With apologies for the slow response....
On Tue, 8 Oct 2002 16:04:05 -0400, Naomi Dushay <[log in to unmask]>
wrote:
> The issue is that Xerces 2.2.0 complains if you assign a
> namespace prefix to the XML namespace URI:
> http://www.w3.org/XML/1998/namespace (though it is still
> fine to import it.) This behavior is currently peculiar to
> Xerces 2.2.0 ... but I'm not sure if Xerces 2.2.0 is right or
> wrong in it's complaint.
Some truffling around on xmlschema-dev returns:
http://lists.w3.org/Archives/Public/xmlschema-dev/2002Jul/0181.html
http://lists.w3.org/Archives/Public/xmlschema-dev/2002Jul/0184.html
which reference
http://www.w3.org/XML/xml-names-19990114-errata
Under NE05 that says:
===========
The prefix xml is by definition bound to the namespace name
http://www.w3.org/XML/1998/namespace. It may, but need not, be declared,
and must not be bound to any other namespace name. No other prefix may be
bound to this namespace name.
===========
That last sentence is the decisive one here, and it means that a namespace
declaration of
xmlns:x="http://www.w3.org/XML/1998/namespace"
is invalid, rather than redundant.
So yes, it looks as if Xerces 2.2 _is_ implementing that errata correctly
here. Recent versions of XSV also now report an error. It's a (relatively)
recent errata, which I guess explains the change in validator behaviour
and also why some other parsers haven't caught up.
FWIW, this has hit other communities too e.g.
http://www.imsglobal.org/developers/ims/imsforum/attachments/IMS%
5FMDBinding%5FChange%5FRationale%2Epdf
I think Naomi's suggested fix works and I've put an amended version
temporarily at
http://www.ukoln.ac.uk/metadata/dcmi/simple/simpledc20021009.xsd
Pete
|