namespaces discussion

Tim Bray tbray at
Thu Sep 10 21:01:48 BST 1998

At 02:48 PM 9/10/98 -0400, Richard L. Goerwitz wrote:
>Peter Jones wrote:
>> [W]hy can't...elements themselves [have] a namespace URI which addresses
>> a DTD entity for that particular element?
>This is exactly what Tim is alluding to.  If the idea can be formally
>defined, and implemented - and incorporated into the namespace spec -
>we could distribute not only processing, but also definition, of XML
>document types.

This will never happen.  The namespace URI is just a name.  It would
be really wrong to assume you can go there and get a DTD, because:

 1. some namespaces (like HTML, MathML, a few others) are going to be
    very widely used, and it would be silly to force one poor server
    somewhere to guarantee a DTD-on-demand for them
 2. some namespaces aren't going to have DTDs or any other kind of schema
 3. some schemas aren't going to be DTDs

For all of these reasons, the namespace name has to remain just that:
a name.  

We are obviously going to need ways to find schemas associated with
namespaces.  SGML/XML have a way of doing this, the <!DOCTYPE declaration,
but it can only really point to one DTD at a time.  Presumably future
schema proposals will have ways of pointing to multiple parallel schemas.

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

More information about the Xml-dev mailing list