[ppml] ripe-55/presentations/bush-ipv6-transition.pdf

michael.dillon at bt.com wrote:
>> Things to encourage:
>> - transparency of operation of the market
>> - blind, double-blind, or N-tuple blind market
>> - separation of the two sides of the market, i.e. those 
>> selling blocks and those buying, by RIR-type entity (who is 
>> also "blind")
>>     
>
> How does this blindness fit with "transparency of operation". 
Any number of ways. But that's an implementation question...
> In any case, it will always be known who was the registered owner of an address
> block, and who is the new registered owner so unless you keep the
> identity of the address secret, there will be no blindness.
Correct.
> And if the
> addresses are secret, then it becomes risky to buy them because not all
> addresses are equal.
>
>   
Which would be why I suggest (in the bit you kept quoted!!!) separating 
the two sides of the market.

Basically, something like an RIR (or actual RIRs, or IANA) would still 
be allocating blocks, just
like today, but with the addition of bids or donations, to fund the 
re-acquisition of blocks *by* the
allocating entity.

The allocation side is blind today - you don't get to choose what block 
you are given. So, that doesn't
change under this split market.

And, I envision the market augmenting existing RIR function, not 
replacing/displacing it.

The split in the market would exist so that addresses would become 
available for both "bidding"
and "requesting (unpaid except for RIR fees)" requesters. It would still 
be possible for someone to
request a block from an RIR and get one which was purchased by the RIR, 
without the recipient
having to directly purchase it.

The "market" would exist primarily on the seller side, between the 
allocation entity, and the holders
of space that is unused or underused. Having *that* operate 
transparently to the community at large,
but blind to the buyer/sellers, ensures that no hanky-panky occurs. SOX 
would probably help there
too, in making sure at worst, there is the ability to audit activity. It 
may even be necessary to have the
buyer's staff (who are directly involved) be bonded.
>> It's a highly non-trivial kind of thing to delve into, but 
>> IMHO something which can work very well, *and* actually has 
>> some chance to improve the aggregation situation in IPv4.
>>     
>
> It sounds like the time is long past when we could have set up such a
> market because it is a non-trivial exercise and we have only two or
> three years of supply left.
>   
The benefit of the market, is in bringing additional supply to the IPv4 
pool. That will continue
to have benefit after the IANA "original" allocations to RIRs has finished.
> In addition, the cost of setting up such a complex and highly regulated
> market has to be weighed against the cost of IPv6 transition, given the
> abundant supply of free IPv6 addresses.
>   
If IPv4 and IPv6 were natively interoperable, that would be true.
However, they are not.

The value of an IPv4 market is *independent* of the availability and 
value of IPv6 space.

Brian