Read thought-provoking article on the future of OPACs, then drove down to Ludlow for soda, thinking thoughts provoked.
Struck me as a bit parochial, library science tends to be a bit that way, I guess.
One thing that jumped out at me was a tendency to differentiate between cataloguers and users, this seems like a really bad idea. The way I look at it, anybody who uses a system is a user, that includes the developers, the admins, the people entering data, the people retrieving data. Not having a clear sense of all likely use cases and user types is a recipe for bad design. A priori assumptions that there being different use cases and user types implies necessary tradeoffs in favor of one or another are a recipe for bad design. There are always going to be tradeoffs due to resource limitations, but the overarching design goal needs to stay optimal usability for all user types in all use cases, any deviation from that should require careful thought and pulling of teeth.
Another thing that bothered me was the idea that centralization of services somehow entailed the death of the library. Libraries are basically information repositories, if the goal is information dissemination, networked repositories are always to be preferred to standalone ones. I think the anxiety is much more about efficiencies equalling redundancies than the demise of the library or the OPAC. Jobs aren't sinecures, folks, the tempo's increasing all the time, we can learn to dance to it & have fun or stand around looking sullen getting our toes stepped on.
Overall impression coming away from this article: librarians would really benefit from closer collaboration with developers & systems people. If there's ice needing to be broken, I suggest getting together informally, having a few drinks & watching Desk Set.