Re: Possible BofF question -- I18n (was: Re: Possible OBF question -- I18n)

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

--On Thursday, May 31, 2018 12:22 -0500 Nico Williams
<[e-mail]> wrote:

> Lack of expertise is not a problem: we can develop it, and we
> can seek out external review.


Nico, this isn't clear to me.  We've been talking about
developing more expertise for years and nothing has happened, in
part because there seems to be little support for those who
would like to acquire the expertise (or career advantages if
they do).   We've also talked on and off for years about
external review but that has not gone anywhere either, partially
because, unless the IESG has internal expertise, it is hard to
determine which external reviewers to believe and how much...
and external review is not a substitute for IETF consensus unles
the IETF community says it is, which we haven't.

> Besides, the IETF has lots of I18N expertise on hand.

I'd question that too.  Certainly we've seen many people stand
up in i18n-related discussions and make very broad statements
based on their knowledge of a couple of languages or even a
couple of scripts, and assuming that knowledge applies to, and
makes them expert on, everything else.  I see that as part of
the problem, even if if only as a cause of lack of energy, but
not as a sign of expertise.  I do see people with expertise not
participating and taking leadership roles (or starting and then
dropping out) because the noise level is just too daunting.

>  Perhaps the IETF has no energy to do I18N work, but that's
> another story.

It should be.  I didn't distinguish it strongly enough in my
note because I'm not sure we know how to tell the difference.   

Perhaps an example will help.  Let's say we charter a WG that is
solely focused on some particular issue in, say, routing.  We
have a strong community expectation that people with little or
no background knowledge in routing principles and prior work in
routing on the Internet will either not show up or will remain
quiet and try to learn.  When those expectations are violated,
we have a vocabulary for talking about the offenders (starting
with "troll") and assorted mechanisms for constraining their
ability to disrupt discussions among experts and, while we hope
that effort to keep the discussions focused will avoid tipping
over into harassment or personal attacks, few people consider
that making sure expertise dominates the discussions to be
inappropriate.

Now compare recent IETF i19n-related WG work.  EAI was chartered
because there was a clear expectation that, if the IETF didn't
do something and do it at least fairly well, we would end up
with a non-interoperable mess and because there were people who
wanted to do the work.  But, while the topic, which involved
some rather complex issues with email and transition models as
well as well as i18n ones and specific interests from specific
language groups, didn't have the same expectations as that
hypothetical routing WG -- "I use email and therefore understand
it' or "I use a language that requires a non-ASCII script"
seemed to be adequate conditions for participation.   Now (I'm
certain no thanks to my tenure as co-chair) that worked out
rather well.  I think a lot of people learned things about what
they didn't know and then proceeded to acquire at least enough
knowledge to fill in the gaps to allow intelligent
conversations.  And the WG came up with a collection of
protocols that have been (and are being) deployed with none of
the experience yet indicating we got anything seriously wrong.
But, at least IMO, it was very painful, some people with
significant expertise on at least one of the topic areas dropped
out, and, by the time the WG was ready to conclude, the number
of people actively doing work was small enough that it is
unlikely that, had that been the level of participation from the
beginning, the WG would have ever been chartered.

A different version of the reason I am wondering whether a BOF
would be useful is, as Peter suggested about one of the
subtopics, to address exactly that pattern and how we get
unstuck from it.

best,
    john