[rfc-i] I-D Action: draft-iab-rfc-editor-model-v2-04.txt

Dave Crocker dhc at dcrocker.net
Tue Mar 13 10:37:29 PDT 2012

On 3/13/2012 7:38 AM, Joel M. Halpern wrote:
>> The charter needs to have simple statement, like the above. Offhand, I
>> would recommend the charter to mandate a chair and tell RSOC to choose
>> from amongst its members.
> The document already says that the RSOC picks its own structure. The point here
> was merely to emphasis that the IAB appointments do not change that. I put that

The charter needs to /specify/ things.  Commentary like the text being 
discussed, tends to add more confusion than clarity, since it is winds up being 
specification not commentary.

To the extent that this text is trying to say "in addition to appointees to the 
RSOC, any IAB member is allowed to participate whenever they feel like it" I'll 
suggest that that is a highly counterproductive model.  I trust that my phrasing 
points in a direction that makes the downsides clear, but will elaborate if it's 
too obscure.

> in as part of resolving some earlier concerns that the text might imply some
> sort of leadership role in the RSOC for the IAB members.

If RSOC organizes itself and chooses it's own chair, then that's pretty clear. 
Elaboration on that is typically counterproductive, IME.

>> That's fine. But that's covered by saying that the IAB appoints RSOC
>> members.
>> It does not need to specify the various pool lengthy and confusing
>> statements about who it might appoint aren't needed.
>> Much worse is to have stray IAB folk wandering in and out of RSOC.
>> Either appoint them as members or don't.
> Fundamentally, the RSOC role is still an IAB responsibility. If the IAB chooses
> to behave in a way that makes of mess of its own job, that is an IAB failure and
> not an RSOC failure.

That's always true of one supervising body over another.  One of the things that 
reduces the potential problem is clarity of boundaries.  RSOC should have very 
clear boundaries, responsibilities, accountability, etc., with respect to its 
supervising body.

> Unless we want to reopen the IAB Charter, this document can not and should not
> declare that IAB members can not be on the RSOC.

That's not in any of the language I or anyone else used in this thread.

Saying that IAM members can be on the RSOC is different from saying "the IAB 
appoints RSOC members, some of whom may be IAB members" and then adding "any IAB 
member may participate in RSOC at their whim"...  The latter is a 
counterproductive arrangement along a number of group performance parameters.

> Hence, to say that the IAB
> should stay away from the RSOC seems a counter-factual statement.

Again, no language in this thread has promoted that statement.

What I, at least, /am/ saying is to distinguish members from non-members and to 
restrict participation of non-members.  My own recommendation is that the 
presence of non-members should be at the invitation of RSOC.

That members are composed variously of IAB members and others is a distinct point.

Stated differently:  Define a group.  Give it a charter.  Give it oversight -- 
reporting, accountability and presumably the ability to reverse decisions. 
Define membership in the group.  Let the group operate with administrative 
integrity; do not build in confusion.

> The IAB should
> be involved, at least via some IAB members, in the RSOC.

fine.  i haven't seen anyone suggest otherwise.

> The goal of the text as written is actually to let the community, and future
> IABs, know that it is important that the IAB be careful not to overstep and
> dilute the effect of creating the IAB.

huh?  1) I don't know what that means.  2) I don't understand how the 
problematic text does that.

> Note that the document does not mandate terms or numbers for the RSOC
> appointees. Thus, a truly foolish IAB could be parachuting in arbitrary members
> from anywhere. This text serves to remind everyone involved that stability and
> avoiding micro-management are important parts of the picture. If we remove this
> text, the parachuting is still the case. But the warnings are gone.

 From what I can tell, it doesn't accomplish that goal.  Perhaps the more 
constructive path is to have language that specifies the core attributes needed 
amongst those who are appointed, whether IAB members or not?  (I'd offer a list 
but I'm not sure what you guys think they should be...)



   Dave Crocker
   Brandenburg InternetWorking

More information about the rfc-interest mailing list