[arin-ppml] simple question about money
tedm at ipinc.net
Mon Jun 9 12:56:24 EDT 2008
> -----Original Message-----
> From: arin-ppml-bounces at arin.net
> [mailto:arin-ppml-bounces at arin.net] On Behalf Of Brian Reid
> Sent: Monday, June 09, 2008 8:20 AM
> To: ppml at arin.net
> Subject: Re: [arin-ppml] simple question about money
> If IPv6 is going to catch on with the technical public, the
> process of using it needs to be sufficiently transparent that
> somebody who spends 3 or 4 hours reading the website to
> figure out what to do needs to come away with a sense of how
> to get things done.
Nobody who has never had any exposure to TCP/IP IPv4 can spend
"3 or 4 hours reading" a website about it and come away with
a sense of how to get things done. Why are you assuming that
it would be any different with IPv6?
I watched many administrators who came out of the Novell NetWare
era who only had exposure to IPX take many months of struggling
to understand TCP/IP.
I remember asking the Teleco manager of Central Point Software,
a really nice guy named Jeff (who is now a karate instructor,
funny the paths people take in their life), back in 1992, if he
had heard about TCP/IP and the Internet, and getting a blank stare.
This was pretty common.
> I didn't. I know that many of you will
> just respond by telling me I'm stupid, but I really did read
> what was there and try hard to understand it.
Your not stupid and please do not get defensive. If IPv6 was simple
then why did it take a team of expert network engineers to dream
> Nowhere did I
> see an explanation that the administration and management of
> IPv6 is (presumably
> by design) different from that of v4 and that the concepts
> don't just move over. In particular, the whole
> supply-and-demand thing is not properly explained or
> documented. There was a scarce resource (v4). It is priced
> high, like many scarce resources. Now there is an abundant
> resource (v6), with millions of times more availability.
> Conventional economics suggests that the more abundant
> resource ought to be a lot cheaper because of the principles
> of supply and demand. It isn't.
I am sure it will be once everyone is using IPv6.
> The mistake I made was assuming that the bizarrely high price
> was motivated by greed or stupidity. I admit that my
> expectations there were set by having paid close attention to
> ICANN's early years. The bizarrely high price is in truth
> motivated by solid operational and technical concerns (and is
> therefore not bizarre), because there still is a scarce
> resource, namely routing table slots. I think it would help a
> lot for ARIN documentation to explain why the usual laws of
> supply and demand
> seemingly do not hold here.
But this is NOT ARIN's concern, as to how many routing slots
One of the "dirty little secrets" in the IPv4 world is that a
lot of legacy assignments were handed out to some of the largest
orgs, pre-ARIN. Those orgs pay nothing for those per the agreements
that created the RIR's. Thus, the cost of administering numbers
is, in fact, lopsided, and is bourne unequally by the orgs that
came later on.
When the automobile was first manufactured, the federal government
required zero safety features. Thus the automakers did not have to spend
any money on these features, and the buying public didn't have to pay
for them. Later on the feds did start requiring many safety and today,
emissions features on automobiles. These add thousands to the cost
of the car. As a result, the latecomers buying new cars today bear
the costs of all these extra safety and emissions features. Anyone
who is still out there driving around a Model-T never had to pay those
costs. Their cost of ownership per mile is a lot cheaper than the
guy with the new vehicle.
That is just how things work. Fortunately, as the shift to IPv6 will
make useless those legacy assignments, ultimately those orgs that are
paying nothing now, will have to start paying for the cost to adminster
IP addressing, and therefore the costs will be more evenly spread, and
should drop for everyone.
More information about the ARIN-PPML