[rfc-i] RSE and REOC: IAB state, developing consensus, and strawman.

John Klensin john+rfc at jck.com
Sat Jan 22 14:47:23 PST 2011



--On Saturday, January 22, 2011 10:14:15 -0800, Ted Hardie
<ted.ietf at gmail.com> wrote...

>...
> I'm not sure that wanting the best of both worlds is what got
> us to where we are; it seems we still want the best of both
> worlds, but we have adapted our sense of how long it will take
> to get to a mature, fully functioning RSE.

Olaf could perhaps have said "wanting to recruit and select for
someone who could reflect all of the important properties of
both worlds on day 1 (or sooner)".  At least from my point of
view, that is what got us into a failure case because we
discovered that the population of people who could satisfy all
of those requirements (and still wanted the job) was null or
nearly null.

>...
>> The first assignment, documented in the SOW, is to develop a
>> vision for the RFC series that takes into account that:

> Sadly, I don't think this works.  We've had repeated comments
> that hiring people
> without clear lines of authority and accountability results in
> frustration, and this amounts
> to hiring someone without clear accountability.  It is a very
> tempting idea, but the way it is described sounds like a
> series of steps:
> 
> 1) Develop connection to community (6-9 months)
> 2) Develop vision of what job entails (order of months?)
> 3) Do job.
> 
> But the reality is every job evolves as it goes, and all three
> items go on concurrently in all jobs.  We have to hire someone
> with a focus on "Do job" in a way that makes them understand
> that 1 & 2 are ongoing parts of the job and that the focus of
> "develop of vision of what job entails" here maps almost
> exactly to "develop vision of how series can evolve".

Agreed.  I see

  (1) Do job
  (2) As part of doing job, explain job to community and listen
  (3) As part of doing, explaining, and listening, develop
extended vision and strategy.
  (4) Repeat.

> What I'm thinking of isn't hugely different from what you
> write, but I think it makes a crucial difference.  If we write
> a statement of work that starts from "create documentation and
> structures that allows continuity for the production house
> function" and "coordinate discussion among streams on Series
> evolution" we have a job description that includes "what to
> do" on the ground day one.  We can go on to say that 1 & 2 are
> fundamental, ongoing parts of the job and that we expect the
> job to evolve as the discussion of Series evolution produces
> new requirements and desiderata.

Right

> But delegating the initial job description to the person we
> hire, especially as job one seems to me like a mistake.  It's
> going to skew our applicant pool to people who are good at
> writing job descriptions, rather than people who can create
> documentation and structures/coordinate discussion. And it
> runs the risk of frustrating our new incumbent if the new
> vision doesn't fit with what we've already developed as
> consensus in this space.

Right, although I do worry that we are moving into territory in
which the person who takes the job needs to have both the skills
and interests needed to define, or at least refine, the role (in
the sense that any first-time holder of a new job always does)
and those needed to operate in a relatively more steady state.
My original hope was that the TRSE plan would help to span that
gap, but things didn't work out that way (largely because the
IAB and the community didn't want them to).

> I think we've actually made a lot of progress  in
> understanding the relationship of REOC to RSE and to getting
> agreement on where we want the final balance of  activity to
> be.  Let's use that progress to write the job description,
> knowing it will evolve with both the new incumbent and the
> community.

Again, agreed, but let's try to keep the boundaries reasonably
flexible so that we can adapt to the person hired, his or her
relationship and mutual trust with the REOC and IAB, and so on.

best regards,
    john




More information about the rfc-interest mailing list