The Code4Lib 2006 Conference started today (2006 February 15) in Corvalis Oregon on the campus of Oregon State University. OCLC is one of the 'Platinum' sponsors, and there are five of us here (counting Eric Hellman of Openly). Looks like there are something like 80+ people here, which is nice to see. When the idea of this was proposed it sounded like a good idea to me; it no longer seems strange to see an email list generate a conference.
The first session this morning was a 'virtual' keynote from the 
Evergreen development team of the PINES consortium in Georgia. This was done by phone with local PPoint slides, and actually worked pretty well using a mike over a speaker phone. PINES is growing to about 300 libraries, with 2 million patrons and 20 million holdings and will be running on an ILS they are writing themselves. Evergreen uses OpenSRF (Open Scalable Request Framework), which uses the Jabber instant messaging protocol for communications. It wasn't quite clear to me why this stack is such a great idea, but the developers seem sold on it for its scalability and flexibility.
Evergreen is planning to have a 'production' release this summer and to go live this fall (2006). Since the current ILS doesn't handle serials processing and acquistions, neither will Evergreen on its first release. It sound like they have around a half-dozen developers involved in the project. They are developing this themselves because they didn't feel either the existing commercial and open source ILS's didn't support the scale of PINES well.
--Th
Art Rhyno thinks our interfaces should be more like Beeb Beeb Buggy (which he actually sold at one time). He's probably right, although I've argued otherwise.
--Th
Dan Chudnov just did a talk about unAPI. Part of the idea of unAPI is to have the simplest protocol possible, since people "just don't get' complicated protocols, or even protocols as simple as OAI-PMH. Maybe I should have been paying more attention (rather than looking for pictures of beep-beep-buggies), but I'm afraid all I got out of Dan's talk was that even the simplest protocol can be 'hard to get', in particular to see what it can do and how you would use it yourself. I'm sure it will all become clear on reflection.
--Th