[ppml] Proposed Policy: IPv4 Countdown
owen at delong.com
Mon Mar 19 18:47:02 EDT 2007
>> c) Even if IANA did maintain this sort of data, if we were to stop,
>> it would mean essentially nothing to the various ISPs and end users
>> that would continue to use IPv4.
>> You seem to have a view that there is a responsible adult who can
>> make the hard decisions. That entity does not exist.
> That isn't true. There is always a "responsible adult" lurking around
> out there, you just want to stick your head in the sand and ignore
> it. That "adult" is called the government, and it's form is of the
> various legislatures and courts and dictators out there who run
> things in the world.
Um, no... THAT is not true. Number one, I would hardly refer to most
governments as responsible, or, adult. Number two, there is no single
government that has authority over IP addressing. In fact, it's
to envision any government having any real authority over any IP
addressing other than specifically what has been delegated them for
their use. Afterall, a government order that says party A owns 188.8.131.52
doesn't mean much if no ISP will route it for them. It means even less
if several other ISPs will route it in other countries for someone else.
> You seem to think that these governments don't give a hoot if we
> all botch up this IPv4 thing and end up with a train wreck on the
> Internet because nobody stepped forward and implemented the hard
I think that whether we botch it up or not, the governments cannot
the situation whether they care or not.
> I can tell you they do. If we don't do it, they will, with
> And the results will not be pretty. That is what happened to the
> DNS ssytem. Do you want a repeat?
You're right. The results of legislation will not be pretty, and, it
that they will have any semblance of what lawmakers intended the results
to be. First, US law is not binding on about 2/3rds of the internet,
believe that puts US law in charge of the largest portion of the
the control of a single national government. Number two, it's pretty
to imagine how one would actually successfully legislate the usage of
a collection of 32 bit integers. Not that the government is smart
to realize this futility. We're talking about the same people that
illegal for me to have an ounce of black powder in my possession for
ejection charges in rocketry without a Federal Explosives User Permit,
require me to store even an ounce of BP in a type 2 or type 3 magazine
if I do possess it legally for rocketry purposes, yet they allow me
up to 50 pounds in my possession with no permit and no magazine if
it is for my firearm hobby. (Yes, this is actually the current state
law on that subject, I kid you not.)
I don't want a repeat, but, I also don't think the government can
achieve a repeat. DNS got mucked up by the trademark and intellectual
property interest groups. The government had minimal effect, but, in
case, the US DOC had clear jurisdiction over certain domains, and, WIPO
managed to do a land-grab over others.
Do you really think that anyone is going to hand global IP management
over to WIPO in any meaningful way? I don't. I don't think anyone
follow WIPOs lead, even if the law said we should.
I think that IP routing and addressing works because people agree to
cooperate and non-cooperating parties tend to get de-routed. I think
that any interference with that process will be treated as damage
and be routed around (paraphrasing old RFCs and related historical
commentary). IP is very different from DNS. It's much harder to claim
you own PI or any other number as property than it is to claim you own
a trademark name.
Frankly, the government/WIPO takeover of DNS was not because it was
not being handled correctly, but, because WIPO didn't like the fact that
domain names were NOT the same kind of namespace as Trademarks
and they were afraid of a meaningful namespace existing that they
didn't control. I would argue that what we were doing RIGHT is what
bothered WIPO the most.
>> The entire
>> addressing system works because people agree it works and cooperate
>> to define how. That's why Lee, John, et al, have asked for
> It doesen't matter if we all agree on a proposal, it matters if the
> proposal we all agree on is carried out.
A proposal cannot be carried out until it becomes policy. It can't
policy until it is proposed. As such, the ability to carry out said
depending on us agreeing to it, I would say it matters very much what
we all agree on.
> For those who advocate a weak numbering authority, which is what we
> now (no offense intended) there is little controversy for them to
> do their
> jobs. But human nature being what it is, when we run out of IPv4 at
> some day in the future, there will be some networks out there who make
> decisions based on what is best for them and damn the rest of the
> A weak central authority won't be able to rein them in.
This is where we disagree. Networks who do what is in their own
interest in a manner sufficiently contrary to the common good will
find themselves without peers to connect them to the rest of the
This has happened historically, and, it will happen in the future.
no need for a strong central authority, and your continued focus on
to scare us into needing one, among other things, convinces me that
our existing rational anarchy is actually pretty robust.
> What do you think the likely outcome will be if a large network
> with many
> very rich online porno customers needs more numbers and makes a
> request of
> ARIN and is denied because there are no more IPv4 numbers?
> a) they will tell their customers "sorry no more servers"
More likely they will tell their customers that they need to get more
> b) they will just grab a random IPv4 block allocated to someone else
> that they don't think is being used.
What makes you think that any of their peers will accept this
for any significant period of time? What makes you think that if
this type of abuse they will continue to have peers?
More information about the ARIN-PPML