Re: [re-ECN] Congestion Exposure BoF request for Hiroshima

Bob Briscoe <rbriscoe@jungle.bt.co.uk> Mon, 07 September 2009 23:42 UTC

Message-Id: <200909072342.n87Ngso2011571@bagheera.jungle.bt.co.uk>
Date: Tue, 08 Sep 2009 00:42:55 +0100
To: John Leslie <john@jlc.net>
From: Bob Briscoe <rbriscoe@jungle.bt.co.uk>
In-Reply-To: <20090907161401.GQ8532@verdi>
References: <20090907161401.GQ8532@verdi>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Cc: re-ECN unIETF list <re-ecn@ietf.org>
Subject: Re: [re-ECN] Congestion Exposure BoF request for Hiroshima
Precedence: list

John,

Thanks - this helps a lot (even tho I've said where I've not taken 
your adivce).

inline

At 17:14 07/09/2009, John Leslie wrote:
> >
> > Attached is my attempt so far. I started again - I'm happy with it so
> > far, but it needs the specifics added at the end, where indicated.
>
>    I commend Bob on what he's written. It's good, even if a bit long...
>
>    For the last three items:
>]
>] TBA: Vision. Relax transport constraints. Congestion != impairment.
>
>    Don't bother.

I've cut out vision, but still included cut-down words on the others.
(I've got the whoel thing a lot shorter, and signposted the last part 
as "Further Background", so people don't have to go there)

Relax transport constraints is what a lot of people want this for - 
one has to play to the crowd.

I find congestion = impairment is the most FAQ I get (unsurprisingly, 
given the word means that).

>] TBA: Proposed working group deliverables
>
>    Hmm... There needs to be a specific document defining usage of bits
>in an IP header. It needs to be either Experimental or Proposed Standard,
>and we need to say which.
>
>    There needs to be a document (quite possibly the same one) saying
>what Congestion-Exposure-aware routers may do with these bits; and
>what we expect of non-Congestion-Exposure-aware routers to NOT do.

I've mentioned this. Thanks.

>    I'm not certain what we need beyond those, though we will need
>Security Considerations somewhere.

I don't need to mention Sec Consids - they will be within the main 
protocol spec (I've got loads and loads and loads of work on this in 
my thesis - I'll have to refer rather than include a lot of it).

>    I expect we're looking for Experimental status; and that suggests
>three deliverables: one stating usage of bits, one evaluating the
>results of the experiment, and one detailing Security issues.
>
>    I don't thing we need to list more deliverables than that in a
>BoF request.

Don't think we need Sec Consids separate. But we do need "problem 
statement/requirements/motivations" as there is a lot to say about 
that, and it doesn't fit well in a protocol spec.

>] TBA: Proposed BoF Agenda.

I've cut this. It can come later.

Bob

>    Copy most anyone else's. ;^)
>
>--
>John Leslie <john@jlc.net>

________________________________________________________________
Bob Briscoe,               Networks Research Centre, BT Research

[re-ECN] Congestion Exposure BoF request for Hiro… John Leslie
Re: [re-ECN] Congestion Exposure BoF request for … Bob Briscoe