Was this tongue in cheek?
I am not an economist, but this is a sure fire way to destroy the internet
as we know it today.
Kris
Was this tongue in cheek?
Nope.
I am not an economist, but this is a sure fire way to destroy the internet
as we know it today.
Why is this so hard to believe? Real estate is mostly a free market, and
that seems to perform pretty well for the most part. How is address space
that different? Please explain how you believe this would destroy the
internet.
Overbuilding office space in San Francisco doesn't *usually* cause real
estate issues in Tokyo.
There's a lot more available real estate than available v4 address
space. That's the biggest one. Second, groups that make the Internet
go aren't necessarily the ones who can afford the address space.
Third, there's no root owner of the address space, so who is going
to sell it?
But really, you just need the first one: small space. The relatively
small pool means that a large company with lots of money could buy the
whole ARIN chunk of the Internet. Speculators would probably buy
address space and leave it unused, much like they do for domain names.
Address space would have to be bought and sold on arbitrary borders,
resulting in massive fragmentation of the tables.
On the good side, address-based lawsuits would revitalize our flagging
litigous economy 
And I agree with what others are saying: It would kick IPv6 into high
gear. I'm not sure that's a good thing, however.
-Dave
I must be missing your point because I don't see how that's relevant.
Think of the 69/8 space as being out in the middle of nowhere (due to the
bogon filters). It's cheap land but it comes with it's problems. If you
invest in it, you can increase it's value, if you don't want to, buy the
more expensive downtown property.
Brandon Ross wrote:
Why is this so hard to believe? Real estate is mostly a free market, and
that seems to perform pretty well for the most part. How is address space
that different? Please explain how you believe this would destroy the
internet.
Aggregation on a per-network basis would go down the tubes. ARIN sets aside a
larger block than a network's initial allocation when first apply. I'm not sure
what the criteria are for this initial large block, but I assume it's around a
year's worth of projected growth. The network is then given adjacent pieces of
that larger block as the need is justified. Once it have used up that block, you
have one nice, pretty entry for the global tables.
If any network had to go to eBay to buy a /22 every time they need one, look out.
You think the BGP tables are too big now ?
Of course, the other alternative is to buy a HUGE block of IP addresses and hoard
them. So much for address conservation and reduction of wasted, unused space.
There's a lot more available real estate than available v4 address
space. That's the biggest one.
Is there? How much address space would be available if it were utilized
efficiently by everyone? Once you put a price tag on it, organizations
become the models of efficiency in utilization. If you can use NAT, you
will, if you can't, you'll only assign unique addresses to
users/applications that absolutly require it. If you don't like those
options, maybe you'll move to v6. I just don't see a problem here.
Second, groups that make the Internet go aren't necessarily the ones who
can afford the address space.
If they can't afford the address space, there's a very good chance it's
because they are inefficient and they should be done away with. Besides,
I just don't see address space being prohibitively expensive on an open
market, as a previous poster pointed out, only a minority of address space
is currently announced in the Internet. I'd bet at least half of that
isn't used at all, and I'd bet another half of that is unnecessary.
Third, there's no root owner of the address space, so who is going
to sell it?
I'll give you that, the initial sale is difficult. Perhaps the
registries/IANA could use the proceeds to setup an annuity to fund address
title registries/other IETF functions for the future.
But really, you just need the first one: small space. The relatively
small pool means that a large company with lots of money could buy the
whole ARIN chunk of the Internet.
There are a large number of organizations with more addres space than they
need. The chances of any one company buying all the address space and
then just sitting on it are nil. Why would their investors allow them to
sit on a asset and not make any income on it? They would have to compete
with the likes of GE, ATT, MIT, and Genuity for sales of the address space
they already hold.
> Was this tongue in cheek?
Nope.
> I am not an economist, but this is a sure fire way to destroy the internet
> as we know it today.Why is this so hard to believe? Real estate is mostly a free market, and
that seems to perform pretty well for the most part. How is address space
that different? Please explain how you believe this would destroy the
internet.
Heh!
Steve
> There's a lot more available real estate than available v4 address
> space. That's the biggest one.Is there? How much address space would be available if it were utilized
efficiently by everyone? Once you put a price tag on it, organizations
become the models of efficiency in utilization. If you can use NAT, you
will, if you can't, you'll only assign unique addresses to
users/applications that absolutly require it. If you don't like those
options, maybe you'll move to v6. I just don't see a problem here.
Ah, but when you buy address space, are you going to be happy about
giving it back when you change providers? Will the provider give you
a fair price on selling it back? Do you think sales departments won't
insist you buy their address space when you connect? And if you can
keep it, it makes the routing table more and more interesting.
> Second, groups that make the Internet go aren't necessarily the ones who
> can afford the address space.If they can't afford the address space, there's a very good chance it's
because they are inefficient and they should be done away with. Besides,
I just don't see address space being prohibitively expensive on an open
market, as a previous poster pointed out, only a minority of address space
is currently announced in the Internet. I'd bet at least half of that
isn't used at all, and I'd bet another half of that is unnecessary.
That's a pretty brash thing to say on Merit's mail server. Much of
the research and coordination of the Internet is done by research
groups and universities. It isn't inefficiency that keeps their
available cash for address space down; it is the nature of their
business.
