textfiles/bbs/FIDONET/JENNINGS/HISTORY/tomj_mrobbins.txt

The History of Fidonet an Interview with Tom Jennings
October 1993

tape 1 <labled 2> side A

<Tom started talking before the tape was turned on>  and I had
managed to steal a modem from the phone company by taking advantage
of the breakup so called which is a long story in itself. But I
had a bell AT&T 212A-something modem, a big clunky thing with a
multikey phone attached and so I thought I know, I'll write a
bulletin board. so my first attempt at a bulletin board was horrible.
It sorta looked like Dos, dos Version 1.25  at the time. And then
after a couple of abortive attempts I realized that this was stupid.
I'll just not remake it, I'll take a good look at cbbs the CP/M-80
based board at the time that was the only thing really worth a
shit. RBBS which was just starting to become available then and it
might have been around for a bit maybe a year. It was written in
basic on an IBM pc.

Marge uhu

Tom  I did not have an ibm pc and basic I will not do anything in
basic except write printer tests or something. So I wrote what
became Fido in C starting November 83. It was probably operationable
about the begining of the year. The begining of 84.

Marge  Do you still have a copy of that software available?

Tom  Ah no

Marge Shucks I would like to play with it.

Tom I would too actually.  I'm curious as to what [they look like
now], but I do have versions 5 and 6, and versions 1 through 6 all
came out in the first year. And they look amazingly similiar.
There's just not that much difference. Some commands are added.
You can guess which ones. Verson 1... well really I don't know what
the first version was.  It probbly wasn't a one. But they were
remarkably similiar. The first usable one only had one message area
and one file area. And everything else was added later.  multiple
areas.

Marge uhu So when did you get the bright idea of getting two of
them to talk to each other?

Tom Well it had been an idea kicking around for some time. Since
the late 70s in what's now the PC world. But that was something
else. There was a system in Andover Mass run by Wayne something?
Andover CNODE it was a CP/M-80 (Zilog Z80) machine he had a whole
bunch of harddisks and a lot of programs to download. and about 8
logical drives like from A: to H: or something. He had a MINICBBS
BBS (stripped-down version of CBBS... utterly minimal) that he
would run as a separate program from it. Some user of his system
had this idea about what if we had bulletin boards that just hopped
a message across the country by going local call to local call to
local call. And while it was an intriguing idea, when you stop to
think about it mindboggling and ridiculous.  But it got some tiny
amount of message traffic and of course we put it away like all
ideas like that, but it kinda stuck with me even though by this
time Fido was basicly... my fido board was rather busy. I was
running a technical board a technical C programmers board...

Marge What was the name of your first board?

Tom  Fido. Fido's board

Marge oh ok

Tom Fido's board. It was a very busy board. It was pretty popular.
and my software was unique in that it was small and compact and it
was not written in basic and it did not have all that crappy
limitations that basic has. People were asking for copies of it
which I gladly supplied. One of them was John Madill in Baltimore
Md. He was just a real sharp guy and we ended up like calling each
others boards back and forth and uploading files and getting latest
versions and chatting you know chat mode and all that kind of crap
and you know, inflating our phone bills

Marge Uhu samething we all do today

Tom Yeah, oh yeah not much is different about it except that phone
calls cost less now.

Marge plus you can get more stuff in one call

Tom Yeah this was at 1200 baud

Marge oooh!

Tom and uh not everybody had it. um So its a long ...I forget what
brought the idea back up.  I thought out the machine to machine
business again as it was also a big giant hack...the hack value
was very high. I just sat down and just sorta did the design work,
apparently nobody else ever got beyond the hoppity hop across the
country with local calls bit.  Well first of all its ridiculous.
Second of all you'd never identify all the local calling areas and
if each of them is ten miles wide you know the US is 3500 miles
wide its crazy so I said well given that long distance AT&T is $11
bucks an hour after 11...

Marge 11 bucks an hour?

Tom 11 bucks an hour, 1200 baud, a page of text, 2000 bytes you
know you can do the arithmatic yourself its pretty straight forward.
It works about to be about 25 cents per call.  So I said what's
the big problem?  That's obviously not it. And I worked out the
basic principles what Fido was about which were written down in
the Fidohist documents I'm not even sure I can recite them anymore.
But basically they're ...

Marge That's ok. I've got those.

Tom Yeah at this point that fido that history is, its actually
pretty accurate It was written in 1985 or 86. So it hadn't hadn't
gone too much adrift by then.  Probably better than my memory.
Which is not always very accurate.

Marge well considering the contradictions I'm getting from everybody
else I'll trust your memory.

Tom Yeah there's a lot of bad... matter of fact theres a lot of,
not just like you know memory drift which is just inevitable there's
people with fucking agendas and I'd like to strangle them all
individually and throw their bodies over a cliff.

Marge Well we'll get into that. But I've got people telling me that
IFNA didn't die until 1991.

Tom Huh? Really?  <Marge laughed> Well you know maybe there was
vestigial legal leftovers didn't drop dead until 91, but it was
quite dead before that. <laugh> I had written them disallowing them
to use the words [Fido and FidoNet] much earlier than that. I don't
remember exactly when. I can look it up. 87 88 89 Something like
that.

Marge The vote to kill it was in the fall of 89. I missed voting
by one nodelist entry.  That was when I joined Fido.

Tom Oh yeah that sounds about right. It's written down somewhere.
I've got some of those docs. I saved the legal letters. The ones
that we sent out with lawyers.

Marge hum..well get back to Fido. We can get to Ifna later.

Tom yeah Where was I? Oh I wrote a design specification and everything
unfortunately I don't even have a copy anymore. One thing I've
never had because I've never had enough money and hardware and I
don't have the continuity the mainframe and university people have
and I don't have a lot of my own files from back then.  Cause I
would change computer systems and they would be completely incompatable
and there would be no mass movement of my data from one system to
another. You know people would go to school and have accounts on
the university machines that stuff's all taken care of for them.

Marge uhu

Tom You know get an account somewhere on some machine..  they change
from a Vax to a some whatever other machine from some company they
get backups of harddisks and they move the data over and they're
done.

