RE: Identification

["Bound, Jim" <Jim.Bound@hp.com>]

Dave,

Assuming nothing.  Asking.  Setting a model. If we all agree on base
principles/assumptions we can move forward. 

But I agree with your interpretations except for one I think?

>> But we cannot hard code identifiers in silicon? 


>There is no _technical_ reason this couldn't be done. People concerned
about
>privacy might get uptight about this though.  

We don't want people to put identifiers in my CD player EIA interfaces
if they exist within the context of location.  The stateless property of
the IPv6 architecture permits that can be changed by processing on the
local network.
Also we want my CD player to be able to support a self healing link
partition if I change the IPv6 scope of identifier, or merge two links
to one link local domain as two examples.

Ergo XXX Electronics cannot burn IPv6 addresses in their chips and
assume permanent identification.  Even identifiers are chaotic and IPv6
architecture implies this is not technically wise.

In addition to privacy nut-cakes who are worried about my CD player and
refridgerator being attacked by the boogie man :---)

/jim