> But really, you just need the first one: small space. The relatively
> small pool means that a large company with lots of money could buy the
> whole ARIN chunk of the Internet.There are a large number of organizations with more addres space than they
need. The chances of any one company buying all the address space and
then just sitting on it are nil. Why would their investors allow them to
sit on a asset and not make any income on it? They would have to compete
with the likes of GE, ATT, MIT, and Genuity for sales of the address space
they already hold.
Who said they would just sit on it? Why not make the Internet
now "The Internet, Brought To You By Company X?" You need Company
X's address space, so you have to provide Company X's services on
Company X's operating system?
It's this kind of thing that keeps areas of the radio spectrum
licensed by the government.
-Dave
If they can't afford the address space, there's a very good chance it's
because they are inefficient and they should be done away with.
capital != efficiency.
Besides, I just don't see address space being prohibitively expensive on
an open market, as a previous poster pointed out, only a minority of
address space is currently announced in the Internet. I'd bet at least
half of that isn't used at all, and I'd bet another half of that is
unnecessary.
it's not the amount of address space in use, the amount of space being
announced, or even the amount of space that is "necessary" which matters.
it's the ratio of the number of prefixes currently being announced to the
number of prefixes, which were they to be announced, would cause lots of
"disconnectivity".
kind've hard to charge against the ratio, since no one body is in charge
of anyone-in-particular's BGP tables.
you can have all the address space you want if nobody is listening to your
prefix announcements.
There are a large number of organizations with more addres space than they
need. The chances of any one company buying all the address space and
then just sitting on it are nil. Why would their investors allow them to
sit on a asset and not make any income on it? They would have to compete
with the likes of GE, ATT, MIT, and Genuity for sales of the address space
they already hold.
you don't have to buy all of it to cause problems. just enough to be able
to keep the price arbitrarily high. and in whose interest is it best to
price small competitors out of the market? exactly those companies and/or
individuals who would be able to do this.
s.
Thus spake "Dave Israel" <davei@algx.net>
Ah, but when you buy address space, are you going to be happy about
giving it back when you change providers? Will the provider give you
a fair price on selling it back? Do you think sales departments won't
insist you buy their address space when you connect? And if you can
keep it, it makes the routing table more and more interesting.
That's already standard practice today, except you're forced to lease PA
addresses instead of buying them and selling them back. However, this
discussion is really limited to PI space, which has completely different
economics.
That's a pretty brash thing to say on Merit's mail server. Much of
the research and coordination of the Internet is done by research
groups and universities. It isn't inefficiency that keeps their
available cash for address space down; it is the nature of their
business.
While I'm sure we all value Merit's contributions to the Internet, both past
and present, you cannot seriously believe they are even remotely efficient
in their use of 35/8 -- nor can one expect them to be without financial
incentive. What boggles my mind is they managed to get allocated another
668,424 addresses in CIDR land -- do a whois and see for yourself.
It's this kind of thing that keeps areas of the radio spectrum
licensed by the government.
And the govt licenses that spectrum to the highest bidder, plus allows
owners to sell spectrum to each other as assets. Pretending the RIRs are
even remotely similar is disingenuous.
S
Stephen Sprunk "God does not play dice." --Albert Einstein
CCIE #3723 "God is an inveterate gambler, and He throws the
K5SSS dice at every possible opportunity." --Stephen Hawking
Thus spake "Dave Israel" <davei@algx.net>
> That's a pretty brash thing to say on Merit's mail server. Much of
> the research and coordination of the Internet is done by research
> groups and universities. It isn't inefficiency that keeps their
> available cash for address space down; it is the nature of their
> business.While I'm sure we all value Merit's contributions to the Internet, both past
and present, you cannot seriously believe they are even remotely efficient
in their use of 35/8 -- nor can one expect them to be without financial
incentive. What boggles my mind is they managed to get allocated another
668,424 addresses in CIDR land -- do a whois and see for yourself.
I didn't say their usage was efficient; what I was expressing was the fear
that they would have to go away entirely, in the face of companies with
tons of cash buying into what was suddenly an unregulated pool of precious
limited commodity that requires no maintenance or upkeep.
> It's this kind of thing that keeps areas of the radio spectrum
> licensed by the government.And the govt licenses that spectrum to the highest bidder, plus allows
owners to sell spectrum to each other as assets. Pretending the RIRs are
even remotely similar is disingenuous.
(*sigh*) I guess I should have been more specific. When I said "areas
of the radio spectrum," I was referring to the areas that were set
apart before anybody learned how to get arbitrary amounts of bandwidth
in a very small range. Specifically, the frequencies on your FM dial
are meted out by the FCC, and while they, like ARIN, find a way to
make gobs of dough off of it, they also make sure the usage is
reasonably fair among the users and that the public airwaves are not
excessively abused. While we can surely find flaws in both the FCC
and in ARIN, that does not mean the work they do does not benefit the
users of their domain immensely.
Sorry, but, the government does _NOT_ auction spectrum space off to the
highest bidder in all cases. If they did, amateur radio would be dead,
as would FRS, CB, and probably a lot of the part 15 allocations.
Owen