Bridging the Gap between Agile and Free Software Approaches

Abstract

Agile sprints are short events where a small team collocates in order to work on particular aspects of the overall project for a short period of time. Sprinting is a process that has been observed also in Free Software projects: these two paradigms, sharing common principles and values have shown several commonalities of practice. This article evaluates the impact of sprinting on a Free Software project through the analysis of code repository logs: sprints from two Free Software projects (Plone and KDE PIM) are assessed and two hypotheses are formulated: do sprints increase productivity? Are Free Software projects more productive after sprints compared with before? The primary contribution of this article is to show how sprinting creates a large increase in productivity both during the event, and immediately after the event itself: this argues for more in-depth studies focussing on the nature of sprinting.

Reassessing Brooks’ Law for the Free Software Community

“…KDE was specifically chosen for this research as it has previously been shown to be a project with more than 10 years of development, a very large and active community [2] producing good quality code [1]. It is therefore reasonable to interpret KDE as a successful project.…”

“…Each of these cohorts has brought to attention one (or several) aspects of this approach [26], criticized [10] or advocated it [22], and, at times, compared it with established approaches [1,30].…”

“…More recently, researchers have started analyzing the fundamentals differences between the established software engineering approaches and the FLOSS development style, comparing it both with the traditional development style (i.e., composed of the phases of requirements, design, implementation, testing and maintenance [23]), and finding differences with the development cycles ( [24,6]), and finding similarities with the Agile paradigm [17,1].…”

“…As a further and one of the most serious challenges that the FLOSS approach still has to face is the issue of trust among end-users: specifically, on one side, the internal product and process attributes of the FLOSS have been analyzed to establish its dependability, reliability and overall quality as an alternative to commercial software, both in research papers [33] and in pan-European projects 1 . On the other side, the chance for end-users to benefit proper, centralized support for software created by the community at large should be assessed [19,3].…”

Software Engineering in Practice: Design and Architectures of FLOSS Systems

“…Generically, it should be established whether well-established approaches, techniques, frameworks and tools from the traditional SE knowledge apply to FLOSS practitioners and developers [2]. Specifically, FLOSS systems and their architectures have attracted significant attention among researchers due to the distributed constraints, modularity and the issue of collaboration.…”