Re: [xml-dev] namespaces redux (was: Re: [xml-dev] [XML Schema]Here's ho

[Liam R E Quin <liam@w3.org>]

On Tue, 2012-10-16 at 16:09 +0100, Andrew Welch wrote:
> On 16 October 2012 15:54, Liam R E Quin <liam@w3.org> wrote:
> > On Tue, 2012-10-16 at 11:47 +0100, Andrew Welch wrote:
[...]

> > One problem is that you can create non-portable documents - software is
> > free (unfortunately) to resolve "config" as a relative URI reference
> > using the base URI of the containing resource, so that the actual
> > namespace URI in the document would vary depending on the document's
> > location.
> 
> ....which is clearly madness, surely, right?

One person's madness is another person's genius.

> > Another is software that will go use that relative URI to fetch a schema
> > of some sort.
> 
> I also really wouldnt want that happening either...
It does, though, fairly often.

> > I was disappointed, as I'd always hoped we (the XMLers) would define a
> > catalog of some sort that could be located at a namespace URI and used
> > to locate schemas, DTDs, stylesheets, documentation, entity definitions,
> > code fragments and so forth. RDDL-style. But the consensus was that such
> > things are neither forbidden nor encouraged, and relative URIs should
> > not be used as a namespace name.
> 
> why wedge it in with namespaces?  Why not some other xml: prefixed attribute?

Because I wanted the concept of validating a namespace - I wanted a
document that gave a list of visible names, and possibly also code to
implement behaviour associated with those names. I didn't win that one,
although I revisited some of it with "automatic namespaces" and
"unobtrusive namespaces" a little.

Liam

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/