[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: ESP revisions straw poll

> From: "Perry E. Metzger" <perry@piermont.com>
> 
> I realize that some other people care passionately about the
> encryptionless ESP issue. Myself, I don't care much, quite
> frankly. I'm happy either way, and I have no personal leanings. I also
> think it makes no real practical difference. *I REALLY MEAN THIS*.
> I have no personal axe to grind here.
> 
> However, I'm a bit of a stickler for following procedure. We had a
> meeting at Memphis and the issue wasn't even close. We had close to
> unanimity against encryptionless ESP. This being the IETF, we follow
> the consensus. It wasn't even a rough consensus -- it was pretty damn
> close to everyone.
>
>   [...]
> 
> The reason the IETF process works is because people agree to follow
> the rules. If we don't follow the rules, it falls apart.


Well, we are closer to agreement than I thought.

I too don't think this issue makes much practical difference, and my
personal opinion is based on an intangible aesthetic preference, not
hard technical requirements.

And I share the discomforting feeling that procedure is being abused.

However, I find it interesting that two people can observe the same set
of events under the same set of ground rules, and still come to
radically different conclusions.

My interpretation:

* The two public IPSEC sessions in Memphis were not making a great deal
  of progress, so the developers-only session was convened on the spot to
  hash out the issues and propose solutions to the WG at large.

* The developers group had nearly unanimous agreement not to support
  ESP without encryption.

* The document editor issued a call for comments on the set of open issues,
  one of the developers objected to including auth-only-ESP in the list
  of issues to be discussed, and Angelos claimed that no one on the list
  (and only two persons in private, one of whom later recanted) wanted it.

* In response to this claim, at least eight people (including Bellovin,
  Orman, Simpson, Glatting, Lynn, Kent, Lambert, Kemp) expressed
  support for it (to varying degrees) on the public list.


The IETF process absolutely requires that any "decisions" reached at
meetings be confirmed by discussions on the mailing list.  In many cases,
the decisions reached at the meeting are confirmed on the list, either
by active discussion or by silent consent.  In this case, the discussion
is active.

It is incorrect to claim that consensus was reached in Memphis, because
by IETF definition no decisions made there can be final.  And it is
incorrect to claim that consensus was reached earlier on the list and
is now being revisited by sore losers, because until the "straw poll"
message, no consensus had been requested.  Many people don't feel the
need to discuss their opinions on the list until an explicit
solicitation is issued.

The process *is* being followed.  When the poll is over, I'll accept
the results, whatever they are.
Follow-Ups: