[ppml] Soliciting comments: IPv4 to IPv6 fast migration

[Ted Mittelstaedt]

>-----Original Message-----
>From: ppml-bounces at arin.net [mailto:ppml-bounces at arin.net]On Behalf Of
>Paul Vixie
>Sent: Thursday, July 26, 2007 2:44 PM
>To: ARIN Address Policy
>Subject: Re: [ppml] Soliciting comments: IPv4 to IPv6 fast migration
>
>
>> ... because the community does not have the balls to allow it to happen.
>
>i don't think it's wise to bet against this community's will or powers.

Why not?  You say later on here that there isn't an internet community so
seems
a safe bet to me...

>
>> If the Internet community had balls, they would appoint a Czar
>and tell all
>> IPv4 holders they had until 2010 to switch to IPv6 and pay the
>fees, to hell
>> with your legacy status.  In 2010 they would aggressivly block
>IPv4 all over
>> the Internet.  In 2012, everyone would have switched to IPv6 and
>you would
>> have 4 or 5 large legacy holders in court, suing
>ARIN/IANA/everyone claiming
>> they were illegally forced to submit to IPv6.  The courts would find in
>> their favor sometime in year 2020 by which time IPv6 would be so
>entrenched
>> and IPv4 so dead, that the wins would have no meaning whatsoever.  And no
>> court would go against the rest of the world and try ordering
>the Internet
>> to stop blocking IPv4 so the legacy holders could get their free
>ride for a
>> few more years.  And even if one did the rest of the world would
>ignore it
>> with the result that a tiny chunk of the Internet would revert
>to IPv4 and
>> become useless.
>
>while i won't address your concern (or lack of same) about lawsuits against
>ARIN, i'm generally in favour of "tough love" positions.  see my 1995 paper
>on domain names (http://sa.vix.com/~vixie/dns-badnames.pdf) for an example.
>
>the big problem with the above proposal isn't lack of "balls", but lack of
>coherency.

Same thing.  Your just using the politically correct terminology.

>there isn't an "internet community" in the sense you mean, and
>there isn't going to be a Czar because there's no way to get universal
>agreement on who it could be.

If there was coherency, ie: "Balls" and a way to get universal agreement
(or mostly universal agreement) there wouldn't be a need for a Czar.

>whatever "we" (the internet
>community or even
>just the ARIN community) do will be by bottom-up consensus,
>period.

If we do anything.  Which right now there is no universal interest in
doing anything other than letting IPv4-to-IPv6 migration just happen
by itself.  Every proposal to either hasten or push out the date
by anybody has been
shot to pieces, sometimes by legitimate bullets, other times by politics.

Clearly, the majority will is that things are fine the way they are.
If some people have problems getting IPv4 allocations at some point
in the future, well tough cookies to them, they shouldn't be bothering
with IPv4 then anyway, they should be using IPv6.

It's the old "I got mine so I don't give a crap if you can't get yours"
mentality.

If the Internet community really did care about people's addressing needs
in the future, most people would be bothered by the idea that there are
legacy addresses floating around out there that aren't being used, aren't
being advertised, but are unavailable for assignment because of paperwork
baloney.  Such as for example the 199.248 block I have brought up before
that
the only reason it's just floating around is because the company that was
assigned to it doesen't even know they have it, and couldn't use it even if
they did know they had it.  Most people would also be bothered with the
idea that there might be legacy allocations currently being advertised that
are way in excess of the address holders needs, but are out there because
the legacy holders got them free and don't want to have to start paying a
yearly fee.  In other words, there would be a whole lot more interest in
cleaning house on IPv4 than just saying the house is a giant effing mess so
we are just going to shut the door on it and build a new house.

The feeling seems to be we are going to spend a lot of effort on migrating
to IPv6 now, because it's easier FOR US to solve the technical problems
of migration, than the political problems of cleaning up IPv4.  While that
might be OK, the issue that is being skirted is how do you handle the
situation in the future, where post-runout, Sally Sue needs addresses, gets
her IPv6, and now needs to connect to Billy Goat who has been on the
Internet
longer than her, has IPv4 he's using that can connect to everyone else,
and hasn't gotten around to deploying IPv6 yet.

It doesen't seem to me that Sally Sue has any compelling enough reason to
induce Billy Goat to go dual-stack all of his stuff.  So, Sally is going to
end up having to go through an intermediary, which is going to put a cost
and hack burden on her that Billy Goat doesen't have.  And as the
intermediaries
get more use, the Billy Goats of the Internet will have even less incentive
to
dual-stack.

I am not sure that even you, Paul, understand.  The Czar isn't really going
to
be needed to force IPv6 migration, he's going to be needed to force IPv4 off
the Internet or we are going to see a huge growth of hacks to try to get
both
addressing to coexist, which is going to impact stability.  And if it gets
bad
enough, and the Internet gets unreliable enough, then the billionaires that
run sites like Google will see it impact their bottom line, and call in the
governments, who are going to take control.

That is how it's worked in other industries where they tried "bottom up"
solutions
in the face of change.

Ted