[ppml] 2005-1 and/or Multi6

[Michael.Dillon at radianz.com]

> It
> takes more CPU to store and sort a table composed of variable length
> elements than it takes with a table of fixed-length elements, and it is
> more difficuly to design hardware to do so as well; not to mention more
> opportunities for bugs. Variable-length elements are a valid option for
> text fields, not for routing prefixes. Speed and robustness vs. storage
> efficiency, speed wins especially when it allows your favorite vendor to
> sell you memory at outrageous prices :-)

Yes, speed wins. And router vendors have introduced more and
more complex designs over time in order to get higher speed.
This includes optimizing software algorithms and shifting
functions into hardware such as FPGAs and TCAMs. 

> Michael, at the end of the day your 48-bit prefix takes the same amount
> of RAM than the 32-bit prefix: 

No, at the end of the day it does not. Today it does, but
we are making an IPv6 policy that should last for many years.
I don't expect router design to stand still during that time
and I also don't expect any router vendors to reveal what
they are working on in their research labs. At the end of
the day, it is possible to optimize by processing fewer bits
and we should not remove that option from the table by
making bad policy and introducing yet another class of
IPv6 address.

> Given that 64k subnets is
> enough for most and that allocations made using RFC3531 will extend the
> initial /48 to a /47 for the few large organizations that require it, I
> don't see any advantages in giving /32s away to sites.

If an organisation has an AS number then their network is not
a "site". 

--Michael Dillon