RE: draft-kurtis-multihoming-longprefix comments

["Tony Hain" <alh-ietf@tndh.net>]

Brian E Carpenter wrote:
> Tony Hain wrote:
> ...
> > People aren't asking for location-independent addresses, in 
> fact they 
> > are asking for the architectural equivalent of a street address.
> > 
> 
> I disagree. They are asking for the equivalent of a cell 
> phone number, on a network that doesn't have a call setup phase.

I think we are trying to get to the same place in terms of real-time
air-link portability, but current implementations of cell addresses
require changing whenever the service contractor changes. This is
clearly acceptable to the individual, even if annoying, but to the
larger entities it represents a provider lock because it becomes very
costly to change. To be specific, they want:
   an unambiguous address that will reach the device, 
   independent of carriage or contractor relationship 
   over time. 
In other words, they should be able to change or have multiple service
providers without requiring any coorelation between the address and the
provider of the transport service. I believe this is closer to a street
address than it is to a cell device.

This is the crux of the discussion here, because that goal is completely
divergent from optimizing a connectivity graph to minimize the size of
the routing table. While our task is to find soulutions that keep the
routing table at a reasonable scale, ignoring the basic requirement as
too difficult is not useful. 

Tony