[ppml] IPv6 initial allocation policy

[Tony Li]

> BGP proxy aggregation
> This discussion from NANOG 11 years ago:
> http://www.academ.com/nanog/september1995/proxy.html
> leads me to suspect that proxy aggregation is a tried
> and tested way of dealing with the future aggregation of
> early assignments, as long as ARIN allocates the address
> blocks so that they are aggregatable. 


As one who was there: proxy aggregation was tried, tested, 
and rejected by the operational community because of the
traffic engineeering implications.  Sound familiar?

To revive proxy aggregation, it would be necessary to put 
into place an operational group of service providers who 
could consider different abstraction naming boundaries and 
appropriate abstraction action boundaries, make a long term 
commitment to implementing them and only then make 
recommendations to a RIR.  Only by careful planning could 
we avoid making the same mistakes again.

Tony

P.s. I'm actually a supporter of doing this, but as always,
I think that you'll find that I'm in the minority.