questions for which more sophisticated analytical techniques may be needed.
Developer and User Networks in OSS Projects
with Ilana Diamant
The capability to acquire knowledge is a key determinant of success for organizations that seek to innovate, and a strong social network can facilitate it. Open source software (OSS) project networks in which relationships are defined by developers working on two projects have been shown to facilitate project success. Users also participate in multiple OSS projects, but the effect of networks defined by user relationships has not been examined. We suggest that users can serve as a channel through which knowledge, especially usability- related knowledge, can move between projects. We draw on open source and the broader IS literature and propose distinct hypotheses for the effect of the network characteristics of user- and developer networks on project performance. Specifically, because the users typically have a less extended and involved type of participation focused on usability concerns, the effect of network characteristics defined by user relationships is expected to be different from the effect of the network defined by the developers, whose participation is likely to be focused more on coding. A sample of projects from sourceforge.net will be used to explore the hypotheses developed. In developing and testing these hypotheses, this research is among the first to suggest that the impact of network characteristics on project success differs with the role of the participant who represents the relationship. This distinction has practical implications for managers seeking to create successful OSS projects.
Learning in Open-Source Software (OSS) Development: How Developer Interactions in Culturally Diverse Projects Impact the Acquisition of Collaboration- and Learning Skills
with Ilana Diamant
OSS/FLOSS development projects often span geographic and national boundaries, as developers from different countries have the opportunity to join those projects. A project team’s cultural diversity creates opportunities for learning from others, but may also lead to conflict and inhibit learning. This research examines developers’ learning in OSS/FLOSS projects. We focus on the acquisition of collaboration skills, and skills about the learning process itself, and examine the impact of the team’s cultural composition on a developer’s learning. Archival and survey data are collected from two large-scale Sourceforge projects. This research can contribute to the OSS literature by examining the impact of a team’s cultural composition on learning in OSS projects. Practically, administrators and managers stand to gain insight into the learning benefits of participation in OSS projects, and thus better assess their value as a training ground for global software development.
User Contribution and FLOSS Project Performance:A Study of Cross Project Participation, Knowledge Contribution and Interactivity
with Ilana Diamant, Laura Dabbish begin_of_the_skype_highlighting end_of_the_skype_highlighting and Ben Collier
The ability to acquire new knowledge is an important determinant of organizational performance. For organizations and work groups that seek to innovate, consumers and end-users are a critical source of knowledge, particularly for feedback on products under development. Free/Libre Open Source Software (FLOSS) projects, because of their open participatory design process, may also benefit from user knowledge. Users can share knowledge drawn from interacting with applications developed by multiple projects. Sharing a user with another FLOSS project may increase access to relevant knowledge, but may increase demands on a user’s attention. In this study we examine FLOSS project user knowledge contributions, and find that an increase in the number of connections to other projects through shared users is associated with a higher rate of exploration-oriented user knowledge contributions in the form of feature requests. We find that interaction between users and developers enhances the relationship between contributions and performance. Exploration-oriented contributions impact market interest in the presence of higher levels of user-developer interaction. We also find that exploitation-oriented user knowledge contributions, in the form of bug reports, impact project task closure more in the presence of higher levels of user-developer interaction. Understanding the impact of user ties across projects and the value they bring has theoretical implications for the study of FLOSS development and user participation, and has practical implications for administrators and project managers.
Sustainability of Open-Source Projects: A Longitudinal Study
with Shoba Chengalur-Smith and Anna Sidorova
This paper examines the factors that influence the long-term sustainability of FLOSS projects. A model of project sustainability based on organizational ecology is developed and tested empirically. Data about activity and contribution patterns over the course of five years for 2,772 projects registered with SourceForge is analyzed. Our results suggest that the size of the project’s development base, project age and the size of niche occupied by the project are positively related to the project’s ability to attract user and developer resources. The ability to attract resources is an indicator of the perceived project legitimacy, which in turn is a strong predictor of the project’s future sustainability. Thus a project’s ability to attract developer and user resources is shown to play a mediating role between the demographic (size and age) and ecological (niche) characteristics of the project and its future sustainability. Our results support the applicability of tenets of organizational ecology related to the liability of smallness, liability of newness, and population characteristics (niche size) to the FLOSS development environment. Implications of the results for future research and practice are discussed.
The Impact of Person-Organization Fit on Turnover in Open Source Software Projects
with Pratyush Sharma and Rachel Chung
Turnover in conventional software development is a critical problem. In this paper we argue that turnover can be a problem for Open Source Software Projects as well. Using Schneider's Attraction Selection and Attrition Framework and the notion of Person-Organization fit we hypothesize about the relationship between the developers' fit with OSS projects and turnover. Specifically we argue that the value fit, needs-supplies fit and demands-abilities fit between a developer and a particular project will have a negative association with developer turnover and that the role of the developer (CVS access) in the project acts a moderator. Since for-profit companies are increasingly leveraging open source software development, implications of our findings will be useful for project managers to attract, select and retain talented contributors in the absence of financial compensation.
- Daniel, S.L., Agarwal, R. and Stewart,K.J.,"An Absorptive Capacity Perspective of Open Source Software Development Group Performance," Presented at the International Conference on Information Systems, December 13, Milwaukee, Wisconsin, 2006.
- Stewart, K.J., Darcy, D. P., & Daniel, S. L. “Opportunities and Challenges Applying Functional Data Analysis to the study of Open Source Software,” Statistical Science, 2006.
- Daniel, Sherae L. and Diamant, E. Ilana, "Network Effects in OSS Development: The Impact of Users and Developers on Project Performance" (2008). International Conference Information Systems 2008 Proceedings. Paper 122.
- Daniel, Sherae L. and Chung, R., "How Anonymous Contributions Impact Digital Innovations: The Case of Open-Source Projects" (2009). Academy of Management Conference.