[ppml] Policy Proposal: Authentication of Legacy Resources

[David Conrad]

John,

On Jul 10, 2007, at 10:27 AM, John Paul Morrison wrote:
> I'm talking about the public Internet - if your printer is still  
> happily running IPv4 within a private network that doesn't count.

OK.

> Assuming IPv6 takes off on the public Internet, it will displace  
> IPv4 just because of the administrative overhead. Who's going to  
> want to maintain two routing protocols in a large network for very  
> long?

The folks who want to connect to the Internet.  IPv6 is not going to  
be universal for a very, very long time as there is so little  
(business) justification for the vast majority of current content  
providers to expend the resources (money, manpower, time) to do the  
conversion.

> So assuming IPv6 replaces IPv4, it's pointless to chase down the  
> legacy users since it will all become legacy.

I personally believe that in its current form, it is highly unlikely  
IPv6 will replace IPv4 in my lifetime.  When the IPv4 free pool is  
exhausted, people are not going to turn off their IPv4-only devices  
and ISPs will do what is necessary to continue generating revenues.   
I imagine there will be business opportunities in mechanisms that  
allow IPv6-only sites/devices to communicate with the (predominantly)  
IPv4 Internet which will tend to prolong the lifetime of IPv4.

However, regardless of the clarity of my crystal ball, it would seem  
prudent to me to have mechanisms in place to clearly and  
unambiguously identify the users of the address space for the (likely  
inevitable) market that will be emerging in IPv4 addresses.

Rgds,
-drc