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

Re: 2.3 URI Ambiguity

From: Mark Baker <distobj@acm.org>
Date: Mon, 1 Dec 2003 13:33:51 -0500
To: "Williams, Stuart" <skw@hp.com>
Cc: "'www-tag@w3.org'" <www-tag@w3.org>
Message-ID: <20031201133351.S4016@www.markbaker.ca>

On Mon, Dec 01, 2003 at 05:54:52PM -0000, Williams, Stuart wrote:
> Hi Mark,
> 
> > Perhaps we should just stick with the other example.
> 
> This indirect reference is a little ambiguous ;-) it would be helpful to
> have an exact reference or quote of the example that you think we should
> stick with.

8-) Sorry Stuart, I was referring to the database/merging example.  I
thought that explained the problem with ambiguity very well.

> This particular piece of text (2.3) has been and clearly continues to give
> us headaches... at our F2F I raised the issue that the case for
> distinguishing between ambiguous use of a URI and indirect identification
> (former 'bad', latter 'ok') had not been made in the text of 2.3.

I figured as much, and was happy with the revised text in that respect.

> The data base example was cast as a case of umbiguous use... however in each
> individual database there was no ambiguity of use. The trouble came with the
> merge operation which did recognise/respect the different uses being made of
> the same URI. One database uses the URI to make reference to a web page, the
> other uses it to make reference to a company. I'll avoid asserting which
> references are direct and which indirect... (since that might depend on the
> URI in question eg. http://www.markbaker.ca). The examples seem to
> illustrate to me that "context of use" (ie. doctype, surrounding content...)
> goes someway to disambiguating the referrent of a reference (made using a
> URI).

True.  So if you point was to suggest that the conference registration
form, by virtue of asking for a registrant's "email address" rather
than their "globally unique identifier", defines the context for
interpretation, I'd agree.  I just didn't see the example explaining
that point.  If it can be fixed, then great, though as I mentioned to
Walden, I'm not sure it can while remaining succinct.

My apologies for the number of exchanges required to get to this
point.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Received on Monday, 1 December 2003 13:31:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:22 GMT