WEBVTT
00:00:00.000 --> 00:00:13.270
34C3 preroll music
00:00:13.270 --> 00:00:17.939
Herald: Back in time, back to the 1990's
where the internet actually made sounds,
00:00:17.939 --> 00:00:22.922
and you could annoy the whole family while
blocking the phone line. He was actually
00:00:22.922 --> 00:00:28.230
heavily involved in these early days of
the internet. He operated and participated
00:00:28.230 --> 00:00:33.713
in these early structures, namely Bulletin
Board Systems and the UseNet. And he now
00:00:33.713 --> 00:00:38.965
takes us back in time, to tell us all
about the time when the internet made
00:00:38.965 --> 00:00:40.951
sounds. Thank you very much LaForge.
00:00:40.951 --> 00:00:47.985
applause
00:00:47.985 --> 00:00:53.437
LaForge: Thank you very much for the
introduction.This is a quite unusual
00:00:53.437 --> 00:01:01.405
setting for me. Typically I give talks
about deeply technical topics. Protocoll
00:01:01.405 --> 00:01:06.739
level details and telecom specs and so on.
Now the first time I speak in the Art and
00:01:06.739 --> 00:01:11.484
Culture track. That is definitely
something new for me. So, why am I here
00:01:11.484 --> 00:01:17.559
and why am I talking about this topic.
First of all, I was involved to some
00:01:17.559 --> 00:01:25.343
extent yes, but for sure I was not
somebody who had any significant role in
00:01:25.343 --> 00:01:31.524
that universe. Neither in the BBS scene or
in the early internet days. I was just
00:01:31.524 --> 00:01:38.360
basically a youngster, a teenager, who had
fun playing with technology and was
00:01:38.360 --> 00:01:43.309
helping others to communicate using
technology. There are many more people who
00:01:43.309 --> 00:01:47.428
have, who are much more qualified than me
to talk about that subject but I ... and
00:01:47.428 --> 00:01:51.920
that's the reason why I'm here and why I
submitted this talk is, you don't really
00:01:51.920 --> 00:01:56.346
see many people speaking about these days
or about those topics anymore. And even if
00:01:56.346 --> 00:02:00.490
you want to research it, I think there's
like one or two books in German on that
00:02:00.490 --> 00:02:07.559
subject, they're very hard to get and also
not very complete. So, I think we have to
00:02:07.559 --> 00:02:15.490
sort of document the history of it for
those people, who have not been around at
00:02:15.490 --> 00:02:19.598
the time. So, this talk will not have as
many acronyms as you are used to from
00:02:19.598 --> 00:02:24.720
talks that I usually give. Still you have
typos in the slides, as you can see in the
00:02:24.720 --> 00:02:28.450
second line already, so that didn't
change. I didn't invent any of the
00:02:28.450 --> 00:02:32.279
technologies covered here. I didn't write
any of the software covered. I was just a
00:02:32.279 --> 00:02:40.390
user and operator or sysadmin. And that's
the world I grew up in from 11 onwards.
00:02:40.390 --> 00:02:46.194
As I said many people lack that history and
to start with that, maybe a quick poll in
00:02:46.194 --> 00:02:51.718
the audience. Who has ever dialed into a
BBS using a modem? Raise your hands. Okay.
00:02:51.718 --> 00:02:58.890
So, I'm preaching to the converted. Okay,
maybe I should invite all of you up to the
00:02:58.890 --> 00:03:05.800
stage and we should make a discussion-
round instead. Anyway. So, circuit switch
00:03:05.800 --> 00:03:10.900
telephony. Well, this is the telephony
from 1876 until about 1988 with
00:03:10.900 --> 00:03:15.416
analog voice circuits over copper wires
and dial-up connections between A and B.
00:03:15.416 --> 00:03:19.185
I guess everybody still remembers these.
Even if you're young, you should have seen
00:03:19.185 --> 00:03:26.780
a classic telephone, I think. And yeah,
you have analog amplifiers possibly in the
00:03:26.780 --> 00:03:31.404
path, but actually the copper wires are
physically switched at telephone
00:03:31.404 --> 00:03:36.486
exchanges. So, this structure looks a bit
like this: We have a telephone at one end,
00:03:36.486 --> 00:03:41.133
we have a telephone at another end, and we
have telephone exchanges or switches,
00:03:41.133 --> 00:03:45.377
which actually switch the circuit - hence
the term circuit switched telephony -
00:03:45.377 --> 00:03:52.260
between A and B. So, you have a copper
wire from your phone to the office, the
00:03:52.260 --> 00:03:56.747
exchange, to which you are connected and
then that exchange again has copper wires
00:03:56.747 --> 00:04:02.735
to other exchanges and so on. And based on
the phone number you dial the call is
00:04:02.735 --> 00:04:08.316
switched to the destination subscriber.
