[ppml] Markets, pricing, transparency, 2008-2 / 8.3.9

[Leo Bicknell]

In a message written on Wed, Mar 19, 2008 at 08:49:57AM +0900, Randy Bush wrote:
> let's get real here and look at over-inflated selves in the mirror.

Sounds like a plan.  Do you have your mirror handy?

> our "self regulation" has led to 75% of recent allocations going to just
> ten organizations (and they keep merging!), a barrier to entry at the
> low end which we justify with words about routing table bloat, and a
> routing table that is half crap, i.e. severely bloated.  i, for one,
> ain't proud.

I don't think the reasons that ten organizations in the ARIN region
have 90%+ of the eyeballs has anything to do with IP allocation
policies.  Phone and cable companies had a monopoly on wires in the
ground prior to the IP thing being of interest.  I am disappointed
that the Internet wasn't able to change that, but that was a long
shot anyway.

As for the barrier to entry; I think we could have done a better
job with IPv4 and should be doing a better job with IPv6.  There
are allocation strategies that make it easier to give out small
blocks now and insure they can grow into larger blocks later and
we seem to mostly not use them.  Sadly, it's probably (10 years?)
too late for IPv4.

> we are on the same exhaustion curve frank solensky drew in the early
> '90s, before arin etc.  i.e. all our "self regulation" has not
> significantly affected consumption of the resource we are supposedly
> conserving.  the only thing which affected things was cidr [0].  i, for
> one, ain't proud.

Looking at what both socialist and democratic countries do to our
natural resources does not lead me to believe we are good at
conversation as a society.  That aside, conservation costs money.
Could someone create SSL based web sites that don't need an IP per site?
Sure.  Could someone design a DHCPeqsue system that allowed prefixes for
end users to move around in response to need for things like WiFi
hotspot networks?  Sure.  But someone has to fund it.

Call it the tragedy of the commons if you like, but I'm not sure what
ARIN could do to prevent it.  I'm also not even sure that "regulation"
= "conservation" in this case either.  Let's look at what our best minds
came up with, from RFC 2050:

1) Conservation: Fair distribution of globally unique Internet address
   space according to the operational needs of the end-users and Internet
   Service Providers operating networks using this address space.
   Prevention of stockpiling in order to maximize the lifetime of the
   Internet address space.

Color me confused, but conservation = fair distribution?  Clearly
the way to conserve oil then is to just divide it all up fairly.
Humm, that doesn't seem right.

Truth be told I don't think conservation was ever a goal.  Consider this
paper from 2000: http://user.it.uu.se/~pekka/ip-over-anything.pdf

IP over everything.  The mantra was get this in the hands of everyone.
It will solve world hunger, eliminate poverty, topple oppressive
governments, and cure diseases.  I don't know that it's done any
of that; but now people living in poverty can pirate pop music and
get plenty of porn, that's still a quality of live improvement,
isn't it?

The only thing this industry ever made any serious effort to conserve
was routing slots, because they cost money.  IP's enabled revenue.
IP's grew the number of interconnected parties increasing value.

> like icann, we have managed to create a global self-perpetuating
> expensive bureaucracy to do the job which used to be done by one
> computer scientist working part time [1].  i, for one, ain't proud.

> [1] - sure it would take more than dr. postel today.  but a thousand
>       times more?

As bureaucracies go I'd say ARIN is pretty cheap.  However, you're way
off the mark.

ARIN's "fact sheet" (http://www.arin.net/about_us/fact_sheet.pdf)
says ARIN has a "professional staff of about 40".  I'm not sure how
going from 1 person to 40 people is "a thousand times more".  Another
interesting measure may be hosts on the internet:
http://www.isc.org/index.pl?/ops/ds/host-count-history.php

In 1997, the year ARIN was founded, there were by this measure
26,053,000 (end of year, adjusted) hosts on the network.  The last
measure is 541,677,360, a growth factor of 20x.

Considering Jon never had a public policy meeting, never processed
a SWIP template, and didn't maintain a web site full of information
in multiple languages.  I doubt he had to respond to subpoenas about
file sharing, spam, child porn and who knows what else.

Now you'll probably cry foul that I didn't include APNIC, RIPE, LACNIC,
and AfriNIC.  In the time Jon was doing what he did many of these
regions did not have significant Internet infrastructure.  He never had
to deal with the language issues in some of these regions.  I think a
Jon to ARIN compairson is the most fair.  But least you don't, the
numbers:

APNIC: 48
RIPE: 100
LacNIC: 16
AfriNIC: 10

Total: 214

So, we still don't make your thousand times more, and that's without
factoring that many of these groups have taken on additional roles,
like ripe developing their database software or running the RIS
project, LACNIC administering a scholarship program, etc.

> btw, the graph <http://www.arin.net/statistics/index.html#ipv4org> is
> not the same in all regions, for example, lacnic, ripe, ...

Back to your original question; it would take a fairly over-inflated
view of the importance of addressing to think the graph should be
the same for all regions.

Monopoly ownership of wirelines, government censorship, per capita GDP,
localization of software; these are factors that change the demand for
networks, and in turn IP's.  ARIN could constrict demand (new policy
proposal: no one gets IP's anymore, thanks!), but can't grow it beyond
giving IP's to anyone who needs them, which is the current state.
Networks are not constrained by ARIN, they are constrained by other
forces.

In short, the graphs are different for reasons wholely unrelated to
addressing policy.

-- 
       Leo Bicknell - bicknell at ufp.org - CCIE 3440
        PGP keys at http://www.ufp.org/~bicknell/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20080318/729a0426/attachment.sig>