Marge That's too bad now

Tom yeah it certainly is. You know I hated to drag around this box
of 8 inch single sided single density floppies for years and now
I can't read them anymore. so I threw them out a longtime ago which
was probably a mistake too. But

Marge I'm not going to contradict you on that one. Anyway

Tom I have a different approach now.  I keep everything on my hard disk
and if I need more hard disks I just have to find a way to buy more and
I back that up on tape and everything just stays live now. No more
offline storage. Offline storage is the thing that killed me.

Marge That is a pain, I'll agree. so tell me about the first fido,
about them talking together.

Tom Well it was pretty exciting. I remember that much. I don't
remember exactly what happened. basicly it was John Madill and I.
I didn't even have enough machines to test it across the room. so
it was a lot of writing into the dark. But the designs are amazing
consistent over time, even with the new wazoo mailers the methodology
is pretty much the same. The first one would dial, send messages
and hang up. It did not do file attach, it did not do lots of
things. It did also multiple messages in one pass. But what it did
not even do (um of course fidonet predates nets and and zones and
regions and all that crap so it initially um) Fido did only point
to point mail I mean literally point to point; if you had three
systems in Boston and a message to each of those systems Fido make
three phone calls. 

Marge so there was no such thing as routed mail from one board to
another?

Tom  Not for the first month or two. But the thing is there is
about ten years of history crammed into like 6 months of developement.
Fidonet has it was defined in like June of 84 was very very
unsophisticated, but by December it was getting pretty damned smart.
And also the terrain, the BBS world, is completely different. It's
not even conceivable now, with the things people now take for
granted. I got away with it because it didn't matter.  There weren't
4 baords you could call in Boston. They just weren't there. There
weren't  4 fidos. Just one. That was it. So some of this stuff that
seems so like completely unlivable today just didn't matter then.

Marge That makes sense.

Tom And as they became important, they got solved very quickly.
In St Louis was where the first serious Fido BBS clustering happened.

Marge ok how did the guys in St. Louis get involved? How did they
find out what you were doing?

Tom Oh I don't know how they found out, they probably just called
my board. Fido 2 was John Madill Fido 3 who was it I think it was
a guy in Atlanta Fido 4 was Tony Clark in St. Louis 5 was oh boy
somebody I forget 5 I think he was Oregon somewhere 6 was of course
Kaplan had oregon 7 was a flaky number that went away 8 flaky number
9 Connecticut we scattered around the country very instantly.

Marge That must have been pretty exciting

Tom  yeah 10, 11 and 12 were all in St. Louis 16, 17, 22 were all St.
Louis There were what um NDC/RCC, McDonald Douglas Radio Control Club
can you hold on a second? I think I have a runny nose.

Marge Sure

Tom sorry I have an allergy this morning

Marge That's alright. This is the time of year for it.

Tom Yeah, its irritating.

Marge Be glad you don't live out here in corn country. I am thoroughly
miserable from the middle of August until past the middle of October
when frost hits.

Tom oh boy yeah that's painful.

Marge I don't go outside much, believe me.

Tom  um

Marge ok so at what point did you start doing routing?  Having one
board pass mail on to another so you weren't all calling each other?

Tom oh what happened was that we there wasn't much call for it for
at the time.  I mean there was just too many unknowns. and the
biggest unknowns were the social ones not technical ones. What
happened was Fido started to cluster [new Fido BBS's distribution
matching areal population density] which was hardly a surprise.
Because they started clustering where people are and of course
there are more people in cities. Well it became obvious that when
I make three phone calls to St. Louis to deliver mail to systems
that are all within local calls of each other that something's got
to be done. So I worked out a scheme and a syntax to do routing.
FOr example, Fidos 4, 10, 16,17, and 22 were all in St. Louis, all
a local call from each other.  I had to have some way to tell Fido
to send all those messages in one call to one of those systems.
And we had to have some way to tell that St. Louis Fido that received
the mail that it was ok to to forward on these messages.  I did
two things. One was to make Fido not really care where messages
came from, just where they wanted to go to. So if a message appeared
either entered by a local human or appeared from from another Fido
it would say ah this message doesn't belong to me this belongs over
there, I'm going to deliver it.  Um there are of course all kinds
of delivery controls, ACCEPT-FROMs, and all that kind of stuff.

The other thing I had to do was make it say, oh, I know all these
systems go in one call.  This was I believe version 10.  And that
was done with this thing called the routing table.  The first
attempt was this hideous method that I kinda of abandoned pretty
quickly and used to be distributed within the nodelist itself.
Its replacement is the basic syntax still used by Fido today.  It
ended by getting implemented as an N by N crossbar switch, where
"N" is the number of nodes in the nodelist, but the table is only
two-N entries long.  Radically different from most mailers today
which are oriented towards echomail. Of course Fido allows any you
to route any piece of mail to go to any other node at random at
any time. And the tables to do so are pretty small. for every
nodelist entry there is 16 bytes or something.

Marge Uhu

Tom 
Um so we had the rudiments of routing down, but we still didn't
have
the hierarchy of nets and nodes. That came later. Which was
basicly the formalization of this informal routing scheme. And
you know its funny. There were two things It was very awkward
when there was about 50 nodes. Fido could do point to point mail
and it could do a file attach and it could bundle up multiple
messages for an area. You know essentially put itself in a net.

Marge Uhu

Tom there's two hideous kludges that went away at the end of 84.
One was, how to define an autonomous network within FidoNet.  For
example, St. Louis is autonomous and how they arranged their business
is none of their business. And we don't really care how they do it
internally. This was one of the basic anarchist principles, deciding
where the control of FidoNet ended, and individual sysops remained.

Marge Tell that to some of the coordinators today

Tom well at that time the net coordinators are actually the best
thing about Fidonet and the least of our problems. I mean if you
live in an area that has a real jerk of an NC you're kinda  screwed.
But the thing is their mental illness doesn't tend to spew over an
area, you know. A bad NC in New York won't really have much affect
in California.

Marge That's true. Actually I'm lucky. My NC's a pussycat. He does
exactly what we tell him to then keeps quiet the rest of the time.

Tom yeah most of them do actually. There's very few troublemaking NCs.
People don't realize that the amount of social engineering going
on. It was quite explicit. but the NCs have a very intimate
relationship relatively speaking with their members than say an RC
or ZC which are late model artifacts that I had nothing to do with.
The only thing that we designed in in March of 85 was NCs.  And
there was an RC function but it was completely different from today.
Today's thing is a complete perversion of it. but basicly we realized
that even an asshole NC has to live in the same city essentially
or the same region.

