--On Wednesday, August 16, 2023 10:29 +0200 Carsten Bormann <cabo@xxxxxxx> wrote: > Hi Rich, > > I'm sympathetic to what you are saying, but there is at > least one point of overgeneralization. (We have to actively > remind ourselves again and again that WGs *are* different.) > > On 15. Aug 2023, at 22:14, Salz, Rich > <rsalz=40akamai.com@xxxxxxxxxxxxxx> wrote: >> >> Is an interim-only WG that never meets at the regular >> meetings acceptable to this IESG? I highly doubt it would be >> acceptable to the larger IETF community, but without >> community involvement in this statement or 2418bis, we'll >> never know. > > It always helps to discuss questions like this not only from > general principles, but also with respect to running code. > > The cellar WG runs exactly like this, and it is generally > considered a rather successful WG. > > This WG needs to draw in some rather special talent, and the > interim-only approach works best for that. > > Clearly, this specialty subject makes cellar's interim-only > proceedings more palatable to the larger IETF community than > it would be for, say, a WG that redesigns DNS. So, if we > wanted it, maybe we would have to make a "meet at IETFs, > occasionally" mandate more specific than summarily excluding > interim-only WGs. Carsten, Sorry for the delay on this... it took some thought. Let me try to clarify some things, parts of which have little to do with the appeal. I don't know if Rich will agree with what I'm about to say, so am responding to your note rather than his response. I'm also going to try to address a few higher-level principles. I have also been repeatedly accused, on and off list, of asking for more and more rules and/or more rigid ones. The is untrue: I am, and have been all along, a firm believer in what Spencer Dawkins have often characterized as "Do The Right Thing" which includes the flexibility to adapt to circumstances and special considerations, to be able to override general procedures and tools when appropriate, etc. But, if we are ever going to make statements akin to "the IETF consensus is..." or "the IETF believes..." and have them be meaningful, there has to be a shared, clear, understanding of what those words mean. Even that understanding should not, IMO, be a rigid rule but, if it is necessary to depart from it in specific cases, we need to be clear about that, not pretend the normal mechanisms apply, Some of what I'm about to say is not part of, or directed to, this particular appeal but, because I'm gotten several off-list notes criticizing and questioning my motives, it may help. I think WGs are different and that it is entirely appropriate for different ones to get their work done in different ways. If those differences mean interims-only for some WGs or, as I have written earlier, no meetings at all for others, under particular circumstances, so be it. However,... Most, I presume all, WGs intend that at least some of their outcomes be standards-track (including BCP) documents or other results that will be presented to the broader community as representing IETF consensus and not, e.g., as the outputs of the FOO group that used some IETF tools and facilities. That, to me, means that there actually are some rules and that a WG who departs significantly from "typical" ways of working takes on some obligations about ensuring openness, transparency, and opportunities for community involvement. Given prior conversations, I suspect you may disagree with some of what I have to say but that is where, I think, the key point of my appeal, Rich's comments, and perhaps the appeal from Ted and Allen all come together: at a very high level, the decisions about boundaries ought to be community decisions, not IESG ones taken in private and announced through Statements, nor WG ones made even more privately and not announced or reviewed externally at all. Does the IESG saying "Suggestions for further changes or clarifications may be made to the IESG directly (iesg@xxxxxxxx) or on another suitable mailing list" make setting policies or reinterpreting procedure-oriented BCPs in statement mode appropriate? I have an opinion but that is a decision that should be made by the community. And I have to admit to some anxiety about the IESG making Statements, doing so without even the level of transparency afforded to a decision to publish an individual (non-WG) Informational document in the IETF Stream, and then acting as judge and jury about any comments made about one of those Statements. So, in particular, * While I believe that WGs are different and should be able to work differently, I do not believe that translates to "every WG is a law onto itself, free to do whatever facilities its work and to do so without oversight" * I do believe (and consider it obvious) that the most efficient way for a group to reach agreement is for the group to be very homogenous in its background and perspectives and to discourage (however unintentionally) involvement and participation, even limited participation, from people with other views and perspectives. I do not, however, consider that their conclusions automatically represent IETF consensus. Especially when the topics are outside whatever constitutes the IETF mainstream and general expertise in the IETF community, I do not believe that the group posting a document that represents their consensus and asking the community to do a two week review --a period that may be too short for any IETF participant who is not already a participant in the WG to dig into the subject matter and reach sound technical decisions about the proposed specification -- as turning the output from that sort of WG into IETF consensus. * I also believe that the idea that the IETF does its decision-making, and most of the critical discussions and reviews leading to decisions, on mailing lists is critical to our consensus process. If we are going to change that, I believe the change must be arrived at by a clear, public, and well-documented consensus process involving the community, not through back doors or the death of 1000 cuts. * I also believe that accountability is important, especially when actions are taken that can reasonably be expected to lead to or enable claims of IETF consensus. Until and unless we change that model, that means accountability either directly to the community or to people, such as ADs, who are accountable to the community via the Nomcom (and, in principle even if not in practice, the threat of recall). With a nod to Dave Clark, I do not believe that ADs have imperial authority answerable only to the deity of their choice, nor do I believe that installing someone as a WG Chair anoints and crowns them as kings over those WGs whose powers include keeping the Responsible AD in the dark about what they and the WG are doing. I hope that at least clarifies where I'm coming from. I think I have been reasonably consistent about it. best, john
