Ed Summers wrote:
It's a good idea. I know there's significant interest here at LC to follow OCLC's lead in making data and services available to the public.
Really? What there'd be the most interest in from _this_ public would be LCC and LCSH authorities, as well as NAF (person/corporate name, and work title) authorities. That should all be accessible via a REST XML API.
I could sketch out what this interface should look like, but it's pretty simple---basically just SRU delivering MARC-XML would be fine. Although it would be nice if the 'references' in the MARC-XML actually supplied the URLs for identifying/downloading the records referenced, instead of making the user construct such a url on their own (or worse, SRU to try and find it without being sure of it's url!). If LC is really interested in providing their core useful stuff.... Jonathan
In particular I know there's a rich API hiding behind the NDNP aka ChroniclingAmerica [1] project that's waiting to be documented and released. I'm personally hopeful that we'll see some of these services and data feeds emerge in the coming year. I don't know if Roy mentioned this when it came up, but Rob Sanderson (azaroth) actually added a simple server [2], and a mod_python handler [3] to that skos code I pointed at in the previous email. So you could run the service yourself. Of course the point of running it at LC would be that it could be kept up to date, and hopefully would be more than just the outline. Maybe a bunch of us could get together and brainstorm about what some services and feeds could look like at the next code4libcon? //Ed [1] http://www.loc.gov/chroniclingamerica/ [2] http://www.inkdroid.org/svn/lcco-skos/trunk/rdfizer/lccn/server.py [3] http://www.inkdroid.org/svn/lcco-skos/trunk/rdfizer/lccn/handler.py //Ed
-- Jonathan Rochkind Digital Services Software Engineer The Sheridan Libraries Johns Hopkins University 410.516.8886 rochkind (at) jhu.edu
