[MLB-WIRELESS] National IP allocation and routing sche me??
Simon J Mudd
sjmudd at pobox.com
Tue May 21 17:07:10 EST 2002
random at sectoid.org ("Random") writes:
> I thought I'd throw my 2c in here, I've been in a few similar
> discussions with others about ip's also for a short while I played with
> the idea of being freenetworks ip friendly (ie, using a small bunch of
> ip's that doesn't conflict worldwide). Modeling wireless networks on the
> Internet is a very good thing for many reasons, but one of them should
> not involve dividing up 10.* (and or 192.*, etc) to use worldwide. With
> all the current trouble with IPv4 especialy!
this is because of the "perceived or real" difficulties in getting
public addresses. If I could ask RIPE tomorrow for a range of
addresses for MadridWireless or RedLibre in Spain I would. I think
that various factors would make these addresses difficult to obtain.
The "solution" of using RFC1918 addresses is IMO the best compromise available.
> But anyway, there definatly should be consistency in Australia at least,
> personally as I have said before on syd-wire I don't see any advantage
> using more than just 10.*, I don't see any use in using 172*/192* for ap
> communications and such. If you want 10.x can be logically divided so
> that you know exactly what is a client, and what is an ap/node, but also
> more importantly as EB just said, route summarisation is critically
> important, especially when you start to envision the complex routing we
> _will_ be dealing with, a semi-mesh network as this is can quickly
> become a routing nightmare, unless initial thought is put into it.
Fine. Using 10.* for both client subnets and AP interlinks is fine.
It's just visually easier to see things if you use a different network
number. Technically you are right: there is no difference between the two.
> Nb, someone mentioned possibility of filtering client -> client comms, I
> hope you relise that half the purpose of these wireless networks is
> client->client comms! ;)
I do, yes.
> I've posted various detailed idea's on how to overcome these issues, and
> how ip's could be efficiently allocated, have a look in the syd-wire
> mailing list archive to see. (middle last month)
It's a real shame there is no global wireless mailing list to discuss
these issues. I'm subscribed to 3 different Spanish wireless lists and
countless other mailing lists or other subjects. I'm sure a lot of
people worldwide aren't aware of some of the things you guys are
discussing, and these are things which really affect them too to some
extent.
I don't know whether seattlewireless, o nycwireless have active
mailing lists but a global list would help solve many global problems
which need to be solved. The local problems become much simpler
thereafter.
Simon
--
Simon J Mudd, Tel: +34-91-408 4878, Mobile: +34-605-085 219
Madrid, Spain. email: sjmudd at pobox.com, Postfix RPM Packager
To unsubscribe: send mail to majordomo at wireless.org.au
with "unsubscribe melbwireless" in the body of the message
More information about the Melbwireless
mailing list