Re: [urn] Important URI-related topic: draft-ietf-urnbis-urns-are-not-uris

[John C Klensin <[e-mail]>]

--On Thursday, May 29, 2014 10:19 -0400 Barry Leiba
<[e-mail]> wrote:

> Andy Newton said...
>> A URNBIS session request has been submitted for IETF 90 in
>> Toronto. We have requested this session to discuss the URNs
>> Are Not URIs topic (draft-ietf-urnbis-urns-are-not-uris). As
>> this can be perceived as a substantial change to URI/URN
>> architecture, face-to-face discussion would be beneficial for
>> gaining consensus and facilitating an important and broad
>> discussion. This will be a great opportunity for participants
>> who have not otherwise been following the discussion to now
>> get involved.
> 
> In preparation for this meeting, and whether or not you intend
> to attend it: please, if you have any interest in the topic at
> all, review and comment on the draft.  Please do that now, not
> later: we'd like the discussion in Toronto to be the
> resolution, not the start of the conversation.

Barry (and Andrew),

Thanks.  I really hope we can get this process moving forward.

I had a couple of interesting, but necessarily incomplete, f2f
discussions this week.  I am going to need to think more about
the issues at their core before commenting further, but it is
increasingly clear to me that, when we say "URN" or make any of
the distinctions about, e.g., naming and location that have gone
on here, we trigger very different and almost independent
discussions.  

The first one is fundamentally more interesting to many of us.
It involves the very nature of naming and identification,
edge-case distinctions about the difference between objects and
their names, even questions about what we mean when we talk
about objects and how (or if) those topics change in the
"information age" (or, if one prefers, with an expansive
definition of the web).  One variation of an old joke about a
fairly early forward-looking information storage and retrieval
system [1] is that, with a traditional library catalog, looking
up "cat" would get pointers to books and articles about cats
while, with systems to come, one would need to be specific about
whether information about cats was wanted or one wanted a sample
cat or two and, if the latter, whether their state of health was
relevant.    All of this leads to very fundamental questions
about naming, knowing, abstraction, and epistemology generally.


The topic is of significant interest and importance to anyone
who believes that the introduction of the web (or Internet more
generally) represents or will cause basic changes in how humans
organize knowledge or think or who is trying to predict the
future in those areas.  From that point of view, philosophical
discussions of some of the topics for the last circa 2500 years
have abruptly become irrelevant -- perhaps the ultimate paradigm
shift.  Or, if one has sufficient convictions or understanding
along those lines, the topic may have become completely trivial
and no longer worth addressing.


The other topic and discussion is completely pragmatic.  The
IETF is supposed to be in the standards business.  There are
communities out there have have (at least in their beliefs and
often based on decades or centuries of experience) real
requirements for real functionality.  One might work through the
first discussion and believe that they and their thinking will
rapidly (or soon) become obsolete, but that doesn't make the
perceived need less real today even if dealing with that need
is, in the long term, nothing more than a short-term patch until
a greater understanding spreads.     As a standards body that
lacks either judicial or divine authority, our choices when
faced with such needs are between responding to them in a way
that promotes interoperability and abandoning the requirement to
others to work out their own solutions.

In the URN space, I still believe that the IETF can do a better
and more realistic job of getting the relevant communities
together and working out solutions that are tolerable for
everyone using the Internet (not just the web).  When I stop
believing that, I'll start pushing to close the WG and leave the
work to other groups or bodies.  We may have an educational role
too -- "doesn't, e.g., the semantic web contain a better answer"
is a legitimate question.   But, from that perspective, the
stated needs of legitimate contemporary communities are much
more relevant than the conclusions of those whose concern is
with the longer-term and more philosophical questions above,
questions that, for URNs, start with whether those needs are
"correct" or not.

best,
     john


[1] See, e.g., http://files.eric.ed.gov/fulltext/ED056732.pdf or
http://dl.acm.org/citation.cfm?id=1476862&dl=ACM&coll=DL&CFID=347915301&CFTOKEN=26330164