[idn] Future of the requirements document

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

As one of the people who (reluctantly) advised the co-chairs to
drop this, let me clarify my position at least...

	* Requirements documents are good. 
	
	* Requirements documents that are either seriously
	ambiguous are not so good.
	
	* When there is a requirements document, protocol
	documents that don't satisfy those requirements are
	really, really bad.

Now, our current situation, as I see it, is:

(i) The WG is long overdue for getting results out.  We need to
finish and wind it down.  Things that block moving forward had
best be really, really, important.

(ii) If there is a requirements document in the queue that
doesn't appear consistent with the protocol documents, it is
unlikely that the latter will move forward before the differences
are resolved.  That is called "more delay".

(iii) The current requirements draft has some writing problems
and dangling references, some technical problems, contains some
requirements statements that the current proposed protocols don't
satisfy.   And we (the co-chairs, the technical advisors, the
ADs) don't believe there is real WG consensus around the current
document (silence or nods from people who haven't read it
carefully don't count).   Fixing it would take significant time
and energy.  The experience of the past two years leads us to
believe that the energy isn't present, even if we were willing to
invest the time (and delay with WG).

On the other hand, one of the most important values of a
requirements document is that it provides a foundation for, and a
record of, a statement about the problem the WG tried to solve.
Having that seems quite important to us, but we think that
writing it and getting consensus on it would likely be a much
quicker process than cleaning up the requirements piece and
getting consensus on it.

That is, again, subject to discussion.  But I'd hope that those
who advocate keeping/ fixing the requirements doc explain why it
is important enough to keep the WG in operation --and hold up the
protocol documents-- for what I think would be many extra months.

      john