Book or paper about functioning of large open source communities?

Question

I would like to know what characterizes large, successful open source communities. Specifically, I am looking for serious research (book/academical paper or similar), which describes common characteristics of successful large open source communities, but also projects that seem to function well without such characteristics.

Some of the tings I expect to read about is:

• How many committers are there usually?
• Are there any other levels of hierarchy apart from committers and others?
• Are they all meritocracies?
• Are the communities slowly corrupted, leading to forking and new healthy communities arising around the same source code?

Answer 1

Martin Krafft's research came to the top of the mind. He was (or is still) studying Debian processes of development, collaboration and such with regards to how tools and techniques enter and leave the processes.

Answer 2

I studied The Linux Kernel Janitor Project when writing my master thesis, and part of what I did was to compare what describes members of that project compared to other open source projects in general as well as how such projects are organized. Several of the papers I read and used answers your questions, but since it is some years since I wrote it, I do not have details of them fresh in memory. But I reccommend that you get a copy of the thesis and look into the community analysis part of it. The thesis is titled Analysis and description of an open source janitor project and is still availiable online. It was later reworked a little and published as a book with the title Someone actually doing systematic non-functional software maintenance.

Answer 3

Although it's a relatively small sample size, Eric S. Raymond's The Cathedral & the Bazaar: Musings on Linux and Open Source by an Accidental Revolutionary discusses the Linux kernel development process and the fetchmail project. The book contains several essays, each of which can be found freely online, in several languages - The Cathedral and the Bazaar, Homesteading the Noosphere, The Magic Cauldron, and Revenge of the Hackers.

The Cathedral and the Bazaar is more technical, focusing on the requirements, design, implementation, and testing of the Linux kernel and Raymond's experiences with the fetchmail project. Homesteading the Noosphere is about more social aspects of open-source development, including project ownership and contributor motivation. The Magic Cauldron is more about the economics of open source development. Revenge of the Hackers is about more recent developments in the open source community.

The sample size is relatively small, but these works provide some insight into open source software development communities.