The Social, Organizational and Disciplinary Aspects of Quality in Free and Open Source Software Communities

Free and Open Source Software (FOSS) is an innovation model that does
not rely on price or formal hierarchies nor alliance agreements. It may be
described by the term “collective invention.” A review of history reveals a
number of examples of private-collective inventions, but most of these has
not survive past the development of a dominant design, but FOSS has.
In FOSS, commitment to a community occurs because of the sense of
autonomy contributors have, feelings of competence that grow as a result
of successful contributions, and social relatedness. The FOSS production
is a highly successful innovation model, and it survives the emergence of a
dominant design, demonstrating it as a new innovation model. Collective
invention in FOSS survives because of motivational aspects of contributors.
Although FOSS has the unique characteristics of being an information
product, a user innovation, and the result of a highly modular design, these
factors do not fully explain why FOSS produces high quality products. It
is understood that quality assurance techniques, methods and tools are
deployed in FOSS development process to control quality. However, these
practices are not the only source of quality. For example, additional factors
that may explain this phenomenon are conditions that foster pro-social
intrinsic motivation. This dissertation asks how do social, organizational
and disciplinary factors contribute to maintaining software quality in FOSS
Communities?
I show that quality in FOSS communities is achieved when the environment
facilitates some social, organizational and disciplinary enablers and desired
features. I identified three enablers and two desired features. Enablers are
qualities or capabilities that contribute to quality in FOSS. Desired features
are intended capabilities, when achieved they created a desired effect which is
maintaining quality. The enablers are (1) personal motivation for quality, (2) governance for quality and (3) the ability to improve. The desired features
are (1) active commercial participation and (2) retention of participants to
sustain quality.
This is a mixed methods study. Mixed methods research is a methodology
for conducting research that involves collecting, analysing and integrating
quantitative (e.g., surveys) and qualitative (e.g., field observations, interviews)
research. I conducted 82 interviews with FOSS contributors and maintainers.
I carried out a Participatory Action Research project in the he Robot Operating
System (ROS) community. I also conducted a survey with participants
(N=387) from 15 FOSS communities.
Software quality is a difficult attribute to achieve. Software is produced with
bugs for more than 60 years now. New solutions to prevent bugs continue to be
developed in both research and commercial contexts. However, these solutions
tend to focus on the technical aspects of software development, while, software
development processes continue to produce bugs. This is, perhaps, caused
by neglecting of other aspects of software development processes (i.e. social,
organizational and disciplinary). It is time to broden the research attention
to the non-technical aspects of quality. This dissertation demonstrates that
quality has social, organizational and disciplinary dimensions that should be
acknowledged, nurtured and studied further.
Software quality is achieved by technical and non-technical instruments.
This implies that managing and implementing software quality necessitate also
managing and nurturing non-technical mechanisms. For example, passion
for developing software should be acknowledged, nurtured and rewarded.
Another example, software development projects should embrace quality and
pitch it as a fundamental believe rather than just merely a checklist.
Keywords: Software Quality, Free and Open Source Software, FOSS Communities,
FOSS Quality.