JISC, Scribd and scholarly repositories
Tony Hurst asks "why doesn't JISC fund the equivalent of Scribd for the academic community?" in a post on the OUseful blog to which one is tempted to ask, "why would they when such things already exist out on the Web?".
Of course, in reality there are good reasons why they might, partly because of the specific requirements of scholarly documents (as opposed to just any old documents) and partly because of assurances about persistence of services, quality assurance, and so on.
I'm minded to ask a different question. One that I've asked before on a number of occasions, not least in the context of the current ORE project, which is "why don't scholarly repositories look more like Scribd?". Why do we continue to develop and use digital library specific solutions, rather than simply making sue that our repositories integrate tightly with the main fabric of the Web (read Web 2.0)?
What does that mean? Essentially it means assigning 'http' URIs to everything of interest, using the HTTP protocol and content negotiation to serve appropriate representations of that stuff, using sitemaps to steer crawlers to the important information, and using JSON to surface stuff flexibly in other places?
By the way, Tony also asks whether there is any sort of cross-search of UK repositories available, to which the answer is that JISC are funding Intute to develop such a thing (a development of the previous ePrints UK project I think). And there are the global equivalents such as OAIster.