[uf-discuss] citation: another example of practice in the wild

[Bruce D'Arcus]

On 8/16/06, Tantek Çelik <tantek at cs.stanford.edu> wrote:

> FWIW, I don't think DC provides a particularly good set of names, far too
> theoretically designed, and eclipsed in practical usage by several of the
> other established formats (OpenURL, Bibtex).

Come on Tantek; that's an entirely arbitrary assessment, without any
real factual basis.  Given the wide-spread use across a whole slew of
formats (HTML, RSS, OpenDocument, the new MS XML formats, Adobe's XMP)
how can you possibly say that it's "theoretical"? More stuff is
described using DC than probably any format or term set in existance.

But rather than argue about this, let's step back and pull out what we
can likely agree on.

The first group is so obvious I really don't think we could possible disagree:

*  dc:title (absolutely need a "title" term, for all kinds of things)
*  dc:date (and the qualfiers issued, etc.)
*  dc:subject (aka tag, keyword, etc.)
*  dc:publisher
*  dc:identifier

And then there's the following, which have one issue or another where
reasonable people can disagree:

*  dc:creator (OK, maybe a little problematic in different ways, but
widely understood and useful, if too broad for most citation needs)

* dcterms:isPartOf and isVersionOf. OK, yes, a little bit abstract,
but they are excellent ways to describe critical relations in
bibliographic data, in ways that don't resume upfront a limited scope
of description. Document isPartOf collection, track isPartOf album,
article isPartOf journal, etc. Am happy if someone comes up with
better terms, so long as they still retain some flexibililty.

Most of the rest isn't that important for citation data, but could be
for other things (different media, say).

Bruce