318 lines
16 KiB
Plaintext
318 lines
16 KiB
Plaintext
FidoNet History and Operation 8 Feb 85
|
||
|
||
|
||
This is a long and convoluted document; it has been sorely needed for months
|
||
now, and it finally got done. FidoNet is growing at a tremendous rate, and
|
||
newer sysops don't have the information that us oldies (pre Sept 84 sysops)
|
||
assume everyone knows; hence the history section here. There is a lot of
|
||
extremely important material covered here that was assumed to be known by all;
|
||
we are finding out otherwise.
|
||
|
||
This also covers some of the dark mysterious secrets about the magical node
|
||
numbers, and how the magical node lists appear from nowhere. Those of you that
|
||
have been FidoNet nodes since way back when, spring and summer of 1984, and
|
||
watched all this develop (such as it was) in full Technicolor, will know most
|
||
of this; if you are a relatively new sysop, much of this may come as a suprise.
|
||
Everyone should read this, experienced sysops, new sysops, and all Fido and
|
||
FidoNet users.
|
||
|
||
FidoNet is no longer just a piece of software; it has become complex
|
||
organism. There are about 160 Fidos in FidoNet right now; this does not
|
||
include Fidos being run as Bulletin Board only systems, just ones that you can
|
||
converse with over the net. If the average number of users on each system is
|
||
300 people, you can start to guess at the scale of things today.
|
||
|
||
|
||
HISTORY:
|
||
|
||
When FidoNet was first tested, there were two nodes: myself here at Fido #1
|
||
in San Francisco, and John Madill at Fido #2 in Baltimore. John and I did all
|
||
of the testing and development for the first pass at FidoNet. Its purpose: to
|
||
see if it could be done, merely for the fun of it, like ham radio. It quickly
|
||
became useful; instead of trying to call each others' boards up to leave
|
||
messages, or expensive voice phone calls, Fidonet messages became more or less
|
||
routine.
|
||
|
||
This was version 7 of Fido sometime in June 84 or so; it did not have
|
||
routing, file attach, retry control, error handling, cost accounting, log
|
||
files, or any of the niceties since added. A packet was made, a call placed,
|
||
the packet transferred, that was it. This was adequate for a month or two,
|
||
when there were less than 20 nodes.
|
||
|
||
In August of 84, the number of nodes was approaching 30; the net was
|
||
becoming clogged, believe it or not. FidoNet wasn't too smart about making
|
||
calls then. With 30 systems, coordination became difficult; instead of a
|
||
simple voice phone call to the (very few!) sysops to straighten out problems
|
||
like modems not answering, wrong numbers, clock problems, etc, it took days og
|
||
et the slightest problem repaired. There were by now six nodes in St. Louis,
|
||
and Fido #1 was making seperate phone calls for each, when obviously one could
|
||
be made. Enter the beginnings of routing.
|
||
|
||
The "original" FidoNet was very simple and friendly; you told me at Fido #1
|
||
that you had a FidoNet node ready, I put you in the list, with your phone
|
||
number, and people called up and downloaded the list; done!
|
||
|
||
Well ... at first, "everyone knew each other"; we were in more or less
|
||
constant contact. However, when the node numbers got into the twenties, there
|
||
were people bringing up FidoNodes who none of us knew. This was good, but it
|
||
meant we were not in close contact anymore.
|
||
|
||
The Net started to deteriorate; every single week without fail there was at
|
||
least one wrong number, usually two. To impress on you the seriousness of
|
||
wrong numbers in the node list, imagine you are a poor old lady, who every
|
||
single night is getting phone calls EVERY TWO MINUTES AT 4:00AM, no one says
|
||
anything, then hangs up. This actually happened; I would sit up and watch when
|
||
there was mail that didn't go out for a week or two, and I'd pick up the phone
|
||
after dialing, and was left in the embarrasing position of having to explain
|
||
bulletin boards to an extremely tired, extremely annoyed person.
|
||
|
||
There were also cases where the new node really wasn't up yet, and the
|
||
number given was a home phone to be used temporarily, but I'd forget that, and
|
||
include it in the list anyways. Or the new node wasn't really up yet, and we'd
|
||
all make calls to it and it would not answer, or worse, the modem would answer
|
||
but the software wasn't running, and we'd get charged for the call.