Marge Right.

Tom You could have a meeting you're actually going to physically meet
this person unlike somebody who is all the way across the country. I
mean that was builtin. I was very explicit about and that still
holds true.  my personal rule of relationships is: the closer you
are to somebody the more information/interaction you exchange with
them. You talk to them, you know who they are, what they look like,
you hear more of their output, you know you talk to your friends
down the street more than you might talk to your relatives across
the country.

Marge Right

Tom  And your relationship to them, all other things being equal,
you know casual acquaintances versus relatives you'll be closer,
you'll have more uptodate knowledge, and more obligations to and
obligations from local people. And we relied on that fact to keep
the NCs in line.  But at the time it was just so things won't get
so disconnected. And that part's proven pretty much true. So 
if you've got a completely irate handful of people in a net and an
NC who's conspiring with them. You show up at a big meeting everybody
sees it.  Whereas if its across the country like our congress
critters, who the hell knows what they're doing.You can't see them.
You never get to meet them and you don't see them in action.

Marge That's for sure.

Tom so anyway we formalized that process. a lot of its actually
written down in the Fidohist Documents. About the way the scheduling
was done. The scheduling used to be done in this nodelist of all
places. Gross! Ben Baker was doing all the routing at the time and
scheduling. He split off a program that did the nodelist and one
that did routing seperately.

Marge So now you had two external programs?

Tom Well at this time we had a whole bunch of external programs. But
this was an external data structure that that turned into um what the
hell's the program? it later became makenl and stuff but at the time it
was called makelist or something. I forget what the hell it was called.
There was also one called routegen. At this point in time Thom Henderson
was starting to come along. what I was doing at the time was basicly
making all the protocols public in the public domain.  I was trying
to document it, not very well but I was trying to document the data
structure even though I never gave source to Fido out I wanted all
the structures to be public for Fidonet.

Marge ok, timewise where were we?

Tom Oh still 84

Marge ok

Tom Still 84 possibly early 85 but about that time. I was badly
documenting this stuff and I forget now when Henderson popped up.
It might have been I think he came by in 85 I'm not sure. But I
think it was within the first year. towards the end of the first
year.  The last 6 months of the first year of actual operation
which would put it I would say spring of 85 plus or minus. He can
correct it.  Its probably written down somewhere. He reverse
engineered Fido which was fine. and um between you and me 
he wasn't really willing to ask questions, he maintained that bad
attitude all these years.  Its really wierd to me. 

Marge ok what did he do to it?

Tom oh I don't mean he reverse engineered the fidonet
program itself, but the protocols,  which was fine because it
probably my poor ability to document them that drove him to that; I
don't have any real problem with that. Its just that he never
bothered to ask which was odd. But he wrote seadog which was a
fidonet interface essentially without a bulletin board which was
great. 

Marge ok

Tom  I just got news that he did this thing that was more graphically
oriented.  And he was definately aimed in the business direction
and that was fine.  there was no problem with that. And then he
contributed a lot actually.  He wrote the arc program. The archiver
and compressor.

Marge Yeah, I knew about that.

Tom Before that the functions were seperate. There was FQ and lu
and lar. I had written lu out of this older program called lar
which was written out of an even older program that came out of
the unix world called archive. And lar was a library archive and
it was a hideous clunky gross disgusting program that didn't really
compile and you know basicly it did arc without compression. It
mashed a bunch of files into a single file. and I wrote a cleaner
one and it was used a lot on bulletin boards.  It seems obvious
now, but it wasn't obvious then to make an archiver that did
compression and archiving at the same time.  He wrote some of the
nodelist processing stuff too. He's cranky, but we were all weird
I suppose. no crankier than the rest of us.  and Ben Baker was
writing code and the nodelist management stuff and Ken Kaplin came
in during all this.  He defined the fidonet as we know it today
and held it together for more than a couple of years.  The IFNA stuff
kinda knocked him out.  He was kinda a strong believer in IFNA
when it all collapsed he fell apart and dropped out. he also got
flamed a lot mostly undeservedly. And even when some criticism was
due it was simply one ofprocess or lack of or whatever. He was
never malicious never never never malicious.  Not once.

Marge  Ok lets back up a little bit. One of the things I've done is
read all the back issues of fidonews, and it seems to me like at some
point the net turned on you guys and did nothing but criticize. Was
there a reason for that?

Tom  That's what happens when you get any bunch of people together.
All you can see is the bad parts and you can't see what went before.
It happens today still. I've learned certainly some obvious technical
things, but boy I've learned a lot about large social structures
wow which I've used since in other projects.  so at least we learned
how to build social structures.  what happens when you put a lot
of people together is it always looks like hell but it tends to be
fairly robust and we were hehe in discovery on this ah um.

Marge  Ok so you started out with nets. That makes sense. How did
RCs come into play? How did they come into being?

Tom 
What we were noticing was that as you if you put little dots on a
map where Fido's were they were appearing in clumps around population
centers.  No big surprise.

Marge that makes sense.

Tom There were many of couse out in the middle of Iowa somewhere
or some smaller city where there was only one node. So what we did
was define this thing called a region, that included all the Fido's in
some geographical area that were not in a city and therefore in a net.
To the FidoNet program regions looked like Nets, but instead of sending
all the mail for a region to one single Fido host for later redelivery,
each would be delivered directly, one by one, like in the original,
pre-net design. This was because in fact, they were not able to take
advantage of the local-call thing, but we wanted ther administrative
function of the net coordinator, to keep their list of Fidos, the FIdos
within a region. Syntactically, they all looked the same, 100/51 was
Fido 51 in net 100, while 10/1 was Fido 1 in region 10. The person
sending email didn't have to know anything.


