Amen. I suspect there’s demographic in library schools (and in libraries), that can remember when libraries really did build their own systems (and universities, at least, sometimes built their own operating systems). The key may be to articulate and prove that Open Source is not a total risk, and maybe also that there is a lot of commonality between what libraries need and what other organizations already use. There are Open Source, high quality, and field-tested, general purpose ledger systems, for example. In the days of the total in-house system, libraries had to build all this from scratch. The software world doesn’t have to work that way any more.
Shared infrastructure and public code powers the Internet. Any serious library vendor or educator would almost certainly have to recommend the Web for the public access side of any library system. There is also a form of software Darwinism that occurs in Open Source, maybe even more so than in commercial development shops, and there are lots of examples where the commercial world has not managed to match the results of the OSS community. Skipping the Linux/Windows debates for a moment, one of the easiest OSS success stories to point to is the Apache web server. In fact, is there any way for a networked organization to avoid OSS? If an organization is connected to the Internet, if it uses DNS, if it sends mail, if it uses the web, then it is almost certainly participating in OSS at some level.
The original point of the ILS was to achieve a level of integration that the technology did not allow at the time. As Eric points out, it’s a much different world now. There’s one point of view that OSS in libraries is only about small systems that don’t have a lot of complex code, and we need these systems, but it’s also an opportunity to do more than our current systems allow. If you cringe about your webcat, ask an accountant or financial guru to wander through your ILS’ acquisitions system and see what happens. OSS can be empowering because it can let you use building blocks crafted by other communities without the kind of licensing issues that a vendor usually has to deal with. We don’t only need OSS for matching what we already have, we need OSS to create better systems all the way around.