Re using pseudo-unidirectional links...

["John Border" <border@hns.com>]

Re http://www.inria.fr/rodeo/udlr/documents/draft-ietf-udlr-lltunnel-00.txt

    Let me apologize in advance for not bringing this (and a few of the
following comments) up at the meeting in LA.  I had laryngitis so I could only
listen at the meeting...

    An observation...  There exist, at least in the satellite world, links
which I would call pseudo-unidirectional.  These links are highly asymmetric
where the return path is so small that running a routing protocol across them
is impossible but they are capable of carrying small amounts of return
traffic.  I don't propose that this working group explicitly address these.
But, it might be worthwhile to describe them in Section 4 of the draft and
indicate that the draft's mechanism can effectively enable routing on these
links by treating them as if they are completely unidirectional re forwarding
routing protocol messages.  What do you think?

John Border/Hughes Network Systems