I written several times about "living where your users are": putting links to your content wherever users are, out surfing on the web. This is what the Common Information Environment blog has to say
I gave a presentation (10Mb PDF download) at the JISC Conference in Birmingham. One of the points I was trying to make was that [library] services should be visible to users where they are, rather than making them consciously enter the library's virtual space. Hardly radical these days, I realise!
from Lorcan grapples with delivery of library services to their users
But then the problem becomes: how do you keep your linking services from becoming stale? What happens to all the great deep links to your content when you redesign your site, or your OPAC vendor changes their query structure?
I talked about the growth of tools such as LibraryLookup, which link deeply into our library catalogues.
As discussed in the post, there is a growing problem in keeping these tools aware of changes to search syntaxes, URL structures and the like, all of which they rely upon in parasitically (not intended negatively, but the best word I can think of to describe the relationship) using the systems of another organisation (such as a library) to deliver that organisation's resources to a user.
In a post this morning, Talis' Justin Leavesley reports on work they've been doing with RedLightGreen to explore solutions to the problem. Their current model appears, from the post, to offer a registry solution, in which potentially volatile 'service' information is recorded once (although nothing would then prevent mirroring or harvesting of that single instance of human record-making). An external product or service wishing access to Library X simply consults the registry via an OpenURL and is directed to the appropriate interface.
Talis say that they are exploring ways in which this approach might be shared with other vendors. It also, clearly, has some synergies with the JISC Information Environment Service Registry and would form a useful resource driving aspects of the People's Network Service.
Standardised descriptions of access points for our databases. Resolution of the ISBN problem (different editions of a book receive a different number from their publisher; it is this number (the ISBN) that drives the sort of search that LibraryLookup undertakes) through OCLC's xISBN or equivalent. Do we have a genuinely useful emergent service on our hands here?
Can we cooperate to build it?
from Hiding the complexity of deep links into our databases?
I think the good news is that computer science has thought a lot about how to manage changing interfaces. In CS, one way to do this is to have a registry, with interface versions, so that as you add new connection capabilities, you still support older, more basic connections.
OpenURLs, permanent URLs (e.g. purl.com) and DOIs, and Web Services may all have a role to play in communicating the metadata needed to make and maintain connections to library content and services that we choose to expose.
Previously:
2005-02-02 librarians - put your content where your users are
2005-03-05 building a better library website
2005-03-06 hacking web content
2005-03-15 take control of the library interface
2005-04-04 Jon Udell screencasts about library web rewriting
Comments