[ppml] Policy Proposal 2007-15: Authentication ofLegacyResources

[Paul Vixie]

> The question is, do we just let people drink as much pear juice as they
> want, or do we ration it so that the supply lasts until the apple orchards
> are in full production of apple juice?

so, to get this thread back on track, is there policy work here?  conrad's
soft landing proposal was a form of rationing.  are there others?  can they
get written up and put into the PDP so that we can all discuss them in ABQ?

(note that in my role as an arin trustee, i'm asked to ponder the fiduciary
impact of a policies that come up through the PDP, but i am never asked
whether i actually like a policy or not, and that's as it should be.  so, i am
not speaking as a trustee on this thread, because as a trustee, there would be
nothing to say, at this stage.)

> > noone should be deploying non-dualstack in this day+age.
> 
> Why not? IPv4 MPLS networks are perfectly capable of providing native IPv6
> services to customers. Any existing service provider has an internal supply
> of IPv4 addresses needed to grow their MPLS network by simply redeploying
> IPv4 blocks assigned to customers who are able to transition to IPv6.

that's a lifeboat scenario, policy-wise.  let's not try to reason from such
corner cases, they aren't illustrative.

> Not to mention RFC 1918, using non-registered addresses, and trying the
> class E range.

none of those would come from ARIN, as things stand today, and so there's no
policy work to be done on them.

in general terms, no new IPv4-only networks should be built using public
resources.  finishing builds now in process is understood.  corner cases like
the IPv4 MPLS example given above would be understood.  but a new build of
core or other infrastructure, begun after August 1, 2007, ought to include
IPv6.  as a trustee i've learned that it's better that i don't author policy
proposals, but i hope someone within the sound of my voice will take this one
as their own and push it.