That's sort of the foundation in terms of
00:04:08.316 --> 00:04:13.914
technology that we're using here. Also
something to document for the
00:04:13.914 --> 00:04:19.000
international audience in Germany at that
time even local calls were metered and
00:04:19.000 --> 00:04:24.942
charged by the minute, flat rates didn't
exist and we had multiple zones so there's
00:04:24.942 --> 00:04:29.219
not just local calls and long-distance
calls but different depending on your
00:04:29.219 --> 00:04:34.780
distance so like up to 50 kilometers or
more than 50 kilometers and so on. And
00:04:34.780 --> 00:04:39.400
given on that and the steep pricing and
not so many people could afford long-
00:04:39.400 --> 00:04:45.308
distance BBSing at least not for a long
time. All of this started with a device
00:04:45.308 --> 00:04:51.156
called the acoustic coupler. It's actually
also how I started even though I'm young
00:04:51.156 --> 00:04:57.946
and I only started in I think about 90 or
91. At 10 or 11 years of age you don't
00:04:57.946 --> 00:05:01.490
have the latest and greatest in
technology. I got a used second hand or
00:05:01.490 --> 00:05:07.331
third hand Olivetti acoustic coupler from
my uncle it had even a battery it could be
00:05:07.331 --> 00:05:12.090
operated mobile it had a battery
compartment with eight Mignon (AA) cells.
00:05:12.090 --> 00:05:15.840
Actually I still own it and I still own
related telephone I just thought: yeah
00:05:15.840 --> 00:05:21.311
don't have to bring it here. But it
still exists. So anyway, here you have to
00:05:21.311 --> 00:05:25.616
dial using your normal phone. You dial the
digits of the phone number and once the
00:05:25.616 --> 00:05:30.859
other side picks up the phone and they put
their receiver onto the acoustic coupler
00:05:30.859 --> 00:05:35.328
and you put your receiver onto the
acoustic coupler, then data can be
00:05:35.328 --> 00:05:40.010
transmitted over the telephone line as
said with manual, dial manual pickup and
00:05:40.010 --> 00:05:49.416
rather extremely low speed. This all looks
like this and the next step in the logical
00:05:49.416 --> 00:05:55.402
progression then was modems, which is sort of you can
think of an automatized method of acoustic
00:05:55.402 --> 00:06:01.109
couplers, where you don't have an air gap
anymore. So in the acoustic coupler you
00:06:01.109 --> 00:06:05.541
literally have a couple of centimeters of
air between the speaker and the microphone
00:06:05.541 --> 00:06:10.016
in the receiver of your phone, versus the
acoustic coupler. So with the modem
00:06:10.016 --> 00:06:16.411
there's a direct connection and also you
have automatic facilities to dial the
00:06:16.411 --> 00:06:20.282
telephone number and to answer the line
and so on. So you don't need a manual
00:06:20.282 --> 00:06:26.540
operator anymore to pick up a phone or
dial numbers. And this thing gets
00:06:26.540 --> 00:06:31.251
transmitted over the telephone
line. This is a stack of various different
00:06:31.251 --> 00:06:35.804
modems – we will see some others here,
some of you will remember the brands or
00:06:35.804 --> 00:06:41.672
the shapes or even the specific models of
those modems. But that's too much level of
00:06:41.672 --> 00:06:50.199
detail for the moment. So let's look a bit
at the speed, or lack of speed, that was
00:06:50.199 --> 00:06:57.999
available. It started with 300 bps. I
actually used 300 bps a couple of times.
00:06:57.999 --> 00:07:05.405
In fact, in like around 1990 of course it
was extremely slow but still it was what I
00:07:05.405 --> 00:07:12.464
could start with at the time. Then the
1200 bps; so this is still rather slow and
00:07:12.464 --> 00:07:16.366
you can slowly read and follow the
text as it's being printed. Unfortunately
00:07:16.366 --> 00:07:20.930
I don't have an animation or something
like that. I'm not such a multimedia savvy
00:07:20.930 --> 00:07:27.411
guy. So yes, then the speeds progressed,
you see the years in which they were
00:07:27.411 --> 00:07:34.769
created. The lines with the asterisk mark
years that I found some secondary sources
00:07:34.769 --> 00:07:38.349
that originally it had been specified
then. But actually the oldest spec
00:07:38.349 --> 00:07:42.464
document for all these earlier ones was
from 1988. So if you go to the ITU
00:07:42.464 --> 00:07:46.676
website, the earliest documents you can
find are from 1988 and none of those
00:07:46.676 --> 00:07:50.163
earlier documents could – at least on the
internet – be found anywhere. Maybe you
00:07:50.163 --> 00:07:55.449
can go to a library or something like
that. Yeah so speeds progressed, different
00:07:55.449 --> 00:08:00.140
modulation schemes were introduced to
squeeze ever more bits into these
00:08:00.140 --> 00:08:07.899
3 kilohertz analog circuit over the
telephone line. And every couple of years
00:08:07.899 --> 00:08:14.460
a new, especially in the 90s, if you
follow this 91 14.400 bps, 93 19.200
00:08:14.460 --> 00:08:21.201
to 1994 28.000 bits per second. And there
were of course also proprietary protocols,
00:08:21.201 --> 00:08:26.861
then you had to have the same manufacturer
of modem that the other side whom you're
00:08:26.861 --> 00:08:32.134
calling and so on, but these are the
official standardized protocols and speeds
00:08:32.134 --> 00:08:38.268
that were used. Which brings us... so okay
we have a telephone system; we can dial
00:08:38.268 --> 00:08:43.287
numbers; we have a modem that can dial
numbers; we have modems that can send bits
00:08:43.287 --> 00:08:50.483
in exceptionally fast speed. What do we do
with this? And this brings us to be BBSs:
00:08:50.483 --> 00:08:55.445
where could you actually dial, and what
could you do there? So what's the BBS?
