[ppml] Policy to help the little guys

[David Williamson]

On Wed, Mar 19, 2008 at 05:08:52PM -0700, Tony Li wrote:
> PI is the portion of the DFZ costs that we can and should fix at an
> architectural level.

I agree with this.  I might generalize it to say that the costs
associated with non-transit leaf nodes should be fixed by a new
architecture.  That covers PI and end site PA, since they seem to be
fundamentally identical.

For those on this list who haven't yet taken a look at what the RRG
mailing list is doing, I strongly urge you to do so.  This is
extrememly important work, and the current best chance for solving some
of the scaling issues.  (Thanks, Tony!)

-David