Abstract
A model is developed to study how the code architecture affects open source software (OSS) development. The model incorporates the resource heterogeneity and diverse motivations of various groups of programmers as well as the strategic interactions among them. We argue that the major advantage brought by a modular architecture of OSS code base is that it reduces both the cognitive cost and the coordination cost associated with OSS development, thus allowing programmers more easily to locate, manage, and contribute to the code base. We show that in OSS development, while modular architecture can potentially increase code contribution, it does not necessarily reduce free-riding; in fact it may well increase free-riding due to the strategic interactions among the programmers. We further empirically test the predictions using the SourceForge OSS development data, and the results confirm our theoretical predictions. The findings bear important theoretical as well as practical implications and provide guidelines for practitioners of OSS development and the collective innovation in general.
Recommended Citation
Peng, Gang; Geng, Xianjun; and lin, lihui, "CODE ARCHITECTURE AND OPEN SOURCE SOFTWARE DEVELOPMENT" (2011). ICIS 2011 Proceedings. 1.
https://aisel.aisnet.org/icis2011/proceedings/economicvalueIS/1
CODE ARCHITECTURE AND OPEN SOURCE SOFTWARE DEVELOPMENT
A model is developed to study how the code architecture affects open source software (OSS) development. The model incorporates the resource heterogeneity and diverse motivations of various groups of programmers as well as the strategic interactions among them. We argue that the major advantage brought by a modular architecture of OSS code base is that it reduces both the cognitive cost and the coordination cost associated with OSS development, thus allowing programmers more easily to locate, manage, and contribute to the code base. We show that in OSS development, while modular architecture can potentially increase code contribution, it does not necessarily reduce free-riding; in fact it may well increase free-riding due to the strategic interactions among the programmers. We further empirically test the predictions using the SourceForge OSS development data, and the results confirm our theoretical predictions. The findings bear important theoretical as well as practical implications and provide guidelines for practitioners of OSS development and the collective innovation in general.