Skip to main content

Amended Appeal Re: grow: Last Call: 'Operation of Anycast Services' to BCP (draft-ietf-grow-anycast) (Dean Anderson; 2006-06-14) - 2006-06-14
Response - 2006-07-10

* To: Dean Anderson
* Subject: Response to the Appeal by Dean Anderson against decision to Last Call draft-ietf-grow-anycast
* From: IESG Secretary
* Date: Mon, 10 Jul 2006 14:31:08 -0400
* Cc: ietf-announce

Appeal by Dean Anderson against decision to Last Call draft-ietf-grow-anycast

  1. This appeal requests the IESG to rescind the Last Call for this
    document on various grounds.

The IETF Standards Process includes an IETF Last Call as a
normal step once a WG has requested standards track or BCP
publication. In practice, the decision to issue
a Last Call is delegated to the Area Director concerned,
after WG consensus has been declared by the WG chair(s).
A Last Call does not imply that the IESG is necessarily
supportive of a document; in fact issuing a Last Call is a
perfectly reasonable way for the IESG to solicit input about
a document's suitability for publication.

The grounds given in the appeal therefore cannot invalidate
the decision to issue a Last Call.

The IESG will take no position on allegations about activities
and publications outside the IETF. The IESG would take the usage
or influence of fraudulent data in IETF WG decisions very seriously,
but is aware of no evidence that this happened in this case.
Any allegations of such matters should be accompanied
by evidence of truth and relevance, and discussed with the
parties concerned prior to any appeal (according to the second
and third paragraphs of section 6.5.1 of RFC 2026).

In fact, according to section 6.5 of RFC 2026, the appeal process
exists to handle disputes that cannot be handled through normal process:

section specifies the procedures that shall be followed to deal with
Internet standards issues that cannot be resolved through the normal
processes whereby IETF Working Groups and other Internet Standards
Process participants ordinarily reach consensus.

The appeals process is not an appropriate tool to use when the normal
process can be used to resolve the dispute. In this instance, the
anycast document was in IETF Last Call at the time of appeal. The
IETF Last Call process may be used to bring up any of the questions
Mr Anderson has brought up against the anycast document, including the
question of whether that document should have been last called. This
appeal was an inappropriate use of the appeals process.

The IESG did not request suspension of the Last Call because this
would benefit nobody. In practice the Last Call generated
significant comment and therefore fulfilled its purpose.

The IESG has heard Mr Anderson's technical message that we need to
carefully consider the impact of anycast on stateful protocols.
When we evaluate this document we will carefully consider this topic
on the same basis as regular Last Call comments.

  1. This appeal asks the IESG to annul the grow WG Chair's decision
    that the draft was ready for IETF Last Call, on the grounds that
    technical issues raised in the WG by Mr Anderson remain open.

On a point of detail, a WG Chair cannot take this decision.
The decision by the WG Chair was that the WG had reached
rough consensus to request publication.

We note that such a dispute should, according to RFC 2026, be raised
first with the WG Chair and second with the Area Director, before
being brought to the IESG. Nevertheless we have chosen exceptionally
to consider this part of the appeal.

We note that the IETF process does not require unanimous consensus
and does not require all technical issues to be resolved to the
satisfaction of the individual who raised them. On the contrary, the
IETF process recognizes that in many cases, only a rough consensus
may be found. We find that Mr Anderson's issues were contributed to
the WG but found little support there, so we see no defect in the WG
Chair's rough consensus decision.

  1. The appeal makes allegations against two members
    of the IESG. The IESG finds these allegations untrue.
    The appeal also suggests that these two members of the
    IESG should recuse themselves. Mr Anderson appears to
    believe that if IESG members are involved in the
    discussion, participate in the process according to their
    roles, or are named in an unsubstantiated allegation,
    this automatically disqualifies them from considering an
    appeal. The IESG does not share this view, and there
    were no recusals.

  2. The appeal refers to "other inappropriate or unethical activities,"
    which is too vague for the IESG to consider.

  3. On an additional matter not mentioned explicitly in the appeal, the
    IESG notes that the AD and WG Chairs considered that Mr Anderson's
    allegations of fraud on the WG list were fundamentally wrong and
    irrelevant and thus disruptive; this is not a matter where the IESG
    intends to overrule the AD's and WG Chairs' judgement. This does not
    imply that statements or questions about incorrect data are
    inappropriate when those data are being directly considered
    by the WG. As noted above, the IESG would take substantiated usage
    of incorrect or fraudulent data within the IETF very seriously.

The appeal is rejected. However, when the IESG evaluates this
document we will carefully consider the impact of anycast on
stateful protocols, along with regular Last Call comments.