El Wed, 05 May 1999, Jonathan NAYLOR escribió:
>With all this talk about RIP-2 on amateur radio, why not look at RIP-98
>instead ?
I have never tried rip98, even didn't remember of its existence ''''':-(,,,,,,,
>It is essentially a cut down version of RIP-2 that consumes much less bandwidth,
>but for all intents and purposes includes the same routing information. RIP-2
>includes a lot of nulls in its data and takes up more bytes than is needed to transfer
>each routing item.
Rip-2 is very simple, it cannot be so bad...
But: does rip98 support classeless subnetting (routes+mask) ? That's very
important. Rip-1 didn't.
Also, rip-2 has a very nice feature: You can define a "password" that must be
in the update to be accepted (and there is also a way to use MD5...).
That's a good way to use "seriously" rip-2 in a radionet, and that it is not
broken by people doing experiments.
>It would be quite easy to retro-fit RIP98 to an already existing RIP-2 program.
My experience with hamradio networks says that a very good way to do ip
networking in packet radio will be using 2 levels of routing protocols:
1) "Intranetwork". Doing that a subnetwork is "coherent", and has internal
connectivity (a set of point to point routes to the hosts in your subnet). That
is very well done by rspf, wich is really 100% adapted to packet radio. Sure,
that routing here must no be too complicate if the subnetwork is defined in a
little geographic zone.
2) Internetwork. Defining well the subnets and having it well
"intraconnected", you can do routing with routes to networks (and interconect
networls) ... rip-2 do that well (?). That means that you don't have to export
millions of routes, only the routes to the networks.
The problem is that implementatios of rip-2 are not thinked for packet-radio
(except the nos).
>Its been in the ax25-utils for years as rip98d.
Where could I fetch the documentation of rip98 ??
>Jonathan HB9/G4KLX
--
Saludos de Julián/EA4ACL
-.-
