[rfc-i] a possible refinement to draft-iab-rfc-editor-model

John C Klensin john+rfc at jck.com
Thu Apr 16 12:36:39 PDT 2009

Hi.  I have been trying to put this list aside for several days
to get some other work done and to see how things would come
together.  I guess I had better respond to a bit of the

> On 12 apr 2009, at 01:38, Brian E Carpenter wrote:

>> "The RSAG is chartered by the IAB."
>> Not if it advises the RSE. If it advises the RSE, the RSAG
>> has to be chartered  by the RSE. If I was the RSE (which I
>> faithfully promise will never happen), I would not accept the
>> existence of an advisory group which I didn't charter and
>> which didn't report to me.

--On Monday, April 13, 2009 09:11:45 +0200 Olaf Kolkman
<olaf at NLnetLabs.nl> wrote:
> [on personal title]
> I think you touch upon the kernel of the proposal. Let me try
> to   explain why I think the committee is an IAB chartered
> activity.
> In my view, the reason why we started to work on the committee
> is to   provide a body that is the carrier of the "RFC Series
> Flame" and while   the IAB has the responsibility of oversight
> the point was made that   the IAB members are currently not
> selected for RFC/Editorial expertise   and that the IAB does
> not have the cycles to consult the RSE in times   of crises.
> Another important goal was to gain long term consistency,   a
> consistency that spans the RSE contract cycles and therefore
> may   also outlive the individual RSEs.

I largely agree with Olaf's comments, but let me try a different
approach to this comment and others.  These comments come with
the disclaimers than I'm a very long-term member of the
community for the RFC series, have been involved on and off in
RFC Series design decisions since before there was an IETF,
serve as a member of the current Editorial Board, was part of
some of the "discussions in San Francisco" to which Leslie
referred, and am now part of the IAB's process and
decision-making in this situation.  I should also be explicit
that I'm trying to figure out how to make a new RFC Editor plan
work, not to, e.g., pile more work on myself or my Editorial
Board colleagues and that I am disinclined at this point to
engage in hair-splitting about question like who "charters"
whom.  And, in this note, I am speaking for myself only

According to Ray's note, there were zero RFI responses that
specifically addressed the RSE role or that the IAOC or the
Committee construed as statements of likely intent to bid on it.
His note also indicates the that IAOC has decided* to use the
SOWs from the RFI, or modified versions of them that it has
decided* upon, both for the Publisher function (which the the
IAOC may decide to sole-source to the Secretariat) and in an RFP
for the Production Center function.  It has also decided (no
asterisk) to recommend to the IAB that the IAB lead a
"community-based selection process" for the ISE and RSE.  The
IAB has not formally agreed to take on those responsibilities or
to use the IAOC-recommended SOWs, although it is not clear to me
that there is going to be a lot of choice in practice.

Because I have a tendency to worry about what can go wrong and
because I think the job of the initial RSE is going to be hard
--perhaps very hard, depending on what some of the words in the
SOW mean-- and very time-consuming, I think the community has to
take the following possibilities seriously.  We could get lucky,
of course, but, given the firm December 31 cut-off, we should be
thinking about the "what if" elements:

	(i) The IAB initiates the community-based process called
	for by the IAOC and gets no plausible candidates for RSE.
	(ii) We have never used a "community-based process" to
	select someone for a position that formally involves
	executive management responsibility (as called for in
	the Model document) nor one that is paid at any level
	above reimbursement for expenses.  Whether the discovery
	is made before or after a selection is made, we discover
	that community-based procedures based on the IAB's
	selection model for IAOC and ISOC BoT members (the only
	models that exist at the moment) are inadequate or
	inappropriate for a compensated position and that the
	IAB does not have time to work with the community to
	sort the details of some other selection plan.
	(iii) There are plausible people who would be willing to
	be candidates for the RSE role, but only if they know
	the "stipend"  or other compensation arrangements in
	advance or understand the terms and conditions under
	which such an arrangement might be negotiated.
	(iv) The IAB makes a selection, but is unable to select
	someone satisfactory, either because it is selecting the
	least-bad candidate from an unsatisfactory lot or
	because someone is given the position who is not up to
	the job.  Please note in this regard that the amount of
	experience on the IAB with selecting people for
	executive management experience and skills with
	publication systems is fairly limited.
	(v) We end up with a candidate who is otherwise
	completely satisfactory, but who has almost zero
	background with the RFC Series.

