HI Brian,
On Tue, Jun 2, 2020 at 3:57 PM Brian E Carpenter <brian.e.carpenter@xxxxxxxxx> wrote:
On 03-Jun-20 10:11, Ted Hardie wrote:
> On Tue, Jun 2, 2020 at 2:56 PM Stephen Farrell <stephen.farrell@xxxxxxxxx <mailto:stephen.farrell@xxxxxxxxx>> wrote:
>
>
>
> On 02/06/2020 22:41, Ted Hardie wrote:
> > And you are convincing me that attempting to settle it on the IETF list
> > will require somebody to judge consensus, since there look to be a minimum
> > of two people with the time and keyboards available to disagree. We
> > apparently, however, disagree on who that should be.
>
> Perhaps not! If you do agree that consensus calling is
> required that seems to imply the LLC is not the one to
> do that. We have a bunch of 14 victims already setup
> to do just that:-)
>
>
> I think the LLC can call consensus on a matter within their remit (just as the IAOC evaluated the feedback on the registration date change policy that I referenced many messages ago). So, I think they are the victims set up to do that in this case.
It's a change to the openness of the standards process, unprecedented since we first started multicasting the audio for free back in the early 1990s. BCP101 defines the LLC's scope:
This isn't one thing. It's a modification of the working methods of the IETF to deal with a crisis (and there's many more aspects of the methods which are changing). It's an effort to provide administrative support for both a specific meeting and the IETF long term. And it is an effort to manage the fairness of impact. For many of us, the lack of airfare, hotel, and reduced costs for this makes this a very cheap option compared to other years. For others it will go up, but they will be on even keel with others. For those who truly can't afford it, donations have made up 100 registration fees, and I'm sure the LLC would be happy for more donations to help those in need.
"The IETF LLC is established to provide administrative support to the IETF. It has no authority over the standards development activities of the IETF."
There's no doubt that the IETF Executive Director *sets* the fees,
As I said in my note to Stephen just now, I meant that the LLC Board, as the relevant community leadership, would call consensus on this. I thought that was implied by the references to "they" and "them", but my apologies for not being explicit.
but IMHO that isn't the point at issue. In this text:
"The IETF Executive Director sets those meeting fees, in consultation with other IETF LLC staff and the IETF community, with approval by the IETF LLC Board."
I don't see any indication of how the ExecD knows the result of consulting the community when there is disagreement. The mechanism we have for that is the IESG determining the rough consensus.
As I said, I think any of our leadership is expected to be able to manage a consensus call, and I don't see why the LLC board shouldn't do it. They are responsible for setting the fees; they use community guidance in RFCs, consultations with the IESG, and public feedback. They are also community selected leaders, just like the IESG. Putting the IESG back in that particular role is a step backward in the division of labor we've set out, and I don't see that its either warranted or appropriate.
regards,
Ted Hardie
I can see nothing in BCP101 that gives the ExecD the power to determine IETF
consensus, although it does require the LLC to respect IETF consensus. Those are two different things.
Maybe this is a tiny gap in RFC8711, where Ted and (Stephen + I) have different interpretations.
Regards
Brian
> Since you referenced the magic number 14, I conclude we still disagree..
>
> I think we do agree that there should be public discussion. I think we do agree that the LLC and IESG should talk to each other about the implications of different strategies to both the ongoing work of the IETF and its financial future. I think we do agree that any conclusion would be revisited in the light of evidence of how it ends up working.
>
> But our disagreement on on who the stuckee is remains.
>
> regards,
>
> Ted
>
>
> Cheers,
> S.
>
