An empirical study of division of labour in free & open source software development: the case of FreeBSD

Author: George Dafermos and Ludo Gorzeman.

More specifically, the presentation reports on the results of an empirical study of division of labour in a large free software project, FreeBSD. The question informing the research is: How is work distributed within a large free/open source software project's development community? To answer that, we analysed FreeBSD's 14 years of development history (1994-2008) as recorded in the project's code repository. To elucidate the division of labour in the FreeBSD project, and its dynamics over time, our analysis tackles, among others, the following sub-questions:

-How many people write code for new functionality?
-How many people report bugs?
-How many people fix them?
-Are these functions carried out by distinct groups of people, that is, do people assume primarily a single role?
-Do large numbers of people participate somewhat equally in these activities, or do a small number of people do most of the work?
-Where did the code contributors work in the code? Was strict code ownership enforced on a file or module level?

The following is a more concise abstract:
-----------------------------------------------------------------------
It has been repeatedly remarked that work in free and open source software projects is not assigned, as contributors undertake the tasks they choose to perform. In parallel, it has been argued that there is no limit on the number of developers that a project can potentially employ, as a consequence of which a tendency of mass participation inheres in these projects. Mass participation, in turn, results in significant cumulative improvement: "given enough eyeballs, all bugs are shallow". The centrality thus attributed to mass participation only reinforces the urgency of posing the following question: How is work distributed within
the development community?: how many people write code for new functionality? How many people report bugs? How many people fix them? Are these functions carried out by distinct groups of people, that is, do people assume primarily a single role? Do large numbers of people participate somewhat equally in these activities, or do a small number of people do most of the work?
-----------------------------------------------------------------------

For an overview of the presentation, see the slides attached. (The slides reflect the state of our research a couple of months ago; but as
the research is still in progress, the slides we intend to present at T-DOSE will be further updated).

Undefined
AttachmentSize
PDF icon dafermos_freebsd_tdose.pdf1.19 MB