Re: Mechanism, not policy [was: Attribute uniqueness...]

> Yes.  You call it a fundamental misunderstanding -- but you
>  will have to admit that it is a consisetnt understanding, that a
> lot of others have it,

No, I can't admit that it is either consistent or that anyone using
namespaces has this understanding.

Check back over xml-dev or cmp.text.tex or xsl-list or the namespace
rec. You will see the question "what do I put at a namespace URI"
comes up from time to time. The answer with 100% consistency by anyone
who has ever cared to answer, including editors and others involved in
creating the spec has _always_ been "nothing" or "it doesn't matter"
or "whatever you want". Never have I seen anyone suggest that what is
identified by the URI _is_ the namespace. The idea is bizarre. Until
yesterday the thought had never struck me that anyone could ever think
such a thing. It has the process backwards: to name a namespace you
take a URI reference, preferably the URI of some resource you control.
The idea that the namespace somehow already has a URI and that's its
name is just so strange.


> >If it were true then you would need the forbid
> >option as relative URI don't refer (in themselves) to resources.

> The step of absolutizing is generally considered a short step
> so your īn themselves could be seen as splitting hairs.

Actually I mentioned that because of your insistence earlier that
relative URI don't have any meaning out of context, and have to be
combined with a base.

> There are two ways of looking at this. Please admit of the other.

I have to admit of the other (now) but in the best part of two years
of using namespaces and following discussions on xml-dev and other
places it had never occurred to me that that was even a possible
interpretation. 

> You seem not to be very familiar with the processing of URIs,

Familiar enough, but what I know or don't know is not particularly
relevant.

> You can propose a totally new system
> of identifiers for namespace names if you like ,

I did not propose it. It was endorsed by the Director of the W3C as a
Recommendation. Several people (including myself) have commented that
using URI as namespace names was perhaps not the best idea as it leads
to false beginner assumptions that dereferecing the name will turn up
something related to the namesace, but its done now and people get
used to it. If Tim Bray posted "a namespace name is just a name"
to xml mailing lists once he did it a thousand times.

> I would like you to really earnestly consider using the URI
> space.

Using "URI space" as you put it is totally unacceptable if relative
URI are used. The idea that XML element names change or become
void as a document is moved is just wrong.

For absolute URI using "URI space" is still bizarre but in that case
the differences do not show up for any namespace processing so we
could agree to disagree in that case. That is for absolute URI
namespace processing is just effectively the literal interpretation.
Namespace processors wouldn't ever care (or look) what resource
was identified by the namespace name. You could assert that the
resource _was_ (some attribute of) the namespace but this would be
relegated to a philosophical argument.

David

Received on Thursday, 8 June 2000 03:38:59 UTC