On 18-Dec-25 18:01, Martin J. Dürst wrote:
Hello Alexis,Many thanks to you for all your work and for the report (most of itdeleted).
Seconded!
Some comments on the editorial stream process below.
Ditto...
On 2025-12-18 09:07, Alexis Rossi wrote:# Observations on the Editorial Stream Process The current RSWG + RSAB model has sent 3 documents to the RPC in the past 3 years, and one more is currently nearing conclusion in the working group. This pace has accelerated over time, as most of these documents are from the past year, but it still seems like a slow pace for relatively straightforward documents. Participation in the RSWG is typically limited to a small number of active contributors, and we’ve seen relatively long lags during those discussions. Additionally, Last Call requests for broader community input have generally not resulted in significant additional participation. This may indicate a lack of controversy over these documents, and/or a lack of wider community interest in Editorial policy.I think it's both. These are issues that affect everybody in the IETF, but not very strongly. Even the people participating have a day job, and participate in actual technical IETF work, both of which usually have higher priority than editorial issues.I wasn’t present for the conversations preceding the creation of this new Editorial Stream model, so there may be concerns or goals I’m not aware of (except as they’re reflected in 9280 or this github repo [1]).I wasn't involved in the situation that ultimately led to the current model, but my understanding of that situation was that more direct involvement directly by the IETF community at large was highly desirable.
Yes, I think that was the primary motivation. However, as for the IETF-specific tools (the datatracker, meetecho etc.) and the IETF process itself (RFC 2026 etc.), it's always going to be a minority interest except when something goes badly wrong. A lot of people just want to get on with their technical work, and tools, processes and the RPC are background noise. So we have now made the discussions open to the community, but the number of people actively involved hasn't increased very much.
However, I think it may be time to consider whether the Editorial Stream process should be adjusted to improve timeliness. For example, I think the IAB’s document process (expert board combined with community review) may be a useful model to consider.First, standards processes are generally always slower than expected. Also, for the editorial stream, in many cases, what's going on 'on the ground' (i.e. how many technical RFCs can get published, in what quality) is more important than how many editorial stream RFCs get published.
Correct. If the RPC wasn't basically working well, or was resisting pressure to do better, we'd have an urgent situation, but neither of those things is true. I'm very impressed by today's RPC, but the priority should be getting technical documents published, IMHO.
The IAB model may look attractive, but the following should be considered: 1) IAB positions hold significant prestige, and the IAB works on important and timely technical topics.
Exactly. In fact the IAB has always (for more than 25 years to my personal knowledge) wanted to prioritize technical topics over admin, and from that viewpoint RSWG/RSAB is admin.
2) The IAB organizes workshops,..., and many documents are the result of such workshops. There are many active participants in the workshops that will occasionally nudge the editors. Still these reports usually take about 1 year to reach RFC status. 3) As far as I understand (sorry, no first-hand knowledge), for many of the IAB members, IAB membership and the related time effort, as well as travel expenses,... is part of their employment. 4) I don't think all the IAB publications are done in a short time. draft-edm-protocol-greasing-06, for example, exists already for about 2.5 years (there may be good reasons for this long time).
Right. As another example, RFC 6709 (September 2012) started as draft-carpenter-extension-recs-00 in October 2006. I wasn't in the IAB then, and persuaded them to take it on, but it took several years.
I am interested in community feedback on whether you think the current model is working well and, if not, what changes might improve efficiency and/or participation. Please feel free to share on list, or with me directly.My general evaluation is that the current model works well enough. The energy that's around (which is as you say not that high) should at least for the foreseeable future be spent on actual work (e.g. the current documents in the editorial stream) rather than to tweak the model again with unclear consequences.
I agree. We still need to clean up some policy matters, I think, but there's
no fire alarm. Ideally, the RSWG and the RSAB would have nothing to do.
A lot of the issues in [1] appear to me to be operational issues, not policy.
Brian
Regards, Martin.[1] https://github.com/rfcseries-wg/new-topics/issues
_______________________________________________ rfc-interest mailing list -- [email protected] To unsubscribe send an email to [email protected]
