Fosdem O'Reilly
2005 Edition Free and Open Source Software Developer's European Meeting






Dev-Room

Calibre Developers' room News

[ Schedule ]

News [ 18-04-2005 ]
Welcome to the CALIBRE room

Learn at the CALIBRE room about academic research performed on libre (free/open source) software. Top researchers in the field will be there to speak about software measurement, project management, software modeling, etc. Find out how libre software works, how it evolves and how it is investigated.

CALIBRE

What is CALIBRE?

CALIBRE is a Co-ordination Action funded by the European Commission for Libre Software Engineering for Open Development Platforms for Software and Services. CALIBRE aims to coordinate the study of the characteristics of libre (free/open source) software projects, products and processes; distributed development; and agile methods. Read more about it at http://www.calibre.ie.

You can also attend the 2nd CALIBRE Workshop on the theme "Libre Software: Which Business Model?" which will be held (in English) Friday March 4th 2005 in Paris (France).

Why should you, probably a libre software developer or a libre software enthusiast, be interested in CALIBRE?

Because you have probably heard about software engineering at university and found out that it has very few to do with 'real' life and almost nothing with libre software. We are interested in changing this and to discover what can be learned from libre software: its way of development, its (social) structure, etc. You will find here technical and socio-technical studies about libre software projects, especially Debian, GNOME, KDE and Linux, its developer population, the interactions between developers, the composition of the community, the evolution of the software, etc.




The CALIBRE DevRoom is a one-day room for Saturday February 26th starting at 13:00 and finishing at 17:00 (see schedule for details). The half-an-hour presentations we have organized are following:

"Evolution of Debian GNU/Linux over the last 7 years: from developers to programming languages"

by Gregorio Robles and Diego Barceló (Universidad Rey Juan Carlos - Madrid, Spain)

Abstract: The Debian operating system is one of today's most popular GNU/Linux distributions. From its birth, more than one decade ago, it has undergone many human, technical, structural and organizational changes. This article tries to study the evolution of Debian in the last years, comparing the last four stable versions of this distribution in size, programming languages and packages. The evolution of the number of voluntary developers (maintainers) in the Debian project is also analyzed and an estimation of the effort in human and economic terms that would be necessary to produce a software of this size is made.
The main evidences that we have found are that Debian approximately doubles its size in lines of code and in number of packages every two years, whereas the mean value of the packages' sizes remains constant. A big majority of the packages of the first version considered in this study 'have survived' in time and are present in more modern versions of Debian, many of them even with the same version number. The most widely used programming language is C, although its importance is decreasing in time.
At last, the results of the evolution of Debian during these last five years are used to make a preliminary prediction of how the next stable version of Debian could be, when it may be shipped and what challenges (limiting factors) Debian will have to surpass in the near future. The contribution of individuals, non-profit organizations like de FSF, universities and corporations during the last Debian distributions will also be shown. More information: Debian Counting.

Slides of this talk (in PDF).

"SimCode: Agent-based Simulation Modelling of Open-Source Software Development"

by Jean-Michel Dalle (University Pierre-et-Marie-Curie & IMRI-Dauphine - Paris, France) and Paul A. David (Stanford University & Oxford Internet Institute - U.S./U.K.)

Abstract: We present an original modeling tool, which can be used to study the mechanisms by which free/libre and open source software developers' code-writing efforts are allocated within open source projects. It is first described analytically in a discrete choice framework, and then simulated using agent-based experiments. Contributions are added sequentially to either existing modules, or to create new modules out of existing ones: as a consequence, the global emerging architecture forms a hierarchical tree. Choices among modules reflect expectations of peer regard, i.e. developers are more attracted a) to generic modules, b) to launching new ones, and c) to contributing their work to currently active development sites in the project. In this context, we are able - particularly by allowing for the attractiveness of "hot spots" - to replicate the high degree of concentration (measured by Gini coefficients) in the distributions of modules sizes. The latter have been found by empirical studies to be a characteristic typical of the code of large projects, such as the Linux kernel. Introducing further a simple social utility function for evaluating the morphology of "software trees," it turns out that the hypothesized developers' incentive structure that generates high Gini coefficients is not particularly conducive to producing self-organized software code that yields high utility to end-users who want a large and diverse range of applications. Allowing for a simple governance mechanism by the introduction of maintenance rules reveals that "early release" rules can have a positive effect on the social utility rating of the resulting software trees.

Slides of this talk (in PDF).

"Software Engineering research in large projects: some examples for Linux, GNOME, KDE and Apache"

by Israel Herraiz and Juan José Amor (Universidad Rey Juan Carlos - Madrid, Spain)

Abstract: Libre software has gained attention from academia in recent years, specially large software projects with high developer attraction as Linux, GNOME, KDE and Apache. This talk wants to give an overview of the kind of research that has been or is performed on this projects and which results and facts have been obtained from them. These views include topics like: software growth, developer integration into projects, social network analysis, complexity, etc. More information: Libre Software Engineering.

Slides of this talk (in PDF).

"Libre projects lifetime profiles analysis"

Krzysztof Kowalczykiewicz (Poznan University of Technology, Poland)

Abstract: A success in the libre software world may have many faces. Different aspects include users satisfaction, developers involvement or even commercial outcomes and media attention. Projects are continually born, they gather some community, they have their ups and downs, stagnation and increased effort phases. From that information some patterns may appear. We've taken historical data available from the SourceForge website to analyse several successful, unsuccessful and average projects. The talk presents our approach to identify successful project curriculum patterns and measures of success, both temporal and overall.

Slides of this talk (in PDF).

"Social links and technical dependencies in free software modules"

by Rishab Aiyer Ghosh (MERIT, University of Maastricht - The Netherlands)

Abstract: New contributors to free software development appear to be socially influenced in their choice of first contribution, preferentially joining modules with a large number of existing developers. This leads one to speculate that social links between modules in the form of common developers may be associated with technical links between modules for example, in the form of functional dependencies. This talk shows data gathered from a detailed study of the structure and composition of the Linux kernel developer community, as sampled through three versions of the Linux kernel. The relationship between the degree of common authorship of module pairs is compared with their degree of functional inter-dependence.

Slides of this talk (in PDF).

"CVSAnalY: An analysis tool for your CVS and Subversion repository"

by Álvaro Navarro and Gregorio Robles (Universidad Rey Juan Carlos - Madrid, Spain)

Abstract: CVSAnalY is a tool written in Python that allows to analyze version control repositories (at the moment both CVS and partially Subversion are supported) used in libre software projects. This talk will show how CVSAnalY works and will put special emphasis in the various results, statistics and graphs that it offers. More information: CVSAnalY project page and some results obtained.

Slides of this talk (in PDF).

"Evolution of Alive Projects in SourceForge: its interrelationships with population size and other attributes"

by Juan Mateos-Garcia and Ed Steinmueller (SPRU, University of Sussex - U.K.)

Abstract: Research on libre software has mostly paid attention to large libre software projects. Our intention is to, by using data from the SourceForge website, focus on smaller projects (both in terms of size and contributors) with the aim of determining the direction and strength of interrelationships between a project's vitality and other parameters such as population size and a set of project attributes which include its intended audience and license, the programming language in which it is written etc. By doing this we will test existing assumptions about, for example, the importance of a project's population size for its survival and evolution, as well as assess the strength of diverse hypothesis regarding the motivations underlying developers' decisions to contribute to specific projects.

Contact person for the CALIBRE room: Gregorio Robles (Universidad Rey Juan Carlos, Madrid, Spain).



 








© FOSDEM 2003-2005 - powered by Argon7