[CST-2] Digi Comms

[Barnaby Gray]

On Thu, May 30, 2002 at 10:51:21PM +0100, Tom Puverle wrote:
> > To explain further, 128.232.2 in binary is
> >     10000000 11101000 00000010 00000000.
> > 
> > and the netmask specified by the /18 is
> >     11111111 11111111 11000000 00000000.
> > 
> 
> I would also like to add that the final bits in CIDR 
> don't necessarily specify a host. In fact it can be
> a cluster of networks on the same geographical site,
> but from a core router they appear to be the same thing.
> (And hence CIDR allows smaller routing tables). However,
> for this to work then networks have to assigned IP numbers
> from a continuous range, which nowadays with IP fragmentation
> may be quite difficult. That's why CIDR is "too little, too late"

Hmm.. to be pedantic the final bits do actually always specify a host
(or a network/broadcast address if they are all 0/1's), otherwise the
IP address wouldn't be unique to that particular host. What you mean
is that you can still have further subnetting and routing based on
this inside that host part, so you can have sub-network blocks  and
sub-host parts (I guess that's what you'd call them), but you do
still essentially need all the way down to the last bit to directly
identify that host.

Barnaby