             Using the Wisconsin Network - Part 24
                   by Andy Nemec, KB9ALN
 
     In the last installment of the series, we talked about
how our packet radio network has been evolving, and talked
about the need to look toward the future. Many of us have
tasted the advanced nature of the internet, and have wondered
if we could do some of what we see there in Amateur Packet
Radio. This time, we will look at the elements of a network
and why we might consider not only other methods of
networking, but of operating as well.

     But in order to do that, we have to break down the whole
networking process and look at what we do with it. We all know
that a network allows us connectivity with other stations. To
aid our analysis, consider what happens when you want to leave
a message to a friend of yours who lives some distance away
from you. If you are in a typical packet radio situation, you
may have 2 or 3 ways to do this:

1) You can connect to your local BBS and type in a message (if
   you both have access to a BBS). While other stations do the
   work you have to remember his BBS address, you have no time
   to correct spelling errors and word your message properly.
   
2) You can connect up to his mailbox and leave a message there.
   Of course, you would have to navigate through a series of
   nodes to get to his mailbox. And if his station is not
   active, you have wasted your time. If you have a shaky path
   to any node in the circuit, you may suffer sudden
   disconnection and your message is lost in the ether. You
   will have to try another time. Again, you have no editing
   capabilities.

3) You could also leave a message in your mailbox, or another
   friend's mailbox. Of course, your mail recipient must
   remember to check for the mail (just like the BBS example
   above).

     If you choose to leave a message in your friend's
mailbox, you also have to remember node aliases, the best
route to take, and might have to wait for the right time of
day to make a connection. To sum it up, the process looks like
this:

1) Connect to your local LAN node.
2) Gain access to the network.
3) Know the alias of his LAN node.
4) Deal with any nodes on the backbone making shaky
   connections.
5) Segment your connection circuit if there are any shaky
   paths.
6) Connect to his LAN node.
7) Connect to his mailbox and leave a message.

     This is a heck of a lot of thought and human intervention
required for a system that is run by computers! In contrast,
consider what happens if you E-Mail one of your friends on the
Internet:

1) You dial-in to an Internet service provider, and leave a
   message. 
2) Your friend dials into his internet service provider and
   picks up his message.

     It can be made even simpler. Some folks have their
computers set up to automatically connect up to the service
provider to send and collect E-Mail. Wouldn't it be nice to do
that on Amateur Packet Radio?

     Well, it is feasible, and many people do it right now
with packet radio E-Mail. The secret is to have an advanced
mail system and an open, transparent network at your disposal,
as the Internet is. What do we mean by an open, transparent
network?  Consider all of the stuff that goes on
behind-the-scenes when making any kind of connection on the
Internet, for any purpose.

     An excellent example of this is what happens when a
person is "Web Surfing". When a person connects up to a "Web
Site", he or she has no knowlege of how the data gets to or
from this site. All they have to know is "//www.bogus.com", an
internet address. This is a transparent connection, computers
that do the routing and relay of the data are not seen by the
user. In a global computer network, there is likely to be a
significant amount of network activity just to allow someone
to connect up to this web site.

     Contrast this with what you had to go through to get a
message to your friend. There is a lot to know, and a lot to
do. You have to know details about your friend's station and
the network that the average Internet user doesn't ever worry
about. So in order to make our mail system more "automatic",
we need the following :

1) A way to send mail with maximim convenience, preferably
   without ever "leaving our own station".
2) A mail system that knows who the recipient of the mail is.
3) Mail systems that know how to work with a network.
4) Network transparency - packets of any kind are routed to
   their destination without human intervention.
5) Some method of receiving the mail and possibly storing it.
6) Ability to collect the mail in a convenient fashion. 

     Amateur Packet Radio has tried to accomplish this in a
variety of ways. It has always been the vision of packeteers
to have "door-to-door" mail delivery. While we will explore
ways to get closer to this with our existing situation, we
will not see mail fully automated with what we are using now.
Although we would like to think of our current network node
system as "auto-routing", it really is not always very
efficient when we ask it to do this. In addition, some of our
network nodes will not pass data that is sent using protocols
other than Net/Rom and AX.25 (remember our discussion of how
some protocols are not well suited for some tasks, but
prefectly to others).

     When discussing a digital radio network, we can further
break this all down to a few basic functions we will discuss
in the next installment. These can be desribed as:

1) A originator of data.
2) A Network Interface Host Computer (comparable to a network
   node we now use, but more functional and automatic).
3) An transparent Auto-Routing network requiring no user
   knowlege or intervention.
4) Another Network Interface Host Computer at the destination.
5) A receiver of data.

     Actually, origination and reception of data can be
combined into the function of Network Interface Host
Computers, further eliminating two of these functions. Next
time we will dissect the functions of each of these parts of
our radio network.

  ***End of Part 24***
