RE: CTP and CARD Drafts

Hi James,

Thank you for the background and the IAB point of view explanation on
seamoby as a group. I guess IAB IESG also works on "rough-consensus"
and is not all clean cut as it looks to outsiders :)

Just one point, we did not spend 3 years on requirements, both the
original problem description and final protocol design went along
without too much fanfare. Unfortunately the requirement phase went through
lengthy debates due to I must say too broad customer-base (quite the contrary
to what it is now :) ) and a major part of the delay (over 6 months)
was due to the requirement draft sitting with IESG for whatever reason.
I understand the load IESG is dealing with, but the WG cannot be
held responsible for that. Maybe we should have started protocol discussions
instead of waiting, or maybe we did need the time to cool off :)

Anyhow, Thank you for carrying the load and getting it all done.

Regards,

Madjid

-----Original Message-----
From: James Kempf [mailto:[email protected]]
Sent: Thursday, March 11, 2004 11:07 AM
To: Nakhjiri Madjid-MNAKHJI1; 'Allison Mankin'
Cc: [email protected]
Subject: Re: [Seamoby] CTP and CARD Drafts 


Hi Madjid,

I'm sure the WG appreciates the efforts of everyone involved in the author
team, certainly I do. Seamoby has not been an easy working group to
participate in. Any good quality work that happens has had to push against
that headwind, and therefore should be doubly valued.

Regarding your question about ROHC review, Allison and I had discussed
having CTP reviewed by someone from the ROHC WG,  and I actually had invited
Carsten to be on the informal Review Board, though he never responded to my
email. So there has been a longer term idea there to have ROHC, as a
customer of CTP, review the design prior to submitting it to the IESG. This
kind of broad, cross area review is considered one of the strong points of
developing a technology in IETF, and that applies to experimental work as
well. It doesn't mean that CTP can be vetoed by ROHC, just that any concerns
they have, as customers, need to be take into account.

Regarding your question about why Seamoby was chartered when there were
fundamental doubts about CT, etc., it is actually fairly atypical on the IAB
and IESG that everyone agrees with a working group charter in IETF if the
chartered work is doing something relatively new that IETF has not done
before (and also in some cases where the work is not new). A recent example
is the CAPWAP work. So the Seamoby CT work is not any different in this
respect. The difference in the case of Seamoby is that the WG spent 3 years
arguing about requirements and was unable to come to concensus on a set of
requirements that made sense, so the IESG has been understandably reluctant
to continue to support the work given the lack of any discernable quality
output. In fact, we are lucky that the IESG, as part of their management
function, didn't simply shut down the WG last year. As a way to summarize
and document the WG's experience with the technologies under discussion and
allow participants to take away a publication, Allison, Pat, and I decided
last year to request the WG to move the protocol drafts to experimental,
which the WG approved, allowing the IESG review to be somewhat lighter. But
this doesn't mean that there is no review at all. It is still part of
Allison's job to see that the documents are reviewed, and any substantive
comments she or any reviewers she recruits uncover need to be addressed.

Regarding your ramping down your work on CT, I understand that you and other
participants on the author team are no longer actively working on CT. This
is a fairly usual occurance in a WG where the outcome is less than what was
originally hoped. As chair, I've got the responsibility to see that any
reviewer comments are addressed so the documents can be published, in fact,
I'm in the process of doing that now for the CARD draft. I will, of course,
pass the results by the WG for comment, but since the drafts are now
effectively out of the WGs hands, we do need to address any comments that
come from AD or IESG review to the satisfaction of the IESG, and not simply
dismiss them, or the drafts will not be published.

Hope that answers your concerns.

            jak

----- Original Message ----- 
From: "Nakhjiri Madjid-MNAKHJI1" <[email protected]>
To: "'Allison Mankin'" <[email protected]>; "James Kempf"
<[email protected]>
Cc: <[email protected]>
Sent: Wednesday, March 10, 2004 10:36 AM
Subject: RE: [Seamoby] CTP and CARD Drafts


> Hi Allison, James,
>
> Thank you James for the clarification and I actually do have the concern
> you brought up.
>
> Allison, I understood the committment to be by Rajeev and Carsten to
review
> the feasibility of CTP for Rohc. Was that discussed as a pre-requisite
> for CTP to go IESG review? especially since CTP is going to be an
> _experimental_ RFC anyway? I guessed I missed that part!!
>
> I for one did give the warning that Rohc of all protocols has the toughest
> requirements on CTP timing and so on, but I guess the balls had already
>  started rolling! Our original TEXT draft (pre-design team) did bring
> up a whole bunch of issues that any CT protocol would need to overcome
> when dealing  with header compression.
>
> A bunch of other protocols were suggested in Wienna for a simpler
> CTP deployment and they were knocked down.
>
> After all that and all the tears, sweat we dropped over CTP in the
> last 3-4 years, it is surprising to see the ADs bring up the discussions
> and doubts they had about CT at time of WG formation!!!
> Why was the WG formed and the item accepted then?
> After all, IETF is based on people doing volunteer work,
> earning daily bread elsewhere.
> IESG should strive keep those people put and motivated rather than pulling
> items out after 4 years.
>
> Finally, I personally had ramped down my CTP volunteer work mid last year
> due to other committments. I applaud other authors, especially Rajeev
> for carrying the load and I have no doubt about the quality of Rajeev's
>  work on Rohc CTP, but I still see this process as a surprisingly
stringent one.
>
> Regards,
>
> Madjid
>
> -----Original Message-----
> From: Allison Mankin [mailto:[email protected]]
> Sent: Tuesday, March 09, 2004 10:18 PM
> To: James Kempf
> Cc: Nakhjiri Madjid-MNAKHJI1; [email protected]
> Subject: Re: [Seamoby] CTP and CARD Drafts
>
>
> All,
>
> Actually, there's been a little bit of misunderstanding of my message
about the
> CTP draft, which is easy to have happen, because we were all very busy
during
> the IETF week.  What I meant was that I had to talk with some ROHC people
about CTP,
> not that ROHC _WG_ review of the document would be required.
>
> Why is this?  It's due to some commitment that was made in the Vienna
Seamoby
> meeting to deliver Seamoby with a use case, which I understood would
probably be
> header compression context.  The minutes show an action item to work with
Carsten
> Bormann to make this concrete.  We discussed the need for the use case
that
> would describe a context very well defined in IETF, with ROHC header
compression
> meeting that criterion.  If contexts for context transfer are just
open-ended,
> and there isn't an clear use case, it's hard to build support for this
> protocol in the broader IETF (not only the IESG - we had a lot* of
questions
> about this charter item when the WG was formed).
>
> So, I've done a little talking and looking at the document.  The example
of a
> service that was supposed to be covered in Appendix C is missing.  There
is no
> companion document describing a use case.  What's up?
>
> Allison
>
> http://www.ietf.org/proceedings/03jul/index.html
>
>
>
>
>
>