[ppml] Policy Proposal 2005-1: Provider-independent IPv6 Assignments for End Sites - Last Call

[Edward Lewis]

>ipv6 swamp can be much smaller, and we should constrain it with all kinds of
>rules about multiple locations, entity-size, requirements for multihoming,
>total allocations per RIR per year.

This is why I swung my vote from 2005-x to the other during the 
tag-team presentation in Montreal.  (Ordinarily I wouldn't attach 
personal names to this, but I've forgotten which is which, so I'll 
say it this way...)

I swung from Owen's to Andrew's proposal during the discussion mostly 
because Andrew's requires the use of IPv4 multihoming and Owen's only 
requires qualifying for it.  (I hope that I've attributed this 
correctly, if not, just reverse the names.)

Ordinarily I think that requiring the use of IPv4 as a gating 
function to use IPv6 is a bad idea.  However, in this case it raises 
the bar to getting the IPv6 space that poses a threat to global 
routing stability.

It seems to me that someone who *is* multihoming in IPv4 is 
demonstrating they have a real need.  Someone who isn't, but 
*qualifies*, doesn't really have the need and therefore shouldn't be 
invited to get the PI space.

The two proposals are similar, there's not a lot of other 
significantly distinguishing elements.  So, something as trivial as 
"having to do versus having to qualify" swung my emotion in the 
meeting.
-- 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis                                                +1-571-434-5468
NeuStar

Nothin' more exciting than going to the printer to watch the toner drain...