Here’s a small report on my attending OSS2007, which is this year’s edition of the most prominent academic conference on FLOSS.
This year’s edition was really very interesting, and the venues in Limerick University were excellent. Adding to that a memorable social event at Knappogue Castle (I’m a king now … more details in pictures on flicker)
Next edition, OSS 2008 should then be colocated with the IFIP world congress in Milan, most probably.
But back to this year’s report.
I’ve been impressed by the true international nature of the conference, and by the very large number of American collegues in particular, compared to 2005, even for a conference held in Europe. Even though other workshops are probably very interesting for the community (for instance the ICSE FLOSS 2007 workshop), I think OSS conference is a must now.
Here’s an unordered list of the most interesting papers for which I’ve attended the presentations :
- Capiluppi, Michlmayr : From the Cathedral to the Bazaar: An Empirical Study of the Lifecycle of Volunteer Community Projects : although the title highlights cathedral and bazaar, it may not be so sure that these two schemes are really at stake here, but a very interesting paper studying evolution in productivity and community size, which may help predict at which time a project is potentially able to radically change its profile.
- Conklin : Entity Matching across FLOSS Repositories : really interesting paper on how to match projects between various repositories, taking as a first step the repositories available in FLOSSMole. Very promising to expect a new range of studies tracking projects among repositories, to produce richer analysis.
- Noll : Innovation in Open Source Software Development: A Tale of Two Features : I liked this paper alot since it presented a set of anecdotal facts, concerning the introduction of new features in two libre software projects (namely metacity and firefox/mozilla). Instead of drawing general figures without making us get a grasp of what one’s talking about like so many other presenters in the conf, this paper was precise and that was a very satisfying thing for the audience, I think.
- Robles, Dueñas, Gonzalez-Barahona : Corporate Authorship of Libre Software: Study of Presence in Debian Code over Time : a very interesting paper which presents an analysis of corporate contribution to libre software projects, taking Debian as a proxy of the whole libre software projects. And guess who’s the winner ? : SUN. I strongly advise the reading of that paper.
Some short papers were also very interesting, I think :
- Dalle, den Besten :Different Bug Fixing Regimes? A Preliminary Case for Superbugs, which investigates the life length of bugs which are still not fixed although an initial patch was made… not definitive results, but promising.
- Michlmayr, Hunt, Probert : Release Management in Free Software Projects: Practices and Problems, which presents the importance of release management in libre software projects. It’s not easy to develop, but coordinating lots of developers in order to be able to issue a software release for a package, or a set of packages (a distribution maybe) is a real challenge which must not be underestimated. Worth reading also
Keynotes were also very interesting :
- The opening keynote by Karl Fogel, about copyright was probably a bit out of scope for most researchers, but very interesting anyway, I think, trying to put in perspective the role of copyright now, compared to its historical background. I didn’t know about the organization Question Copyright which was endorsing the presentation. Very complementary (and more radical maybe) to Libre software advocacy groups, or other Copyright reform initiatives like Prof Lessig’s Creative Commons…
- The keynote by Ari Jaaksi from Nokia was really very interesting on how things happen in companies trying to adopt an opensource policy. He told us bits of the story on how the N800 product was made, and the models for collaboration with communities, and the many challenges for the company’s middle management facing a “sandwich” situation : top management really enthusiastic for OSS, and developpers also, of course, pushing alot, and in between, the middle management trying to make things work, and projects to deliver. Really delightfull
The workshops were very interesting, I think, although we couldn’t stay until the end of the last day. I attended the 2nd International Workshop on Public Data about Software Development (WoPDaSD 2007) which discussed lots of elements on public repositories of metadata on FLOSS projects (like FLOSSmole and others). I’m particularly interested since there are links here with what is present in the actual software repositories (forges) that we’re trying to improve.
Maybe one of the most surprising sessions of the conference was actually related to these public repositories of metadata for FLOSS research. There was a debate between panel of researchers, discussing a resolution about requirement that data used to produce research on FLOSS be made available publicly as a requirement. Very very interesting discussions, and audience votes. The resolution was voted against in the end. It was really a mind challenging debate for such a conference, I think.
Finally, Debian is everywhere in Ireland it seems