Eggert, Lars <lars@xxxxxxxxxx> wrote: > On Mar 4, 2013, at 13:18, Eric Burger <eburger@xxxxxxxxxxxxxxxxxx> wrote: > >> I will say it again - the IETF is organized by us. Therefore, this >> situation is created by us. We have the power to fix it. We have to >> want to fix it. Saying there is nothing we can do because this is the >> way it is is the same as saying we do not WANT to fix it. > > what is "the fix"? There is an obvious place to look for ideas: the directorates. See: http://www.ietf.org/iesg/directorate.html We see that there are a number of differing implementations of these, even if the explanations try to follow RFC 2418. Variety is good! > The IETF is set up so that the top level leadership requires technical > expertise. It is not only a management job. Indeed -- we reject traditional management almost as much as kings and presidents. ;^) > This is a key differentiator to other SDOs, and IMO it shows in the > quality of the output we produce. The reason the RFCs are typically > of very good quality is that the same eyeballs go over all documents > before they go out. (which doesn't scale...) > This creates a level of uniformity that is otherwise difficult to > achieve. But it requires technical expertise on the top, and it > requires a significant investment of time. Wearing the hat of senior Narrative Scribe, I have watched the IESG deal with the increase in the number of documents they must review. I see a lot of variety, including folks saying, "I'm not reviewing this one." I've been seeing a lot more reliance on the RFC Editor to work with authors based on "notes" attached to approved drafts. I will agree we need a broad range of expertise at the top; but we are learning to work around holes in that broad range. This does not bother me -- I date back to when there was _no_ barrier to publishing an RFC. Other folks worry about this more than I do -- that's fine. But wishing does not make it so. Directorates _could_ fill some of these holes; but it's not obvious that they always do. I suggest that those of us who worry about the holes put more effort into _how_ directorates could fill them, since we have long passed the point where fifteen experts can cover the whole field. > I don't see how we can maintain the quality of our output if we turn > the AD position into a management job. (I really don't think there's much danger of that.) > Especially when technical expertise is delegated to bodies that rely > on volunteers. We're _all_ volunteers! > Don't get me wrong, the work done in the various directorates is > awesome, but it's often difficult to get them to apply a uniform > measure when reviewing, How important is that, really? > and it's also difficult to get them to stick to deadlines. They're > volunteers, after all. I don't think we really "believe in" deadlines. I suspect we think of deadlines as "damage we need to route around." ;^) Nonetheless, we could improve the overall result with any of a number of management tools. (Fundamentally, there will be a typical distribution of time- in-queue for directorate reviewers; and ADs waiting for reviews could prod before the scheduling becomes critical. Also, a system of noting review opportunities to the whole directorate, and accepting partial reviews from multiple members would improve information flow...) > And, as Joel said earlier, (Joel has pointed out in private email how I misunderstood what he was saying... my apologies.) > unless we delegate the right to raise and clear discusses to the > directorates as well, I don't think anyone wants to go there! > the AD still needs to be able to understand and defend a technical > argument on behalf of a reviewer. Indeed, that is a common belief... but it has scaling problems. We might note that almost every DISCUSS by IETF Chair Russ Housley refers to a particular Gen-ART reviewer. I can't offhand recall a single case where Russ was asked to "defend" a technical argument of a Gen-ART reviewer. (Sometimes, he does add his own description: does that count?) The General Area is the most obvious place where scaling has hit us: the IETF Chair has grown so far beyond full-time that something has to give. Russ, I believe, reads Gen-ART reviews, not the original documents, and points out areas that rise to DISCUSS level. He asks for text to "address" these issues, and tends to clear his DISCUSS once the issue is better understood. (I should perhaps note that today's IESG has made great progress in trusting each other to put significant concerns in RFC Editor notes instead of continuing to block documents.) > If there is a controversy, the time for that involvement dwarfs the > time needed for the initial review. I don't believe that's entirely true. Perhaps some IESG members can offer more information here. > There is no easy fix. Well, maybe the WGs could stop wanting to > publish so many documents... ;^) -- John Leslie <john@xxxxxxx>