Marge allright

*********

tape 1 <labled 2> side B

Marge Right

Tom and the figure to the right of the slash was your old fido
number which were assigned sequencially starting with one. until
we did a switchover. Then we defined magical address 0 as the entry
point for that net. Every net by defination has a 0 node and it
would be physically the same as one of the numbered nodes in the
net. So in St Louis's case I think it was 22 I'm not sure. 
100/22 was Ken Kaplin's board but 100/0 was the logical entry point
for the net. It wasn't necesary physical but it usually mapped onto
one of the existing fidonet nodes. So we had a mathimatical model
that was independant of physical reality. And that's all neat
and clean. That's all fine, what to do with the region the outlying
areas. So what we did was, we defined them as large giant geographical
regions. And um you pick a region host who has a 0 address like
you know region 10

Marge Yeah

Tom The other function of NCs was to maintain their fragment of
the nodelist which was sort of an obvious functon and that's one
that's worked out pretty well. It was a reasonable decision at the
time and has remained so. Incoming host is just a big paper throne.
Their machine gets to handle incoming mail and make more phone
calls and they get to maintain their little fragment of the list.
So I decided, its funny none of the software in the world does it
anymore except Fido, to syntactically make them look the same
net 125 and region 10 but the software behaved very differently.
If my system has a message destined for 100/22 Fido
knows to it says 100 that's another net, that's not my net. So it
goes to 100/0 and it makes a packet and sends it to 100/0.
So there's a lot of messages that go for net 100 you know 2 22 16
50 whatever the hell they would all go to the same bundle the same
file that would be delivered to 100/0 Who would take it apart and
distribute the messages.

The regions however, even though they look the same you know like 10/1
or 10/2 the software knew that it was a region and it would not do
that bundling. You know it would deliver each message individually.
It would make a phone call for every message it had in region 10
for example. 

Marge  ouch

Tom And the way it knew the difference was in the raw distribution
nodelist different keywords defined nets and regiosn.

Marge right

tom surprisingly little software does this today. Um it was
complex enough that people didn't understand it. 

Marge sounds kinda simple to me, but go ahead.

Tom um yeah its very odd. Fido was all point to point mail There
was no echomail at the time. That was a relatively new invention.
I don't know exactly when I put file attaches in. That was one of
John Madill's suggestions. You know as soon as he mentioned it
seemed obvious.  and it got done pretty quickly.  Basicly before
June of 85 we had a flat system that would send messages and files,
it could pack up mail for multiple sites into one phone call but
it did so by these explicit maps.  So after June we had something
that looked very much like the modern Fidonet.  Minus echomail.
Then there was just a whole bunch of lateral growth. Things got
bigger.  Wynn Wagner III was working on some bulletin board 
on his own.  Appariently he was actually working on a similliar
thing to fidonet.  But when Fidonet came by and it was obviously
just like this large thing already existing he had written a lot
of the wazoo protocols.  He just wrote the fidonet compatable stuff
and plugged it into his bulletin board. And it started talking on
FidoNet.  And his was a much more full featured system
than mine.  He had different ideas than me. I was going after the
lowest common denominator.  The most common denominator universality.
He wanted to go after the vertical features. You know graphics and
all that kind of stuff.  But that's fine. That's what its all about.
And you know Henderson was doing his own thing with sort of a
personal mailer.  The Saint Louis gang was maintaining the nodelist.
They were massively improving the accuracy getting methodology on
how to put the nodelist together, how to qualify nodes and all that
kind of stuff. Somewhere here I met Randy Bush via FidoNet.
he's a crusty crumudgeon with a very strong lefty background.  He
likes code. He does stuff.  He doesn't care about posturing so much
as getting things done.  He liked FidoNet because it was an extremely
low cost grass roots communications tool.  He was like me um
sufficiently paranoid to not want to rely on a centralized
organizational scheme.  when he saw the documentation problem I
was having he jumped in and offered to document it. And he did. He
did FSC-0001 and we're using it to this day.  And basicly without
those documents Fido wouldn't be where it is today.  A lot of people
can not stand him He is crusty.  And he got that whole technical
standards committee seeded.  And that's the document all the later
stuff has been written from.

Marge We're still following that one today aren't we?

Tom oh yes absolutel Hell yes Because it is a very informal committee
of a bunch of developers who have these documents and there is no
authority.  Its this very delicate balance of credibility. And most
of the developers recognized that standards are necessary.  And
while some of them were stupid there's problems here and there 
the kind of obvious stuff.  they also recognized that it's got to
be this way. Some think, this is stupid, I'll just write the best
program in the world and everybody will use it and occasionally
they cause a lot of mess, but usually their withers away.

Tom so let me think where we were

Marge let me ask you a digression question here. There's been a
lot of moaning and groaning lately about the technical standards
committee is not doing anything and we're stuck on the least common
denominator as opposed to upgrading the standards to the latest
and greatest. How do you feel about that issue?

Tom They simply don't know what going on. And they don't know how
its supposed to work.

Marge The average sysop doesn't, but that's beside the point.

Tom well the thing is the standard's committe is not supposed to
design features. The technical standards committee does not design
things, they document.  They are merely defining, by writing down
what's already being done what the common denominator is.  It has
nothing to do with features.  Features are done by developers. And
yes they move exceedingly slowly because the lowest common denominator
has to move slowly and it does move slowly. Its all too trivial
and easy to just write super features continuously continuously
continuously that's not the problem. the problem is to have a
network scattered across the planet still be able to talk to itself.
Because there people just now starting to write programs for new
hardware some of it's really low end. And at the same time there
are programs for like Dos and Windows and whatever that have been
around for 5 close to 10 years exceedingly well developed and
refined,  that the new people will never achive because the other
people had a five year headstart.  So if you let the leading edge
of technology drive things it will collapse into chaos.
So I think people just don't get it.  I don't know why people are
waiting for the standards committee to tell them to do things
anyway. Its another case of people wanting others to do their work.
It doesn't stop anyone from writing features That's for sure.

Marge that's true  Ok tell me about regions. I think that's where
we are in the story.

Tom oh yeah well what happened was we defined this logical entity
called a region which on paper looks like a net. 
what we did was divide North America into 13 regions.
so that there'd be as many people here in California/Nevada as
there is like out in the East coast. We put the regions where the
population was.

Marge ok so regions were supposed to function as nets?

