Lijo Thomas <[email protected]> wrote: > We agree your point that once the packet leaves the RPL network, then the > deadline information has to be carried in hop-by-hop header. In fact we > mentioned required text for calculating deadline time, while traversing > through different time zones.
Sorry to pick nits, but it's about whether the packet leaves an RPL network.
RPL runs across a multitude of layer-2s!
It's when it leaves an RFC8138 compressed LLN!
The difference in terminology leads to a difference in thinking.
Pascal> Also we always claimed that RFC8138 is an encoding and that we can
Pascal> always turn a packet to uncompressed and back. Deadline creates an
Pascal> exception to that rule, which changes RFC 8138 into a sub IP
protocol as
Pascal> opposed to a compression.
To emphasize Pascal's point, and for many people who were not at the IETF
back in 2003, "subIP" was an entire area like Routing and Internet, which
addressed/created things like MPLS, ATM adaption layers, etc. That's what
we are trying to avoid, because it has political and process issues.
Pascal> All in all, I think that an IPv6 header (e.g., a new option in a
Pascal> hop-by-hop header) should be provided, even if for now it appears
to be
Pascal> for completeness only.
I haven't read the deadline draft at this point, so I don't know if a new hop
by hop header makes sense, but it sounds right.
--
Michael Richardson <[email protected]>, Sandelman Software Works
-= IPv6 IoT consulting =-
signature.asc
Description: PGP signature
_______________________________________________ 6lo mailing list [email protected] https://www.ietf.org/mailman/listinfo/6lo
