Re: Security requirements for identification

[Dave Crocker <dhc@dcrocker.net>]

Christian,

CH>  The proper statement is
CH> that multi-addressing, combined with a procedure to dynamically change
CH> the addresses used in an association, can provide additional benefits,
CH> such as using redundancy for surviving network re-homing events.


Absent a long-standing convention on the definition of "multiaddressing" I am
using the basis in

  <http://brandenburg.com//specifications/draft-crocker-mast-analysis-00.txt>(*)

In fact, one of the major difficulties in discussing this topic is the absence
of consistent and precise terminology, in any of the fora I have been
tracking.

All of the multiaddressing proposals encompass both the use of multiple
addresses and a means of adding to the set of addresses.

Separating the two constructs does seem interesting, from a formal standpoint,
but I'm not sure how it is relevant from the standpoint of engineering any
solutions or the standpoint of any current proposals.

d/

(*)  I submitted the draft to the internet-drafts folks twice over the last
2 weeks but it not seen it announced, yet, though the I-D folks say it is
coming out today.  So the link to my own site is simply as a backup
--
 Dave Crocker <dcrocker-at-brandenburg-dot-com>
 Brandenburg InternetWorking <www.brandenburg.com>
 Sunnyvale, CA  USA <tel:+1.408.246.8253>