- From: Clark C. Evans <cce@clarkevans.com>
- Date: Thu, 15 Jun 2000 16:20:26 -0400 (EDT)
- To: Tim Berners-Lee <timbl@w3.org>
- cc: James Clark <jjc@JCLARK.COM>, Henrik Frystyk Nielsen <frystyk@microsoft.com>, John Cowan <cowan@locke.ccil.org>, David Carlisle <david@dcarlisle.demon.co.uk>, David Turner <dturner@microsoft.com>, XML-uri@w3.org, Andrew Layman <andrewl@microsoft.com>
I warmly put forth the following suggestion:
1. We set a deadline of 60 days for resolution.
2. If resolution is not obtained, then the default
resolution is to deprechiate relative URI usage.
3. Any proposed resolution should be accompanied
by an open source implementation (preferably
an uncopyrighted diff to JClark's expat or xp )
I would ask for an implementation since our "words"
are vague -- what is required is an actual machine
description so that we don't go off on philisophical
tangents or agree but with different meanings for
the same words (ie, a paper over resolution).
I choose the "deprechiate" option since it seems the
most compatible. If we go literal, then it breaks
those who have to absolutize (if this case exists).
If we go absolutize, then it breaks those from the
literal camp who are not using absolute URIs. In
short, the only solution I see is to deprechiate
relative syntax ... until a time when the identity
operator can be unambigouly defined.
I choose 60 days since it will take at least 15-30 days
for an implementation to be developed, and another 15-30
days for people to become comfortable / suggest and
incorporate improvements with it.
Here's hoping,
Clark
Received on Thursday, 15 June 2000 16:13:36 UTC