I think it's dumb, but they're asking me, so I'm looking. There are real problems with the idea of locating all the data on one of their own systems... that means that they're taking on maintaining a full time broadband network to the actual library, and keeping a machine run and administered full time.

I'm mostly looking to put together a good presentation that shows the best possible scenario... and why google is both better and cheaper for them. Given the tool Crazy mentioned, and the setup and dev time involved in making it work for them, I think that it should be an easy presentation to make.