NC> Also, how exactly does subnetting down work out?
When a host transmits a packet, it includes in that packet the IP address of 
the destination host. The IP address is really a network number and a host 
number, concatenated, and the subnet mask defines how many bits are the 
network and how many are the host. 255.255.255.0 defines 24 bits for the 
network and 8 bits for the host. Every host on a single wire is supposed to 
have the same network number, and a unique host number, e.g., 192.68.1.1, 
192.68.1.2, 192.68.1.3, etc. are all hosts on a single segment if the mask is 
255.255.255.0.
If the network portion of the transmitted IP address doesn't match the 
network address, any devices on the segment configured to route packets and 
connected to other segments will look to see if it knows how to find the 
destination network. If it does, it forwards the packet onto some other 
segment.
This should make it obvious that every host on a segment must be in agreement 
on the subnet mask used, must use the same network address and must use a 
unique host address.
That explains straight IP; the industry has been inventing dozens of 
technologies (DHCP, IP Masquerade, proxy servers, and many many others) to 
break every single rule mentioned here.
Jeff
--- GoldED/2 2.42.G0615
---------------