|
||
|
||
This obviously could not go on. We had to have some way to make sure that
|
||
at least the phone numbers were correct! I started a new policy; before giving
|
||
out a node number and putting it in the list, I had to receive a FidoNet
|
||
message from the new node, directly. This verified that at least the new Fido
|
||
was half way running. At the time, Fido had a provision whereby Fido #1 could
|
||
set the node number remotely; I'd send a message back, and presto! a new node
|
||
was up.
|
||
|
||
Well, this didn't work properly either; at the same time, the Fido software
|
||
was changing so rapidly, to accomodate all the changes (literally a version a
|
||
day for a few weeks there) that I was losing new node requests, wrong numbers
|
||
caused by illegible handwriting, all sorts of problems. Out of laziness I
|
||
would still assign nodes "word of mouth", and got in the same trouble as
|
||
before.
|
||
|
||
The people in St. Louis (Tony Clark, Ben Baker, Ken Kaplan, Jon Wichman,
|
||
Mike Mellinger) had their local Fidos going strong, and understood what FidoNet
|
||
did, how it worked, and what it was about. They volunteered to take over the
|
||
node list, handle new node requests, and leave me with the software. They
|
||
tightened up on the FidoNet message requirement, and in a few months, had the
|
||
"error rate" (wrong numbers, etc) down to practically zero, where it is today.
|
||
|
||
Though I did the programming, Ken Kaplan, Ben Baker, and the crowd in St.
|
||
Louis did much of the design and most of the testing of routing, forwarding,
|
||
and local nets. They still remain the experts on the intricacies of routing,
|
||
and help sysops set up local nets.
|
||
|
||
Please keep in mind the entire process, from two nodes to over 50, took only
|
||
three months! Fifty nodes is more than it sounds; at that level it becomes a
|
||
large scale project. FidoNet went from about 50 nodes in Sept 84 or so, to the
|
||
current 160+ in Jan/Feb of 85.
|
||
|
||
FidoNet today is a network quickly approaching the levels of complexity of
|
||
commercial networks, and has many more capabilities than many "mini" networks,
|
||
such as USENET, which has no routing or hosts. Only ARPAnet has some of the
|
||
features of FidoNet. The southern California local network is three levels
|
||
deep, with hosts in Orange, LA, Ventura, San Berdino and San Diego counties.
|
||
|
||
FidoNet is just too large today to run as an informal club. The potential
|
||
for error is just too high to include numbers at random within the node list.
|
||
I imagine we are in a predicament today what the radio ameteur operators had a
|
||
number of years ago.
|
||
|
||
The requirements for new FidoNet nodes are pretty minimal, and they appear
|
||
to be arbitrary and harsh if you aren't aware of what's going on. This is to
|
||
spell them out in detail, so everyone will nderstand the process.
|
||
|
||
FidoNet'S PURPOSE:
|
||
|
||
Very simple; it is a hobby, a non-commercial network of computer hobbiests
|
||
("hackers", in the older, original meaning) who want to play with, and find
|
||
uses for, packet switch networking. It is not a commercial venture in any way;
|
||
FidoNet is totally supported by it's users and sysops, and in many ways is
|
||
similar to ham radio, in that other than a few "stiff" rules, each sysop runs
|
||
their system in any way they please, for any reason they want.
|
||
|
||
THE STIFF RULES:
|
||
|
||
Actually, not as bad as it sounds; basically, politeness as a rule:
|
||
|
||
1. New nodes, see below.
|
||
2. If your system is going to be down for a week or more, please let Fido
|
||
51 know. They can take you out of the list while you are gone, so
|
||
other FidoNet sysops won't be wasting phone calls.
|
||
3. If you change your phone number, or decide to stop running Fido, let
|
||
them know, so other FidoNet sysops won't be wasting phone calls.
|
||
|
||
The thing to keep in mind is that FidoNet's telephone calls to send mail are
|
||
costing someone money; if you are down just for a night or so, don't worry
|
||
about it, just make sure your modem doesn't answer.
|
||
|
||
THE NODE LIST
|
||
|
||
Obviously (if you are a FidoNet sysop that is) the node list is a text file
|
||
containing aal the names, phone numbers and other things on each node, and as
|
||
distributed by Fido 51, routing information for the many local networks. It is
|
||
a very compact list, and so there is no clue as to how that list is made.
