[ppml] [arin-announce] Legacy RSA

John Santos JOHN at egh.com
Fri Oct 12 17:37:48 EDT 2007

On Fri, 12 Oct 2007, Ted Mittelstaedt wrote:

> >-----Original Message-----
> >From: ppml-bounces at arin.net [mailto:ppml-bounces at arin.net]On Behalf Of
> >Kevin Kargel
> >Sent: Friday, October 12, 2007 7:26 AM
> >To: ppml at arin.net
> >Subject: Re: [ppml] [arin-announce] Legacy RSA
> >
> >
> > That's a pretty good deal.  I wish I could get it in my contract that
> >my rights would never be lessened even if policy changed.
> >
> Kevin, don't forget ARIN defines legacy holders as IPv4 holders.
> IPve ONLY.  In ARIN's definitions, there is no such thing as an
> "IPv6 Legacy Holder"

So far as I know, no legacy holder has obtained or asked to obtain
IPv6 addresses (or additional IPv4 addresses) without signing an
RSA and becoming a non-legacy holder.  I believe that some legacy
holders may also hold non-legacy addresses, but I don't know that
for sure.  In any case, this is straw man.

> I'll be happy to write you a contract that states unequiocably that
> you have permanent, perpetual rights in how to configure any one
> of a box of Latticenet cards I happened to see in a junk store
> a couple years ago. ;-)  Or maybe Arcnet cards?  ;-) ;-)

Lots of contracts contain perpetual rights clauses.  So what?

> My only concern with the "Legacy RSA" is that somewhere within it, there
> is a statement that the term "Legacy holder" means "IPv4 only holder"
> That way there is no chance in the future that some court could
> misinterpret the contract and use it to extend over IPv6 assignments.

IANAL, but this seems extremely unlikely to me.  The only way to
obtain IPv6 assignments is to sign an RSA specifically applying to
those resources, and no court is likely to ignore that.

> This really comes down to your position on moving to IPv6.  The
> official word is that IPv4 runout is a fact, and that IPv6 will
> replace it.  There are, unfortunately, a lot of people out there
> (like Dean) who apparently think that they can manipulate the system
> into making the Internet some sort of permanent shared IPv4/IPv6
> environment - if that were to happen, the assignments of the Legacy
> holders would become a permanent, unpaid, drag on the Internet.

There are a lot of incorrect assumptions here.  1) IPv6 will *not*
replace IPv4, at least not for many years, if not decades.  The
"Internet" will be mixed IPv4/6 for a long time to come.  2) It
is not a matter of "manipulation" to make the Internet a permanent
shared IPv4/IPv6 environment, it is a natural consequence of a
slow migration to a new technology.  The only alternatives are
either creating an entirely new, unconnected IPv6-only network
(with virtually no users, providers, no Ebay nor Amazon nor YouTube
nor CNN nor Kazaa nor emails of pictures from your brother-in-laws'
cousin's bowling tournament), or a global flash-cut, where everyone
has to toss out billions (trillions) of dollars worth of suddenly
obsolete equipment.  3) I can't recall Dean (or anyone else)
actually saying they want to acquire and think they have the
right to acquire IPv6 resources without signing an RSA for them
and without paying normal end-user fees (i.e. $100/year) just
because they are a legacy v4 holder.  Some people have proposed
*offering* legacy v4 holders IPv6 and/or reductions in fees as
an inducement to getting them to sign an RSA and return unneeded
IPv4 assignments, but that is not the same thing at all.  4) Legacy
IPv4 assignments can never increase, only decrease (as the holders
go out of business, abandon their holdings, fail to find an ISP
which will route for them, or move to IPv6.)  It is fundamental
mathematics that a monotoniclly decreasing function will tend
toward zero, so the whole "permanent, unpaid drag on the Internet"
argument is wrong.  5) At one time, the Internet consisted solely
of legacy assignments.  Are you saying, despite Moore's law operating
over a decade and a half, it costs more now than it did 15 years ago
to handle legacy traffic?

> But in the last analysis, the so called "rights of the (IPv4) Legacy
> holders" are only of value as long as the rest of us continue to keep
> routing their legacy traffic - ie: their IPv4 traffic.  Ask yourself,

This is yet another false assumption.  I have a legacy class C (/24)
that is of great value to my company, despite the fact that it doesn't
get routed.  We have 3 private-line connections to various customers,
and if we renumbered to RFC1918 addresses, we would have to co-ordinate
the usage with all 3 customers, and an unknown, much larger group of
potential future customers.  (The issues are routing, firewalls, traffic
segregation, etc. and grow geometricly with the number of participants.)

I've mentioned this before and lots of people have said they have
similar situations.

> what is a reasonable expectation as to how long the rest of the Internet
> will be willing to continue to do this, before telling the Legacy
> holders (and, indeed, ALL IPv4 holders) that they must switch to IPv6.
> That is the true length of time that these so called "rights" will have
> any actual value.

Eventually, the last v4 host will be shut down.  Long before that, the
last v4 router in the DFZ will be shut down or converted to v6 (or v-next
or whatever) only.  So what?

> Ted
> _______________________________________________
> You are receiving this message because you are subscribed to the ARIN Public Policy
> Mailing List (PPML at arin.net).
> Unsubscribe or manage your mailing list subscription at:
> http://lists.arin.net/mailman/listinfo/ppml Please contact the ARIN Member Services
> Help Desk at info at arin.net if you experience any issues.

John Santos
Evans Griffiths & Hart, Inc.
781-861-0670 ext 539

More information about the ARIN-PPML mailing list