00:08:55.445 --> 00:09:01.257
Fundamentally, it's some computer – any
hardware, any operating system, any
00:09:01.257 --> 00:09:06.570
software. Some computer that accepts
incoming calls attached to a modem and
00:09:06.570 --> 00:09:11.334
offers some kind of interactive service to
the people who dial into that BBS. And if
00:09:11.334 --> 00:09:16.165
you wanted to operate a BBS, you had to
have a separate dedicated computer for
00:09:16.165 --> 00:09:21.369
that. Because, at the time, most of the
BBS software – and most of the software
00:09:21.369 --> 00:09:25.998
that people used in general – predated
multitasking operating systems. So when
00:09:25.998 --> 00:09:30.209
you ran the BBS, the computer was busy
running the BBS; you couldn't do anything
00:09:30.209 --> 00:09:35.659
else at the same time. So you had to
invest quite a bit into a separate second
00:09:35.659 --> 00:09:40.850
computer, or third or fourth, to actually
operate that BBS. You had to have a
00:09:40.850 --> 00:09:45.505
separate telephone line. Because if you
operate the BBS into which people dial
00:09:45.505 --> 00:09:49.997
into, of course any time of the day or
night people will dial in there, so you
00:09:49.997 --> 00:09:53.304
cannot use your normal phone line that you
use to make phone calls but you had to
00:09:53.304 --> 00:09:58.654
have a separate dedicated phone line. And
of course the system had to run more or
00:09:58.654 --> 00:10:03.772
less 24/7 so people could dial in and
reach it. Luckily, on the user side there
00:10:03.772 --> 00:10:10.649
was not so many requirements in terms of
technology that you needed. Your computer
00:10:10.649 --> 00:10:15.734
of course you only power when you use it,
and you can share the regular phone line –
00:10:15.734 --> 00:10:20.186
with the side effect, as in the
introduction has been mentioned, that your
00:10:20.186 --> 00:10:26.511
family might have gone angry if you
occupied it too long – but otherwise no
00:10:26.511 --> 00:10:31.754
additional infrastructure other than a
modem required. Now you dial into the BBS
00:10:31.754 --> 00:10:39.249
– what kind of content do you get? What do
you do in that BBS? And the name BBS in
00:10:39.249 --> 00:10:45.560
English is a Bulletin Board Service,
that's actually the acronym expansion. So
00:10:45.560 --> 00:10:49.869
there were Bulletin Boards, message boards
where you could exchange messages and
00:10:49.869 --> 00:10:55.352
texts with other people, other users of
that BBS or the so-called sysop, the
00:10:55.352 --> 00:11:01.007
system operator, the guy running that BBS.
You could also chat with the system
00:11:01.007 --> 00:11:07.106
operator, which, well, didn't exist before
– the ability to chat with somebody else
00:11:07.106 --> 00:11:14.174
remotely over a text-based terminal. There
were also multi-user games, text-based, as
00:11:14.174 --> 00:11:19.021
well as so called file areas where you
could download files. And downloading
00:11:19.021 --> 00:11:26.053
files, given the speeds back then and so
on and so on, of course it was primarily
00:11:26.053 --> 00:11:31.375
text documents or small programs or
something like that. Mp3 didn't exist of
00:11:31.375 --> 00:11:37.200
course, at least until 95 or whenever it
came out, so maybe some mod files for your
00:11:37.200 --> 00:11:42.386
module tracker, something like that. And
of course, last but not least, ASCII and
00:11:42.386 --> 00:11:48.237
ANSI artwork, which basically is an entire
subculture and scene and community in
00:11:48.237 --> 00:11:56.618
itself, creating artworks and drawings
using the character set that was used by
00:11:56.618 --> 00:12:04.386
ANSI.sys, which was the DOS, you could say
display driver, in quotes, in a certain
00:12:04.386 --> 00:12:09.796
character set and you could draw graphics
like this. We will see some more. And
00:12:09.796 --> 00:12:15.722
people were putting a lot of effort into
this, and sort of competing who could who
00:12:15.722 --> 00:12:21.589
could make the best representation or the
most expressive artwork given the limited
00:12:21.589 --> 00:12:29.521
resolution and the limited characters and
colors available in this domain. So, what
00:12:29.521 --> 00:12:33.825
kind of software did one use? Or what kind
of technology was used? Well, we already
00:12:33.825 --> 00:12:39.573
had the computer and modem, you needed
some software. So on the BBS side, BBS
00:12:39.573 --> 00:12:46.024
software, there's an unlimited number of
different BBS software programmes, and
00:12:46.024 --> 00:12:50.663
extensions, and modifications thereof, a
lot of them are freeware or shareware.
