[arin-ppml] The role of NAT in IPv6

[William Herrin]

On Thu, Apr 15, 2010 at 6:37 PM, Gary T. Giesen <ggiesen at akn.ca> wrote:
> I'd argue that the collective costs of implementing and supporting NAT
> far outweigh the costs of more routing slots (though neither is a small
> number, I'd argue NAT is at least an order of magnitude higher).

Gary,

It's entirely possible that more money is spent designing software
around NAT's limitations than is spent supporting routes in the DFZ.
Nevertheless, comparing the two is a logical fallacy.

In a free society it is perfectly OK for me to choose to buy something
whose use is incompatible with also choosing to buy and use your
product. That's called an "opportunity cost." It's quite unreasonable
to suggest that I shouldn't be allowed to buy the other product merely
because you will then have to spend money altering your product before
I'll be willing to buy it.

On the other hand, when I announce a route, I directly and immediately
consume resources on your router, whether or not I'm communicating
with your customers or taking any action that causes you to earn
revenue. I'm not merely declining to purchase your product, I'm
actively taking money out of your pocket. Since this otherwise
antisocial behavior turns out to be critical to the overall function
of the Internet, it makes sense to regulate it through an application
of public policy.

In summary: my declining to purchase your pink dress *is not* a valid
public policy concern. My taking your wallet *is* a valid public
policy concern. Try to understand the difference.

Regards,
Bill Herrin


-- 
William D. Herrin ................ herrin at dirtside.com  bill at herrin.us
3005 Crane Dr. ...................... Web: <http://bill.herrin.us/>
Falls Church, VA 22042-3004