Re: [Roll] RaF vs RAN as a TLA

[Michael Richardson <[e-mail]>]

Carsten Bormann <[e-mail]> wrote:
    > “Host” is a great word for a thing that is not a router.

Point taken, and I want to go further with this.

    > If you call RPL-unaware things “hosts”, you reclaim the term “Leaf” for
    > routers that don’t forward (“leaf router” in long form).

The key point in useofrplinfo was that the *Host* was something that could
not understand the IPIP headers that we need in some places.

From reading roll-unaware-leaves, there is some ambiguity between Hosts
that choose not to participate in the routing mesh, and those that are
incapable of understanding the RPL artifacts.  I feel that a Host is
in the later category.

I am specifically thinking about a subsystem that might use ethernet
or PPP between Hosts, but have an edge device that speaks LLN.  The internal
nodes would be hosts, but the edge device would be RUL, but still a router.
The lack of connectivity might be more about range rather than power
(such as the window smash detector)

    > Look, ma, no abbrevs.

I'm specifically offering to change useofrplinfo's acronyms to match
roll-unaware-leaves.   It's at the IESG, but that change in terminology would
likely not have a signfiicant effect.

I resisted calling ~RAF == Hosts because I was concerned I'd confuse things worse.
I was trying to get processing of IPIP headers (where both IPdst=self) into
the revised IPv6 Host requirements document, but it seemed like a really big
fight, and I was busy.  Had I succeeded, then calling them Hosts would have
been good.

--
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-