|
||
|
||
Here is the current process for new nodes to obtain a node number, and get
|
||
into the node list. This assumes you want to run a public access Fido;
|
||
specialized systems are covered seperately, below.
|
||
|
||
SET UP FIDO
|
||
|
||
Of course, you should get your Fido running first; no sense in trying to run
|
||
mail if your Fido doesn't run! In your FidoNet area, enter a message for Fido
|
||
#51, and include the following information:
|
||
|
||
1. Your boards name
|
||
2. City and state
|
||
3. Sysops name
|
||
4. Board phone number
|
||
5. Maximum baud rate; 1200 assumed otherwise
|
||
6. Hours of operation; 24 hrs assumed otherwise
|
||
7. Way to contact the sysop during the day. This is
|
||
not absolutely necessary, but it makes it easier
|
||
if there is some problem.
|
||
|
||
Most of this is pretty obvious. The sysops voice phone number will be kept
|
||
secret; it will not be given out. It is only use@X*_" -Y9e is some problem,
|
||
andd a FidoNet message can't be sent for some reason.
|
||
|
||
For Fidos that want to run with an unlisted phone number, a few other things
|
||
are needed:
|
||
|
||
8. A public FidoNet to act as mail host
|
||
9. The systems actual phone number
|
||
|
||
A host is required for an unlisted number, so that youucan receive mail.
|
||
(If you don't want to receive mail, then there is no reason for you to be part
|
||
of FidoNet!) The host system will have to have the unlisted phone number, of
|
||
course.
|
||
|
||
Fido 51 needs to have the phone number also, but it will be kept secret.
|
||
This is so that they can contact you directly if there is any problem, such as
|
||
a known bug or a question, or if your host drops out of the network, so there
|
||
is some way to contact the local nodes.
|
||
|
||
GETTING A NODE NUMBER
|
||
|
||
This is the part that seems so arbitrary if you aren't aware of what's
|
||
happening. What happens is: you send Fido 51 the message described above.
|
||
When they receive it, they put the stuff into the node list and fido list, pick
|
||
you a node number, and mail a copy of it to you the next weekend.
|
||
|
||
This tests your system at the same time; you have to be able to sucessfully
|
||
send and receive mail in order to get the node number. Out of it, you get a
|
||
copy of the latest lists.
|
||
|
||
NOTE: Fido 51 does not mail out copies of the lists to everyone on a
|
||
regular basis; it would mean too many phone calls ($$$ ...). You can get the
|
||
new node list Friday evening at Fidos 10 and 51, or Fidos 1 and 2 later that
|
||
weekend or early the next week, and usually most any other busy Fido.
|
||
|
||
If it all works, then 1) you know your system is workkng 2) Fido 51, the
|
||
node list keepers, knows it's working 3) the other 160 or so Fido sysops know
|
||
that your system was working at least as recently as the last node list. Print
|
||
out the last few weeks nodelists; compare all the changes, not just the
|
||
additions.
|
||
|
||
This is why node numbers aren't given out "word of mouth", or at other
|
||
sysops request. It has to be done directly, as a test.
|
||
|
||
HAT FIDO 51 REALLY DOES
|
||
|
||
Making the node list is more than just typing in the information; they make
|
||
sure that the information in the list is accurate as possible. This frequently
|
||
means voice phone calls to double check, or calls to the new system to see what
|
||
the problem is; sometimes it is as simple as the wrong baud rate, the time
|
||
wrong on the new system, so that it is not running FidoNet at the right time.
|
||
|
||
Ken Kaplan and Ben Baker do the node list work when they have "spare time";
|
||
please be patient! As the number of new nodes increases every week, response
|
||
time goes up. Currently, the node list is done once a week; new node requests
|
||
must be received in Wednesday nights mail (by Thursday morning) so that they
|
||
can work on it Thursday night, and send it out on Friday night, so that you
|
||
will have it over the weekend. The volume of mail is such that it may take a
|
||
few days to get out.
