On Thu, Mar 19, 2009 at 04:04:50PM -0000, Pete Johnston wrote:
> 1. dcterms:coverage:
>
> "Where appropriate, named places or time periods can be used in
> preference to numeric identifiers such as sets of coordinates or date
> ranges."
>
> This says the resources (places, periods) "can be used in preference to"
> literals which identify/label resources, which I don't think really
> makes sense. The range already tells the implementer that the value is a
> place or period as non-literal value, not the name of the place or
> period as literal value. Again the comment is introducing an element of
> confusion between things on the one hand and names of things on the
> other. A thing and its name aren't "interchangeable"; they are two quite
> different things(!).=20
>
> I think the intent here is to say that a "value string" for a
> non-literal value may consist of the name/label of a place/period,
> rather than a set of coordinates or a date range. Here, the comparison
> is (depending on how "a set of co-ordinates" is modelled, I suppose)
> arguably more "like-with-like", but then it raises similar issues to
> those discussed in the creator/contributor/publisher proposal: why are
> we discussing literals if the value is a non-literal?
Pete,
In reviewing this for the call, I'm wondering if the problem
is indeed only with the notion of "numeric identifier" (given
that sets of coordinates and date ranges can be modeled as
non-literal resources). If so, then one might proceed as
with the CCP properties and simply delete the words "numeric
identifiers such as" from the comment. Instead of:
Where appropriate, named places or time periods can be used in
preference to numeric identifiers such as sets of coordinates or date
ranges.
to read:
Where appropriate, named places or time periods can be used in
preference to sets of coordinates or date ranges.
Tom
--
Thomas Baker <[log in to unmask]>
