On 17. Jul 2023, at 22:45, Eric Vyncke (evyncke) <evyncke=40cisco.com@xxxxxxxxxxxxxx> wrote: > > On 17/07/2023, 19:54, "ietf on behalf of Keith Moore" <ietf-bounces@xxxxxxxx <mailto:ietf-bounces@xxxxxxxx> on behalf of moore@xxxxxxxxxxxxxxxxxxxx <mailto:moore@xxxxxxxxxxxxxxxxxxxx>> wrote: > >> Last Call is far too late to fix most problems that crop up. I don't >> like throwing out people's hard work, but we need better ways of making >> sure that WGs produce output that really is technically sound and >> respectful of a broad spectrum of interest. LC is necessary but >> insufficient to do this. > > Alas you are correct... One mitigation is to request several "early reviews" by directorates, it can help a lot. Definitely. But this observation doesn’t detract from what I said: We have well-defined milestones where we check consensus. If the documents are not properly prepared for those milestones, we get late surprises and other inefficiencies. There is no reason why we should let documents pass that don’t have community consensus(*). (The parallel discussion about RFC 8252 indicates that this does happen — we should maybe move our attention to that discussion rather than searching for the reasons in the way we have been running interims. [This does not relieve us from the obligation to run them properly.]) Grüße, Carsten (*) for a definition of community consensus that is based on merit, not on big egos or deep pockets of specific players.
