[
Lists Home |
Date Index |
Thread Index
]
This seems to be one of the definitions where the fractures occur in the concrete.
"Markup is fundamentally the identification of information. Marking up a document is effectively a naming process, identifying labels that are appropriate to particular pieces of information."
Markup is:
o Identification of information (uniquely naming)
o Classification of information (labeling from a set of labels)
One can both identify and classify with markup. The rub of
using http in the names is that one must accept that the owner
of the domain name is the arbiter of what is appropriate. We
accept the authority or we don't. The premise that I believe
many who started the web subscribed to was that the authority
could be loose and almost uninterested in the consequences of
ownership of authority. That proved quickly not to be the case.
The premise of the ISO groups that created SGML was that ownership
of authority was a primary concern, so identification and classification
were made separable in system IDs and Public IDs. I assert
the SGMLers had it right. One must explicitly "opt in".
len
From: Simon St.Laurent [mailto:simonstl@simonstl.com]
Tim Bray wrote:
> If somebody wants to make concrete proposals for changes in the
> infrastructure to address these problems, that would be helpful, and
> it might add to the quality of discourse around here too. -Tim
Concrete shoes might be a good idea for URIs...
But as for concrete proposals, see:
http://monasticxml.org/namespaces.html
I'd love it if the specs came around to such concrete practice.
|