What happens when the bazaar grows: a comprehensive study on the contemporary Linux kernel development model

• Autor(a) principal: Melissa Shihfan Ribeiro Wen
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Biblioteca Digital de Teses e Dissertações da USP
• Texto Completo: https://doi.org/10.11606/D.45.2021.tde-07092021-041136

The popularity and consolidation of many Free/Libre Open-Source Software (FLOSS) projects in the information technology (IT) market keep industry and academia interested in identifying practices that can be beneficial to the software development process. Two decades ago, a set of practices observed in the Linux kernel development was used to characterize the FLOSS development model as a noisy bazaar. However, since then, the FLOSS ecosystem diversified its forms of development. The Linux kernel project has also undergone notable transformations in its community and development processes toward professionalism and civility. FLOSS projects usually have a community supporting its development and organically producing plentiful information to describe how, when, and why a particular change occurred in the source code or the development flow. Although the existence of several studies on the FLOSS phenomenon and its development, these essential sources of information have been overlooked due to the informality and socio-technical challenges for data collection and analysis. Neglect of these resources may have led some studies to outdated and shallow results regarding FLOSS development practices. Bearing this in mind, we considered the great wealth of open-access materials and the Linux project relevance and protagonism on FLOSS phenomenon to mitigate the distance between what is investigated by academia and what is observed in practice on the development of the Linux kernel. We designed a multi-method investigation to cover academics and practitioners perspectives on the projects socio- technical aspects. We used a multivocal literature review, examining peer-reviewed papers and grey literature, to accurately map the Linux kernel development communitys current characteristics. We included the participant observation on the development community as a third perspective to discuss our findings and nuances involved in community-based development. We also synthesized a set of research strategies to review FLOSS community publications. As a result, this research summarizes the state-of-the-art and state-of-the- practice of the Linux kernels contemporary development model. As an adjoining outcome of this work, we present a combination of research methods that could boost and guide future FLOSS ecosystems research.