Re: Namespace names and URIs

-----Original Message-----
From: Dave Hollander <dmh@commerce.net>
To: xml-uri@w3.org <xml-uri@w3.org>
Date: Wednesday, May 17, 2000 12:01 PM
Subject: Re: Namespace names and URIs


>On Tue, 16 May 2000 01:57:40 -0400 (EDT) Tim Bray wrote:
> >
> > I apologize in advance for the length of this message. We're all
> > busy but this is important. But here's a summary:
> > 1. URIs are sound in their design, just as as TimBL claims

yes

> > 2. Namespaces set out to solve the problem of naming things, no
> >     more, and they succeeded


yes

> > 3. It is reasonable to want more from namespace names, and for
> >     this reason, the fact that they are syntactically URIs is good, as
it
> >     leaves open the door for the building of the Semantic Web


yes

> > 4. It is wrong to compromise the basic utility of namespaces by
> >     imposing strict URI-ness on them


not sure what that means.

> > 5. The use of relative URI references as namespace names is wrong
> >    and dangerous and should be, at the least, deprecated


disagree.  I have argued why elsewhere. People do it where it is in fact
safer than using absolute URIs,
as in self-referential schemas. Real world examples have been quoted.


>I agree with Tim Bray.  I can accept a softer warning about the use
>of relative URI references as I trust the community will strengthen it in
>practice.

>As for point 3, I think namespaces *should* be more than just unique
>names, but the namespace rec does not provide it.

It says that dereferencing was not a goal of the spec.
I read this as saying that the NS spec does not require you to dereference a
spec.
Had I thought that the meaning was to discourage the use of the NS URI to
point
to a schema, then I would have objected most strongly at the time.

>  Remember the
>public reaction to the introduction of the namespace rec? This reaction
>included lots of dire warnings that namespaces are underspecified.
>I believe this issue is just another example. The namespace Rec leaves
>namespace semantics up to the application.


Maybe we should at this point put it down clearly on the table that the
namespace URI is taken to identify the semantics of hte namespace, and
that an important direction of technological development is toward
langauges for defining the syntax (xml-schema) and in certian ways semantics
(rdf-schema, ...) in a machine redable way.  I wouldliketo point out that
though not at all
mandatory it can be very useful to make such descriptive documents available
in response to requests for the namespace resource.


[...]

>We don't have the foundation for specifying how namespaces are anything
>but a bunch of names. Now that some of us have been using namespaces
>for a while, perhaps we can develop a proposal for namespace semantics.
>I think that might be interesting and may be willing to participate if
others
>consider it is a good use of time and effort.


We have a pretty good schema language for describing the syntactic
properties of a namespace and that is a start.  (The XSV validator
demonstrates the power of that).   RDF  (which originated the request for
namespaces) has a language for introducing limited semantic information
(subproperty etc).  We also have a lot to gain even from putting a
human-readable document there.  I disagree that we have no foundation. We
have the whole URI system -- so longas we leave it unmolested. We have two
schema languages.  There is a lot of momentum. Will people please stop
pulling the rug out from under it all?

Noah M pointed out to me that we have to come clean and point out this
direction.  It is too expensive to have the community split between those
who now take it for granted and those who are resistant to it.   Can we
agree on this, please?

Tim BL

>Regards,
>Dave
>
>
>__________________________________________________________
>Dave Hollander
>Co-Chair W3C XML CG and Plenary
>Co-Chair W3C XML Schema Working Group
>dmh@info2000.net
>970-613-0605
>__________________________________________________________
>"Of all the things I've lost, I miss my mind the most"
>Sheila Landy, mother of 2
>
>

Received on Thursday, 18 May 2000 02:33:55 UTC