--On Wednesday, July 27, 2022 11:47 -0400 Jay Daley <exec-director@xxxxxxxx> wrote: >... > There was a long conversation roughly two years ago on > manycouches and tools-discuss about discontinuing the live > audio streams, and while you mentioned privacy in passing in > that discussion, there was no reference to this being a > consensus requirement. The general feeling of those > participating was that a) audio streams and jabber were > sufficient for anonymous participation; and b) fee waivers > handled the issues of exclusion, and so the audio streams were > retained and enhanced. Jay, Again, I'm not trying to make a case that the decision was the wrong one, but I think your explanation above helps illustrate the point I was trying to make and may help me clarify it. The IETF tradition for decades is that WGs, teams of various sorts, and other groups (often self-selected) hold discussions and, where relevant, reach consensus. We recognize that consensus in those groups represents agreement among the people who are most interested, who have the time and energy to participate at that level, and, ideally, are most expert in the subject matter. However, we avoid assuming that agreement within such groups is equal to community consensus (or even community discussion) precisely because people make decisions about where to commit their time and there may be significant perspectives that are not adequately represented in them. If we could assume that equality between WG agreement and IETF consensus, we would not need, e.g., IETF Last Calls on Standards Track documents. Taking the Meetecho discussion as an example, I don't even remember raising the privacy issue although, if you say I did I have no doubt that is true. There are people in the IETF who are passionate privacy advocates, some of them apparently believing that privacy consideration should trump any other considerations. I'm not one of them -- I see privacy, as I see many other things, as something that has to be evaluated in terms of tradeoffs and priorities-- but I think we should all be very interested in and pay attention to what they have to say. AFAICT (and recall), none of those passionate privacy advocates were active in Meetecho. So my bringing it up was to see if anyone who was participating actively wanted to discuss it (apparently no one did) and as a placeholder or warning that there might be an outcry when the issue came up for IETF-wide review as a Manycouches (or SHMOO) output . For many of these topics, that never really happened, at least partially because, two years ago, we were very much in the middle of an emergency situation. However, today, substituting (e.g.) a Manycouches discussion and "general feeling of those participating" for IETF agreement and consensus is an example of a trend I hope does not become permanent. Coming back to the question of when (or whether) there was a formal consensus call and how it was documented, what I'm about to say is a personal style question and, as the IETF has evolved, I may now be in the minority. I believe we should have our discussions around what The Right Thing is To Do rather than trying to cut off or direct discussion by citing legalistic precedents. Taking this issue as a (not very good) example, I'd rather have a discussion about privacy (or, in that case, listen while others have it) than about earlier decisions. If, in that discussion, someone notices that things are headed in a way different from earlier community discussions, they should by all means say "we seem to be changing our minds; why?" or the equivalent but that should not, IMO, be where the discussion starts or a way of ending it before then. To summarize my original comment in that light and the light of my response to Lars: (1) We used to do this in a particular way and there was (unless I completely dreamed it) some significant community discussion about why that was important. Whether there was a formal consensus determination (I doubt that there was) is not really an issue. (2) We are now changing it with, AFAICT, no prior notice or warning to the community (and I am not talking about one WG or a team or two) and no realistic opportunity for members of the community to question the decision (even those who participated in the test sessions would not have been able to notice this issue unless they somehow thought to ask). (3) Emergencies aside, is this how we want to make (or have made) significant decisions in the future? Obviously, if I am correct in believing that the community has no interest in micromanaging or even reviewing every small decision, a different but related question is how we have a discussion around how to separate what is significant from what isn't. But, unless we want decisions that some people in the community might consider significant and important (even if only symbolically) to increasingly emerge from WGs without IETF review and/or small-team discussion, then it may be time to start doing the work. (And, yes, I --and probably others-- have some ideas but see no point in contributing them except in some context where they would be welcome and a public discussion might be expected to occur.) best, john