00:12:50.663 --> 00:12:55.785
Some of them are public domain, some
actual free software, some are
00:12:55.785 --> 00:13:00.741
proprietary. For any operating system, for
any computer architecture, people were
00:13:00.741 --> 00:13:09.143
writing BBS software. Whether you had an
Amiga or Atari or you had Apple or DOS PCs
00:13:09.143 --> 00:13:16.883
or you name it, software was written, by
hobbyists primarily. One concept that you
00:13:16.883 --> 00:13:21.861
will find in BBSs is the concept of so-
called doors. You can think of it as
00:13:21.861 --> 00:13:27.713
similar to CGIs in web. So basically, the
BBS software could call an external
00:13:27.713 --> 00:13:32.639
programme, which would then take over the
input and output to and from the user. So
00:13:32.639 --> 00:13:36.668
you could have sort of plugins to your BBS
software which would add additional new
00:13:36.668 --> 00:13:44.865
games or add chat software or messaging or
whatever. On the user side you had a
00:13:44.865 --> 00:13:50.897
primarily so-called terminal program. It's
called terminal program because actually
00:13:50.897 --> 00:13:57.399
it emulates a serial terminal, which is a
dedicated hardware device with a keyboard
00:13:57.399 --> 00:14:02.397
and a screen and a serial line, but not a
general-purpose computer and in order to
00:14:02.397 --> 00:14:07.448
make a general-purpose computer behave
like a terminal you had a terminal program
00:14:07.448 --> 00:14:12.982
on dos which I was using at the time. It's
primarily telex and telemate I think were
00:14:12.982 --> 00:14:20.668
the favorite ones at least on this side of
the planet and you started that program,
00:14:20.668 --> 00:14:25.333
you had a serial port, the serial port
attached to your modem and from there you
00:14:25.333 --> 00:14:29.350
dialed and the terminal program then was
responsible for displaying the texts and
00:14:29.350 --> 00:14:34.932
the ANSI graphics and so on and exchanging
files of a variety of different protocols,
00:14:34.932 --> 00:14:42.419
which we will also cover later but before
we go on let's do a quick demo of how this
00:14:42.419 --> 00:14:48.809
looks like. Now as a note I don't have a
modem here I'm not emulating a modem I'm
00:14:48.809 --> 00:14:54.507
not emulating a serial port, these days
you can get the same experience by using
00:14:54.507 --> 00:14:59.466
telnet over the internet but you can
actually telnet into BBSs, I just want to
00:14:59.466 --> 00:15:10.488
basically show how it looks like. So this
is the terminal program and we have now
00:15:10.488 --> 00:15:15.758
connected to the BBS this is sort of a
introductory graphic that we see before
00:15:15.758 --> 00:15:22.465
even logging into the the box yeah some...
of course the scrolling was much slower
00:15:22.465 --> 00:15:29.655
back then, so now we can scroll back up to
actually see what was there. Yes, some
00:15:29.655 --> 00:15:33.893
more graphics. You still haven't seen the
login prompt yet, as you can see a fairly
00:15:33.893 --> 00:15:39.919
graphics heavy BBS. Then you can choose
the theme of the BBS, a user interface,
00:15:39.919 --> 00:15:48.617
I'm going to go for the classic ANSI here.
Finally, I come to a login screen and I
00:15:48.617 --> 00:15:57.915
can log into the system where I have to
enter my handle and the password which is
00:15:57.915 --> 00:16:03.774
now in clear-text over telnet. For those
of you interested in this, not that
00:16:03.774 --> 00:16:07.380
there's anything useful I just registered
this morning at the BBS so there's nothing
00:16:07.380 --> 00:16:13.637
associated with this account. Yeah some
more graphics. Finally, we are at a
00:16:13.637 --> 00:16:19.862
message board and we see as I said I just
logged in or registered this BBS today. We
00:16:19.862 --> 00:16:26.504
see there is a message number one from
Hawk Hubbard, "Welcome", so if I want to
00:16:26.504 --> 00:16:31.176
look at that message I could basically say
"I want to read it now". This is the
00:16:31.176 --> 00:16:36.284
message reader I go in here, then here,
"Welcome to forge" and so on.. So he
00:16:36.284 --> 00:16:45.242
welcomes me to the BBS now let's go to the
main menu of the BBS, which in this case
00:16:45.242 --> 00:16:53.704
looks like that and you have different...
the file areas, where you can download
00:16:53.704 --> 00:16:57.760
files, you have the door games that I
mentioned, you have an ANSI gallery, a BBS
00:16:57.760 --> 00:17:02.878
list, you can look at the last callers who
has called this mailbox and you can see
00:17:02.878 --> 00:17:08.443
this... Well, yeah three test calls from
me this morning, but you can see actually
00:17:08.443 --> 00:17:14.429
other people are still logging into this
BBS and it's 2017, so it's not... to me
00:17:14.429 --> 00:17:18.491
this is mostly history but during the
preparation of this talk I discovered that
00:17:18.491 --> 00:17:22.502
some people, for some people it is still
the present and I'm very happy to see
00:17:22.502 --> 00:17:28.076
there's still such an active community
around BBSs and which enables me to show
00:17:28.076 --> 00:17:33.963
all of this without firing up some
emulators and so on. So yeah, we also can
00:17:33.963 --> 00:17:40.095
look at one-liners, here's some messages
that people can leave to other people,
00:17:40.095 --> 00:17:46.314
other users in the BBS, again with some
quite a graphical... We don't want to
00:17:46.314 --> 00:17:52.265
leave any additional words here, but what for
example we can look at the ANSI gallery
00:17:52.265 --> 00:17:58.325
just very quickly, can try to select
something here, I have no idea what I'm
00:17:58.325 --> 00:18:06.980
looking at so... Ok... so here you have a
sort of a viewer that, yeah... So it will
00:18:06.980 --> 00:18:17.140
show you the sections of a sort of longer
artwork in this particular case... Yeah...
