In the market for a new model
November 14, 2008
I presented my thoughts on my research to my research group and received some good feedback that I intend to act on.
In the market for a new model
On Saturday Nov 8, I presented my thoughts on my research (a condensed version of what I’ve said in this blog) to my research group, the Social Learning Software Laboratory (SL2) of the School of Information Systems and Technology (IST) at Claremont Graduate University, and was generally well received. I got some very good feedback, which I intend to follow up on: I have a dissertation to read, I may need to look for a different model of power networks, and I should better define the kind of open source software project I’ll be looking at, and also what members of the community I’ll consider as part of the project.
Lorne Olfman, my advisor, suggested (not for the first time) that I read a recent IST dissertation on project management in open source software projects, and I think that I’m nearly ready to read that now. If I remember correctly the author analyzed forum archives of open source projects, looked for messages or groups of messages that were decision points for the project, and mapped them to project management phases, tasks, or milestones. These same decision points may be indicative of people exercising power within the community.
Terry Ryan (IST faculty, and now Dean of IST), who had read some of what I’d written on my topic back when I was still thinking of working on traditional firms, thinks I should see if I need to find a different model of power networks: not only may people in OSS communities think about power differently than in those in traditional firms, but they may hold different kinds of power than are described by the Crozier and Friedberg model I plan to use. Immediately upon reconsidering Crozier and Friedberg, I can see that hierarchical power (one of the four kinds of power in their model) may prove problematic, because there’s no explicit, institutionalized hierarchy in an OSS project.
I was asked what kind of OSS community would I be looking at, and are there enough of them to make my work worthwhile? Initially I said that the interesting ones are the Linux, Apache, Mozilla sized projects, and there are probably enough of those (I can think of maybe 25 without much effort). But now, I think that the interesting communities to me are those that pass a certain threshold of understandability. Beyond a certain number of people, the number of relationships and roles can be difficult to track, especially when all one has to track them with is e-mail archives and version control logs. So I’m interested in communities beyond that threshold, whatever it is. Probably around 15 people. I think I also need to define what I mean by a community member. I’ve been mapping them to developers in my head, but there are influential users also who may hold power: those who test software and submit bug reports, or use the software but need specific features, for example.
