This thesis by Karin van den Berg was part of her Master’s program at Tilburg University in the Netherlands. It does a nice job of boiling down many of the previous attempts at developing criteria for evaluating open source projects (see her references for the extensive list) and then uses as a case study the field of course management systems to try out the evaluative framework. Somewhat unsurprisingly to me, she hones in on Moodle and Atutor as being the top two contenders when these additional open source considerations are factored into the equation; it is nice though to have some substance to back up ones’ instincts.
This piece was quite close to my heart for a few reasons, and thanks to Stuart Yeates posting on the Educause community blogs for finding it. It’s meaningful to me first off because I’ve been looking at these CMS thingies for far too long now through the Edutools project. It was meaningful to me because she cites Edutools numerous times throughout the thesis, and it is nice to get some academic ‘props.’
And finally it is meaningful because for the last year I have been going around giving a presentation that I think basically says a lot of the same thing, though I frame it slightly differently. The jist of that presentation is that making good open source choices is all about picking projects that are a suitable fit with the capabilities and maturity of your own organization, and trying to educate people on what some of the qualities of OS projects are that they can base those judgements on. The presentation was actually a summary of a funding proposal for an “open source suitability decision making tool,” a proposal that didn’t get funded. And I am really glad it didn’t get funded. Not because such a tool wouldn’t be useful. But instead because, as one of the reviewers astutely pointed out, and as this thesis backs up, the judgement of an open source project’s “maturity” is too multivariate (and evolves too quickly) for it to be serviced well by both such a small one-time grant (we were really not asking for much) but also from such a centralized research model as we had proposed. Still, it is really vindicating for me to see basically the same set of criteria brought out in full in this thesis, as it makes me fill slightly less half-cocked (so, what, two-thirds cock’d?) – SWL