[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Content Negotiation not enough for schemas (was Re: Namespace URI addres

• From: "Rick Jelliffe" <ricko@a...>
• To: "'XML Dev'" <xml-dev@i...>
• Date: Fri, 18 Jun 1999 04:40:52 +1000

From: Murray Maloney <murray@m...>

From: Rick Jelliffe

>Using the namespace URI to be locate a schema does not preclude
>the possibility of there being any number of schemas associated
>with that URI.

As I mentioned.

This is just the self-describing/header-described problem again:
should an XML document
assume that all resources are made available over the Web, or should it
support unplugged systems also? In the case of the XML encoding PI,
it was decided that relying on Web mechanisms was not enough, because
an XML document can have a valuable life before and after a WebServer.

In any case, the only resolution system currently around for URIs is
content negotiation, which relies on all schemas being at the same
location and being of different MIME types, as I understand it.
What about when the schemas are on unrelated servers or use the
same schema language or when you want the schemas available from
local "file:" URLs?

>Nobody is proposing that a document has only *one* schema.
>The mechanism is known in the biz as 'content negotiation'.
>This is known to work for multiple languages and devices.

See above.

The other trouble with content negotiation is that it has to be
supported
at the browser level: the requesting agent has to send its preferences
to the server. There is no (markup) mechanism to make these preferences
part of the document.

Has any vendor announced that they will support multiple schema
validation through content negotiation in their browser or database
product?  Is it likely?

Since we can assume that vendors will not support plurality of
schemas in different schema languages at different locations
suitable for different processing model (especially if the browser
or datavase vendors are also schema-tool vendors), user will have put
in PIs (or similar) in their code to indicate schemas for home-grown
tools. If they want to provide open schema access, then
these PIs will need to be in some common form.

Rick Jelliffe


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

• Follow-Ups:
  • Re: Content Negotiation not enough for schemas (was Re: Namespace URI address resources)
    • From: Dan Brickley <Daniel.Brickley@b...>

• Prev by Date: Re: More CDATA questions
• Next by Date: Another reason why namespaces URLs should not be schemas
• Previous by thread: Re: Top-down or bottom-up?
• Next by thread: Re: Content Negotiation not enough for schemas (was Re: Namespace URI address resources)
• Index(es):
  • Date
  • Thread