Re: [Roll] [roll] #132: draft-ietf-roll-trickle-mcast-04 - Clarify scope value of 3 - subnet-local

[Michael Richardson <[e-mail]>]

Ralph Droms (rdroms) <[e-mail]> wrote:
    >> Also, I am in favor of Ralph's proposal on using PAN ID for MPL scope
    >> 3.  I don't see how any automatic configuration could take place if we
    >> can't identify a concrete identifier for the scope.  The other
    >> alternative (and to be honest the one I thought we were going to use)
    >> is DODAG ID.  This would allow your scenario where a subnet of
    >> different link technologies could support MPL domain 3.

    > Don - at present, MPL is, as far as I know, independent of RPL.  In
    > particular, MPL can be used in a mesh that does not use RPL.
    > Therefore, there might not be a DODAG ID to use as a MPL domain
    > identifier.

While I agree with you, that MPL is separate from RPL in the same way that
HTTP can run over V32bis or ITU-T Rec. X.214 (OSI TPx) if we want it to,
if one want to use MPL in another environment one will have to specify
the boundaries somehow.

Using PANID limits one to 802.15.4-ish networks: what is the PANID of
ethernet?  ROLL is specifically chartered to work across a multitude of
different layer-2 protocols.

--
Michael Richardson <[e-mail]>, Sandelman Software Works