ietf
[Top] [All Lists]

Re: draft-housley-iesg-rfc3932bis and the optional/mandatory nature of IESG notes

2009-09-03 00:25:45
In fact, I do not think the presence or absence of a note means what you describe.

First, remember the Independent Submission do get reviewed. They do not get IETF review. But they do get technical review by senior technical participants in this community. This review can be thought of, as another note pointed out, as being comparable to the peer review required for scholarly journals.

Secondly, given that one of the points of the stream is to provide a mechanism for informed critique of our work, it is unsurprising that the IESG has objected to such critiques.

Third, it should be remembered that the IESG review being discussed is not a technical review. It is a review for the relationship between this work and the IETF work. If the IESG finds that there is a relevant relationship, and provides a reasonable note saying so, then the ISE is going to publish it. (There may be negotiation about the wording, but the editors responsibility to be factually accurate means that such accurate notes will be published.) If IESG members in reading the document find technical problems, then they should tell the ISE, just the way any other reviewer would.

Next, given that the review is not technical, and given that the review is done by a small number of people, I would not want a reader to draw conclusions about technical competence based on the presence or absence of an IESG note. Such conclusions should be based on reading the document.

And let's be careful about asserting that IESG review is an assurance of quality. There have been plenty of marginal documents that the IESG has approved over the years. Sometimes they approved them for very good reasons. Note that even 15 years ago the IESG was under interesting publication pressures some of the time. I am sure it still is.

If you want to abolish the Independent Stream, then we can have a discussion (separately from this one) about who would get to do that, and whether that would mean that the IETF would have to find a different outlet, or the Independent Stream. The mix of history, practice, and branding would make for a very confusing situation. Personally, I think the Independent Stream is valuable. (That's why I help review them.) So I do not want that stream or its independence weakened.

Also, remember that there are actually 4 streams. I presume that no one is trying to suggest that the IESG should get to put IESG warning notes on IAB documents or IRTF documents? Yet IRTF documents are actually sometimes less reviewed than Independent Stream documents. It is not uncommon for an IRTF Working group to produce multiple reports when the group can not come to agreement. (Note, I do not think that the degree of review correlates with the degree of quality. That is a further confusion about these notes.)

Yours,
Joel M. Halpern


Richard Barnes wrote:
Being a relatively short-term IETF participant, I lack the history that many on this list have, but since Jari asked for comments, I'll provide some.

Stated briefly, I agree with Steve Kent, Adam Roach, Ben Campbell, and others that it makes sense to have IESG notes be mandatory for the ISE to include in independent stream RFCs.


Stated at more length:

What is clearly going on here is that our branding is out of sync with the expectations of our customers. Whatever their historical meaning, RFCs are now interpreted by the broad community as documents that have the been reviewed and approved, to a greater or lesser degree, by the Internet community. I think we all agree that documents that go through the IETF or the IAB can more or less legitimately claim that imprimatur.

Independent submissions clearly cannot. Given that, it's not clear to me why the independent stream exists at all, other than for historical reasons.

Given that the abolition of the independent stream doesn't seem to be an option at this point, the next best thing to do is to require that independent-stream RFCs alert the reader to two things:
1. That this is not a document that has received IETF or IAB review, and
2. If the Internet community has any serious concerns, what they are

Clearly the first point is an issue for Headers and Boilerplates. The second point is represented in the current process by IESG notes; if the Internet community has concerns about a document, they can be included in the document as an IESG note. Given that the IESG is selected through a community process, I'm comfortable with this delegation, though requiring IETF consensus would clearly add some assurance.

The other implication of the above paragraph is that the *absence* of an IESG note indicates to the reader that the community has no serious concerns, which means that enabling the ISE to reject IESG notes effectively enables the ISE to speak on behalf of the community. Given the choice, I would prefer the IESG to speak for me than the ISE.

So I agree with Jari's option (b), that IESG notes should be something that is always applied to the published RFC.

--Richard





Jari Arkko wrote:
I would like to get some further input from the community on this draft.

But first some background. This draft was brought to a second last call in June because several IESG members felt uncomfortable with the IESG notes being used only in exceptional circumstances. I asked Russ to prepare the -07 version. This version allowed notes to be used at the IESG's discretion and suggested that the linkage (or lack thereof) to IETF work would typically be explained in the note. This version was taken to the second last call.

While the number of comments we received was small, after the last call was over I determined that the consensus was against this change. As a result, I asked Russ to prepare the -08 version. This version goes back to the "exceptional" wording from -06, but incorporated a number of editorial corrections that had been made in interim. I also took the draft back to the IESG telechat last week. The IESG was not extremely pleased with the new version, but my understanding is that they were willing to accept the changes. However, a new issue was brought up: one of the changes that Russ and I felt was editorial highlighted the fact that the document makes the IESG notes a recommendation to the RFC Editor, not something that would automatically always be applied to the published RFC. Some IESG members were concerned about this, and preferred the latter.

And now back to the input that I wanted to hear. I would like to get a sense from the list whether you prefer (a) that any exceptional IESG note is just a recommendation to the RFC Editor or (b) something that is always applied to the published RFC. Please reply before the next IESG meeting on September 10. Some e-mails on this topic have already been sent in the Last Call thread -- I have seen those and there is no need to resend.

(For the record my own slight preference is b. But I have to say that I think the document has been ready to be shipped from version -06, and its unfortunate that we're not there yet, particularly since this document is holding up the implementation of the new headers and boilerplates system for independent submissions, IRTF submissions and IETF submissions. I will exhaust all possible means of getting this approved in the next meeting, as soon as I know what the community opinion is.)

Jari Arkko

_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf

_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf

_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf