W3C home > Mailing lists > Public > www-tag@w3.org > July 2002

RE: TB16 Re: Comments on arch doc draft

From: Bullard, Claude L (Len) <clbullar@ingr.com>
Date: Wed, 3 Jul 2002 09:56:01 -0500
Message-ID: <2C61CCE8A870D211A523080009B94E430752B63A@HQ5>
To: "'Patrick Stickler'" <patrick.stickler@nokia.com>, ext Tim Bray <tbray@textuality.com>, WWW TAG <www-tag@w3.org>

I agree on that point.  My point was that the 
security issue goes elsewhere to be resolved. 
Business will work that one out.  The Semantic 
Web isn't doomed to failure, but I suspect that 
it will be an uneven success for other reasons. 
Anyone who sits down to do even simple DTD 
design is aware just how much naming the names 
becomes a football.  Deep classification will 
be even tougher regardless of the resilience 
of the technology that supports the process. 

As the song goes, "you have to fight for the 
right to party."  I assert that the application 
of dereferencing is always a local information 
owner decision, and that the decision should 
be as easy to make as possible, and that decision stated 
IN THE CONTENT.  Otherwise, standards governing 
content aren't as helpful as they can be.

"I'd like to look at the reality of the modern Web. 
Standards-compliant browsers with beautiful rendering, 
ADA section 508, the Internet Explorer monoculture under 
assault from Gecko and non-PC-form-factor devices, Web 
Services. I'd like them to conclude that the only sane, 
sensible, affordable way forward is to go standards-based 
on all their content.  --Tim Bray"

No disagreement here, but again, sauce for the goose, 
sauce for the gander.

Your issue points to the ever growing debate 
about names, rights over names, and the governance 
of the Internet via decisions made 
over architecture.  At that point, our two 
issues intersect.  The TAG is an architecture 
recommendation committee.  The problem of the 
term "SHOULD" is that it is a means of social 
engineering.  That isn't something, in my 
opinion, that is the TAG's responsibility, 
but then it is a question of who's it is 
and can it be avoided here.  SHOULD is about right 
because it is an enabler and doesn't remove 
the option.   The rest of the rant over 
nefarious, incompetence, etc., is simply 
over the top and has a paralyzing effect 
given a consortium that has consistently made 
one decision (namespace names are just 
names) then later come back to augment 
that with a social position (namespace 
names SHOULD be dereferenceable).

The WWW was fielded stupidly.  We will 
have these issues for many years and they 
are affecting levels of society that 
apprarently the inventors could not imagine 
or did not feel empowered to resolve.  Spilt 
milk, but to be cleaned up and the question 
is, does the decision to insist on dereferencing 
make that easier or harder.  Governance at 
the level of the architecture is what Lessig 
asserts no one notices, but in truth, many 
of us have known for a very long time would 
be the ultimate outcome of the lab experiment 
escaping the lab.


From: Patrick Stickler [mailto:patrick.stickler@nokia.com]

> This isn't an issue.  It is a philosophical debate

I would consider the adoption of ambiguity as to
what a URI or URI reference actually denotes to be
a rather critical issue, and one that can severely
impact the functionality of the Semantic Web.

If a given URI or URI reference does not in fact
denote one and only one thing, then the Semantic
Web is doomed to failure.

Having namespace URIs that resolve to namespace
documents represents such ambiguity, and the TAG
saying that namespace names SHOULD or even MIGHT
resolve to anything is an invitation for even more
ambiguity in the interpretation of URIs in other
Received on Wednesday, 3 July 2002 10:56:42 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:32:32 UTC