namespaces vs. schemas

[John Cowan]

Dean Roddey wrote:

> Hmmm. Oh well, I guess the idea is still valid though in terms of DTDs and
> Schemas in general. As they proliferate and go through versions, the ability to
> assign some form of universal ID in such as way that applications can get that
> information from the parser would be a very useful way of dealing with some of
> the potential confusion that could arise. It would be pretty simple, avoid any
> need for centralization, etc...

You could insert a PI like

	<?DTDMatch hash="93FEA302CAB32932-23923BAC203FEA38"?>

at the top of the DTD.

> It would be particularly easy to apply it to Schemas and XML files, because it
> would only require the agreeing on of a particular element type of attribute
> type or extra blah="duh" statement in the <?xml ?> line perhaps. Since DTDs
> don't have anything like that (do they?), I'm not sure how it would applied to
> them, though they are the files that would best benefit from such a thing.

DTD files (entities containing external subsets, that is) can begin
with "<?xml ..." PIs specifying the XML version and the encoding,
just like all other external entities.  Only the document entity
can have a "standalone" pseudo-attribute, though.

> Ok, so I'm confused. Maybe its just a V vs. NV thing or something. If I define
> a namespace, don't I also define a set of tags that are valid within that
> namespace?

Yes, but no particular way of doing this is prescribed: it can be
DTD, XML-Data, XSchema, prose in English or Japanese, or what you
will.  Furthermore, the namespace name URI doesn't necessarily point
to this description.  In the previous ns draft, the namespace PI
could have "src" pseudo-attributes, potentially more than one, which
defined actual schemas.  This facility has been removed from the
current draft.

> I always assumed that the URL that the namespace mapped to would be
> more than just a human readable unique identifier of some sort, and that it
> would define the tags that legally belong to that namespace.

It ain't necessarily so....

> So am I missing something here? Do you mean its just not required, or that its
> not been thought of at all?

It has been thought of, but is apparently considered too large a
problem to handle as yet.

-- 
John Cowan	http://www.ccil.org/~cowan		cowan at ccil.org
	You tollerday donsk?  N.  You tolkatiff scowegian?  Nn.
	You spigotty anglease?  Nnn.  You phonio saxo?  Nnnn.
		Clear all so!  'Tis a Jute.... (Finnegans Wake 16.5)

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)