[arin-ppml] 2009-1 comment

[michael.dillon at bt.com]

> Consider an organization with a large assignment that they've 
> been using for a number of years.  Over that time, it has 
> been subnetted and allocated across their network in a manner 
> that made sense at the time, but today would be considered 
> inefficient.  The organization could renumber their internal 
> infrastructure to consolidate their addresses into a smaller 
> portion of their original assignment, and free up the rest.  

I'd be surprised if an organisation with such a large 
assignment could achieve this renumbering in much less
than 2 years. The transfer market won't be in existence
long enough for them to actually sell these addresses 
unless they take a risk and start renumbering today.

But when you balance the risks/rewards of renumbering
to free up and sell an address block, versus deploying
IPv6, it seems to me that the IPv6 effort wins out.

In an era where every ISP has taken on the extra work
involved in preparing for the IPv6 internet, and where
budget is tight due to the overall economy, I don't see
how anyone could justify a business case to sell IP addresses.
Fact is that the IPv6 Internet will be upon us in about
two years, and we expect the economy to be improving in
about the same timescale. If a company is ready with
IPv6 services in two years time, they could ride the wave,
but if they have wasted that time in freeing up IP
addresses, then they can only make one sale, and that's
the end of it. Bad decision.

-- Michael Dillon