Tom they were supposed to function only administratively as nets.
The NC would be this person who will help you get online ,
he's got the files you need and all that garbage. But if it happened
in the middle of the California desert for instance where's that
guy going to go? He's not in  a net.  He would just fall into the
geographical area for the region.  The idea was as you plunked down
more nodes the region host would have more and more nodes but as
soon as they got enough to get in a clump that little clump would
split off and form a net and the region host load would drop a
little bit. And that's pretty much how it went.  Like Fremont Calif
which is a medium sized city, but its time was not you know it was
not populated with Fido's.  As it got 2 3 and 4 Fidos they split
off from the region and the region host went from having 40 nodes
under his list to having 35 or 30 so that load didn't go up fast.
And this was all designed in March of 85 What happened was the real
problem was the nodelist was getting to be large. It was getting
to be to be like 64K long and I've forgotten when that happened.
Something like late 85 or 86

Marge Ok

Tom um it had to also do with echomail making things very popular.
Also machines were getting cheaper and modems were getting cheaper
and all this in hindsight is obvious.  The NCs and RCs would make
their own little fragments of the nodelist and would fidonet mail
it to Saint Louis where Ken Kaplin would compile it into one big
nodelist, then Ken would mail out a copy of that file to all the
NCs. And that worked fine for a while until it ended up
sending a lot of 64K files at 1200 Baud

Marge you know that could do a real number on the phone bill

Tom yeah and it was also one of the drives to form IFNA to help
pay for nodelits distribution phone bills and stuff like that. And
this was getting to be a full time job.  So anyway Ben Baker did
the diff process which instead of mailing out the whole nodelist
they could mail out a difference from last week's nodelist which
was vastly smaller.    Then even that got to be too many phone
calls, too.  So we decided to tier this whole thing, for regional
distribution.  So the RCs seemed the logical choice.  At this
point I was sort of out of the loop of running this stuff. I did
not flag it as a bad decision.  I don't remember it as such. I
might have, I might not have.  Well what happened was they ended
up building a heirarchy.  It unloaded a lot of work and 
things they just went along. The nodelits was getting to be such
a big burden that quick solutions got done that probably weren't
best in the long run,  but nobody knew ...How the hell could you
know how to do this stuff. Its not like you could look at all the
other amateur networks out there and see what they're doing.  Or
other national groups without money that had severe technical
coordination problems.  So the IFNA thing came by. I think it was
Ken Kaplin's idea but I was right there.  It was like gee what if
we had an organization like the Radio Amateur Relay League.  Its
off to the side, it doesn't control anything Its utterly not
mandatory to join. Its a highly optional club you could join, get
a newsletter and little badge you know and pay dues and you know,
all that kind of junk. And what it would do was would pay for Ken
Kaplin's phone bill, send representatives to Electronic Mail
association meetings and all that kind of stuff.

Marge that makes sense

Tom Basicly good stuff.

Marge Ok had anybody written any policy at this point? Policy 1
policy 2 policy 3? When did that stuff come into play?

Tom Yes it did and 1 and 2 I don't have any files before policy 4.
I would love to have a copy of earlier policy documents

Marge I would too.

Tom The thing is its another one of these revisionist things. It
always happens. I don't think there were actually polices 1 and 2.
I don't think they were called that. They were just somebody wrote
a net etiquette file and then somebody just went from there. IF
you go backwards in history you don't find a straight line you find
this messy cloud. the fidohist documents were used for a lot of
things. I'm not sure if I quoted the thou shalt not annoy thou
shalt not be easily annoyed in there.  If I did, I did not make it
up. Ken Kaplin or Ben Baker who made that up. Or maybe even Tom
Henderson which was kinda of ironic.  But that was one of the
earliest ones. Ken Kaplin was certainly a proponent of it.

Marge So those history docs are actually a fore runner of policy?

Tom Also the early policy docs were not Policy docs. That actually
began with policy 4 Its not like policy. Its policy and procedure.
The procedure part is what everybody sort of wants to have. 
How we qualify what you have to do to be in the nodelist.  And its
not completely arbitrary. There's good solid reasons for it.  Even
if occasionally the reasons were wrong. The intent was to eliminate
errors and technical problems. And all of the requirements are
technical only and there's not much of a problem with the procedural
 But basicly the procedural stuff is great. Its what should be in
the document and what it used to be 99%. Policy 4 is about 2/3
bullshit policy and 1/3 solid procedure.

What's happened was we just didn't know enough. Probably I screwed
up by not defining these things well enough. And things got blurred
and instead of things getting thought out clearly so policy and
procedure got all mushed together in a hideous mess.  We're working
on a solution today. which is well kind of amusing.

Marge If I turn the tape off will you tell me about it?

Tom Ah later <laughter> Its not done yet.

Marge Ok

Tom Its Randy Bush wrote it up once in a document called the
Revolution 9 and its around.  Its a way to automate things that
are considered policy now. I think its going to go out with a
whimper and a bang and nobody will notice after it goes through
that its done any other way. So you know the policy documents were
everything was sort of the world was sort of benign at that time.

Marge So when did things get messed up?

Tom Well right from the start. There were whiners and nutcases
right from the ten node level.  politics is what happens when you
greater than 2 people. when you have 100 people or more your bound
to have people who are just completely dissatisfied with whatever
the hell you're doing.  Cause that's standard they want something
else. They don't have the control, they're not programmers or 
whatever the hell and off it goes.  and the only real problem is
everyone acts as if its this horrible unexpected thing like we're
so disappointed that somebody's upset and nobody likes what we're
doing. its the nievity that's really the bad part. It just goes
with the territory.

The other thing that I learned was that a well run thing is not
necessarily a smoothly run thing. Fidonet is not smoothly run.
Its chaotic, its backstabbing its whining and complaining and things
crashing all the place. But this is normal and healthy. I think
Fidonet is gettinga little bit unhealthy at the moment and I think
this has to do litterally and absolutely the ZC/RC structure and
nothing else. And maybe the EC guys in some cases.

Marge What do you think the solution is? We're definately
digressing, but this is interesting.

Tom We're going to get rid of them, lock stock and barrel.

Marge Are you sure the tape should be running?

Tom What?

Marge I said are you sure the tape should be running at this
point?

Tom   Oh I've been saying this for years. <laughter>

Marge It sounds to me like you're actually plotting something
now.

