[ppml] Research on transfer markets, was: RFC 1744 and its discontents
tedm at ipinc.net
Tue Apr 22 19:37:54 EDT 2008
> -----Original Message-----
> From: ppml-bounces at arin.net [mailto:ppml-bounces at arin.net] On
> Behalf Of Randy Bush
> Sent: Tuesday, April 22, 2008 3:46 PM
> To: Matthew Pounsett
> Cc: arin ppml
> Subject: Re: [ppml] Research on transfer markets,was: RFC
> 1744 and its discontents
> [ i do not disagree (strongly) with much of your message, but
> i really can not let the following go ]
> > Considering that "fair and equitable distribution" is what
> we think we
> > have now
> some may think that. the screwees we barred from entry at
> the lower end sure do not, and have not for a decade. have
> we not read their pain on nanog and this list?
> what we have is a highly jury rigged distribution based on
> how market incumbents thought the game should be played. if
> we don't think it is jury rigged, why is policy so long and
> complex and why do we keep making up more?
The ARIN policy is really not any more complex than the Parker Brothers
Monopoly game rules. See:
> let me repeat
> > look how long a community of vested incumbents has kept barriers to
> > entry at the low end. cathy screamed at me in public when
> i suggested
> > small allocations in the first arin denver meeting; when we had an
> > actual victim there whose business was being severely
> hampered by our
> > bottom up policies. now, when the end is in sight, we are finally
> > talking about sharing the last orts here in the restaurant
> at the end
> > of the universe. how magnanimous of us.
> > to be clear, the bald-faced trade we made was protecting the
> > incumbents' router capital costs by making it hard for folk
> at the low
> > end.
> and, at the other end, 75% of the address space has gone to
> ten entities in the last couple of years. and take another
> look at the internal statistics,
> perhaps we should understand that some people might not
> consider this "fair and equitable distribution?"
There is a reason for that which has nothing to do with IP space
assignment, and has a lot to do with how the US FCC punted on
the issue of requiring networks to be open.
Tell me, how many ISPs are choices on ANY given cable TV network
in the US?
Thank the FCC for that.
Tell me, how many ISP's are choices on the Verizon fiber deployment?
Thank the FCC for that one as well.
And the DSL providers are not far behind.
> and i do not blame arin hostmasters, management, ... they do
> a great and often thankless job. we, the community did it
> [ab]using arin's philosophy of thinking with our bottoms up.
The idea that policy in IPv4 assignment has significantly affected the
trend of consolidation is pure fiction.
Consolidation is in MANY markets. It is the NORMAL AND NATURAL
progression of corporations in any capitalistic market.
THAT IS WHY WE HAVE ANTI-TRUST LAWS
Left to themselves, just about ALL markets would collapse into
And the idea of protecting incumbents router costs?
Baw ha ha ha ha!!!!!
Geeze Randy, ex-runner of gated software, please, please lecture
us on how ALL ISPs MUST USE HARDWARE ROUTERS that have super expensive
fixed cards in them.
We ran gated and later Zebra quite well for years on FreeBSD,
and would be still doing so to this day if Ebay hadn't made
used Cisco gear available for LESS than the cost of a fast
>From a TECHNICAL perspective, an ISP can be EASILY built from
purely commodity, cheap PC hardware that will serve quite enough
customers to qualify the ISP for a /22 (IPv4)
And if your too incompetent to build one (an amazing thing if
your planning on running that large of an ISP) you can go to
Imagestream and buy a PC off the shelf with it preconfigured
as a router and have at it.
The problem of building an ISP with that many customers is a
SALES AND MARKETING one. NOT a technical one.
When you are competing against a cable provider that has the
city covered with a sole monopoly ISP, and has the bankroll to
LOSE MONEY for YEARS at a time to build up a customer base, it
is no wonder that small ISP's can't get started these days.
More information about the ARIN-PPML