Capwap Mailing List: Re: Process in CAPWAP

[Bob O'Hara (boohara) (boohara]

Dan, David,

I certainly appreciate the difficulty that our chairs have to navigate the path 
to consensus necessary for a successful standard.  I do not envy them their 
task, particularly in this area where we believe that the success of the CAPWAP 
standard is critical to this market area and the customers in it.

There are certainly enough open issues to resolve, that asking for expert 
advice on this particular issue is not likely to adversely affect the time for 
completion of our work.  I think that this will be very helpful to us. 

Thank you.

 -Bob
 
-----Original Message-----
From: Romascanu, Dan (Dan) [mailto:dromasca [at] avaya.com] 
Sent: Thursday, June 15, 2006 1:07 AM
To: Bob O'Hara (boohara)
Cc: capwap; david.kessens [at] nokia.com
Subject: RE: Process in CAPWAP

Hi Bob,

The ADs looked in the issue that you raised related to the consensus on the Mux 
vs Multiport issue # 115. We analyzed the traffic on the list and we also 
discussed with the WG chairs.  Our opinion is that although the chairs 
suggested a solution for consensus the WG did not agree with it. Despite the 
chairs and WG efforts to reach a timely conclusion according to the current WG 
schedules, more work needs to be done to reach at minimum rough consensus on 
this point.

The chairs have the responsibility to drive the consensus process, and the 
criteria for reaching the decisions should be clearly explained. 

In order to overcome the dead-end we recommend to the WG to bring in the 
discussion opinions about the consensus criteria (QoS, scalability, 
compatibility with routing and switching infrastructure), their priority and 
recommendations from external experts that were less involved in the 
discussions until now, for example by consulting the Routing Area.  We offer to 
approach the Routing Area ADs to have an expert designated to this purpose.

We recommend that the WG continues its work and capitalizes on the effort made 
until now to close a large number of open issues in the next version of the 
draft. This issue together with the other open issues should be brought for 
discussion on the list and in Montreal, and we hope that a conclusion based on 
rough consensus can be reached soon after the Montreal meeting, so that the 
impact on the Working Group schedules be minimized.

David and Dan

 
 

> -----Original Message-----
> From: Bob O'Hara (boohara) [mailto:boohara [at] cisco.com] 
> Sent: Saturday, June 10, 2006 3:16 AM
> To: Romascanu, Dan (Dan)
> Cc: capwap
> Subject: Process in CAPWAP
> 
> Dan,
> 
> A few days ago Dorothy Gellert sent an email to the working 
> group on behalf of the chairs, announcing a decision on a 
> technical issue in the working group.  Specifically, she 
> declared that the chairs decided the discussion of the use of 
> a mux header vs. the use of separate ports for the 
> differentiation of control and data traffic in the protocol 
> to be over.  She also indicated that the chairs had chosen 
> the mux header as the method to be used in the protocol and 
> set a deadline of today for closing the issue.
> 
> As soon as I read her email, I replied and asked that she 
> document the process that the chairs used to arrive at their 
> decision.  My background is in the IEEE, where the process is 
> clear, transparent, and open to all participants.  CAPWAP is 
> my first IETF working group.  My understanding of 
> decision-making in IETF is that it is based on rough 
> consensus and that the chairs have responsibility to 
> determine that consensus.  With that understanding, I asked 
> Dorothy to document what the chairs used as evidence of rough 
> consensus on that issue, since I see evidence in the postings 
> to the list that there is more support for separate ports 
> than there is for the mux header.  At a minimum, there is no 
> consensus on this issue.
> 
> Since sending that email two days ago, neither chair has 
> responded.  I find that quite disturbing for several reasons.  
> 
> First, this is an issue that has been debated extensively 
> (and still is being debated, regardless of the pronouncement 
> from the chairs).  The working group deserves to know how 
> this technical issue was resolved.  
> 
> Second, standardization by fiat of the chair does not seem to 
> be in keeping with the open process requirements of the IETF. 
>  Perhaps I am being naïve, but I don't believe there is a 
> feted inner core (FIC) that is actually developing all the 
> IETF standards.  
> 
> Third, the time allowed for any response to the email was 
> only three days.  This is an absurdly short time, given that 
> it is the start of the summer vacation period in the northern 
> hemisphere.  It is quite possible that many participants will 
> not even see her email until after the deadline expires.  
> 
> Finally, if there is no consensus on the issue, the chairs 
> are expressing an engineering opinion and should be required 
> to justify that opinion just as any other member of the 
> working group.  I don't believe that the IETF anoints the 
> chairs of any working group as expert and able to make 
> technical decisions for the working group.
> 
> I would appreciate your thoughts, as the AD, and response on 
> these items.
> 
> Best regards,
>  -Bob
> 
> Bob O'Hara
>