[arin-ppml] Effect of ARIN's Letters

[michael.dillon at bt.com]

>> they said "there appears to be no business case for IPV6".  

> That's because *today* there isn't. In fact, there will *not* 
> be a business case until lack of IPv4 addresses make it hurt, 
> and believe me it *will* hurt.

This just isn't true, at least not in Europe.

For starters, in the UK, there is an ISP who provides broadband services
which support IPv6 <http://www.aaisp.net.uk/broadband.html>. This
company clearly sees a business case for IPv6.

My own company has customers in several countries in Europe who are
buying IPv6 Internet access from us. It's only available in a few
limited cities and for special customers right now, but there is a
business case or we wouldn't be doing it. In addition, some large
companies now include IPv6 in their RFPs, not as a checklist
requirement, but because they only want to sign contracts for service
with an ISP who is serious about IPv6. For us, the business case is that
we can get these contracts for IPv4 services today, because we have a
serious plan to deploy IPv6 before the v4 addressing crisis becomes
real.

I'd like to emphasise this point. The business case for IPv6 is to have
a smooth and steady transition through the IPv4 addressing crisis in
order to sign contracts for IPv4 services with companies who want to
INSURE themselves against problems during the crisis. There is an awful
lot of money at stake here.

Some people want business models handed to them on a platter so that
they can just turn the crank and make money. Today there are no
simplistic IPv6 business models like that, but there are definitely
revenue-positive business cases to be made for people who can think out
of the box.

Note that the IPv4 addressing crisis will also not be a simple thing. On
the one hand we are running out of v4 and ISPs may not be able to get
more addresses, while on the other hand RIR actions may bring the crisis
to us sooner than pure numerical analysis would suggest. At the same
time, there are transfer policies in place that will attract
speculators, regardless of the rules. This will muddy the waters,
contributing to the crisis. There are a whole raft of minor and major
technical issues with v6 support in certain pieces of hardware and
certain software packages. If you lived through the mid 90's with IPv4,
you will have an idea of what I mean. Some deployments of v6 will be
stable, but others will not due to these technical issues. This also
contributes to the crisis.

You cannot rely on other people to solve this crisis for you, or make it
go away. The only reasonable strategy is to prepare contingency plans
today, which includes doing a lot of technical work testing and trialing
IPv6. But it also includes planning what products to discontinue in
order to recover IPv4 addresses so that the network can continue to grow
and sell the more profitable IPv4 products.

--Michael Dillon

P.S. Some people on this list are confused and think that I speak for
some company or other. To be perfectly clear, the above opinions are
mine and not the opinions of some company or other.