xml-dev - Re: Irony heaped on irony

[Dan Connolly <connolly@w3.org>]

David Megginson wrote:
> 
> "Simon St.Laurent" <simonstl@simonstl.com> writes:
> 
> > >I could move the schema, but that would break lots of _other_
> > >schemas, including the schema for schemas, which depend on it.
> > >
> > >Seems to me having something of mime type text/xml at the namespace
> > >URI for XML is not something we should have to apologise for.
> >
> > But it does seem that such a sweeping change in namespaces best practices
> > is worth an explanation or preferably a full-blown trip through the W3C
> > process, complete with working drafts.
> 
> Tim and Simon are right.
> 
> I will freely admit to not being aware of all of the minutiae of XML
> Schemas, but Henry's approach seems dead-on wrong to me.

Why?

As to Simon's point about process, the content of
http://www.w3.org/XML/1998/namespace is at the discretion of the W3C
webmaster;
it's not the result of any WG decision or Recommendation or anything.

As to Tim's point about the loss of the ability to get something
that works with a web browser, I agree that's unfortunate and
I expect to fix it. But I don't expect to remove the schema
in doing so.


>  The schema
> for schemas (and others) should reference an XML schema for the xml:
> Namespace using the xsi:schemaLocation attribute, as in
> 
>   xsi:schemaLocation="http://www.w3.org/XML/1998/namespace
>                       http://www.w3.org/XML/Schemas/xmlschema-20000518.xsd"

Why? Why use schemaLocation when there's no need to?


> Please remove the schema from http://www.w3.org/XML/1998/namespace and
> put it somewhere else, then update the schema for schemas to follow
> this (better) practice.  This also has the advantage that users can
> refer to the specific version of the xml: schema that they want to use.

Hmm... issuing an address for each specific version of the schema
spec that we put on the web server is a reasonable idea, but
removing the schema from http://www.w3.org/XML/1998/namespace
removes the ability to use namespaces in schemas in the most
straightforward way.

> > We've spent over a year on XML-Dev and elsewhere explaining to the
> > world that Namespace URIs are just identifiers, battled over the
> > three/one namespaces for XHTML issue, and now it seems that
> > namespaces are indeed supposed to point to schemas.  (And packaging?
> > Is that gone?)
> 
> I think that the XML Schema WG got it right with the
> xsi:schemaLocation attribute,

sure, it's a useful mechanism when you need it. But when
you don't need it, i.e. when the same party decides the
namespace name and the mechanism of publishing a
definition of it, why bother?

There was a urn attribute in HTML:
	<a href="http://..." urn="uuid:....">...</a>
but nobody bothered to use it and I think it didn't
survive into recent HTML versions.

> and I'm puzzled that they've failed to
> follow their own recommendation (that's small-r "recommendation", not
> "Recommendation").

use of schemaLocation is a MAY, not a SHOULD (i.e. not
a recommendation), I think... yes:

	"The xsi:schemaLocation and xsi:noNamespaceSchemaLocation
         attributes can be used in a document to provide hints as to the
physical
         location of schema documents which may be used for validation."

	-- 2.6.3 xsi:schemaLocation, xsi:noNamespaceSchemaLocation
	http://www.w3.org/TR/2000/WD-xmlschema-1-20000407/#xsi:schemaLocation


I think that on the contrary, it's important that the schema spec
leads the way toward the self-describing web.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************