Namespace URI address resources

[Jonathan Borden]

Andrew Layman wrote:

>Recent mail has suggested that a namespace URI could be used to attempt a
>retrieval of a schema, either using the URI directly or in some modified
>form or via content negotiation.  Other recent mail has objected, saying
...
>
>This argues that the first position is correct, namely that using a
>namespace URI to retrieve a schema is reasonable and is different in kind
>from retrieval of other possibly-related resources.
>

    It is true that schema definition being akin to class definition has a
unique position in the document grove. Yet so does a namespace and
overloading the mechanism by which a schema is attached to a document onto
the mechanism by which a namespace is attached to an element yields
difficult problems.

    The essential difference is that the namespace is defined by the URI
whereas the schema is not defined by the URI rather the contents of the
resolved URI. How might this create a problem? Suppose I have a document
with a schema at a particular URI resolved by the "http" protocol e.g.
http://www.microsoft.com/schemas/x.scm
now suppose I wish to e-mail this document behind a firewall, or place the
document into a database... behind the firewall, the "http://" protocol is
blocked due to corporate policy. Similarly if decades or centuries pass, the
data within the database remains valid but the http://www.microsoft.com
address has become inactive. The namespace remains the same. The schema is
unavailable.

    Suppose I solve this problem by encapsulating the XML document in a
multipart/related MIME message, placing the schema at a location labeled
"Content-ID: schema". Now the schema is accessed via the URL "cid:schema",
all is good, we can resolve the URL ... but the namespace changes!

    This problem could be solved by use of a PI to declare the schema
<?xml-schema href="URL" ?>. Gateway software would thus be able to
encapsulate and transmit XML documents by different network protocols,
keeping such protocols independent of the document namespace. All that need
change in the document is a PI.

Jonathan Borden
http://jabr.ne.mediaone.net (for now :-)



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/ and on CD-ROM/ISBN 981-02-3594-1
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)