Re: [tcpm] WGLC on draft-ietf-tcpm-ecnsyn-07

[Alfred Hönes <[e-mail]>]

Sally,
folks,

First of all my apologies for having missed the WGLC because
of the misbehaving upgraded Mailman instance that eventually
has been fixed today by the Secretariat ...

I generally agree that this document has a very useful direction,
and I do not really want to hold it off in any way.

However, as I had written in November off-list to the authors:

> I have quickly followed up to your revised ECN SYN/ACK draft,
>     draft-ietf-tcpm-ecnsyn-07.
> An interesting evolution of results!
>
> Since I guess there will be more follow-up research and more updates
> to the draft, I defer another full pass over the document until later.
>
> For the moment, only ...   /snip/

My assumption about continued research efforts had not been opposed
by the authors in November, but I have not heard of any additional
results announced to the wg since.

Have there been more results in the meantime confirming the results
that lead to the -07 modifications?
Are we sure in the meantime that other scenarios and modified
parameters will not lead, perhaps in the short term, to another
shift in the observations from simulation results?

I would not have any concerns if the draft were targeting
Experimental, but it aims at Standards Track.

Last year, at the first WGLC, all seemed rather stable, but now?
Will the IESG challenge the soundness of the proposal, due to
the recent change?
How can we convince the IESG that the proposal is ripe for PS?

Personally, I do not seriously expect additional research killing
the arguments for the proposed algorithm, but additional results
might still change the perspective slightly.
I would appreciate to hear more than one voice from the research
community regarding these considerations.
A WGLC based on a single 'heavy-weight' +1 might not suffice to
provide evidence of "strong support" to the IESG.


Thus, the question arises whether the *normative* part of the
proposal should better be split off into a relatively short PS
document, accompanied by an Informational document with the
detailed presentation of the motivation, quantitative research
results and related work, etc.
The former would be more concise and perhaps be preferred by
implementers, and the latter could more easily be updated or
amended if additional results became available.


I'll try to resume the incremental review that had been paused
in November quickly now and, if necessary, report more editorials
to the authors within a few days.

I do not insist on the split proposal made above, but I would
appreciate it being discussed and the decision for progress
be made based on consideration of the possible consequences
of new results.


Kind regards,
  Alfred.

-- 

+------------------------+--------------------------------------------+
| TR-Sys Alfred Hoenes   |  Alfred Hoenes   Dipl.-Math., Dipl.-Phys.  |
| Gerlinger Strasse 12   |  Phone: [phone], Fax: -18         |
| D-71254  Ditzingen     |  E-Mail:  [e-mail]                     |
+------------------------+--------------------------------------------+