Owen DeLong owen at delong.com
Tue Apr 12 19:23:44 EDT 2011

On Apr 12, 2011, at 2:35 PM, Milton L Mueller wrote:

>> -----Original Message-----
>> On Tue, Apr 12, 2011 at 04:00:14PM -0400, Owen DeLong wrote:
>>> In a market where higher prices lead to greater production,
>>> speculation brings liquidity. In a market where increased production
>>> is not possible, speculation leads to higher prices and increases
>>> the probability of market manipulation.
>> Who knows, in our context "production" might be interpreted as
>> increasing the rate of adoption of IPv6. Certainly if IPv4 addresses
>> were to become prohibitively expensive, IPv6 will look a lot more
>> attractive. Food for thought.
> Exactly. Address trading of v4 blocks would start to reflect the true scarcity value of the addresses, as well as operators' judgments about the time horizon of their value. And that could help surmount the migration hump by putting a lot of adjustments into motion.  
> There is a weird schizophrenia in RIR list discussions of this problem which reveals that many people haven't thought this through very carefully. On the one hand, you get comments like this: 
> "Who cares, IPV4 is dead --" 
> Interspersed with comments suggesting that:
> "Hoarding of IPv4 addresses by speculators will drive up prices to levels that will [pick one] i) destroy the internet ii) force us all into the arms of a merged AT&VZ; iii) be really, really bad
I suspect that more represents the diversity of the community and the forward-looking nature of the comments
than representing any particular level of failure to think the matter through properly.

> Is it just me, or is it not possible for both of those sentiments to be valid? 
Oh so tempting...

It would not be possible for one individual to hold both of those sentiments, but, since both are
predictions of the future depending on disparate courses of action advocated by different
parties in entirely different contexts, I think it is fair to say that they are quite likely equally
valid in their respective contexts.

> If IPv4 is really dead, it doesn't matter at all how we handle the trading or need for v4 blocks. Indeed, you should support hoarding and speculation because the hoarders and speculators will be defeated and lose their shirts. 
Obviously, IPv4 is not yet dead. IPv4's future is obviously rather limited and there are many of us who believe
that down the path of attempting to extend its useful life lies a great deal of madness.

> If the "IPv4 is dead" meme is wrong (and I think it is _dead_ wrong) then it matters a lot how we handle trading policy. We have to define policies that adjust to the growing scarcity in the most efficient and socially beneficial way. Reiterating 1996-era best practices as if they were religious principles doesn't help. 
I think you take it far too literally and well out of context. I think that for some definitions of the term, IPv4 is dead. Depending on growing your network in the IPv4 realm is a definite recipe for failure at this point.

However, on the flip side, there is a reality that servers will still need some additional IPv4 addresses for some time  in order to preserve existing functionality to existing IPv4 users.

I think the transfer market is going to progress rapidly to irrelevance in some combination of the following ways:

1.	There will be no sellers at any price (after an initial burst) and the market will
	push everyone to IPv6 to sustain their businesses as as result..

2.	There will be an abundance of sellers, prices will remain relatively low, trading
	will be massive and the IPv4 routing table will collapse instead. This will push
	everyone suddenly to IPv6 in order to restore basic routing functionality.

Either way, I consider the market to be a necessary short term exercise and little more.


More information about the ARIN-PPML mailing list