« Finally! Somebody Got It Right! Whew. | Main | Ridiculous Amounts of Time-Wasting Fun: »

June 20, 2002

Course Management Systems, Open Source, and the Interest Horizon:

homoLudens ponders, "[I]t's about what CMS shore you first land on. Problem is that so many are getting shipwrecked on high-priced, user-infantilizing Microsoft / Blackboard / webCT coastlines. I still don't get why there's no major educational foundation funding development of open source (or cheap source, a la Manila or p-Machine or whatever) for schools. [link via SiT]

Pat makes the erroneous "CMS leap" -- while Manila is a content management system (CMS), Blackboard and WebCT are course management systems (also, confusingly, CMS) .... which by the way are not the same thing as an LMS (learning management system) or LCMS (learning content management system). Too many acronyms in the industry. Course management includes content management, but vice versa is not true.

In any event, there is one obvious shining example to PatD's question about open source alternatives: MIT's Open Knowledge Initiative. But the obvious example (or, more appropriately, exception) of MIT aside, the reasons educational institutions don't take on open source development of CMS's are fairly simple and straightforward. (Caveat: I work for Blackboard, one of PatD's examples, so while these comments are my own personal ideas and don't necessarily represent my employer's position, they may be biased. Salt to taste.)

  1. As online teaching and learning becomes more important to the mission of educational institutions, the insitutions want scabable, secure, cross-platform, usable, enterprise-class software. When you have potentially tens of thousands of users (or more, in some cases) relying on a mission-critical system, it doesn't make sense to rely on v0.94b of Joe's Online Course System that you pulled off SourceForge last night just because it's open source. (Note: before you get your or Richard Stallman's knickers in a twist, this statement isn't meant to imply that open source software can't be scabable, secure, cross-platform, usable, or enterprise-class. Just that there aren't any open source course management systems (yet) that meet those criteria.)

  2. Scalable, secure, cross-platform, usable, enterprise-class course management systems are notoriously resource-intensive to develop. This ain't a three-man job. It either takes an organization with existing knowledge and resources of an MIT or a big, honking chunk of cash to pay for those resources. (Note: again, before your free software panties start bunching, this is not meant to imply that the open source approach can't generate scalable, secure, cross-platform, usable, enterprise-class software because of resource constraints. E.g. witness Apache. However, let's not forget that the resources that go into open source projects are not free-as-in-free-beer -- they are donated, either out of someone's personal time or out of an employee's time -- so there is a cost associated. That's important to the next point.)

  3. Educators are interested in course management systems; open source developers, in general, are not. I believe the massive failure of open source software to compete effectively in user-facing markets is due to lack interest from open source developers, not lack of skills. "User-facing markets" are those markets where the user of product is a typical consumer, e.g GUI operating system interfaces, office suites, browsers, and, yes, course management systems. Most open source software depends on developers wanting to "scratch their own itch" -- e.g. they have to be self-motivated, because no one pays them to do it (because no one has found an effective way to make revenue off open source yet!). E.g. see Kottke's comments regarding Mozilla, the open-source progenitor of Netscape 6.

    This is what Clay Shirky has called the "interest horizon." (Yours truly also ranted about this.) Commercial software projects have a "resource horizon" -- a time or money limit by which the development is no longer economically feasible for the company. Open source software has an "interest horizon" -- since it's frequently volunteer work, the development is limited by the developers' interest in the project. Put simply: if it bores them, they're not going to code it. That's why we have great open source operating systems, compilers, databases, etc -- those tools interest the geek set who programs them. Spreadsheets and virtual classrooms don't.

The exception to this is organizations like MIT, where there are both enough resources and interest to fund and drive the development of an open source initiative that would never emerge organically on its own.

So to summarize: there are no scalable, secure, cross-platform, usable, enterprise-class course management systems because it's (1)complex, (2)expensive, (3)not of interest to the vast majority of unpaid open source developers. :-/

Posted June 20, 2002 08:04 AM