Re: [Rfced-future] Comments on draft-iab-rfcefdp-rfced-model

[Peter Saint-Andre <[e-mail]>]

Hi Murray, thanks for your review. Comments below.

On 3/10/22 10:57 AM, Murray S. Kucherawy wrote:
> Hi Eliot,
> 
> On Thu, Mar 10, 2022 at 7:49 AM Eliot Lear <[e-mail] 
> <mailto:[e-mail]>> wrote:
> 
>     On 10.03.22 16:37, Murray S. Kucherawy wrote:
> 
>>     On Wed, Mar 9, 2022 at 11:10 PM Eliot Lear <[e-mail]
>>     <mailto:[e-mail]>> wrote:
>>
>>         The philosophy behind CONCERNs is that they should be rare,
>>         and that the
>>         community really should be taking responsibility for the
>>         content of a
>>         document.  Thus, unless something truly rises to a threat to a
>>         stream or
>>         the long term health of the series, the RSAB should stay out
>>         of the
>>         way.  Those two CONCERN criteria are sufficiently broad to permit
>>         members to act if the policy truly looks like it will cause
>>         harm in some
>>         way.
>>
>>
>>     If we are to think of the RSWG as a typical working group (which
>>     the document otherwise suggests is how we should think of it),
>>     then I can't discount the possibility that something might be
>>     handed to the RSAB that is not really ready for consideration, for
>>     whatever reason.  As I read what you're saying here, this proposal
>>     presumes that can't (or won't) happen.
> 
>     I think the way to look at it is risk-based, and that authority is
>     balanced such that the risks of authority abuse by the RSAB are
>     prioritized certain classes of errors made by the community.
> 
> 
> This document suggests to me that abuse concerns are handled by the 
> appeals process, not by the balloting phase.
> 
> I think my review of this is informed by my experience on the IESG.   
> The IESG has a ballot position called "DISCUSS" that lets us require 
> further discussion in several situations before the document under 
> review can advance to publication.  Four prime examples of this are: (a) 
> a document that is trying to present a good idea, but does it quite 
> poorly; (b) a document that is trying to advance a bad idea, even if 
> clearly; (c) a document whose development process was not properly 
> followed; and (d) a document that was tainted by a process abuse of some 
> kind, such as a conflict of interest or improper use of discretion.
> 
> As I read the proposed "CONCERN" criteria, they cover some but not all 
> of those possibilities, and that's what caught my attention. 

The Program discussed the CONCERN criteria in fair detail. My read is 
that (a) and (b) would be solved by having RSAB members involved in RSWG 
discussions and that (c) and (d) could be appealed.

> Now, it 
> could be the case that "serious harm" and "serious problem" in the 
> "CONCERN" criteria are meant to cover all of those possibilities, but 
> that wasn't evident when I read it.  I suggest that be made explicit if 
> that's the intent.

IMHO "the proposal would cause serious harm to the overall Series, 
including harm to the long-term health and viability of the Series" is a 
more specific and clearly defined criterion than "is trying to advance a 
bad idea" in your (b) clause.

As to your (a) clause, if the proposal is defined so poorly that it 
can't be implemented then I would expect an RSAB member to raise a 
CONCERN that the proposal causes serious harm, as above.

> It's been pointed out to me separately that there's an expectation that 
> the RSAB will be more directly involved with the development of a 
> proposal in the RSWG than would be the case with similar work in the 
> IETF stream.  If so, then a "CONCERN" would indeed be unique (just like 
> me putting a "DISCUSS" on a document that I approved at the WG level 
> would be pretty weird), but it also means this expectation of direct 
> involvement was not sufficiently evident in the document when I read it.

The document says this several times, specifically in Section 3.1.1.2 
and in Section 3.2.2 points 3, 9 & 10. Do you feel it should also be 
mentioned elsewhere (e.g., in the section about the RSAB)?

Peter