00:18:17.140 --> 00:18:25.454
well... And the artwork... to me there
always was a lot of similarity between the
00:18:25.454 --> 00:18:36.604
sort of, between the ANSI art artists and
the people doing... Now I'm lacking the
00:18:36.604 --> 00:18:42.829
word, street art basically I think there's
a lot of similarity between that. Okay
00:18:42.829 --> 00:18:46.479
good, that was just a very quick demo of
course I could now look at more messages
00:18:46.479 --> 00:18:52.618
and write messages and play blackjack and
do whatever I want, which I don't in this
00:18:52.618 --> 00:18:58.512
case, so we will log off. And again some
more graphics and you can leave a comment
00:18:58.512 --> 00:19:05.112
to the sysop if you want or you can just
basically... Log of... Ok, that for a very
00:19:05.112 --> 00:19:10.240
quick demo of the look and feel. Now since
I'm such a technical person and looking at
00:19:10.240 --> 00:19:15.288
protocol stacks, I tried to draw a
protocol stack diagram for BBSs, which
00:19:15.288 --> 00:19:20.036
ended up at this. So basically at the
lower layers we have the pots, the plain
00:19:20.036 --> 00:19:25.525
old telephony system or ISDN, which we
will get to in a few slides. We had modems
00:19:25.525 --> 00:19:30.440
on the analogue telephone system, we had
other things on ISDN. In the end at some
00:19:30.440 --> 00:19:35.800
point you always have rs-232, a serial
port, either emulated or real, and then
00:19:35.800 --> 00:19:40.670
either you had a terminal program directly
on top of that or, for example to transfer
00:19:40.670 --> 00:19:46.689
files, you have used X modem or Y modem or
Z modem, which added error correction and
00:19:46.689 --> 00:19:50.408
retransmission and block transmission so
you could safely transfer files without,
00:19:50.408 --> 00:19:58.365
or at least with less, corruption. The
checksum algorithms were not so scientific
00:19:58.365 --> 00:20:05.554
in many cases. Here we then have well some
other things, FTN, Point what does that,
00:20:05.554 --> 00:20:10.030
UUCP we will cover that later. Basically
you could run different protocols and
00:20:10.030 --> 00:20:14.500
different systems on top of that. One
curiosity that I still want to mention is
00:20:14.500 --> 00:20:20.200
that, which I actually I forgot until on
Twitter somebody reminded me a couple of
00:20:20.200 --> 00:20:25.599
days ago that this existed, and I went "oh
yes, RIPterm, I used that quite some time
00:20:25.599 --> 00:20:32.000
ago", so instead of having these text-
based user interfaces some people, company
00:20:32.000 --> 00:20:38.147
called TeleGrafix came up with a language
called RIPscript which was a fairly
00:20:38.147 --> 00:20:45.080
compact language of textual commands, by
which the BTS could control a vector
00:20:45.080 --> 00:20:49.142
graphic renderer on the client side in
your terminal program, and you could
00:20:49.142 --> 00:20:54.502
actually draw VGA resolution graphics like
the one that's presented here on the slide
00:20:54.502 --> 00:21:00.770
from the VBS on the screen of the user,
which was quite a big change compared to
00:21:00.770 --> 00:21:08.522
the ASCII art or ANSI art that you've seen
before. Yeah, so we're still at BBSs and
00:21:08.522 --> 00:21:12.599
BBSs that are isolated, so you can
participate in those bulletin boards and
00:21:12.599 --> 00:21:16.880
you can read and write messages and
exchange ideas and recipes and thoughts
00:21:16.880 --> 00:21:22.161
and cheat codes and whatever you want to
exchange. Users log in at different times,
00:21:22.161 --> 00:21:27.991
the BBS is busy if it has only a single
line while it's being used by some other
00:21:27.991 --> 00:21:32.314
user. Of course you can add as a BBS
operator, as the sysop, you can add more
00:21:32.314 --> 00:21:37.473
modems and more phone lines, which is of
course expensive, together with the multi-
00:21:37.473 --> 00:21:43.134
port serial cards and and everything that
was required. You can have time limits for
00:21:43.134 --> 00:21:49.140
each user, but in the end it's sort of,
there's a limit to how far you can scale a
00:21:49.140 --> 00:21:57.879
single BTS sort of - not a BTS, a BBS,
jeez, a single BBS... Well also there's a
00:21:57.879 --> 00:22:04.235
scalability limit for BTSs, but that's
another talk, so, yeah. Which brings us to
00:22:04.235 --> 00:22:11.750
one method of more efficiently engaging
with BBSs for exchanging messages which is
00:22:11.750 --> 00:22:18.040
a concept of points or offline message
reading. So as we have just seen in this
00:22:18.040 --> 00:22:26.416
example we log in to the VP... the BBS and we
have an online interactive session with
00:22:26.416 --> 00:22:29.904
the BBS while we read and write the
messages and of course it means we occupy
00:22:29.904 --> 00:22:36.477
the telephone line for an extended period
of time and it's not used very efficiently
00:22:36.477 --> 00:22:41.154
because humans typically read slower than
at least a fourteen point four or twenty
00:22:41.154 --> 00:22:48.372
