I'm very happy with the proposal that we change RFC #1 so that Relation
identifies a related resource, using either a formal identifier such as
a URI or a less formal string.
We then add dcq:RelationType for qualified DC. Relation is, in this
respect, no different from other elements whose qualified versions allow
Types. [Decisions]
[Decisions] http://www.mailbase.ac.uk/lists/dc-datamodel/files/decisions.html
Misha
> Looks like we've painted ourselves into a corner here. The accepted mantra,
> I believe, has been that the Relation element is basically meaningless in
> its unqualified form. Therefore, we have succumbed to faking it by kludging
> these funny parseable free text strings for "unqualified" relation such as
> "isBasedOn http://foo" <http://foo> . I admit to never having liked this
> but, as with many things in DC and out of DC, have come to grudgingly accept
> it. Eric's new definition, "The identifier of a related resource" backs away
> from this. This appears to be a significant dumbing down of the definition.
> Also the term "the identifier" seems to negate the notion some of us had
> accepted a month or so ago that it didn't have to be "an (formal)
> identifier", it could be some "free text string" that (sort of) identified
> the related object (e.g., that white statue of the guy on the horse in the
> first floor of the Musee d'Orday.
>
> Carl
> ----------------------------------------
> Carl Lagoze, Digital Library Scientist
> Computer Science Department
> Cornell University
> Ithaca, NY 14853 USA
> Internet: [log in to unmask] <mailto:[log in to unmask]>
> WWW: http://www2.cs.cornell.edu/lagoze/lagoze.html
> <http://www2.cs.cornell.edu/lagoze/lagoze.html>
> Phone: +1-607-255-6046
> FAX: +1-607-255-4428
------------------------------------------------------------------------
Any views expressed in this message are those of the individual sender,
except where the sender specifically states them to be the views of
Reuters Ltd.
|