W3C home > Mailing lists > Public > www-talk@w3.org > November to December 2001

RE: What is at the end of the namespace?

From: Dan Brickley <danbri@w3.org>
Date: Fri, 16 Nov 2001 09:59:46 -0500 (EST)
To: <Patrick.Stickler@nokia.com>
cc: <roconnor@Math.Berkeley.EDU>, <www-talk@w3.org>
Message-ID: <Pine.LNX.4.30.0111160954240.15883-100000@tux.w3.org>
On Fri, 16 Nov 2001 Patrick.Stickler@nokia.com wrote:

>
>
> > -----Original Message-----
> > From: ext Dan Brickley [mailto:danbri@w3.org]
> > Sent: 16 November, 2001 16:10
> > To: Stickler Patrick (NRC/Tampere)
> > Cc: roconnor@Math.Berkeley.EDU; www-talk@w3.org
> > Subject: RE: What is at the end of the namespace?
> >
> >
> > On Fri, 16 Nov 2001 Patrick.Stickler@nokia.com wrote:
> >
> > >
> > >
> > > > -----Original Message-----
> > > > From: ext roconnor@Math.Berkeley.EDU
> > > > [mailto:roconnor@Math.Berkeley.EDU]
> > > > Sent: 16 November, 2001 03:26
> > > > To: www-talk@w3.org
> > > > Subject: Re: What is at the end of the namespace?
> > > >
> > > >
> > > > -----BEGIN PGP SIGNED MESSAGE-----
> > > > Hash: SHA1
> > > >
> > > > > > Yes, URIs *may* denote abstract resources. No, HTTP URLs
> > > > may *not*.
> > > > >
> > > > > Patrick, as an author of both of those specifications, I can
> > > > > definitively state that what you are saying does not match
> > > > what I intended
> > > > > when I wrote the sections to which you have referred.
> > > >
> > > > I must admit that I always thought that HTTP URIs were
> > some retrivable
> > > > resouce, but after looking at the introduction to  RFC 2068:
> > > >
> > > >    Practical information systems require more functionality
> > > > than simple
> > > >    retrieval, including search, front-end update, and
> > annotation. HTTP
> > > >    allows an open-ended set of methods that indicate the
> > purpose of a
> > > >    request.
> > > >
> > > > It seems that Sean is right.  HTTP URIs seem like they could mean
> > > > anything.
> > >
> > > I don't read it that way at all! I think you are reading your
> > > own interpretation into the language, not judging what it
> > > actually says.
> > >
> > > A 'request' means that something should be provided as a response
> > > to that 'request'. You can't 'request' an abstract entity. You
> > > can only reference it.
> >
> > You can request a representation of an abstract entity.
> >
> > Dan
>
> But your not saying that the representation of the abstract
> entity and the abstract entity are the same. Surely not!

I am not saying that.

> And let's not leave concrete but non-web resources out of
> this. You don't mean that some photo image of you that has
> an HTTP URL denotes *you*, do you? Then how do you make
> statements about you versus statements about the photo?
> You can't.

Use different URIs.

I've written RDF that says, in effect, "there is a person whose mailbox is
danbri@w3.org and who is depicted in a photograph which has a digital
encodeing whose sha1 is blahblah". I could've also given HTTP URIs for the
person. the photo, and its encoding. If I'd given the same URI for the
person and the digitization of the photo I'd have been mistaken, since one
has a size-in-pounds, the other has a size-in-bytes. It's a mistake to
inspect a URI string and try to reason from the fact that it begins with
the string 'http' that it is necessarily the name of a 'document' of some
kind.

I do however think that some URI schemes have been deployed in this
fashion, ie the owners of those URI subspaces have said 'only use these
names to denote things of a certain type' (eg. telephones, java
classes...). To the best of my knowledge, http: names don't work that way.

There's room for both approaches, and for clarification...

Dan

> Sorry, nice try, but (with all due respect ;-) I don't buy it.
>
> Patrick
>
Received on Friday, 16 November 2001 09:59:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 27 October 2010 18:14:27 GMT