[MLB-WIRELESS] What has happened to all of technical discussion

[Roger Venning]

This is exactly what the Routing / Addressing working group is planning. 
One of the issues is that the OSPF routers must run in 
point-to-multipoint mode because if A can ping B and B can ping C does 
not A can directly ping C... The issue with this is that I haven't fully 
resolved the bugs in the free package, Zebra, in order to get it working 
in point-to-multipoint (which was an unsupported mode before I started.).

The end goal is that backbone nodes will have a single IP address per 
WAN interface. This will run OSPF. It will be configured in ad-hoc mode, 
on the same frequency, with the same ESSID as all other backbone WAN 
interfaces (exceptions only in regions that have high node density where 
interference and congestion becomes an issue, only channel allocation 
gets past these). The OSPF daemon will run in point to multipoint mode. 
The end result is that a node will not need configuration especially for 
individual neighbours. Mesh routing will occur.

Roger.

Steven Haigh wrote:

>That's what's confised me too... what is there to stop us using ADHOC on
>*all* nodes, and then using routing (OPSF or whatever) to figure out who is
>connectable?
>
>I think this is probably the closest we could get to a mesh.....
><snip>
>

-- 
-------------------------------------------------------------
Roger Venning	\ Do not go gentle into that good night
Melbourne        \ Rage, rage against the dying of the light.
Australia <r.venning at telstra.com>                 Dylan Thomas