Re: Sergeant-at-arms engagement model

[Keith Moore <[e-mail]>]

On 10/11/19 11:40 AM, IETF Sergeant-at-Arms wrote:

> RFC 3005 establishes the charter for the IETF discussion list, 
> [e-mail] <mailto:[e-mail]>. It provides for the IETF Chair 
> to appoint Sergeants-at-Arms (SAA) to help manage postings to this 
> list. Matthew Miller and Dhruv Dhody are the current sergeants-at-arms.
>
> The SAA team (Matthew, Dhruv, and Alissa) have been working on 
> documenting the SAA engagement model, operating procedures, guidelines 
> and email templates to help us make our communications consistent and 
> clear. That guidance is collected in the GitHub repository < 
> https://github.com/ietf/saa >.
>
> We will be using this engagement model from now on. We consider it to 
> be a work-in-progress, though, and will be looking for ways to improve 
> it as we gain more experience with it. We would welcome you to file 
> GitHub issues or send us email at < [e-mail] 
> <mailto:[e-mail]> > with your feedback.

Thanks for posting this.

IMO, it's dangerous for the SAAs to be defining their own rules for IETF 
list speech without the backing of community consensus. And yet, reading 
the "rules", I also see and appreciate that there's an attempt to anchor 
them in consensus documents where such documents speak to the issues of 
concern.   In the absence of sufficient direction from consensus 
documents, documenting the SAA's assumptions and interpretations of 
consensus rules could be constructive.   I do expect, however, that 
these rules be discussed and reviewed publicly (not on a private list, 
and certainly not on a proprietary service like GitHub), and that formal 
IETF community review and consensus needs to be obtained going forward.  
I don't believe that the "rules" should be enforced or considered 
binding until such consensus has been obtained.

It is disturbing, and I believe inappropriate, that the email message 
announcing these "rules" requests feedback through private and 
extra-IETF channels.    GitHub requires people to sign up with GitHub 
before filing issues, which facilitates tracking by GitHub and perhaps 
other parties, and is detrimental to the privacy of participants in the 
discussion.

Perhaps more importantly, the effort to divert discussion of this issue 
of tremendous importance to the IETF list, away from the IETF list and 
from the IETF itself, seems likely to have the effect of "burying" a 
discussion of tremendous importance.

One of my larger concerns about both RFC 3005 and the notes on GitHub is 
that there's no effective oversight or accountability of the SAA for 
their actions.   While the nature of SAA actions is that they should be 
kept confidential, it has become apparent that the SAA seems to want to 
act on its own, stretching the bounds of IETF consensus ruling, and also 
that it can have a chilling effect on IETF discussion.  Therefore I 
believe that record-keeping of SAA deliberations and actions has become 
essential.   Such records should be confidential by default, but made 
available to IAB at any time for review (whether or not in response to 
specific complaints).   A retention policy for such records should also 
be part of the record-keeping.

RFC 3005 says that complaints about SAA actions should be referred to 
IAB, but doesn't explain what power IAB should have to correct actions 
of the SAAs.   This situation should be addressed.

I appreciate that the notes on GitHub acknowledge that "professional" is 
a very ambiguous term.   I support efforts by the IETF community (NOT 
merely the SAAs) to define "professional" in IETF context.  Similarly, 
"disruptive" is also ambiguous and needs clarification.

The words "They also include criticizing an idea in an insulting or 
excessively hostile manner" are troubling.   I realize that one may 
criticize an idea in such a way as to effectively be critical of the 
person proposing the idea.  And yet, it is essential that people be able 
to discuss ideas candidly, and sometimes to criticize ideas 
emphatically.   I believe it's inappropriate to impugn a participant's 
motive without supporting evidence of that motive.  But I don't believe 
it's wrong to point out any problem with an idea itself, nor with 
potential ill effects of an idea, nor even with the appearance of an 
idea.   Sometimes this is a fine line, but it's essential that the SAAs 
not interfere with vigorous discussion of relevant ideas.

Keith