On Jun 30, 2011, at 9:46 PM, Martin Rex wrote:
Keith Moore wrote:
On Jun 30, 2011, at 1:09 AM, Martin Rex wrote:
(a bunch of stuff in defense of NAT)
Rather than having another of an endless series of discussions about
the merits of NAT or lack thereof, can we just agree that it should
not be pre-ordained that this WG should assume NAT as a solution?
You absolutely want to have fairly fixed addresses within
your home network, and you absolutely want to have a short-lived
ephemeral IP-Address assigned on your internet side of your
home gateway for the purpose of privacy.
No, *you* want these things. I do not, and imposing them is not in the
interests of users in general.
Otherwise the number of very unpleasant surprises, including
stuff like this:
http://www.theregister.co.uk/2011/06/30/xbox_swat_police_rait/
Nowhere in this article does it say that the user had a static IP address. And
as I indicated earlier, even with a dynamic IP address, there are frequently
other ways to find a host's IP address. If you force all hosts on a home
network to have dynamic IP addresses, you break applications that need stable
addresses. If you don't force all hosts on a home network to have dynamic IP
addresses, those that don't need stable addresses can still get ephemeral
addresses via privacy addresses, DHCP, or other means.
You keep arguing for the perpetuation of bad hacks because of accidental
properties of those hacks. I'd rather have well-designed mechanisms that are
tailor made to suit particular purposes.
Keith
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf