On 25 Mar 2025, at 20:46, Benjamin Kaduk wrote: > On Wed, Mar 26, 2025 at 09:36:14AM +1300, Brian E Carpenter wrote: >> On 26-Mar-25 09:23, Salz, Rich wrote: >>> From a quick scan of old proceedings, there were 206 drafts submitted to the IESG prior to IETF 119, 120 and 121. There were 188 Last Calls prior to IETF 62, 63 and 64. In other words, the workload has really not changed over the last 20 years. We have to look elsewhere for what has changed. >>> >>> Document count seems much less accurate than page count. >> >> True, if you can find the data (and don't count things like MIB or YANG modules, because surely ADs don't check them in detail). But anyway, I think this aspect is a red herring: as Stewart Bryant said, society (and the industry) have changed around us. If there are new problems, that's probably where they lie. > > As a categorical statement your assertion about MIB and YANG modules is > surely false. [1] Even as a merely general statement I suspect it remains > false, since I think I remember ADs other than myself finding internal > inconsistencies in such modules during my time on the IESG. > > But I am inclined to agree with you that it's a red herring and the > change in expectations that arise from the surrounding societal context are > a bigger factor. And that is not necessarily bad; if you go and look at > RFCs from the 1980s and 1990s they just simply would not meet the bar for > publication given the current expectations for document quality and > completeness, which in general result from hard-earned experience about > what is needed in a specification to ensure interoperability, correctness, > security, and reliability. Document complexity does appear to be increasing over time: drafts go through more revisions before publication as an RFC, and take longer to complete, the number of references they make to other drafts and RFCs is increasing, as is the number of RFCs that update or obsolete other RFCs, and the number of RFC 2119 keywords per page is increasing. I’d expect drafts to become more complex over time, since there’s more deployed base to consider. But I’d also guess the workload on the IESG increases as the documents they have to review become more complex. Data at https://sodestream.github.io/paper-characterising-the-ietf-through-the-lens-of-rfc-deployment.html (Figures 3-8) Colin
