Katz, Stephen (GILW) <[log in to unmask]> wrote:
> I have been following these discussions with interest, but I wanted to make
> a more general comment to raise the question of whether "Namespaces" is in
> fact a stable and completely ripe standard....
No standard is ever completely ripe, but namespaces are required for the
work we're doing and they seem to have no serious gaping problems. Let's not
throw water on a standard like this that is so widely implemented. If we're
going to use DCMI stuff in the XML and RDF worlds, we need namespaces.
> Namespaces does not provide any help at the moment as to what to do with
> each concept that is defined.
Of course not! Namespaces are a very basic concept, upon which we build more
powerful things (RDF, file formats, etc.).
> "the namespace draft doesn't care what (if anything) they point at."
>
> To me this is frightening news for any project that strives to establish
> automated procedures using namespaces. What is the point of arguing about
> what version of what element/qualifier/attribute/language should be in what
> namespace if there are no rules within the namespace specification that
> unequivocally tell us what we will find at the namespace resolution
> address??
Why do you care what's at the address? Sure, it would be nice to put up a
description of the namespace, etc. (I currently prefer an
XHTML/RDDL/RDF-Schema combination) but we can discuss this after we decide
on namespaces. What's important that we have namespaces and that they are
described in Dublin Core material, etc.
> URL of a namespace, what will be the practical impact of discussing (and
> perhaps even agreeing on) what namespace should contain what DCMI content?
That XML and RDF formats will achieve interoperability, of course!
> The decisions taken within the DCMI in this respect will end up being a kind
> of DCMI implementation of Namespaces.... Perhaps this (and other similar
> approaches to using namespaces) can eventually be fed back to the W3C
> process to further refine the namespaces specification?
There's certainly been more than enough discussion about namespaces given to
the W3C -- they're a very hot issue and I doubt that the W3C will be likely
to fix things any time soon.
--
[ Aaron Swartz | [log in to unmask] | http://www.aaronsw.com ]
|