[DNA-BOF] Replacement term for L3 Link?

[Greg Daley <greg.daley@eng.monash.edu.au>]

Hi Eric, JinHyeock and Spencer,

There's some difficulty in finding an appropriate term,
especially because of the usage of link as an L2
designator.

As you know, the term link is already used in IPv6,
for on-link determination for router discovery. Also,
it is the boundary for the link-local scope.

So essentially, we have already got ambiguity.

I believe that as Spencer says, we're trying to
get to a point where we can describe the reachable
nodes which can be reached without forwarding or
hopping.

However, I think also that it may be awkward to use
the term 'hop' since most IP configuration messaging
never 'hops' at all.

Since we're talking about the domain where local delivery
works, there was a suggestion here that we call it
an 'IP neighbourhood'.

Essentially, this would be the range where link-local
traffic is delivered, and consequently (since ND/RD is
based on link-local packets), the range where the
configuration is valid.

Change of neighbourhood therefore entails IP
configuration change.

Please tell me what you think.


At this stage, we've submitted the charter to the ADs
for pre-circulation to internet area WG chairs, so
we really have to get the charter nailed down (today?
yesterday?)

Of course, I've got a couple of (previously planned)
amendments to put in as indicated by Eric and JinHyeock.

We're getting ready for the next IESG conference call,
so the charter has to be ready, approved by the ADs,
checked by the WG chairs and available to the IESG
for inspection at least week before the call
(call is on the 22 Jan).

So if we can live with L3 Link in the charter as it stands
(with the minor amendments) then that would be easiest.

In that case, any terminology we agree on would go into
the first round of charter amendments when the group
next re-charters (Charter issue #1).

Since this may not be acceptable, I'll create a new
charter to show how this term could be used as a drop-in
replacement for L3 Link.

I'll post this shortly.

Greg





NJEDJOU Eric FTRD/DMR/REN wrote:
> I agree with keeping the term "link" for L2. Even if people are not obliged to using the charter terminology, i think the charter is a place where terms should not carry any form of ambiguity. 
> 
> -----Message d'origine-----
> De : Spencer Dawkins [mailto:spencer@mcsr-labs.org]
> Envoyé : jeudi 8 janvier 2004 13:39
> À : dna@eng.monash.edu.au
> Objet : Re: [DNA-BOF] Modified Charter available.
> 
> 
> Yes, please pick terms now, otherwise you'll spend time explaining
> ideas to confused people later.
> 
> My suggestion is "link" for level two/link layer links, and "IP hop"
> for level three/network layer links - if you decrement TTL/hop limit,
> it's an IP link, if you don't, it's something else. "IP hop" seems
> silly, but isn't that what the IP hop limit is limiting?
> 
> Please insert your suggestions for better names here, of course!
> 
> Spencer
> 
> ----- Original Message ----- 
> From: "NJEDJOU Eric FTRD/DMR/REN" <eric.njedjou@francetelecom.com>
> To: <greg.daley@eng.monash.edu.au>; <dna@eng.monash.edu.au>
> Sent: Thursday, January 08, 2004 4:49 AM
> Subject: RE: [DNA-BOF] Modified Charter available.
> 
> 
> 
>>Hello folks,
>>
>>
>>>Description of Working Group:
>>
>>>For the purposes of detecting network attachment, an
>>>L3 link is defined by the range within which IP
>>>packets may be sent without resorting to forwarding.
>>>In other words, a link is the range where a given IP
>>>configuration is valid.
>>
>>In my opinion talking about "L3 link" is not a good idea at all.
> 
> People
> 
>>will very soon get confused between L2 link and L3 link. The
> 
> traditional
> 
>>"network layer" or "IP layer" expression was very convenient.
> 
> 
>