[I18nrp] Time to take a half-step back? (was: Re: charter -01)

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

Hi.  I think it is time to return to, and think about, the
purpose of this effort and see where that leaves us.

I think I can safely claim to be one of the small number of
people who have been doing a disproportionate share of the work
in/for the IETF in this loosely-defined area in the last
quarter-century [1]. I've gotten frustrated enough by the
relatively recent inability to have useful and constructive
discussions about and to process documents that I initiated the
discussions that led to the BOF.  I decided to be fairly quiet
during the BOF and to remain silent so far in the charter
discussion to see where the list would take things.  I think it
is time to say something.

At least to me, the problem here is that we haven't been able to
process documents.  I am frankly much more concerned about
i18n-specific documents than I am about protocol specifications
that include i18n-issues or provisions, if only because, if the
foundations are not firm, the references and adaptations cannot
be satisfactory.   In recent years, we have seen that WGs don't
get (and, more important, keep) a critical mass of active expert
participants.  Reviews on IETF Last Calls either get silence or
a horrible S/N ratio, neither of which helps convince the IESG
that there is informed community consensus.  We have been
talking about the importance of recruiting more expertise and
activity but, with a few notable individual exceptions, have
made little progress.  In particular, attempted educational
efforts have either fallen flat or been declined entirely.
While we are making little progress in growing expertise and
participation, we manage to have the same disfunctional,
ignorant, and/or irrelevant discussions over and over again and
do so to the point that some of the existing pool of experts
have started asking themselves whether we _really_ have to go
through those issues, and make time-consuming and often
fruitless efforts to explain them, yet again.  If even a few of
those experts answer that question with "no, not another time"
and walk away from the IETF or those issues, the effect is to
further reduce the pool of available expertise [2].

I'm not optimistic about a directorate helping very much, at
least unless it comes with clear IESG agreement about more
status and authority than the typical "review team" and, in
particular, the ability to move work forward that much of the
IETF community does not understand [3], not just catch glitches
or push back on particular efforts.  It would be, IMO,
irresponsible on the part of the IESG to give it that status in
the absence of clear plans about how people are recruited for
(or excluded from) the directorate.  I also have some concerns
about whether a group that is clearly "in" the ART area and
whose function is to "assist[s] the Area Directors of the
Applications and Real-Time Area with regard to
Internationalization" is going to have sufficient leverage with
the IESG in those years in which, by happenstance, none of the
ART ADs are both committed to and comfortable with this work.

However, if the directorate is the best idea we have at this
point, then let's give it a try -- certainly nothing else seems
to be working.

That said, I believe the recent discussion illustrates part of
the underlying problem with internationalization work in the
IETF (and perhaps other IETF work -- maybe this is not a unique
problem).  Charter-01 was posted on the 13th.  We have now had
discussions on two topics -- language that attempted to clarify
scope wrt liaisons as not interfering or changing existing
relationships and responsibilities and questions about what 
"internationalization" really means or includes and what is
excluded.  In the latter case, we have survived for many years
without precise definitions and boundaries.  If we discover that
we need better ones (I'd be surprised), probably the right way
to do it is to put a revision to RFC 6365 on the agenda and see
if we have expertise and energy to write and process it, not to
hold up this charter.  For both issues, if we can't trust the
directorate leadership and relevant ADs to keep the directorate
from going off into the weeds, or if the directorate goes into
protocol lawyer mode about its charter, I suggest that the
directorate idea should be dead in the water because that is
just one of many possible serious failure modes.  A directorate
thus preoccupied is unlikely to be moving documents forward
(again, we have been here before).

Maybe the lack of comments about anything substantive -- such as
how the directorate will be staffed and actually work, how it is
going to recruit and train people, etc. -- means that those
questions can safely be left to the ADs and directorate
membership.  If so, perhaps we are done.   But, unless there are
issues that would affect the directorate plan in significant
ways that I haven't understood yet, let's not spend a lot of
time debating characteristics or t8y (or perhaps even the
boundaries between t8y and i7e, s7y, or i9y [4]).

best,
     john



[1] For those who don't know or don't remember, the effort that
led to MIME, the SMTP extension model, and media types was
initially organized to give us a way to handle non-English,
non-ASCII email without the recipient having to guess what was
going on.  And then there is not only RFC 20 (cited in the
draft) but RFCs 205 and 373 but they predate the IETF.

[2] I observe that, of those who have made significant
contributions in this area by chairing key WGs or authoring/
editing or contributing key text to documents, a non-trivial
proportion seem to have dropped out of the IETF or moved on to
other work in the IETF and largely dropped out of its i18n
efforts. Independent of how one assesses the expertise or
contributions of those people, they were at least exerting
significant effort over extended periods and their departure
doesn't make the expertise pool any larger.

[3] Every time someone stands up in a WG and says something that
might be interpreted as "I don't want to know about all of those
details; just tell me what to do" and is not immediately
escorted away from the microphone and reeducated, it probably
sets up back. It certainly impedes more helpful conversations.
At least we know what to do if similar people write "just use
UTF-8" into their documents.  Those who can identify an issue
with one language they read or speak and either assume that is
the only significant problem or that all other problems can be
extrapolated from it are not much more helpful and might be less
so.

[4] triviality, ignorance, stupidity, irrelevancy...