[address-policy-wg] 2007-08 New Policy Proposal (Enabling Methods for Reallocation of IPv4 Resources)

On 29 okt 2007, at 15:47, Nigel Titley wrote:

>>> The best solution is NO market and reclaiming address space that  
>>> is unused.

>> How much of the currently allocated IPv4 address space do you  
>> believe is unused?

About half the ~ 40 legacy /8 assignments don't show up in the routing  
table.

>> How much do you think a reclaim programme would cost to run?

Don't know; don't care too much. Let the people who want the addresses  
pay for it.

> And we've tried appealing to people's better natures to return  
> unused space. The result confirmed what most of us thought already:  
> people mostly don't have better natures.

So appeal to something we know they do have.  :-)

> So the next phase is to encourage them to return addresses into the  
> pool (and remember, it doesn't matter whether they return via the  
> RIRs or not, as long as they become usable) by allowing them to  
> sell, buy or barter.

Just you writing this gives them a reason to not return the space.

The negative consequences of trading address space outweigh the  
positive ones.

We're going to run out, the only question is if it's going to be a  
year or two sooner or later, and wheter it's going to be "we're all  
out" or "you want buy some IPs, I give you real good price, /8 for  
only $10/IP". Trading something that's in demand but has no supply  
will lead to hoarding, reducing availability or at the very least  
making it unpredictable.

> IPv4 space has a limited life anyway. Once we hit 51% of traffic  
> being IPv6 there will be a rapid flip-flop and IPv4 will be dead. A  
> market in V4 addresses will at least allow network designers to put  
> a real cost on not switching to IPv6 and may actually result in  
> business cases being built. This will speed the adoption of IPv6.  
> This is one of the aims of our proposal

I'm against address trading, but there are bad ways to do it and much  
worse. I suggest that we first come to a world-wide consensus on  
whether we want to do it. If so, then we can talk about the mechanism.