[arin-ppml] Address Range transfer question

Ted Mittelstaedt tedm at ipinc.net
Wed Aug 20 00:34:14 EDT 2008

> -----Original Message-----
> From: wherrin at gmail.com [mailto:wherrin at gmail.com] On Behalf 
> Of William Herrin
> Sent: Tuesday, August 19, 2008 3:29 PM
> To: Ted Mittelstaedt
> Cc: Stewart Dean; arin-ppml at arin.net
> Subject: Re: [arin-ppml] Address Range transfer question
> On Tue, Aug 19, 2008 at 5:26 PM, Ted Mittelstaedt 
> <tedm at ipinc.net> wrote:
> > Your was ASSIGNED to you by PSInet.
> > It was NOT "sold" to you.  PSInet had no legal authority
> > to "sell" blocks to customers at the time, and in
> > fact, NEVER had such authority.
> Ted,
> If was at the time a legacy block (given the 
> range in which it falls, I assume it was) then the legal 
> situation is a lot more murky.

That is true.

> No legal structure existed 
> which would preclude PSInet from disposing of its legacy IP 
> addresses in whatever manner it saw fit.

Nothing precluded PSInet from SAYING anything that they wanted.
Wether a letter that they wrote to Bard College would be
accepted by the rest of the community, -particularly- by the
people running Whois at the time, is a different matter
entirely.  I do not believe that it would have been.  And,
quite obviously, NEITHER did ARIN since it proceeded to
allocate the entire block that contained the /24's that
PSInet supposedly disposed of to Bard, to cogentco.

Of course, all of this is interesting historical speculation,
nothing more.  You may disagree but I believe that the
important thing is for Stewart to understand that letters
that were written by allocatees pre-RIR have practically NO
legal weight today.

We are very likely going to see more of this, of people
trotting out 15 year old documents and such that claim
they have a stake on a piece of the IPv4 landscape, in
the future.  I don't think it is wise to encourage it,
as it distracts from the more important discussion of
how to best proceed forward.

Stewart could spend the next year e-mailing and writing
various people attempting to try to stake claim on
these numbers - and for what?  To avoid a renumber, a
very simple thing.  If he simply does as I recommended
and proceeds forward with abandoning his existing holdings
and getting a fresh block, he will be done before he
knows it.  And not only that but he will have a 
contiguous block and can make some logical routing
arraingements, such as advertising aggregated routes
inside his network.

> On the other hand, in theory those 12 class C's constitute a 
> /21 and a /22 which are not necessarily invalid end-user 
> assignments under current policy. If you can justify their 
> use under today's ARIN policy, Bard College may be able to 
> ask ARIN to assign the blocks as if new under the existing 
> policies. Particularly if you can get Cogent to sign a letter 
> to the effect that they "don't oppose" the assignment.

William, those 12 "class Cs" are part of -
which constitutes a total of 256 "class Cs" according to
Stewart.  As Cogent has undoubtedly requested and
obtained additional IPv4 blocks from ARIN since that assignment,
those PSInet IPv4 numbers have been included in their
utilization mathmatics.  In short, they had to justify
utilization of all the IP numbers that were part of the
block that were NOT assigned to Bard to get more of them.

I would be amazed if Cogent was given further assignments
from ARIN while allowing a total of 244 /24's to sit unused.
I would assume those subnets are assigned to other customers.
Why would Cogent be willing to break up a nice, clean /16
that they can advertise with a single routing entry, into
multiple advertisements, just so they can slide these 12
subnets out of their allocation?


More information about the ARIN-PPML mailing list