Earlier, Patrick Frejborg wrote:
> Ah, then ILNPv4 and hIPv4 are not that far from each other...

Correct, and that is not a coincidence.  Both HIP and ILNP
have their roots in the IRTF NSRG.  There are limited ways
to use off-the-shelf mechanisms to carry identifiers in IPv4.
The ILNP designers have gone to some lengths to document
that HIP-like mechanisms could fit inside the architecture
(NB: This required no protocol changes in ILNP, just clarity 
of documentation.)

Multiple ILNP papers have made it clear that one can also
use a cryptographically-generated ID with ILNP.  That might
be a CGA-like thing.  It also could be a HIT-like thing.

(As a user, I do NOT want to use cryptography on all or
even most of my packets, so I find that HIP's focus on 
cryptography isn't well aligned with my personal operational
requirements.  Other folks mileage may vary.  Similarly,
I find the CGA arguments non-persuasive and would not want
to use a CGA for my own packets.  ILNP leaves the ID 
generation flexible, so users can make choices sensible 
for local environments.)

Later, he wrote:
> ...give them options go either to IPv6 (ILNP) or
> to something that is backwards compatible with
> the current deployed stack.

ILNP already *IS* backwards-compatible with the current 
deployed stack.  So the quoted phrase, especially the word "or",
is very confusing.

RFC-1687 says many things (e.g. convergence with ISO CLNP
would be nice).   It wasn't obvious which part of that RFC
Patrick was referring to.

Yours,

Ran

_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg

Reply via email to