>From my experience, the term "L2.5" (and "MAC layer", although MAC tends
to be a more media-specific term, and sometimes "Framing layer") is
generally used to refer to behavior that is specific to a given
_combination_ of a L2 media type and an L3 protocol.  

L2 would generally be behavior that is independent of whether the L3
protocol is IPv4, IPv6, IPX, LLDP, or whatever else.

I think this distinction is important.

-Dave

> -----Original Message-----
> From: Alexandru Petrescu [mailto:[EMAIL PROTECTED]
> Sent: Monday, March 26, 2007 9:54 AM
> To: Roberto Baldessari
> Cc: Templin, Fred L; [EMAIL PROTECTED]
> Subject: [Int-area] Re: [Manemo] the layer below IP
> 
> Roberto Baldessari wrote:
> > Hi Fred,
> >
> > Thanks for writing this down.
> >
> > As I said privately, I agree. "sub-IP" is definitely more general
and
> > less confusing than 2.5.
> 
> I'm sometimes comfortable calling that a MAC layer instead of sub-IP.
> 
> Exceptions are some links which don't have MAC layers.
> 
> Alex
> 
> >
> > regards,
> >
> > Roberto
> >
> >> -----Original Message----- From: [EMAIL PROTECTED]
> >> [mailto:[EMAIL PROTECTED] On Behalf Of Templin, Fred L
> >> Sent: Thursday, March 22, 2007 10:24 PM To: [EMAIL PROTECTED] Cc:
> >> [EMAIL PROTECTED] Subject: [Manemo] the layer below IP
> >>
> >> I made a rather strong statement with an individual during a
> >> meeting this week that really should more appropriately have been
> >> vetted with the rest of the community instead - so, here I am.
> >>
> >> In the classic model, we have traditionally considered the layer
> >> immediately below IP (L3) as the link layer (L2). But, in some
> >> common scenarios there is actually quite a lot that goes on below
> >> IP but above the datalink. Some have called this "L2.5", but I find
> >> that term to be misleading because it seems to imply that there is
> >> one and only one functional layer between L3 and L2 and that does
> >> not accurately capture all scenarios.
> >>
> >> I wrote a draft that tried to address this by re-introducing the
> >> legacy OSI terminology which specified a 3 sublayer decomposition
> >> of L3 (the internet sublayer (L3c), the subnet enhancement sublayer
> >>  (L3b) and the subnet access sublayer (L3a)). But, even this more
> >> specific terminology suffers from hazy distinctions between what
> >> belongs in one of the sublayers as opposed to another.
> >>
> >> So, can we simply call it the "sub-IP layer" as opposed to the
> >> other alternatives? "Sub-IP" makes no statement about any layering
> >> structure that may occur immediately below IP, and as such it
> >> allows for any level of mechanisms that may need to be implemented
> >> below IP down to and including the (classical) link layer. The term
> >> can also be used even if there is no intermediate mechanism at all,
> >> since L2 is by definition a "sub-IP" layer.
> >>
> >> In summary, I belive there is no one precise way to call the
> >> mechanisms that might occur immediatley below IP so can we just
> >> collectively call them:
> >>
> >> The "sub-IP layer"
> >>
> >> Thanks - Fred [EMAIL PROTECTED]
> >>
> >> _______________________________________________ Manemo mailing list
> >>  [EMAIL PROTECTED] http://www.mobileip.jp/mailman/listinfo/manemo
> >>
> >
> > _______________________________________________ Manemo mailing list
> > [EMAIL PROTECTED] http://www.mobileip.jp/mailman/listinfo/manemo
> >
> 
> 
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please visit http://www.messagelabs.com/email
> ______________________________________________________________________
> 
> _______________________________________________
> Int-area mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/int-area


_______________________________________________
Int-area mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/int-area

Reply via email to