[rfc-i] path forward with RFC 3932bis

Jari Arkko jari.arkko at piuha.net
Sat Sep 19 01:34:42 PDT 2009


As you may recall, my conclusion of the discussion was that while 
opinions were split, a dispute resolution model emerged as a potential 
compromise. A week ago I promised that we would come up with a specific 
proposal. Russ, Olaf, Harald and myself have now worked on this. In the 
process we have realized that the devil is in the details, but we do 
have a proposal that we believe addresses the various interests in an 
acceptable manner. The full updated draft is at 
http://tools.ietf.org/html/draft-housley-iesg-rfc3932bis-09 but the 
important part is copied here for your convenience:

   Experience has shown that the IESG and the RFC Editor have worked
   well together regarding publication recommendations and IESG notes.
   Where questions have arisen, they have been quickly resolved when all
   parties become aware of the concerns.  However, should a dispute ever
   arise, a third party can assist with resolution.  Therefore, this
   dispute procedure has an informal dialogue phase followed by a formal
   arbitration phase if the matter remains unresolved.

   If the IRSG or the RFC Editor has concerns with the content of a
   particular IESG note, then they should contact the IESG with a clear
   and concise description of the concern.  Alternate content may be
   suggested.  Informal dialogue is desired.  At the end of the
   dialogue, the IESG can reaffirm the original IESG note, provide an
   alternate IESG note, or withdraw the note altogether.

   The dialogue should not take more than six weeks.  This period of
   time allows the IESG to conduct an IETF Last Call to determine
   community consensus if desired.

   If dialogue fails to resolve IRSG or RFC Editor concerns with the
   content of a particular IESG note, then they can take the matter to
   the IAB for a final ruling.  The IAB review will occur according to
   procedures of the IAB's own choosing.  The IAB can direct the
   inclusion of the IESG note or withdraw the note altogether.  Unlike
   the IAB reviews specified in RFC 4846 [I3], in this situation, the
   IAB decision is binding, not advisory.

The rationale for choosing this model is first of all the fact that 
normal discussion should be given an opportunity, and only if that fails 
should the dispute resolution be invoked. We have chosen a model where a 
third party, the IAB, helps resolve the conflict. We believe the use of 
a third party is a necessary part of the compromise. We also believe 
that this model allows the independence of the RFC Editor to be retained.

An alternative that we considered during discussion was a two-party 
model where the RFC Editor still made the final determination about the 
requested note, but was required to ask for an IAB opinion before 
ignoring the request. We are not sure if this model would work as a 
compromise, because the two party model may not satisfy those who felt 
that the RFC Editor should not be able to decide on this on its own. 
However, the alternative does raise the bar for ignoring a request for 
an IESG note. An advantage of the alternative model is that it can be 
described purely as an application of the rules in RFC 4846. If we were 
to choose this model, the last paragraph would read as follows:

  If dialogue fails to resolve IRSG or RFC Editor concerns with the
  content of a particular IESG note, then they are required to acquire
  an opinion from the IAB.  The IAB can direct the inclusion of the
  IESG note or withdraw the note altogether. As specified in RFC 4846,
  IAB's opinion will be advisory.

In any case, the decision on what to do rests again with the community. 
We are asking the IETF community, the RFC Interest list, and the IAB to 
think about our proposal and provide feedback and/or alternative 
suggestions. I will wait for this feedback from the IETF until October 
1st. Given that this matter concerns the boundary between the IETF and 
RFC Editor operation, I will also ask the IAB to make a decision on 
whether they are comfortable with the model going forward.

Jari Arkko



More information about the rfc-interest mailing list