[ppml] back to the policy issue again

[Phil Howard]

On Thu, Jan 09, 2003 at 12:03:39PM -1000, Barbara Roseman wrote:

| The current policy was developed over many meetings and in coordination 
| with the IPv6 working groups in APNIC and RIPE. The policy as it stands has 
| not impeded adoption of v6 in those regions, and I doubt that it is the 
| main reason companies are being slow to move to IPv6 here in the ARIN region.

There are reasons IPv6 will generate interest in those regions.  One
of them is that the ARIN region already has a more substantial portion
of the IPv4 allocation and is thus facing less pressure to find solutions.
Another is that internet growth has slowed in the USA but is increasing
in other regions of the world right now, and thus they are seeing new
deployments with address shortages and newer router and server software
that is already ready to go for IPv6.


| There is always room for improvement in our current policies. Working 
| through the issues is a good thing, but it works best when there is a 
| specific goal in mind, such as easing adoption of v6 for small to mid-sized 
| companies, or adopting subnet boundaries that conform to a multi-homing 
| standard (not yet in existence). Just saying we need to generally change 
| the policy to encourage adoption of IPv6 doesn't get us very far along, nor 
| will it dramatically change the v6 situation in our region, IMHO.

I think there needs to be a distinction between goals and methods.  The
methods should then address business realities.


| John Brown has pointed at a distinct issue: the current policy makes it 
| difficult for small companies to request v6 addresses without "gaming" the 
| application process. He is understandably reluctant to do that.

Size shouldn't matter.  That's something that IPv6 was expected to offer.


| One specific policy change we might consider is whether the 200 customers 
| in 2 years is an unrealistic expectation for early adopters. This issue was 
| also raised at RIPE several meetings back, and it was decided that even for 
| academic institutions, this might be an acceptable number. Perhaps we need 
| to revisit this qualification in light of operational experience.

That expectation is biased.  Some of those who want to be the earliest
adoptors may not even have 200 IPv4 customers today.  Those who do have
the greatest customer base probably won't even be interested in putting
any money into it (and without money, it won't happen) until the demand
level is there.  And the demand level they need will have to be one which
will return profits from their investment in a reasonably short time frame,
such as 3 years.  The required numbers might be in the 1000's.  So those
who could qualify have their own business limitations, and those who do
not have the business limitations won't qualify.  Maybe you'll get lucky
and find a few people in between.


| This is the kind of specific issue we need to explore and discuss to remain 
| on topic.

If the question is what should the policy be, then I have to ask if the
policy is going to address a business world.

-- 
-----------------------------------------------------------------
| Phil Howard - KA9WGN |   Dallas   | http://linuxhomepage.com/ |
| phil-nospam at ipal.net | Texas, USA | http://ka9wgn.ham.org/    |
-----------------------------------------------------------------