[arin-ppml] IPv4 Depletion as an ARIN policy concern

[Roger Marquis]

Kevin Kargel wrote:
> NAT started out as a kludgy local workaround and will always pretty much be
> a local workaround. NAT is nothing more than a silly router trick.

Whether or not NAT is "kludgy", a "workaround", or "silly" is an opinion.
It is an opinion not supported by fact i.e., the market for end-node
network gear without NAT, which is effectively zero.  As a result arguments
against NAT are irrelevant.

> Just because NAT is not "built in" to the IPv6 protocols does not mean it
> can or will not be done.

This wasn't the case for v4, why would v6 be different?  Applications like
SIP hack (workaround, kludge, ...) the ISO 7 layer model in ways it was not
intended to be hacked, and will need to be accommodated.  To keep NAT
implementations and libraries from being numerous, and to avoid
incompatibilities, a standard is required.

To state this another way, those of us who manage end-nodes will not be
purchasing IPv6 network gear that fails to support a standardized version
of NAT.  Manufacturers will in turn not offer such equipment knowing it
will not sell.  How anyone can continue to argue that IPv6 does not require
NAT in light of this is beyond me, but I do hope you work for a competitor.

Roger Marquis