Tom ah yeah eventually the idea is to eliminate the preceived need
for them. And um the nodelist is the center of it.  That's the
center of their power. Like the ability to kick people out of it
like out of the net that he doesn't like. That's complete perversion
and corruption of what the hell supposed to be going on with that
stuff. Oh the original idea and this was original idea this is from
March of 85 was that an inverse proportional relationship thing.
Well you know those who were close had more power. Those who were
far away had less.  And it was arranged that it was so. You know
an NC in San Francisco has no effect on a guy in St. Louis.
Absolutely none.  You know all the posts are on paper anyway.  If
an NC was a total jerk you threw him out. Not much problem to get
rid of him you just sort of stopped using him. You just make your
own list you mail out to everyone else ignore the idiot screaming
over there This guy is the NC signed supporting sysops.  its a done
deed. And that happens more than once and its a fine thing.

What happened was we ended up with this backdoor heirarchy put in
essentially by the St. Louis guys but not out of maliciousness but
just sort of out of everybody figured that we just had to get stuff
done. And they basicly built this heirarchy of RCs and later ECs
that had no checks on them. No balances whatsoever. The original
balances were put in with the original RC and NC stuff. There were
no ZCs. Cause then there were no zones. Even though it was flawed,
highly flawed at least it had been thought out. The way to limit
power.    These guys that were doing the nodelist stuff even though
they tended to be good guys orginally and of course it was guys it
was 95% male and had been for a long time It just like other things
tends to freeze itself into place.  And at least with the early
stuff, the early defined things people understood and perceived
there to be paper posts and no actual power Everyone knew that was
supposed to be how it worked. Even though NCs occasionally get away
with murder, you know reading people's mail and stuff that tends
to not be the rule cause its sorta outrageous. But the RCs and ZCs
could do whatever they want because they were just making things
good. Nobody wrote down what they were supposed to not be doing.
So that's what happens.

Marge Ok tell me about IFNA

Tom Well it was sort of a good idea that was kinda ill formed 
What was designed by Ken Kaplin and me and other people was a pretty
benign thing. Two things happened. Of course nievity naivity was
a huge part of it again. Let me think what exactly happened.
Lawyers first of all. two different lawyers. As conceived by the
original gang of co-conspirators it was basicly like I said like
the AARL voluntary club informal representative.  that part was
fine. Nobody had any real problems with that. At this point the
conspiracy seekers were starting to see efforts to control in places
were there really weren't any.  We were starting to talk about what
we could offer tshirts coffee mugs insurance cheap modem deals ....

Then we had what was basicly the first Fido conference in Colorada
Springs Colorado. I think that was 87. Pretty sure that was in 87.

Marge I probably have that written down some place.

Tom  Yeah 86 or 87 We were going to announce the full plan for this
IFNA thing.  And its wierd I had audio tapes of the whole proceeding.
It was so hateful I erased them. Oh how I regret that.  <laughter>
Oh my god do I regret that. There's no other record of it.  I had
it on tape. Don Kulha made them. Well ok it was kinda ill formed
in some ways but it wasn't unrepairable. It was certainly had no
mandatory or control aspects to it. But what happened was the people
that who hosted the conference one  of them was a lawyer. And he
caught wind of this and I don't think he even had viscious things
in mind he was just a lawyer and he liked to set up complicated
structures.  So he set up this horrible horrible thing that was
just awful; the word railroad job was used a lot. At the tim I
thought they were wrong, but in hindsight yeah they were right it
was a railroad job.  For starters, you werent' allowed to vote on the
thing unless you paid to join it first.  Which was PR disaster
and probably illegal but all he cared about was getting it ramrodded
through making the club really big so he could take credit for
having made this big organization rather than thinking about what
the people involved wanted.  So it started off on the worst possible
foot.  People were screaming and yelling paranoia it seemed like
paranoia at the time but it was actually pretty accurate. I have
to admit

Whereas the people like me and Ken Kaplin envisioned like this club
you could join or not. And if you didn't there would be no ill
effects whatsover. It ended up being billed by this lawyer guy and
some of his cronies as this thing that will run Fidonet for you.
So all hell broke loose. Then what happened is shit attracts flies.
More shit attracts more flies I started to get disillusioned with
this thing. I forget when I dropped out of it.  But I dropped out
of it pretty quickly. This is not a good idea.  We should drop it.
Ken Kaplin and a bunch of other people said no no we just need to
fix some of these bad problems. and with the nieve optimism of
trying to go ahead and do some of this stuff 

***************
begining of tape 2 <labled tape 1>

Marge Hey Tom if you want the tape stopped at any time just say
so.

Tom I was a little nieve.  To put it mildly. So anyways the IFNA
thing became very distasteful, I dropped out. I had written to this
other lawyer named Thomas Marshall scumbag you can put that on tape

Marge <laughter> Don't worry its on there.

Tom Oh my God what a pig, what a horrible well anyway what he had
done was gotten involved in IFNA so fast with dollar signs in his
eyes over  this idea of mandatory membership.  You know to control
to operate fidonet. He saw Fidonet as a gigantic commercial
possibility which is something I was 100% sensitive to the whole
time. And still am, though I'm not too much worried now. He wanted
to make everybody pay money to be in the net. The next year a
terrible terrible thing happened in Virginia. I think it was the
next year. He got involved and he started playing off all sides of
all fences he was my lawyer for trademark stuff and ripped me off
royally. He was Thom Henderson's lawyer over trademark issues and
he was IFNA's lawyer. He was running deals for all of us. And that's
all fine when everybody's friends. You know, you just sort of do
each other favors.  But he was working it hard. He was a real piece
of shit. What his was doing was probably illegal as well. He charged
me three thousand bucks for trademarks and dragged on for a year
and a half and did a lousy job. He did similliar things to Henderson.
IFNA had exclusive right to use the trademarks Fido fidonet, and
the dog with the diskette which was a disaster. I mean definately
horrible things could happen. Luckily with Ken Kaplin who was pretty
altruistic and definately honest he was not a problem. But 
Marshall was definately behind the scenes, sometimes in front of
it trying to manipulate stuff. It got really really hateful.

We ended up fighting corporate politics, over the control of the
nodelist <laughter> in Fidonet.  It got really bad.  We got a lawyer,
they got a bunch of lawyers actually. We had a little secret cabal.
No one knows who it was.  And I'm not going to tell you any more.
I love conspiracy. But its a benign conspiracy. Really. 

Marge   Ok

Tom We're on our side.

Marge <laughter> and that automaticly  makes you the good guys?

