RE: Broadcast IP Network Model Problem Statement

["Walmsley, Clive" <Clive.Walmsley@dsto.defence.gov.au>]

Michael,

thanks for the article. I have the following comments and observations
to make on your posting entitled 'Broadcast IP network model - Problem
Statement' to the udlr WG.

1) In previous posting to the group I rasied the question of :

>>> wrt the recent draft  RFC 'A link Layer Tunneling Mechanism for
Uni-directional 
>>> Link', has any body considered or given thought to the  IP
addressing 
>>> plan/structure that would be used or needs to be developed when
operating over a
>>> uni-directional service?

I suspect that the question you raised and mine are similar. I had half
hoped a reply indicating  the use of multicast IP addresses (is this the
class D network addresses?) could be used. 

As I see it the biggest question I see is whether the work should
provide a mechanism for a point to point IP network service over a
broadcast media or whether it should try to adopt a point to multipoint
solution. The thought of a multipoint to multipoint system makes me
knees quiver!


R/ Clive