Re: call for consensus on proposed semantic change

Melinda, 

First of all, thank you for clarifying the scope of change (to MIB draft only).

I agree with Juergen in that the "end-to-end input" variant is not quite
end-to-end, unless you are talking about just one NAT (traditional NAT) in
between.

Now, as for the "hop-by-hop input" variant, I would suggest renaming it to
"realm-specific input". Hope you dont mind my using the new term for the
reminder of the e-mail. 

Realm-specific input variant is a superset of the two variants in that
realm-specific is not restricted to traditional NAT. The change proposed is
simple, and has big bang for the buck, I believe. NAT vendors can relate to
realm-specific input and can implement midcom mib easily, no matter what type
of NAT they implement. The concept is a natural for NAT vendors and is in line
with the terminology used in NAT MIB and other NAT RFCs. Application vendors
also benefit because the realm-specific variant will work with all NAT flavors.


NAT vendors - Hope you agree with what I am saying. Do chime in. Thanks.

regards,
suresh


--- Juergen Quittek <[email protected]> wrote:

> Melinda,
> 
> I want to add a clarification.
> 
> The proposed change (the "yes" option) is a hop-by-hop variant,
> but the current version (the no option) is not necessarily
> end-to-end.
> 
> It is end-to-end if there is only a single NAT on the data path.
> 
> With more than one NAT on the data path, a NAT will typically
> not receive address information of both ends of the communication.
> 
> And, not surprisingly, I prefer the "no" option.
> 
> Thanks,
> 
>     Juergen
> 
> 
> --On 21.11.2004 10:22 h -0500 Melinda Shore wrote:
> 
> > There's been some discussion about whether or not midcom's semantics
> > should be changed so that the PER and PRR inputs would contain hop-by-
> > hop descriptions rather than end-to-end descriptions.  If you support
> > this change please post "yes" to the mailing list; if you do not support
> > this change please post "no."
> >
> > The change would be reflected in changes to the existing MIB draft but
> > not to the published semantics RFC.
> >
> > Thanks,
> >
> > Melinda
> >
> >
> > _______________________________________________
> > midcom mailing list
> > [email protected]
> > https://www1.ietf.org/mailman/listinfo/midcom
> 
> 
> 
> 
> 
> _______________________________________________
> midcom mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/midcom
> 


=====