Re: Should we share?

[glg@balrog.k8lt.ampr.org (Gary L. Grebus)]

Since we've spent the last two weeks untangling static routing tables
here in New England, I'm pretty much in agreement with what Brian
says.  Assigning separate subnets to each physical LAN looks
attractive initially, but eventually it becomes hard to manage.
(Routers on the MA backbone are currently carrying routes for about 
50 subnets).

The idea of doing some sort of routing at the AX.25 level isn't so
different from what's done in the wire network world.  The networks
I'm familiar with don't use an IP subnet for every physical segment.
They include routing below the IP level, for example Ethernet
bridges, to consolidate physical segments.  Subnets are assigned based
on a combination of administrative convenience, physical connectivity,
and special requirements (e.g. security).  I don't think routing at
the AX.25 level will scale, but that's not its purpose.  Granted,
there isn't any existing implementation.

RSPF really attempts to address all of the issues that folks have
mentioned: dynamic routing, subnets split over multiple physical
segments, mobile users, acquisition of marginal routes, etc.

Some people have been successful in making the current RSPF 1.0
implementation work, despite known bugs and limitations.  It would be
interesting to know the network topologies.

Is anyone interested in collaborating to try to implement the RSPF 2.2 spec?

	/gary
	K8LT

P.S. I think Brandon's idea of using DHCP (dynamic assignment of IP
addresses) is a great one, not to solve routing problems, but to make
it easier for new users to get started.  I wouldn't mind being able to
automate some of the IP address assignment job.  Of course this basic
idea has been around since at least 1987, and nobody has run with it
yet.