W3C home > Mailing lists > Public > www-tag@w3.org > July 2003

RE: "On the Web" vs "On the Semantic Web"

From: Bullard, Claude L (Len) <clbullar@ingr.com>
Date: Fri, 25 Jul 2003 15:45:17 -0500
Message-ID: <15725CF6AFE2F34DB8A5B4770B7334EE022DC6AF@hq1.pcmail.ingr.com>
To: "'Norman Walsh'" <Norman.Walsh@Sun.COM>, www-tag@w3.org
Cc: "'Tim Berners-Lee'" <timbl@w3.org>

When no consensus is achieved for clarification of a 
requirement, the chair has three choices, IME:

1. If no consensus is possible, these are not the same 
requirement.  Most likely the case here.

2. If consensus is possible but clarification has not 
been achieved, the chair must push the work back to 
the subcommittees where they can either resolve it 
and present the case for an updated document, or let 
it die there.

3. If clarification has been achieved in the opinion 
of the chair, and a minority is resisting the definition, 
it is time to vote and move forward.  Significance is 
it the eye of the beholder.

Process can be a lovely and useful thing.

Easier said than done, but having read these same 
arguments since before the web was the web (basic 
issues of ontologies), I don't think this issue 
should keep the web architecture document from 
moving forward unless a clear case is made that 
the current architecture fails to function acceptably.
If a future architecture can fault, that is future 

TimBL writes: "The concept of 
identity has to do with what is invariant about the measurable 
qualities of something, and in this case they are the information 
content, not the subject."

No identity without identification (a measure needs 
both a ruler and an eye to use it).

That is also the definition of a class which can have 
members with variant properties.  Those variant 
properties enable new members or in Tim's terms, 
division of the space although that smells of 
subclassing.  One can't conflate instances with 

"The architecture is *not* one in which the classes are deduced from 

Yes, but the subclasses are contexts because they are 
based on the variant properties, not the invariant 

<rant>A definition that does not include context 
of use will not scale for universal labeling.  
A measure is the definition of the context and 
a system of measures is a context-evaluating system.

The URI is a dereferenceable name.  One 
may get an error.  That is all the web 
architecture knows; that is all it needs 
to know.


From: Norman Walsh [mailto:Norman.Walsh@Sun.COM]

/ "Bullard, Claude L (Len)" <clbullar@ingr.com> was heard to say:
| Norman writes:
|>I don't know if that's a useful point to raise or not, but I am very
|>deeply concerned that we appear to have at hand an issue that is
| It isn't intractable.

Well, perhaps I could have been clearer. It's not intractable in any
technical sense. That is to say, engineers will continue to deploy
useful software irrespective of what anyone says or even what the
right answer is.

But from a committee chair's point of view (I'm wearing the chair hat
for this meeting after all :-), if you have a significant minority
that say something is X and a significant minority that say something
is Y and neither side can be persuaded to accept that the opposing
side's X or Y is the right answer and no compromise position Z that is
acceptable to both can be articulated *and* you want to move forward on
a consensus basis, that's an intractable problem.

| We really really should have stuck with 
| PUBLIC and SYSTEM identifiers.  I never 
| encountered the nuttiness with those 
| that URIs seem to provoke.

That's a opportunity to reopen the names vs. addresses permathread :-),
but I'm going to resist. Must...send...message...quickly...

                                        Be seeing you,

- -- 
Norman.Walsh@Sun.COM    | The perfect man has no method; or rather the
XML Standards Architect | best of methods, which is the method of
Web Tech. and Standards | no-method.--Shih-T'ao
Sun Microsystems, Inc.  | 
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.7 <http://mailcrypt.sourceforge.net/>

Received on Friday, 25 July 2003 16:45:24 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:56:00 UTC