MAJOR ISSUE: "Concentration of power"

[Brian E Carpenter]

Hi,

Basavaraj.Patil at nokia.com wrote:
> 
> >> As in any organization there needs to be a system of checks and
> >> balances. At this time there does not seem to be any such for the
> >> power of the IESG.
> >
> >I don't think that's true. There are at least 4 "features" of the IETF
> >proces that act as checks and balances on the IESG:
> 
> Okay... I guess that is true.
> 
> >
> >1. The IAB as an advisory board for WG formation (and in practice for many
> >major IESG decisions)
> 
> For WG formation, yes. What kind of *major* IESG decisions does the
> IAB get involved? Are there instances of the IAB challenging and
> over-ruling the IESGs decisions (other than in the case of
> appeals). Does the IAB get involved only when a view is sought by the
> IESG or does the IAB act proactively?

My experience is out of date since I left the IAB more than a year ago,
but in my time on the IAB, there were several occasions a year when
the IESG would ask the IAB informally for advice on how to unblock
a situation or decide about a contentious document. And there were
certainly cases where the IAB stuck its nose in. I felt this was
the IAB's job, if it noticed a case where it thought the IESG was at
risk of going off track.


> 
> >2. The appeals process
> 
> The appeals process causes fear in some of being ostracized by the
> IETF community. So not vey effective. Looking at the number of appeals
> in the past and outcome, one would be discouraged to take the appeals
> route. But I agree that the process provides the capability. I am not
> sure how well this works in reality.

I think it has been very useful. Of course, for the IESG and IAB, an appeal
is a nuisance and looks like a distraction from "real work", and it would
be a problem if there were too many appeals per year. But a small
number of appeals (won or lost) keeps everyone aware that the process
rules deserve respect.
> 
> >3. The recall process (never tested, for some reason)
> >
> >4. Payback time, i.e. the NomCom process.
> 
> How? By providing input to the nomcom? Does this work?
> And looking at the results of the last few nomcoms, I dont see this.

The nomcom gets contradictory input, so their decisions are not always
to everybody's liking.

> 
> >Now, it's possible we need *more* checks and balances, of course.
> 
> Or *more effective* ones than what exists now :)

Can you phrase that as a precise problem statement?

   Brian