> -----Original Message-----
> From: ext Roland Schwaenzl
> [mailto:[log in to unmask]]
> Sent: 07 November, 2001 20:44
> To: [log in to unmask]
> Subject: Re: [POLL] What is at the end of the namespace?
>
>
> Dear Patrick,
>
>
> did you have a chance to traverse
> [http://www.dublincore.org/documents/2001/09/17/dcmi-namespace/] ?
>
>
> rs
Looks great, except for two things, both concerning the
statement:
"All DCMI namespace URIs will resolve to a
machine-processable DCMI term declaration for all the terms
within that namespace."
1. Namespace URIs are not themselves required to resolve, nor do
namespace URIs denote vocabularies. Namespace URIs are only
punctuation and do not function in their full URI role. They
are merely opaque, globally unique strings.
Thus even though in the case of DC, as the namespace URIs are HTTP
URLs which *could* be resolved, doing so promotes the misconception
that namespace URIs should resolve; and the document suggests
that the namespace URIs denote the three distinct sub-vocabularies,
when in fact, they simply partition names (yes, this is a fine
pragmatic distinction, but which interpretation you chose has
significant impact on complex vocabularies which utilize multiple
namespaces, or namespaces which "host" multiple vocabularies).
Thus, IMO, the above statement should be removed, and no resolution
should be defined for any DCMI namespace URI.
2. It is ambiguous whether, if the namespace URI resolves to
"something" (which I don't think it should), each term URI also
resolves to "something".
I applaud DCMI in not adopting URI fragment reference syntax in
URIs, which serves to further complicate numerous issues in the
interpretation and use if URIs in general, though it is possible
to interpret the above statement as carrying over to term URIs
as well, such that e.g. http://purl.org/dc/1.1/title should
also resolve to "something".
Though I am of the opinion that neither the namespace URI
nor the term URIs need resolve to anything -- as they are just
terms in a controlled vocabulary denoting abstract concepts,
and do not denote retrievable web resources -- the specific
intent of the DCMI in this regard could be stated a bit
more clearly.
Best Regards,
Patrick
--
Patrick Stickler Phone: +358 50 483 9453
Senior Research Scientist Fax: +358 7180 35409
Nokia Research Center Email: [log in to unmask]
|