--On Thursday, June 29, 2023 11:07 -0400 Warren Kumari <warren@xxxxxxxxxx> wrote: > On Wed, Jun 28, 2023 at 5:10 PM, George Michaelson > <ggm@xxxxxxxxxxxx> wrote: > >> Apart from reducing WG there is the underlying need to also >> decline work. Not just how many points of organisation: >> decline to address all submitted and almost any problems. >> > > Well, there is also the signal that we keep hearing from the > community that we should be encouraging and fostering new > works and newcomers to the IETF. Warren, Based on a series of recent difficulties, I suggest that, when that signal gets close to particular proposals, some questions should be asked. For example: (1) Does the proposed work lie well within the IETF's expertise? My perception is that, if there is not already significant expertise in the IETF, the result is often of poor quality (or at least lacking any perspective other than that of the authors), takes up much more time than when there is more expertise, or both. If the work lies at the IETF's edges, how do the proposers recommend bringing in sufficient expertise and perspectives to allow the IETF to do good, quality-controlled work? Or what do they recommend as an alternative? "Trust us, we understand this and are the experts" will never equate to IETF community consensus. (2) Have the proposer(s) considered other venues where there might be better support and evaluations and/or complementary other work? Is there a clear answer to "why the IETF?". (3) How mature is the proposed work and what are the expectations of the authors? If they are looking for a rubber stamp on work they consider finished, why should the IETF invest community resources (including AD time) in doing that? If the work, and proposed specs, are already deployed and the installed base will become a reason to resist IETF-recommend changes, why not just propose an Informational document describing what is implemented and deployed to the ISE rather than trying to pull the IETF in and creating more work? (4) If the work in done in the IETF, standardized, and deployed, will it have a significant effect on making the Internet better or will it serve only a small and/or specialized community? I think that, if we were better at asking those questions, we would be taking on a lower proportion of work that requires too much time and energy and that often comes out with poor quality anyway. It does not eliminate the challenges of work that is really important, not obviously better done elsewhere, and for which IETF expertise is low: I think we need to be thinking about better ways to handle that type of work because we are not dealing with it well now. But it would likely improve things, including AD workload, overall. That approach of course carries the risk of the questions being weaponized to keep unpopular topics and the people proposing them out. Maybe the IESG should be working with the community to create forms of the questions that could be more or less objectively scored or other safeguards. Or maybe we can trust the IESG on this one. But I contend we should be encouraging and fostering new work that the IETF can do well and trying much harder to recruit newcomers who are interested in contributing to the IETF's ability to do good work and create a better Internet. If someone shows up looking to standardize bandwidth and throughput mechanisms and associated measurement specific to avian carriers, I hope they would be asked whether the IETF has sufficient expertise in ornithology and, if not, where the expertise will come from as well as whether improvements in those areas will make the Internet better. best, john
