[idn] The IRNSS BOF and its implications for IDN

[John C Klensin <klensin@jck.com>]

I've just posted a long-ish note to Liana about IDN identifiers,
languages, and similar issues.  Since I assume some of the WG is
tired of that thread, I'm posting this separately to make an
announcement and provide some additional context.

First, in case it hasn't been clear, the "layered search" or
"DNS search" or "jack up the applications and insert a directory
layer" (all the same thing, but I've concluded that "directory"
terminology may be misleading) is showing up at the Salt Lake
City IETF in the form of the "IRNSS" BOF on Monday morning.  The
mailing list is available:

	ietf-irnss@lists.elistx.com

	See  http://lists.elistx.com/ob/adm.pl for subscription
	and archival information.

Now, a few suggestions and an offer:

draft-klensin-dns-search-02.txt is the core reference document
for that BOF.  I don't expect to let anyone speak, even to raise
questions, who is obviously unfamiliar with it.  If the draft
isn't in the archives by later today (circa 12 hours from now),
I'm going to post it to that mailing list (it is 62Kb long, so
I'm not going to do that unless it is necessary).

The agenda for the BOF is not firm yet. The one that was posted
earlier this week is a placeholder.  A real agenda will be
posted  during the first half of next week.  I am willing to try
to give anyone agenda time to discuss either

	* Reasons why the framework is inadequate and proposals
	for fixing it, i.e., a critique of the document.
	
	or
	
	* Proposals for actually implementing search layer two
	or search layer three.

If you want to do either, drop me a note with a description of
what you intend to talk about, and an estimate of how much time
you need, by Sunday.  Note that I will probably adjust the time
estimates when I do the agenda.

And, _please_, I'd like to get contributions and suggestions
from those who have been pushing approaches in the IDN WG that
seem to be a bit off its target or that require information or
facilities that the DNS lacks.

Giving you the microphone has two additional conditions,
however.  If the pre-meeting conditions are not met, you will be
removed from the agenda.   And, if the meeting conditions are
violated, you will be rudely cut off and asked to sit down.

Condition 1: Meeting time is for discussion, not
"presentations".  If you intend to present a proposal, your time
request needs to point to a document.  That document may either
be a posted (or in queue) Internet-Draft (I-D) or sent to me in
I-D form with explicit permission for me to post it on your
behalf if you don't submit an updated I-D within one week after
the meeting.  The latter path is an exceptional procedure -- the
document will need to be in my hands by 5PM US ET (2200 UTC), 6
December.  No exceptions, even for delayed email.  And I will
read it for content: if it doesn't seem to be written in a
framework that fits into the BOF, or if it isn't a complete
enough proposal to permit evaluation, then I'll return it and
you come off the agenda.   Similarly, if you intend to use a
slide show, I want the slides in hand on the same schedule.

Condition 2: I will open the BOF by doing a very brief
presentation on history, context, and the general model.  That
is going to be the only "presentation".  Attendees will be
expected to have looked at the mailing list, read documents,
etc.  And speakers will introduce their documents or issues only
to the extent needed to establish context, then be prepared for
discussion and answering questions.

For those who have become accustomed to other things, this
approach is fairly close to the traditional IETF way of doing
things.  Meeting time is a valuable resource, and we need to
spend it on the type of back-and-forth discussion that doesn't
occur well on mailing lists, not on reprising things that are
better understood and studied in written form anyway.

I think the session is being multicast.

    john