Re: [httpRange-14] What do HTTP URIs Identify?

/ "Jonathan Borden" <jonathan@openhealth.org> was heard to say:
| Tim Bray wrote:
|> Roy T. Fielding wrote:
|> ...As such, allowing the
|> > identifier to identify more than one concept is a non-starter.
|>
|> Er... and how do you disallow identifiers from identifying whatever
|> people think they identify?
|
| A slightly subtle point, but when we say that a URI necessarily identifies
| only one concept, this is saying that the _name_ of the concept is the URI.
| Now people may assert different things about the concept, and these
| assertions may be contradictory, but they are identifying the same concept
| if they use the same name for the concept, i.e. the same URI.

Uhm, really?

  I assert that http://example.org/xyzzy identifies my car.
  I assert that http://example.org/xyzzy is green.
  I assert that http://example.org/xyzzy is a pickup.

  I assert that http://example.org/xyzzy identifies my wife's car.
  I assert that http://example.org/xyzzy is green.
  I assert that http://example.org/xyzzy is a sedan.

I'm never quite sure when I'm wandering off into the semantic weeds,
but it seems to me that both of those sets of statements are
internally consistent without being simultaneously consistent. And the
center of the inconsistency is that they are identifying two different
concepts using the same name for the concept.

I understand Tim Bray to be saying that such inconsistencies are a
simple fact of life and nothing can be done about it. I have to agree.

| Across the broad information space of the Web, the problem is sustaining any
| "promise" that what each person says about the URI will be consistent. But
| again the name of the concept is the URI so each use of the URI by
| definition talks about the same concept.

I don't see how the assertion that "each use of the URI by definition
talks about the same concept" can be justified.

| <rdf:Description rdf:about="http://example.org">
|     <ex:Color rdf:resource="#Red"/>
| </rdf:Description>
|
| and somewhere else
|
| <rdf:Description rdf:about="http://example.org">
|     <ex:Color rdf:resource="#Red"/>
| </rdf:Description>
|
| The _thing_ i.e. http://example.org is the same, even though the statements
| made might be contradictory.

Well, you've used the same name. To the extent that URIs are absolute
and context-independent, you can say they're the same, but I thought
the point of statements like this is that they could be used to make
assertions about things, not just the names of things.

|> ...The
|> Semantic Web has to be able to tolerate the fact that you can't know
|> what a resource is, and thus different parties may not have a shared
|> perception of this, just like the Web needed 404 to work. -Tim
|>
|
| On the Semantic Web you might not ever know everything there is to know
| about a resourse, but you can know some things.

Including some inconsistent and mutually exclusive things?

                                        Be seeing you,
                                          norm

-- 
Norman.Walsh@Sun.COM    | So, are you working on finding that bug now,
XML Standards Architect | or are you leaving it until later? Yes.
Sun Microsystems, Inc.  |

Received on Saturday, 3 August 2002 10:09:03 UTC