Re: httpRange-14 , what's the problem

Tim Berners-Lee wrote:

> >
> > DC: RF, so you conclude I can point to my car with an HTTP URI?
> > RF: Yes.
> > TBL: I have in my mind a consistent model where HTTP URI points to a
> > document about a car. I don't have a consistent system where HTTP URIs
> > designate cars.
>
> Yes.   This is indeed the issue.  I was asked (in the TAG)  to explain my
> position and
> have done my best in the time available in
>  http://www.w3.org/DesignIssues/HTTP-URI.html
>
> This explains why I find alternative positions which have been posted
> so far unacceptable.
>

I've started to read through this. Without committing one way or the other
to this viewpoint, there are two associated, (somewhat) practical matters
that I would like clarification on:

1) If HTTP URIs necessarily identify documents, what does

http://www.w3.org/2001/XMLSchema

identify?

a) The XML Schema Definition _Language_
b) The specification that describes the XML Schema Definition Language

It seems to me that an XML Namespace is not a document, and so by your
reasoning we should REQUIRE that XML Namespaces MUST have a '#' which leads
me to the second question:

2) Can a URI _reference_ that starts with "http:" identify something other
than a fragment of a document? Now you may say that as long as the
content-type = application/rdf+xml, it might identify anything, but what
about content negotiation? Suppose a URI returns either application/rdf+xml
or text/html depending on conneg? Can an isolated URI reference ever
identify something that is not a fragment of a resource representation?

I am not opposed to saying that a URI reference might identify anything, I
am just a bit unsure about saying that the range of a URI + fragment-id is
greater than the range of the URI itself.

Jonathan

Received on Monday, 29 July 2002 13:42:53 UTC