Namespaces are dead.

[Rick Jelliffe]

From: Jonathan Borden <jborden at mediaone.net>

>Why ought namespaces be dead? Is Biztalk that important? Just don't use
>Biztalk or the XML Data namespace if you don't want it, this is a free
WWW!

Unlike Steve Newcomb and like Tim Bray, I think namespaces do solve a
different problem to architectural forms. Namespaces provide a way to
treat a local name as a public identifier.  I like namespaces; they are
convenient but XML needs to have available ASAP something to prevent
them from being used by vendors to capture data.

The big door that architectural forms opened to allow multiple schemas
operating on the same data: whether for structure description, data
description, competitive schema languages, or any other reason.

All along people have perceived clearly that the namespaces were (if
tied directly to schema specifications) a way to prevent "open"
extensiblility and instead to capture data to single, probably
proprietary schema tools and systems.  The extensibility of XML not only
refers to adding new elements, but in preventing data capture (this is
why PIs have targets, for example.)

The stylesheet PI is good, because multiple PIs are possible: you dont
need to generate different data depending on the other end (in the
pleasant future). Tieing particular schemas to namespaces is bad,
becuase it forces a document to use XML data or XML schema only: we are
right back in slack old HTML land. It is disgusting.

>... You can always use a "urn:" based namespace URI to prevent
>linking the namespace to a schema.

But then I still have to reprocess my names to if I want to send BizTalk
or use XML Schema.

> One issue to consider, however, is the impact of either mechanism of
>schema/namespace association on how a document containing elements from
>multiple namespaces ought be validated.

I think the open/closed content model directly impacts that: some places
the schema will allow you to extend without being particularly
interested in what happens. This is pretty much in architectural-forms
territory. I cannot see why a schema-selecting PI could not also allow
architectural-forms, but that is not a requirement for this problem.

Rick Jelliffe


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)