RE: Action item from yesterday's meeting

> > This is not a MIB design discussion.  This is a pure semantics
> discussion.
> > Certainly, we could implement the semantics you are discussion in
the
> MIB
> > with some effort.  But first the WG would have to agree on changing
> semantics
> > that were discussed and agreed on about a year ago.
> >
> [suresh] Agreed. I ran this by Christian Huitema brifly during the
IETF
> and I
> believe, he was agreeable to the change. The change, after all, doesnt
> break
> the functionality for traditional NAT and simply extends the semantics
for
> twice-NAT and other types of NAT.

Let's be clear: I don't support the change in semantics, and I believe
that specifying a flow by end-to-end addresses is just fine. However, if
Suresh can insert the functionality without making the protocol more
complex, I have no objection.

-- Christian Huitema.