Tom Of course. Doesn't it with everyone else? <laughter>

Marge should I play this tape for the entire net and let them
vote on it?  Actually probably very few people are going to hear
this tape.

Tom I'm not worried. You know everybody's got people they work
with and so that's that.

Marge No the point is anything on this tape I publicize I'll
clear with you first.

Tom <laughter> Ah ok well thanks. <more laughter> We can talk
more about that later.

Marge Well that's only fair.  That way If you say something you want
reworded we can do that too.  there's probably only 4 or 5
people I'll play the tape for because  they're helping me write
the history.

Tom yeah well that's fine.

Marge Ok

Tom I forget when IFNA died. I guess it was 89.

Marge it was fall of 89.

Tom yeah 89 in San Jose actually. It was a fidocon that was the
final the final legal stuff got done. Oh wait a second though there
was another there was another event that I attended in Oregon either
before or after the 89 Fidocon Um how does it go? Luckily these
things are dated. I can find them out. I'll endure it for the
moment.Its a minor point, not a critical one.

Marge yeah we can work on the timeline later.

Tom yeah. the IFNA debacle.  Tom Marshall had taken me aside in 88
in Virginia at the fidocon. oh my god it was really gross. They
had this big dinner, Ken Kaplin went for these Fidocons all out
big convention hotels you know 80 bucks a night and this heart-attack
food. You know these big buffets with potatoes and meat and canned
vegetables and all that kind of crap So we get into this thing and
its all fido people, just a bunch of goofball nerd computer
programmers and you know its not an umm classy crowd.  I like it.
Its good old solid middle America, you know.

Marge Uhu

Tom working class stiffs. Here we go into this you know this
cheesball ritzy hotel with polished brassplated steel tubing and
um we go into this big room. And there's the IFNA board up on a
raised platform with their little table seperated from everyone
else. And boy was that the wrong thing to do with this crowd.
<laughter>  Oh boy people were pissed off. It was great though.
They didn't even see their own folly. And they just you know Fidonet
is relatively speaking lowbrow. I mean its not pretentious.  Its
not a professional organization with a capitol P profession.

Marge Uhu

Tom People don't wear suits usually. They're just you know for
the most part quasy apolitical working class middle class nerds.
This was just a bad thing to do to them because it said we're
seperate from you, we're better than you. We're on a raised
platform. And some people who were given seats up on the raised
platform like Hank Wever from Holland and myself and Wynn Wagner
who actually sat up there for a whle than came back down made
it very plain we were not going to take part in this farce. So they
all did the usual dinner stuff of having speakers and clapping of
hands patting everybody else patting everybody on the back. They
gave me this little plastic plaque and said make a little speech
and I got up and said oh my god it was horrible, I was terribly
embarrised to speak I was embarrised at what I said, but I said If
anybody goes along with this IFNA stuff you deserve what you get.
its not supposed to be centralized. Its supposed to be grassroots
what do you think you're doing here? This whole thing is a farce
and an embarrissment and no clapping apalled silence red faces,
and Ken Kaplin said Tom well that about wraps it up

Marge <laughter>

Tom oh it was so much fun. It was great. That got written up in
OMNI magazine Of all places. That was fun.

Marge it sounds like you shocked them.

Tom Oh I was furious. I was absolutely furious. I was just that
livid. Thomas Marshall had taken me aside before the dinner. That
was the scumbag lawyer guy and said um we're going to do this thing
and you know we're going to control Fidonet and develop it
ocmmercially.  People were saying will what if the Acme Corporation
had a bulletin board and they were part of Fidonet.  Are they
commercial or not? Most of them were run by an employee who manages
to scam a phone line and computer and they soft of nominally support
the company's products or not and well its the Acme Company's
bulletin board, its really just another bulletin board. It has
public callers and files and you know like a little bit of data
sheets on the products, but mostly its just a bulletin board.  I
personally don't call those commercial bulletin boards.  And most
people at that time didn't. But that was the kind of stuff that
was being decided.

Well Tom Marshall said we want to charge anybody associated with
any business $1800 a year to be in the nodelist. And if you look
go down the list, you'll find that about half of them in some way
are associated with a business like the sysop's a consultant or is
an employee of a company like the Acme Corporation running a bulletin
board sort of on the sly.  In otherwords he saw a huge pile of
money here. And I had the last laugh a few years later, me and a
bunch of other people and a whole bunch of legal fees later. So um
yeah that was the begining of the end.

The next year was San Jose I think and before or after that was
this thing in Portland which was, I think, region 16 con well it
was pretty civilized. It was nice and friendly. It was in sort
of a small-c christian retreat in the outskirts of Portland. Really
nice place, really great people. And you know we sat around doing
tall tale telling. And sitting around telling some history as all
social organizations do. There was more machinations there, phone
calling, conspiring. We actually had to work out another 
lawyer's letter to stomp on these guys, basicly threaten them with
a lawsuit if they didn't desist. I had withdrawn permission to use
the trademarks from IFNA at this point. And they continued to anyway
which was a good sign because it was the begining of the end I mean
obviously they were running out of stuff.

And it finally got killed entirely I guess it was 89. When I wrote
letters to everybody saying you can not use the trademarks change
your name immediately you know in 30 days.  They started a proxy
war. We got more votes than they did at the last minute.

Marge Ok question Fabian Gordon, in net 107 had been accused of
infiltrating IFNA and killing it from within. Is there any truth
to those accusations?  Fabian Gordon he was NC of net 107 back
then. Deathnet.

Tom   Oh yeah Well a lot of us did exactly that thing. Of course
he did. So did I. They were doin what I think was very clearly a
very evil controling thing and we used the same scabby weapons they
did to kill it. what they did was wrest control of the nodelist
from the fidonet. they had exclusive control over the trademarks
and then by this meeting in 89 they walked in armed with proxies
literally under their belt. You know this one person walked in with
4 or 5 proxies. And were going to change the bylaws so that they
could change the bylaws anytime they wanted to without requiring
a vote of the membership.

Marge  Oh geeze, that's dumb.

