Informal Collaboration in Commercial Open Source Development: An Empirical Study of Embedded Linux

The use of Linux in embedded devices has increased enormously in recent years. Software vendors offer Linux distributions for embedded systems, device manufacturers employ Linux in their devices, and component manufacturers increasingly make drivers for Linux available.

Most of the publicly available code for embedded Linux is developed by commercial firms, not by hobbyists. While it is true that also for standard Linux many contributions come from IBM and other large firms, the situations differ. IBM pursues the strategic goal of establishing Linux as a widely used operating system. In contrast, embedded Linux is an integral part of their products for hardware manufacturers, and the core market offering for specialized software firms.

This raises the question if and how the development process of embedded Linux differs from that of other open source software. After all, companies working with embedded Linux have a legitimate interest in protecting their competitive advantage when the code helps to differentiate their market offering. How can this interest be reconciled with the open source culture and the requirements of the General Public License (GPL)? The GPL stipulates that software based on code under the GPL (e.g., Linux) be again licensed under the GPL, and that each receiver of the software is entitled to obtain the source code and is allowed to use, modify, and redistribute it. And if firms voluntarily make code public beyond what is stipulated by the GPL – what are the benefits they derive from doing this, and how important are these?

The present empirical study aims at understanding these issues. Among others, it addresses the following questions: Who contributes to publicly available code for embedded Linux? To what extent are different types of companies involved (software firms, device manufacturers, component manufacturers)? How important are contributions from hobbyists? Do companies make some of their developments related to embedded Linux public? How much? What types of developments? What are the reasons to make code public or to keep it secret? What means are used to keep important code secret? Is there a company policy in place regarding the publication of code? How often do developers interact with one another in public open source projects? In what way? What impact has a developer’s personal attitude towards open source on his behavior? What determines the amount of code that is voluntarily made public?

From a more general point of view, the study addresses the issue of the optimal level of IP protection for innovation. Since the legal means to protect one’s developments in open source software are rather restricted, the question arises why (and when) commercial firms choose to use Linux at all for their embedded systems. Finally, does embedded Linux overall evolve at a good pace, or is there underprovision because of a public good problem?