eight kilobits per second. So people
invented something called points or
00:22:48.372 --> 00:22:52.198
offline message reading and different
concepts different systems different
00:22:52.198 --> 00:22:57.142
standards different technologies. What
they did in the end is they compressed and
00:22:57.142 --> 00:23:04.890
batched all the messages for you into
files and you on your client-side you were
00:23:04.890 --> 00:23:08.596
writing your messages offline and also
compressing and batching the messages that
00:23:08.596 --> 00:23:12.662
you've written and then you make a call,
you quickly exchange those files in both
00:23:12.662 --> 00:23:18.603
directions even in full duplex if the
system supports it and then you terminate
00:23:18.603 --> 00:23:23.185
the connection again. So during a very
short call you can exchange much more, many
00:23:23.185 --> 00:23:27.500
more messages and you have all the time to
read through those messages without having
00:23:27.500 --> 00:23:33.030
to look at the phone meter or your phone
bill all the time. So, more scalability,
00:23:33.030 --> 00:23:37.829
more users, shorter connection time, lower
cost for everyone involved. Definitely an
00:23:37.829 --> 00:23:45.314
interesting technology, but still sort of
scalability is limited of a single BTS
00:23:45.314 --> 00:23:52.334
which, eh, BBS which brings us to BBS
networks, store-and-forward networks which
00:23:52.334 --> 00:24:00.437
basically extended the ability to exchange
messages beyond a single BBS, but so
00:24:00.437 --> 00:24:04.890
basically the bulletin boards or the
message groups that you had at a BBS were
00:24:04.890 --> 00:24:09.578
replicated over different protocols that
were invented by various different people
00:24:09.578 --> 00:24:16.281
over time, so not only one BBS had all the
messages of a given bulletin board but all
00:24:16.281 --> 00:24:21.526
the other BBSs participating also were
receiving these messages and replicating
00:24:21.526 --> 00:24:28.884
them all over the network. Also for
personal mail, which is like email, right,
00:24:28.884 --> 00:24:35.444
between two participants, you could route
those messages across the network. The two
00:24:35.444 --> 00:24:40.276
users exchanging messages didn't have to
connect to the same BBS anymore. So much
00:24:40.276 --> 00:24:45.440
more scalability and also you could use it
efficiently for message routing to reduce
00:24:45.440 --> 00:24:51.660
the need for long distance calls and so
on. So let's look at a couple of these BBS
00:24:51.660 --> 00:24:58.776
networks and the technologies they used.
One large and very popular example of
00:24:58.776 --> 00:25:05.302
course is the Fido Network which consists
of two parts, net mail and echo mail.
00:25:05.302 --> 00:25:12.750
Net mail is the private personal mail and echo
mail are public message boards or message
00:25:12.750 --> 00:25:19.651
groups. Fido had some, the technology used
by Fido called FTN Fido technology
00:25:19.651 --> 00:25:24.062
networks were used also by other networks.
They were using the same protocols, but
00:25:24.062 --> 00:25:30.110
they were not the same group of BBSs or
the same content and so on. Treknet for
00:25:30.110 --> 00:25:35.880
Star Trek fans was one, Gernet in Germany
was an example for that. And there also
00:25:35.880 --> 00:25:42.368
were other technologies and other networks
such as Z-Netz, where they called it
00:25:42.368 --> 00:25:49.190
"Bretter" actually, so boards, the
individual message groups. And again they
00:25:49.190 --> 00:25:53.929
had other offsprings that used the same
technology but have different groups and
00:25:53.929 --> 00:25:59.460
different policies and different
structures such as T-Netz or CL-Netz. And
00:25:59.460 --> 00:26:08.328
then there was the big faction of people
who did UUCP, the UNIX to UNIX copy, which
00:26:08.328 --> 00:26:13.180
we will look at a little bit. And MausNet
is another german example here originating
00:26:13.180 --> 00:26:21.363
from the city of Muenster, which was used
to up to 120 BBSs here. Let's look at Fido
00:26:21.363 --> 00:26:28.717
a little bit more. Started allegedly in
1984. Of course I was not involved at that
00:26:28.717 --> 00:26:37.644
time at the age of 5. It reached a limit
of 250 nodes in 1985 because apparently, I
00:26:37.644 --> 00:26:41.911
suppose probably, a single integer UINT8
was used for the node number or something
00:26:41.911 --> 00:26:45.992
like that and then about 250 should be
sufficient for everyone. I don't know what
00:26:45.992 --> 00:26:51.239
the other 5 are for. And then they
introduced in '86 hierarchic regional
00:26:51.239 --> 00:26:57.984
routing and addressing that was more
scalable and in the end at the peak of the
00:26:57.984 --> 00:27:06.384
Fido net propagation it was 39,000 nodes;
that's BBSs not individual users but
00:27:06.384 --> 00:27:12.199
39,000 BBSs were interconnected with an
estimated 2 million users worldwide and
00:27:12.199 --> 00:27:18.882
that's for a you know hobbyist amateur
network is I think quite impressive.