Tom yeah well That's what they were about to do. So what we did,
we caught wind of this we got ourselves with faxes and phone calls
we got us like one more proxy than they did. We appointed me
president or something and pulled a coup on them. Then to dissolve
this ugly mess we basicly got them to agree to put it to a vote of
the fidonet nodelist at large.  We took a copy of the nodelist and
froze it then said ok this is the voting list. Right? And we
intentionally worded it to require a positive acknowledgement of
IFNA. not a default yes, a default no.  I forget how the wording
was. It was something to the effect of if you want IFNA to exist
and run Fidionet send in a yes vote.  Of course no one will do
that. And they didn't.  And the IFNA dropped dead.

Then the board voted itself to dissolve.  The legal disolution
might have happened in 91 they were quite dead after that vote.
They had no credibility left whatsoever. And there was a lot of
hard feelings about it. Um There had been fallings out among people
like Ben Baker and Ken Kaplin was totally burned out by this. 
ALl those who just loved money and did not give a shit, they would
say anything and they would do or say anything to appease the
anti-commercial people. They thought I was a total fool. I don't
really care about that. they just wanted to make a lot of goddamed
money. And we were saying fine not here.  And yes we intentionally
attacked IFNA and killed it.  As far as I was concerned it was for
the better.  I mean it was inarguably better.  unless you were just
trying to make money.  <Laughter>

Marge Ok, I can see that.  Now let's drop politics and go back to
history. I'm very curious about how echomail as opposed to netmail
got started. And where in hell did the ECs come from?

Tom um  <laughter>

Marge and that part of the tape I'm definately not going to play
to a few people.

Tom up from the depths of hell. <laughter> 

Marge now watch it. Some of my best friends are ECs.

Tom well same here but some of the bad ones are really bad. 

Marge I know

Tom Mostly they're completely cool. they just don't understand what
they're messing with. They don't understand that they are obligated
to transmitt people's mail is the thing they don't get.  

Marge But in the begining wasn't it just mostly netmail?

Tom yes, it was only netmail and what happened was Jeff Rush AKA
the shadow one of the millions of shadows I'm sure. I mean how many
shadows are there in the world?

Marge well I've got two of them in an echo I moderate and it can
be a lot of fun.

Tom Well anyway Jeff rush was in the dallas area they were basicly
having like these pizza based meetings,  arranging thins over
netmail using crude mailing list things.

Marge Uhu ok

Tom  Well it was barely adequate. So he wrote scan and toss as were
two seperate programs. It was not called echomail yet. It had lists of
what systems to send a message to, and used netmail as the transport
mechanism.  It was fairly basic but it was definately an interesting
use of Fidonet.  Well what happened was um it spread like wildfire
because it allowed mailing lists for the first time, basicly. And
then it killed Fidonet. The traffic was just enormous.  It was just
beyound the ability of Fidonet. Um because everything was writtin
for occasional point to point mail you know it wasn't really meant
for these huge floods. File attaches were working at this point.
Then somebody took Henderson's ARC, it could have been Rush it
could have been Henderson, or somebody else entirely,ran ARC over
toss/scans files then caused them to be delivered as a file-attach.
This saved all the overhead of the FidoNet bundling.  the whole
world breathed a sigh of relief

and now suddenly with all these files around you had to coordinate
a little more highly than you do for just plain messages because
the unpackers in Fidonet would just take care of the messages
whereas you had to detect the presence of a an echomail
bundle, unARC it, copy the messages and all that kind of junk. So
a bunch of procedures and tools sprung up around what's now known
as echomail. I don't know when it got called echomail.  
But very shortly thereafter plain old scan/toss got banned banished
cause it was murder on Fidonet.   There's still vestigial pieces
of it left in confmail and zmail and all that stuff where you could
allow or prevent inbound fidonet point to point mail to become or
come from echomail.  Its still useful. I mean if you have small
systems that let people send messages and have it appear in echomail.
Especially in small and private nets. 
So it needless to say utterly revolutionized Fidonet. The point to
point mail was all well and fine but in fact the only people who
could use it were sysops because you had to have access to that
rather internal overhead message area in the fidonet and 2 because
it required relatively speaking sophisticated knowledge of
network topology was. You had to understand about delivery, you
had to understand about other nodes and nets and all that kind of
crap.  With echomail you don't have to know anything.  
you log into a local bulletin board you don't even have to figure
out for 6 months how the whole thing works. You can take part in
a conversation and that's what made fidonet useful.

Marge I don't know if my users are dumber than the average or not,
but some of them never figure out that when they're talking in the
echos they're talking to people all over the country.

Tom yeah and the thing is it doesn't matter. They think gee there's
an awful lot of people calling from Iowa here. That's all fine.
Its like when you walk in a room and turn on a light switch I mean
you don't want to have to sit through a lecture on basic electricity
to do that. You want to flip the damn switch and the light go on.

Marge Right!

Tom Am echomail was good at that. The good thing is more people
have access. The bad part is what's going on gets lost and twisted.
And in the case of light switches its not a problem. In the case
of echomail it gets abused and it still should require some 
knowledge of what's going on 

Marge it seems like this might be a good stopping point. How about
I call you sometime next week after I've talked to a few other
people and we'll resume?

Tom oh sure

Marge OK thanks a lot

Tom that's a good idea.

Marge I sure appreciate your help with this project Tom.

Tom That's ok. So good luck with it. So what are you going to do
with it?

Marge I haven't decided yet. We're having a region 14 con in a
couple of weeks and I'm going to take what I've got to the meeting
and discuss it with the guys who are there. Right now I'm just
collecting material.

Tom yeah ok

Marge so I may or may not play the tape. It depends on how many
people show up.  But that's as far as it goes. Anything that gets
into Fidonews or gets published I'll clear with you first.

Tom Ok, I do have some files just scattered at random. I was
going to mail them to you I know I promised, but I haven't.

Marge well if you can give me instructions on how to fTP there I
can have JJ pick them up.

Tom yeah well right now they're just buried in a bunch of
sub-directories. I've got to go dig them out and see what's there's
some stuff that's private in there. There's some stuff that's not.
I just have to find the 1/2 dozen files that I just don't want
they're like oh embarrissing rants between various people and
personal matters that I don't want..they're just all mixed together.

Marge Ok

Tom then you can have the whole bag.

Marge ok, great.  Just let me know how to get there and one of us
will pick them up.  I'm working on getting an internet account one
of these days.

Tom oh ok yeah we'll stay in touch.

Marge ok thanks a lot Tom.

Tom ok bye

Marge Bye