Re: [Terminology] Update to TERM charter text

[Pete Resnick <[e-mail]>]

On 13 Apr 2021, at 18:09, Mallory Knodel wrote:

> I am offering my view that there appears to be a choice in processes 
> and we may have chosen badly if I do a quick calculation:
>
> Current process: IESG+community approval of a WG charter + production 
> of a document + WG+AD+IESG+Community approval.
>
> That’s three milestones and several rounds of approvals (6?).

Close enough. No community approval of the WG charter is needed, and 
community approval of the document (i.e., Last Call and coming to 
consensus) comes before IESG approval.

> IS: production of a document + GA/AD sponsorship + IESG+Community 
> approval.
>
> That’s three milestones and 3? approvals.

Again, pretty close. The AD still has to approve the document. And 
remember that community approval (i.e., Last Call and coming to 
consensus) is normally twice as long for AD-sponsored as it is for WG 
documents.

In effect, everything after the document is produced and approved by 
whoever wrote it (WG or individual) is the same, as documented here:

https://datatracker.ietf.org/help/state/draft/iesg

With the exception of the Last Call period being different. For any WG 
document, you're adding on the chartering requirements and the WG coming 
to consensus ("approving") of the document.

> IS just seems more lightweight, and ideally the AD and willing members 
> of the IESG could work on production, cutting down more hurdles.

AD-sponsored documents (best not to use "IS" to distinguish them from 
Independent Series documents directly to the RFC Editor) can be more 
lightweight, but as I said earlier, only if the document is simple 
and/or non-controversial. If it is complex and/or controversial, then 
you're simply folding the work that would have happened during the "WG 
coming to consensus" step into the "community approval" step. That ends 
you up with an unbounded (or only ad-hoc bounded) IETF-wide discussion 
of all of the controversial or complicated issues on the IETF list, 
moderated by the sponsoring AD, which tends to be far less efficient 
than having a WG sort those issues as best it can before then bringing 
it to the IETF list.

> Now I think IS isn’t ideal if it wouldn’t represent community 
> consensus. (I’ve heard others say this but surviving last call would 
> mean consensus, would it not?)

I think people were confusing them with the Independent Series 
documents, which are not community consensus documents.

> But since the IESG approves it, this is rather more satisfying ime due 
> to DEI initiatives needing some healthy degree of leadership in order 
> to break reactionary impasses, sometimes, as a peril of majority rules 
> communities.

First, a good WG chair is in a much better position to facilitate 
breaking impasses than an AD on the main IETF list. Second, IMO rough 
consensus is not majority rules. (See RFC 7282.)

I think on a controversial topic, a WG producing a document is actually 
more efficient and more likely to get the rough consensus of the 
community than an AD-sponsored document brought immediately to Last 
Call.

pr