00:27:18.882 --> 00:27:23.862
The addresses looked like this. That's
actually a node number that I used around
00:27:23.862 --> 00:27:33.019
'95 in Nuremberg at the time. Z-Netz
started as Zerberus-Netz - and I'm not
00:27:33.019 --> 00:27:37.340
sure if padeluun or Rena or any of the
people involved in the audience if then I
00:27:37.340 --> 00:27:45.608
hope I represent the history correctly -
which is a network technology created in
00:27:45.608 --> 00:27:50.827
Germany. The standards are inspired but
different than the Usenet and UUCP
00:27:50.827 --> 00:27:54.785
protocols and there were all kinds of
flame war about who understood the specs
00:27:54.785 --> 00:27:59.567
wrong and whether there's an improvement
between ZConnect compared to the Usenet
00:27:59.567 --> 00:28:05.855
standards or not. But anyway it was
different and there was one program called
00:28:05.855 --> 00:28:09.980
CrossPoint which was the most popular
point software at the time I think at
00:28:09.980 --> 00:28:15.739
least on DOS for Z-Netz and also for other
technologies. The screenshot here at the
00:28:15.739 --> 00:28:21.138
bottom actually is a cross point
screenshot. And cross point in the early
00:28:21.138 --> 00:28:28.015
90s already had features that I'm still
missing today in any email client that I
00:28:28.015 --> 00:28:35.824
have found. Right? Imagine you have a
thread that crosses multiple folders,
00:28:35.824 --> 00:28:42.204
multiple news groups, multiple whatever
and you have threading like the tree of
00:28:42.204 --> 00:28:46.038
the thread across folders and news groups
and so on. I mean that's something that
00:28:46.038 --> 00:28:50.397
you cannot do with any of the
software still today. Maybe you have you
00:28:50.397 --> 00:28:54.600
have an answer which software today
supports this but for sure nothing I have
00:28:54.600 --> 00:29:00.143
found has the kind of features and
functionality. Unfortunately it was
00:29:00.143 --> 00:29:06.540
written in Pascal and it had a line length
limit of 255 characters per line which
00:29:06.540 --> 00:29:10.687
made it not very compatible to Usenet
standards where lines could have different
00:29:10.687 --> 00:29:17.883
lengths so one couldn't continue to use it
in today's time and age at least not
00:29:17.883 --> 00:29:29.633
easily. Usenet is another network of these
BBS days where messages were exchanged by
00:29:29.633 --> 00:29:35.379
a system called UNIX to UNIX copy. UNIX to
UNIX copy predates the Usenet it was used,
00:29:35.379 --> 00:29:39.341
well as the name implies, to copy
something between UNIX machines - file
00:29:39.341 --> 00:29:44.016
copying - and some of those files that
people were copying were internet mail at
00:29:44.016 --> 00:29:51.123
the time. And then the Usenet news format
was invented. The format is quite similar
00:29:51.123 --> 00:29:55.730
to internet mail, which we still know
today, but it's not a personal mail
00:29:55.730 --> 00:30:00.805
between person A and person B, but it, you
could post it to a so-called news group
00:30:00.805 --> 00:30:06.099
and there was a hierarchy of news groups
which replicated and flooded messages
00:30:06.099 --> 00:30:11.412
across the entire network, across the
globe. And it was a flooding mechanism
00:30:11.412 --> 00:30:17.199
involve to make sure that the messages get
replicated and the duplicates get detected
00:30:17.199 --> 00:30:23.432
and duplicates are not basically
transmitted again or rather shown again
00:30:23.432 --> 00:30:32.660
and so on. The routing was originally
defined in route maps in UUCP which is a
00:30:32.660 --> 00:30:39.777
quite a bit odd over time because it's
basically a static source based routing
00:30:39.777 --> 00:30:46.174
for the UUCP mails. News as I said
they were flooding anyway. Usenet was
00:30:46.174 --> 00:30:53.119
quite popular until well into the 90s. I
was news master of two news servers for
00:30:53.119 --> 00:30:58.411
some time basically doing system
administration of those boxes. And just to
00:30:58.411 --> 00:31:02.869
give you an anecdote again; into this
context we will get to Kommunikationsnetz
00:31:02.869 --> 00:31:07.833
Franken, which is a nonprofit organization
in the area of Franconia in southern
00:31:07.833 --> 00:31:14.713
Germany, where I was active. And at the
time internet - like when we actually got
00:31:14.713 --> 00:31:21.233
to IP, at some point, IP traffic was so
expensive that it was rather difficult to
00:31:21.233 --> 00:31:25.729
get a full newsfeed over IP because you've
wasted a lot of your expensive bandwidth -
00:31:25.729 --> 00:31:30.928
wasted in quotes - but you used it for
news and so what we did actually is, we
00:31:30.928 --> 00:31:36.051
put up a satellite dish at a building in
Nuremberg and we had satellite feeds from
00:31:36.051 --> 00:31:42.287
the US. So there were US companies that
were streaming compressed Usenet batches
00:31:42.287 --> 00:31:48.175
up to a geostationary satellite which has
a downlink over Europe and then we got two
00:31:48.175 --> 00:31:54.790
megabits of compressed batched news net
news in, I would say, let's say 95ish or
00:31:54.790 --> 00:32:00.350
something like that, so that was
definitely a big improvement. So we we had
00:32:00.350 --> 00:32:07.581
a full news feed coming directly from the
US without having to pay for all the
00:32:07.581 --> 00:32:12.475
International data transfer. Another
curiosity is the Floppy Poll/Point. Now
00:32:12.475 --> 00:32:19.157
nobody is laughing yet. Well not everyone
had phone lines in the 90s, particularly
00:32:19.157 --> 00:32:23.910
in eastern Germany. Phone lines were still
a rare commodity after reunification
00:32:23.910 --> 00:32:28.976
happened in 90. It took some time until
people could get connected to the
00:32:28.976 --> 00:32:33.159
telephone network. And so what people did
is actually they exchanged daily floppies
00:32:33.159 --> 00:32:38.560
by postal mail. So basically rather than
sending your compressed batches of
00:32:38.560 --> 00:32:46.429
messages over modems, because well for a
modem you need phone lines, you put a
00:32:46.429 --> 00:32:53.147
floppy - I would assume 3.5 inch at the
time, not so much four and a quarter inch
00:32:53.147 --> 00:33:01.333
- but you put a floppy in an envelope you
send it to your BBS and the guy
00:33:01.333 --> 00:33:05.927
opens the envelope and puts it in the BBS
and he sends you a floppy in return. So
00:33:05.927 --> 00:33:09.060
you add one day or something to your
transmission but then well the
00:33:09.060 --> 00:33:13.708
transmission speed of messages in those
networks at the time was sort of one to
00:33:13.708 --> 00:33:17.685
two days or maybe even three days anyway
so if you add another day what does it
00:33:17.685 --> 00:33:23.750
matter? It was such a big advantage that
you could get messages like worldwide
00:33:23.750 --> 00:33:30.895
messages at all in such a short time and
for basically no cost whatsoever. Okay
00:33:30.895 --> 00:33:37.639
getting to the internet, yeah. How did I
start to access Internet, how did people
00:33:37.639 --> 00:33:41.939
start to access the Internet at the time?
