Introductory comments by Brian, then discussion.
Brian Schottlaender and Constance Malpas, in the afternoon session; photo by Dana Peterman
Brian started with two ideas:
* the "collective collection"
* "networked infrastructure"
We are charged with managing two different kinds of collections, digital and print. It's not at all clear to me that if either one of them were taken away from us we'd be happy.
The collective collection is multi-type -- electronic & non, books & journals, primary & secondary, etc. In addition, the institutions that manage the collections are multi-type.
The collective collection is not several collections, or a collection of collections (like the UC libraries). What distinguishes the "collective collection" is the application of a greater will to manage it for the greater good, and a network infrastructure.
Network infrastructure is necessary b/c the collection is also *distributed*, both on small & large scale. This network infrastructure can be visualized as a series of nodes, like other networked services. In this context, the RLFs, for instance, are each nodes, as are individual libraries or perhaps even the UC as a whole.
Think about what makes a trusted system. The definition of trust that Schottlaender likes best: "it does what you expect it to, and it does not do what you don't expect it to do." When you turn the light switch, the lights actually come on, not that the curtains fall or the building explodes or the lights don't actually come on. When we build "trusted archives" will they do what we want them to? In the case of some sort of catastrophic failure, can we recover with our current archives?
(Ed. note: it seems that by this definition libraries in general act as trusted systems: people trust us to be open, answer their questions, stock their research materials, etc... and when that doesn't happen everyone gets all flustered because libraries, in general, are pretty reliable. We have set that expectation for our users).
Posts
