Re: [urn] Summary of WG decisions Re: Feedback on draft-ietf-urnbis-semantics-clarif-03

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

--On Tuesday, May 31, 2016 15:39 +0200 Julian Reschke
<[e-mail]> wrote:

> On 2016-05-31 15:31, Andrew Newton wrote:
>...
>> Just for my sanity, I reread
>> draft-ietf-urnbis-semantics-clarif-03.txt. It seems to me
>> that this draft does exactly the thing you have requested.
>> ...
> 
> It might *intend* to do it, but (a) it's very unclear what the
> actual change is, (b) whether it is needed, and (c) whether
> that it tries to change actually is consistent with what the
> URN spec attempts to do.

Julian (and Andrew), 

Just so you know (and to repeat what I've at least hinted at),
I've been looking at  draft-ietf-urnbis-semantics-clarif since
Julian's first posting on the issue to see if I could find thing
that could reasonably be improved.  I've found several [1] and
put new text into a working draft for -04 but have not posted it
pending input and direction from the co-Chairs [2].  

I don't believe those changes will satisfy Julian's concerns.
The reasons are connected to why I've been looking for
instructions from the co-chairs and summarized in Andy's note.
>From my point of view, this document updates 3986 _for URNs
only_, it is not an attempt at a general update (that should
have been quite clear from the first sentence of Section 4; -04
will be even more clear on the subject).  In principle, we could
do a paragraph by paragraph review of what portions of 3986 are
affected by that, but it would be a lot of work, it would make
this document longer and more complicated, and I don't believe
that anyone would read it except for the purpose of finding
things to quibble and pick nits about.  I see that as
potentially a huge source of further delay and not particularly
useful; if you disagree, please work with Andy to convince the
WG.  

Similar comments apply to "whether it is needed".  Before the
decisions that led, first to the "URNs are not URIs" proposal
and then to this document, the WG spent seemingly-endless months
bogged down in discussions about whether a particular mechanism
or usage was consistent with, or permitted by, 3986.  For each
case, some people argued that it was, others argued with equal
passion, that it was not.  We were not converging on subjects
that were arguably outside the WG's scope, such as whether 3986
needed work or was clear enough.   While there are technical
arguments having to do with differing layering abstractions that
I believe justify the separation, my impression of the WG's
conclusion was that we simply needed to separate ourselves from
those arguments if we were to make progress.   So, if the
question is "whether this is needed to make URN work proceed"
the answer is that the WG decided "yes" and Section 2 of 
draft-ietf-urnbis-semantics-clarif (at least -02 and later) is
intended to say that.  If the question were, instead, "can the
WG prove to the satisfaction of all readers and interpreters of
3986 than it is impossible to develop a functional URN model
that meets the needs of the communities the WG has identified
without some separation from 3986", that answer to that question
is probably "no", or at least "no within the level of energy
that exists", but I hope the WG doesn't need to do that.

best,
   john

p.s. Sean, the two typos that you spotted were repaired in the
working draft for -04 in February or earlier.  I don't remember
whether someone pointed them out to me (on or off list) or
whether I just spotted them and fixed them.  Unless directed
otherwise by WG Chairs, I generally don't post new drafts just
to fix obviously typographic and other trivial editorial errors.
But thanks for spotting them again and having me go back and
confirm the repairs.


[1]  Not a surprise.  I don't think I've ever looked, months or
years later, at a document I had finished or put aside for a
while and not found areas of possible improvement.  

[2] I believe I did say on-list that I was going to wait.