Now, in any of those unfortunate --and, I hope, improbable--
scenarios, we are going to need sources of significant and
substantive advice.  Under some of them, providing that advice
to the RSE should be fine.   Under others, that advice should go
directly to the IAB, unmediated by the RSE.  And, under still
others, it should involve a discussion with the IAB and IAOC
about things that are not working well and need to be changed,
unmediated by either the RSE or the IAD.  Logically, I can see
only the following possibilities for that advice:

	(i) Hire Bob Braden as a consultant, assuming that he is
	not involved in the Production House bid.
	(ii) Hire Alice and/or Sandy as consultants, assuming
	that they are not involved in the Production House bid.
	(iii) Hire Joyce Reynolds as a consultant, assuming that
	she is not involved in the Production House bid.
	(iv) Recruit a committee of experts to provide that

I note that neither the "RFC Editor Model" document nor BCP 101
seem to allow for any of the first three options, even if the
community, IAB, or IAOC concluded that they were needed.

With regard to the fourth option, there is only one organized
group with the needed expertise and perspective, and that is the
current Editorial Board (or whatever subset of them can be
persuaded to do this).  The other possibility is to try to
recruit such a group de novo, as some have proposed in various
parts of this process.  The difficulties with doing so are that 

	* the IAOC has really left the community no time to
	organize a recruiting process and 
	* trying to change contractors/editors and
	organizational model at the same time is inherently and
	severely risky, trying at the same time to create a new
	body to carry forward institutional memory would, IMO,
	border on the insane.

and, if you want the Editorial Board to do this, I think logic
dictates that you don't want to start replacing that group via
an untested (and, indeed, still undesigned) procedure until
there is reasonable confidence that the new RFC Editor structure
and interrelationships are stable ... or until it is clear that,
whatever the issues are, the Editorial Board is not helpful in
resolving them.  Longer-term, I agree with Olaf's observations
about institutional memory and consistency, especially if people
assume that we are unlikely to ever again have only two RFC
[Series] Editors in a forty year period.  Anyone who believes
that the cutoff date between a temporary and transitional
arrangement and that longer-term one can be set arbitrarily at
six months or some other predetermined short interval has, IMO,
either never been involved in a project that slipped longer than
expected or never learned anything from it.

Now, if we can agree that the above describes, more or less,
what we are trying to accomplish, then I really don't think it
makes a lot of difference what the group is called.  It makes
even less difference who "charters" it.    On the other hand, if
we can't agree on what this group is supposed to do and why,
then we might as well call it a "Rubber Duck" for all any title
will do for its effectiveness.


* "decided".  I can find nothing in BCP 101 that authorizes the
IAOC to "decide" on the Statements of Work to be used in
contractual processes.   Even if the community had seen those
SOWs before in the context of preparations for an RFI, it
appears to me that BCP 101 (and certainly the discussions that
led up to it and the discussions during the Minneapolis plenary)
make it clear that the community is expected to be consulted
again.  When the SOWs are actually modified by the IAOC, without
any community consultation prior to their being decided upon and
announced, that appears to me to be a fairly clear violation of
at least the intent of the processes.  To reprise that plenary
discussion as well as the pre-BCP 101 ones, the IAOC conducts
bidding processes, determines successful bidders, and negotiates
contracts and, for obvious business reasons, is likely to need
to keep those stages confidential.  But the documents on which
bids are requested, job descriptions and statements of work,
etc., are all expected to be exposed for community review and
comment _before_ decisions are made that lead into the
confidential parts of the process.

I have been deliberating as to whether to turn this into a
formal appeal but have been reluctant to do so because of
concern that the time it would take would lower the odds of
having a functional RFC Editor on January 1 (or would be used as
an excuse if the existing process does not lead to one).  But I
believe that the process that has been followed here should be
of great concern to the community, at least unless the community
is far more comfortable than appeared to be the case during the
1995 discussions, with an IAOC that works in silence and is, in
practice, accountable only to its membership.

More information about the rfc-interest mailing list