Mikael Nilsson wrote:
> This is stated again and again, and still I don't see anyone in the
> library world seriously considering the most widely deployed protocol in
> existence - HTTP.
>
> It works. It's reliable. W3C has guidelines for deploying information
> about abstract entities over HTTP. What is lacking?
>
Well, HTTP is a transport and sure enough, libraries do a lot of
transporting of data and metadata using HTTP (e.g. OpenURL, SRU/W). It's
not the transport that's the issue, it's what it resolves to. The W3C
guidelines for linked data say to resolve to RDF. RDF is a structure. We
still haven't gotten to content. It's the lack of a *content standard*
that is holding libraries up, and that's what we need to work on.
Libraries need a standard representation of *their* vocabulary(s) in a
web-friendly format.
The list of vocabularies in the W3C document shows, I believe, that few
communities have yet gone this route -- libraries are not the holdouts.
Also, the vocabularies that are listed there are generally pretty narrow
(FOAF and CC) and much simpler than what libraries need. In the first
pass at DC/RDA we have 256 properties and 50+ vocabulary lists. Compare
this to the DC terms and think about the work that took!
What this means that until we have a web-based vocabulary for library
data there isn't much that libraries can do. Some of us are exploring
web-based metadata schemas for libraries, like the DC/RDA and the work
to express LCSH in SKOS. The impact, however, will be low unless we can
measurably demonstrate that the benefits outweigh the costs. We have to
*show* that library data can be expressed in a web-friendly format and
that there is something tangible to gain from that. We have quite a bit
of work ahead of us.
kc
--
-----------------------------------
Karen Coyle / Digital Library Consultant
[log in to unmask] http://www.kcoyle.net
ph.: 510-540-7596 skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------
|