[rfc-i] new draft summarizing updated Transitional RFC Editor recommendations now available

Ted Hardie ted.ietf at gmail.com
Tue Nov 23 16:36:02 PST 2010


Reply in-line.

On Tue, Nov 23, 2010 at 2:11 PM, Craig Partridge <craig at aland.bbn.com> wrote:

> Institutionally, putting representatives of the streams in an oversight or
> supervisory capacity to the entity that is to lead arbitrations among them
> is structurally unsound.

As may be clear from previous discussions, I think "lead arbitrations
among them"
is also the wrong model, with "foster consensus among them" being my
preferred view of the role.  More on that in a later email.

In this model the IAB's role in hiring and firing is summarized as:

   The Series Editor
   is designated by the IAB, and may be removed by the IAB, subject to
   contractual requirements.  The Series Editor reports to the REOC
   (Section 5.1.3).

but this document seems to have been delegated this entirely to
the REOC, as set out in section 5.1.1; the only liaison clearly
set out in this is the liaison from the IASA.  There no clear
liaison relations even to the IAB, which is the responsible
party.   Having some IAB liaison to this to maintain
continuity seems to me the minimum sensible method
for oversight of a duty which it is delegating in this way.
A voting member appointed by the IAB also serving on the
IAB does not seem to me to be a problem (note that the
staggered appointments may also raise the possibility
of the NomCom appointing a member of this committee
to an approving body).

But I think the overall reason for avoiding stream approving
body members on this committee is not simply because they
are all overworked to start with, but because of the text in
section 6.2:

"6.2.  Series Editor Review of Inter-Stream Conflicts

   The streams are encouraged to resolve conflicts on their own.  Any
   stream approver may request a Series Editor review of an inter-stream
   conflict at any time.  Review by the Series Editor must include
   assembling a review committee of four disinterested REOC members plus
   the RSE, who will chair the committee."

We have been told time and time again that the RSE has
nothing to do with the content approved by the streams, yet
this review is not limited to editorial issues.  It could be
read to provide a different appeals process to the one implied by
BCP 92 (currently RFC 5742) (where IESG decisions would be
appealed to the IAB per RFC 2026, for example).  That is a very
substantive re-assignment of responsibility to a group which has only
just been invented, and it is not a good idea.

If the RSE/REOC does not have this role for content-related
inter-stream conflicts (and I think it should not), then a big
part of the reason for removing the stream approvers from
the pool of potential committee members goes away, at
least in my opinion.  Advice not to assign this to busy people
seems to be more appropriate here (and I agree with your
idea that previously serving members of review bodies are
more likely to have the time to do this right).

On a related note, the document currently allows the
IAB to waive a fairly serious inherent conflict of interest:

"Series Editor must not be from an organization that
provides RFC Editor services.  The IAB may, however,
override this provision in specific instances, but only after reviewing
the matter with the REOC and IAOC and informing the community."

allowing that and absolutely forbidding any member of the
current IAB, IESG, IRSG from serving seems at best inconsistent.


Ted Hardie

More information about the rfc-interest mailing list