On Tue, Mar 31, 2020 at 01:28:43PM -0500, Pete Resnick wrote: > On 31 Mar 2020, at 12:56, Alissa Cooper wrote: > > I think what you suggest is a recipe for either delegitimizing the IETF > > or causing it to enter a state of dysfunction. It assumes there will be > > consensus at the end of four weeks. > > It does not assume that, and I don't think you've actually run through all > of the reasonable outcomes. First of all, this has been discussed for two > weeks already. I think the IESG has a pretty good idea of the leaning of the > community and can draft something that has a high likelihood of achieving > rough consensus. (And remember, rough is good enough.) But let's assume the > worst case scenario you pose: Moreover, if there is no consensus at the end of the LC, the IESG could call it anyways and anyone who feels strongly enough about it can appeal the call. That would put the final outcome in exactly the right place: the IETF Chair and the IAB. Whereas the IESG just making a decision w/o a Last Call based on a claim to authority that is -presumably?!- not subject to appeal... strikes me as even more negative an outcome than an inconclusive LC. The risk of lack of consensus is overstated anyways. The options to choose from are reasonable. We'll have individual preferences, but none of the options should be particularly upsetting. > > But if the community can never reach consensus, then either the nomcom > > chair will have to figure this out on his or her own > > A bad choice for the reasons you mention below. The NomCom chair should not > be put in that position. Indeed. The appeal authorities should be instead. > > or the nomcom can never be seated > > A non-choice. That's simply "The IETF self-destructs". We all agree, not an > acceptable outcome. True. Supposing it could happen, that would be a reason to possibly eschew whatever path might end up in that outcome. However, I don't think it's possible, and the IESG calling a consensus (subject to appeal) even with an inconclusive LC would be just fine. Nico --
