Main menu:

Site search



August 2016
« Sep    


The Librarians and the Chocolate Factory: OCLC Developer Network Day

I felt like I won the golden ticket for geeky librarians last week. I was one of the lucky folks invited to OCLC headquarters in Dublin, Ohio to discuss their new “Grid Services” and an OCLC developer network that they’re trying to get off the ground. The guest list included several of my code4lib buddies, too: David Walker, Peter Binkley, Jeremy Frumkin, and Jonathan Rochkind. Roy Tennant, Don Hamparian, and Andrew Houghton participated on the OCLC side of things.

Roy started us off with an assurance that there were no non-disclosure agreements to sign and that we could blog and make public anything we saw or heard during the visit. He also told us that this would be the first of two meetings, with an upcoming meeting to be held in Europe.

In a nutshell, OCLC is launching a new… well, I was going to say product, but that doesn’t really cover it. Nor does service, exactly. They want to make many of the webservices that OCLC uses internally available externally to library developers. But not only that, they want to foster a network of people who are using these services, and give them feedback channels to let OCLC know what’s working, what isn’t, and what webservices we want to see, and to share with each other and the world the ways we find to mash-up these data streams.

This is a fantastic idea. I’m trying to throttle back my enthusiasm a bit, because I’ve had my heart broken so many times by ideas like this that pitch beautifully and then come off as mediocre in the actual delivery, but my early impressions are that they are doing a lot right. The working name for this endeavor is “WorldCat Grid.” The Grid group’s aim is to build a consistent interface for OCLC network services, and to provide developers with “one neck to choke” (their words) if something breaks.

I really perked up when they said that the grid portfolio does not have “explicit revenue goals.” That means this group is not expected to turn a profit. Instead their goal is to “expose data at the services layer to see if we can make the data work harder, and maybe in ways we don’t image yet. The whole purpose of web services is to allow things that we haven’t thought of.” Like I said, I’m trying to throttle my enthusiasm, but this made me want to stand up on my chair and cheer. From a business point of view, this makes a lot of sense to OCLC, because when you give people more ways to mash up their data, you increase the value of an OCLC membership and you provide an incentive for libraries to upload their records to worldcat, not to mention gaining yourself publicity, innovative new uses for data you already have, and free QA testers for web-services that OCLC is developing for internal use anyway.

They hope to have something official rolled out by January 2008, but for the moment the services they envision providing fall into four categories:

  1. Network Services
    • search services, e.g., worldcat api / search, open URL
    • metadata crosswalk
    • identity management, e.g., user profile, authentication
    • payment services, e.g., interlibrary fee management
    • delivery / fulfillment
    • social networking, like recommendations, rating systems, and user created lists
  2. Registries
    • bibliographic registries, e.g., worldcat, worldcat identities
    • clusters, e.g., FRBR clusters, xISBN, xID
    • resolver knowledge base, e.g., openly one gate
    • worldcat registry of institutions, including policies (e.g., ILL), OPAC registry, resolver reigtry, IP addresses, etc.
  3. Reusable components
    • toolbox of programmable components
    • some released externally via developers’ network
  4. Developers’ Network
    • Open to participation from anyone, not just OCLC member institutions
    • We need to make IP and licensing explicit for any code contributed
    • Participation in the developers’ network gets you early access to OCLC prototype services, access to OCLC developers, and more stuff yet to be defined.

Unfortunately, they haven’t really fleshed out what they mean by many of these proposed services, and I don’t know much more than the fragmentary notes I’ve written here. We managed to extract a promise that they would start rolling out services as soon as possible for testing and play-time, and I hope that this happens. Of course we’ll have to see whether the implementation can live up to the promise of this program, but I came away from the meeting feeling quite positive about the possibilities, and I’m looking forward to playing with any services they can make available.


Comment from Ross
Time: October 4, 2007, 10:00 am

So Peter’s allowed back in the country now?

Comment from bess
Time: October 4, 2007, 10:03 am

Yeah, I guess he finally got his passport straightened out. It will be good to have him back at conferences. :)

Comment from pbinkley
Time: October 4, 2007, 5:36 pm

Great summary – I’ve got to get mine written up. (And actually, the OCLC campus was ceded to Canada for the day, to avoid diplomatic entanglements.)

Thinking back, I think the best part of it was the evident intention to get the whole package right, and not just roll out some more-or-less cool web service ideas. They’re thinking about communications and IP and community-building as well as technical interoperability right from the start.

Comment from custom research
Time: October 9, 2007, 8:30 am

Wonderful Bess, Where do you find the time…?

Comment from K.G. Schneider
Time: February 6, 2008, 3:59 pm

Would love to read a follow-up assessment of this…

Write a comment