> On Jul 16, 2023, at 4:50 PM, Carsten Bormann <cabo@xxxxxxx> wrote: > > >> >>> For information, a WG is there to get some work done. >> For your information, a WG is there to BUILD CONSENSUS. > > Yes. Which takes work to build. Hint: It’s called a working group. I don’t disagree. But the output of a successful working group isn’t merely a protocol; it’s a protocol that’s backed by consensus of various kinds - including, ideally, consensus to implement, deploy, and use it. A protocol spec that isn’t supported by such consensus is arguably worse than no output. > > Bonus points are available after building consensus if your consensus result > > * doesn’t destroy the Internet > * doesn’t disadvantage the actual end users (RFC 8890) > * can be implemented at a reasonable cost > * has a deployment strategy so it is technically possible to put it in > * has the deployment incentives so it actually gets deployed > * doesn’t have barn door sized security holes > * achieves any of the original objectives > * has at least a rudimentary understanding of its security objectives and how they are met > Of course, though I might set a couple of those bars a bit higher. But my point is that manipulating the group to produce a particular output, that doesn’t considerthe needs of the broad spectrum of interests that is the internet, does various kinds of harm and is a misuse of IETF. > etc. etc. > > Which all also takes work. > >> discard all or part of the earlier work. > > I cannot help noticing that you are very fixated on discarding, destroying. You seem to be cherry picking. But how else do you fix a WG that has been rigged? If there’s no penalty for rigging, it will happen more and more often. In case you can’t tell, part of my goal would be to salvage the portion of the group’s output that consensus says is useful. But rough consensus should be a necessary condition for publication of an IETF Consensus document, and I don’t consider “consensus by exhaustion” as meeting that criterion. Keith
