ADs who are also WG chairs

[Brian E Carpenter]

"Hallam-Baker, Phillip" wrote:
> 
> > What's to stop Big Bad Company stuffing an Oasis WG with ringers?
> 
> What is to stop them doing that in an IETF WG? Since everyone is
> theoretically participating as an individual I can't see how the WG Chair
> could claim consensus if big bad company wanted to block something.

This is the point about IETF consensus that you truly don't seem
to get. The chair, or the IESG, when judging consensus, is entitled
to decide that the 100 hands waving are equal to only one, because
of the knowledge that Big Bad Company has sent 100 people to the
meeting or mailing list. You may not like the fact that we give
this power to the chairs and the IESG, but that *is* the IETF's
answer to this question.

> Oh yes, sorry I remember the WG chair can claim consensus because at the end
> of the day it is only his opinion that matters. See DNSEXT for details.

We shouldn't argue from one case where you happen to disagree with the
chair and have chosen not to appeal. We have ~100 WGs and I don't hear
100 claims of this kind, or even ten a year. (That doesn't mean your
complaint is invalid, or valid, but we have no evidence that it's
typical.)

> > What's to stop a commercial grouping pushing something through
> > an Oasis WG that is good for them and harmful to the Internet?
> 
> Same as when the IETF proposes something that would be harmful, the
> community of users rejects a protocol that it does not want.

There is no connection between my question and your answer, and no 
link between the first and second halves of your answer.

The equivalent question is "What's to stop a commercial grouping
pushing something through an IETF WG that is good for them and harmful 
to the Internet?" The answer is supposed to be "checks and balances
applied by careful WG chartering, Last Call and IESG review." This
could of course be failing - that's a legitimate topic for this list -
but what I was asking is how does Oasis deal with this risk?

The market is of course always able to reject something it doesn't want.
That is orthogonal to something that is harmful. Users don't understand
what is harmful to the commons, they understand what is good for them.

> But it is a good question since there are far more companies with market
> power who are now participating in OASIS than are currently fully engaged in
> IETF process. So if there is a risk that 'evil' protocols come from OASIS
> and 'corrupt' the Internet then that is going to be happening regardless of
> what the opinion is here.

That may be true. And if it's true, let's consider that it may be so *because* 
the IETF has been successful at deflecting 'evil' work. But actually, I
suspect that Oasis is mainly working on topics that IETF would consider
out of scope (and that W3C might have expected to handle).

> > How can Roberts' Rules be applied on a mailing list?
> 
> Roberts rules are not directly applicable to a mailing list since
> it is asynchronous conversation. They are applicable to conference
> calls and in person meetings.
> 
> Even so we have had many successful polls through mailing lists,
> in many cases using Quaker polls and other consensus finding
> processes rather than the more divisive up or down questions we
> see in IETF process.

The biggest problem is that RR allow the chair to drive to a decision
on exactly one question at a time, or to push that question on the
stack until a subsidiary question has been resolved. Serialising
debate in that style is essentially impossible on a mailing
list.

> 
> > (I'm not trying to make debating points - I really am concerned
> > that these are serious weaknesses in the model you describe. And I
> > spent two years chairing a Board that supposedly followed Roberts'
> > Rules and used email extensively, and I can't answer the third
> > question.)
> 
> And your evidence for thinking that the IETF approach is working at any
> level would be what?

Finding out what isn't working in the IETF is exactly what this
WG exists for.

   Brian