|
||
|
||
(Please note that Fido 51 is an unattended node; there is no one there to
|
||
answer Y)ells unless someone happens to walk by. The machine is located at
|
||
Data Research Associates, who kindly donated the phone line, and runs on a DEC
|
||
Rainbow 100+, donated by Digital Equipment Corp.)
|
||
|
||
Fido 51 is an extremely busy system; they receive 125 messages a week
|
||
through FidoNet alone, so please be patient.
|
||
|
||
CHANGES, MISTAKES AND UPDATES
|
||
|
||
If you ever find wrong information in the node list, please send the
|
||
information to Fido 51; they will include it in the next list.
|
||
|
||
If you become part of a local net, ie. you have an incoming host, notify
|
||
them, and it will be included in the node list also. Other changes might be
|
||
baud rate (got a new modem!) hours of operation, board name or sysop, etc.
|
||
|
||
SOME OTHER THINGS ...
|
||
|
||
If you have questions or problems with any part of Fido or FidoNet, please
|
||
ask. Here's where to go for problems:
|
||
|
||
HARDWARE, SOFTWARE, PERFORMANCE OR INSTALLATION TROUBLES
|
||
|
||
Call or FidoNet to Fido #1, me, Tom Jenningg. FidoNet is best, if possible;
|
||
that way, I have your "address and phone" handy. If not, then call Fido #1 and
|
||
leave a message. If you leave it at G)oodbye, when you call back looking for a
|
||
reply, remember to check in the ANSWERS area; Fido will NOT tell you if there
|
||
is mail for you, you have to search for it.
|
||
|
||
Fido #1 always has the latest versions of Fido for all hardware supported,
|
||
available for download. Fido #1 ALWAYS runs one revision later than the
|
||
released version; it is used to test new featurS BV IVW$V^YY.so that whhe
|
||
released they will be working. Check the FIDO download area for the current
|
||
Fido version.
|
||
|
||
I have nothing to do anymore with maintaining the node list, nor do I hand
|
||
out node numbers.
|
||
|
||
ROUTING, NODE LIST, LOCAL NET QUESTIONS AND PROBLEMS
|
||
|
||
Fido 51. Since they keep the list, they're the ones to contact for node
|
||
list problems. If you want advice on how to set up a local net in your area,
|
||
they can offer help and advice.
|
||
|
||
|
||
SPECIALIZED SYSTEMS
|
||
|
||
If you are setting up a private network, and it is to be truly private, what
|
||
you do with it is your own business. If, however, there is any possiblility
|
||
that members of your private network may wish to communicate with any members
|
||
of the public network, you should contact Fido 51 for the allocation of a block
|
||
of node numbers to be assigned by you to the nodes in your network. This is to
|
||
avoid node number conflicts upon receipt of FidoNet mail in the public network.
|
||
|
||
LOCAL NETS
|
||
|
||
Neither I nor Ken Kaplan nor Ben Baker "run" FidoNet; local networks such as
|
||
the one in Southern California and Massachusetts are entirely the
|
||
responsibility of the sysops in the area; the only thing we ask is that the
|
||
designated "incoming host" for that area be somewhat reliable, for the obvious
|
||
reason that it will be receiving lots of phone calls from across the country.
|
||
|
||
As a matter of fact, you are encouraged to form local networks, or join one
|
||
that exists locally. IT makes it cheaper for other systems to send you mail,
|
||
and generally streamlines FidoNet operation.
|
||
|
||
Other than that, local nets are totally standalone; that is what they are
|
||
for! For instance, SoCal can run their net anyway they please; it is their
|
||
hardware, their phone lines, and their phone bills. It is their investment in
|
||
work, and they should reap the benefits. If there is a "FidoNet policy", this
|
||
is it.
|
||
|
||
AND SO ON ...
|
||
|
||
I hope FidoNet is a bit clearer now; if you have any suggestions, or want to
|
||
volunteer to help, please let us know. Our only interest is in keeping the
|
||
node list correct and up to date; this simple list is what ties the entire net
|
||
together.
|
||
|
||
Ken Kaplan Fido #51 314/432-4129
|
||
Tom Jennings Fido #1 415/864-1418
|
||
Ben Baker Fido #10 314/234-1462
|
||
Call The Works BBS - 1600+ Textfiles! - [914]/238-8195 - 300/1200 - Always Open
|
||
|