RESTful Repositories?

Andy Powell has made notable contributions to best practices concerning persistent identifiers for quite a long time.  I have always found his recommendations practical and free of ego.  I almost said free of ideology, but of course we all suffuse our musings with the aggregate of experience and beliefs which round to ideology.  The ideology of others tends to disappear to the extent that it matches our own.  Protective coloring rears its barely-discernible head.

Andy's opening keynote at the VALA2008 conference in Melbourne a fortnight ago exposed a theme of his ideology which serves the community well in the domain of persistent identifiers, and which he brings to bear on the evolution of repositories.  I paraphrase:

Deviation from mainstream Web idioms reduces uptake and quenches the natural interconnectivity which underlies the richness of the Web.

Andy's own summary of the issues includes the following:

our current preoccupation with the building and filling of 'repositories' (particularly 'institutional repositories') rather than the act of surfacing scholarly material on the Web means that we are focusing on the means rather than the end (open access). Worse, we are doing so using language that is not intuitive to the very scholars whose practice we want to influence.

One way to think about repositories is as the bookshelves of the digital library.  They are designed to impose order and facilitate management of content.  We don't ask scholars, having just published an article or book, to 'go to the library to find the most appropriate place for it... and don't come back until you do!'  Not a perfect analogy, but it speaks to the issue of mandating overhead to authors in order that their work is fixed in the scaffold of their discipline's knowledge stores.  Still, we have bookshelves for a reason, and something like them is necessary to support the management of digital assets as well.  It is hard enough to look after digital resources in a persistent way.  Current repository technology is not yet mature, for sure, but it isn't the case that we don't need what it is trying to deliver. (I don't think Andy would disagree with this).

Andy goes on to say:

our focus on the 'institution' as the home of repository services is not aligned with the social networks used by scholars...  As a result, we resort to mandates and other forms of coercion in recognition that we have not, so far, built services that people actually want to use. We have promoted the needs of institutions over the needs of individuals.

Well, yes, but it isn't as though we don't see this all the time.  It is a rare case when the institutions I have administrative dealings with tailor their procedures and requirements to my needs.  Instead, procedures are designed to increase management efficiency, often at the time-expense of individuals.  I myself have been known to whine about just such impositions (duh), but presumably, the gains in efficiencies of such requirements redound to the general benefit of all.  Thats the theory, anyway.  Sometimes its even true.

The question of where the natural home for repository functionality might be is tricky.  Lorcan Dempsey refers to institutional reputation management -- a natural and important piece of the puzzle.  Publishers are loathe to lose control of the content, but their time is passing... open access is simply too compelling a juggernaut to be resisted.  OA is a when question, not an if question. Professional societies want to play, and some of them sit, somewhat uncomfortably, astride the roles of domain advocacy and commercial publishing. Witness the American Chemical Society doing the splits as the open data boat slowly slides away from the commercial asset management pier.

It is still possible that another entirely different model will emerge... more in-the-cloud.  A distributed model does seem to complicate curation, (and that institutional reputation thing), but I wouldn't count it out just yet.  Still, some institution has to take care of this stuff... responsibility involves the attachement to artifacts, even if they are bitstreams.

Andy goes on to appeal for more RESTful architectural design, and in this I think he is dead on the mark:

the 'service oriented' approaches that we have tended to adopt in standards like the OAI-PMH, SRW/SRU and OpenURL sit uncomfortably with the 'resource oriented' approach of the Web architecture and the Semantic Web.  We need to recognise the importance of REST as an architectural style and adopt a 'resource oriented' approach at the technical level when building services.

There are some details of Andy's perspective that I'm happy to contend, but as usual he forces our attention back to design for the Web, of the Web, by the Web.  Sounds pretty much right to me. An excellent keynote to start off VALA 2008.

Post Script: Speaking of architectures, I see that Roy Fielding, the prime progenitor of all things RESTful, has tired of endlessly explaining the same things on list-after-list, and has started a blog.  This is welcome news indeed.  And you gotta love his colorful domain name.
-----

Public sculpture along the Southbank Promenade of the Yarra River in Melbourne

 

Comments

[simonfj]

"One way to think about repositories is as the bookshelves of the digital library". So we might be thinking about a DDC number to classify them.

Another way to think about repositories is as a TV station. No reason the remote control couldn't have the same DDC numbering system.

The common problem with both models is that the we know we are talking about interactive libraries or TV stations. Having someone come in, rip a file or a few pages from a book, and put a new compilation back on the same shelf is a bit of a concern.

The TV model is possibly more appropriate in a WWWeb where people scratch the surface of a social network to find the libraries which interests a (professionalized) group, and to which they can be seen to allocate varying degrees of value; like (E.g) the reputation you and your peers place on Andy's writings.

So I'm trying to understand why, as with Roy Fielding who "has tired of endlessly explaining the same things on list-after-list", and "many of them questions that I have answered a dozen times or more, but are simply lost in the thousands of email messages archived in too many places around the net for me to even think about tracking them down and forwarding", your social network wouldn't have seen the utility of running a forum, and treating to as the door or reference desk to your peers' inter-institutional libraries.

It would sure make a nice change from watching you and your peers broadcasting from your own blogged channels. It would also act as a nice easy way to create a context for your Q&A's, and links to your data, and leave an interactive archive of them behind when you've finished.

As you say "we have bookshelves for a reason, and something like them is necessary to support the management of digital assets as well". Very true. But if less time were spent trying to classify "the assets" and more time spent solving Roy's problem, which is every peer's problem, I'm sure learners like me would make our own decisions about who's opinions are useful and who's are not, as we always do.

And your group could measure success by just counting the eyeballs. Either they will come back or they won't.