Well mail and news was sort of the
00:33:41.939 --> 00:33:52.571
Internet in the beginning via UUCP, which
is nice and fine, but it's not IP, yet. So
00:33:52.571 --> 00:33:56.880
what you could do is you could, instead of
dialing into a BBS, you could of course
00:33:56.880 --> 00:34:02.554
use your modem to dial to the serial port
of the TTY of any UNIX machine that's
00:34:02.554 --> 00:34:06.151
somewhere else. If you have a UNIX
workstation somewhere, that's connected to
00:34:06.151 --> 00:34:12.482
an IP network using 10base2 or whatever
was the network technology at the time or
00:34:12.482 --> 00:34:20.775
FDDI or whatever, x21... then you could
attach a modem to a serial part of such a
00:34:20.775 --> 00:34:26.105
UNIX box and you just get the login prompt
when you connect with the modem to that
00:34:26.105 --> 00:34:30.451
box. Like you sit in front of your Linux
system today, you have your login prompt.
00:34:30.451 --> 00:34:35.851
And then on that workstation you basically
you could remotely use that workstation
00:34:35.851 --> 00:34:41.160
and then you could run FTP clients or IRC
clients or telnet, gopher, whatever on the
00:34:41.160 --> 00:34:48.862
text console. That was mostly available to
people in the academic sector of course
00:34:48.862 --> 00:34:53.700
because they had some UNIX machines at
universities. I was too young to be at
00:34:53.700 --> 00:34:59.860
university, so I had to use FTP mailers
for quite some time. So what's an FTP
00:34:59.860 --> 00:35:03.802
mailer? Well it's basically some FTP
client that runs on a remote machine
00:35:03.802 --> 00:35:07.747
somewhere that's connected to the Internet
and that has email access and you can use
00:35:07.747 --> 00:35:13.345
input/output over email. So if you want to
FTP to some FTP server you send an email.
00:35:13.345 --> 00:35:19.847
It says "ftp ftp." and an "ls" and
then some hours later you get a response
00:35:19.847 --> 00:35:25.649
with the list of the files, yeah? And then
after you've got the list of the files you
00:35:25.649 --> 00:35:30.394
do the first CD to change into a directory
and then you get again the response. And
00:35:30.394 --> 00:35:34.556
then finally you know which file you want
so you issue a get command over the file
00:35:34.556 --> 00:35:41.516
and then you get this long series of
UUencoded mails. UUencode is a method of
00:35:41.516 --> 00:35:50.271
sending binary 8-bit messages over mails
before MIME existed. The MIME format which
00:35:50.271 --> 00:35:54.245
we use today for email attachments and so
on. That didn't exist at the time, so it
00:35:54.245 --> 00:35:58.366
was UUencode before, so yeah. So hours or
days later you got that and it
00:35:58.366 --> 00:36:04.460
worked perfectly fine, I mean, I was quite
happy to be able to use that at the time.
00:36:04.460 --> 00:36:09.783
Now, then, if you had dial-up access to
UNIX boxes, you could also do something
00:36:09.783 --> 00:36:19.216
called SLIP, which is a serial line IP. So
you could transport IP over the modem line
00:36:19.216 --> 00:36:26.662
and as a result you have IP at home in
your apartment! Unbelievable! it was later
00:36:26.662 --> 00:36:30.518
superseded by PPP which introduced
features such as auto-configuration,
00:36:30.518 --> 00:36:34.554
authentication, compression and so on -
well there was a compressed SLIP, but yeah
00:36:34.554 --> 00:36:39.811
not quite as compressed as PPP - and
popular software stack at the time - and
00:36:39.811 --> 00:36:47.412
I'm talking about early 90s, mid-90s - is
basically Trumpet Winsock on Windows with
00:36:47.412 --> 00:36:52.441
NCSA Mosaic as a browser, because Windows
back then didn't have TCP/IP, so you had
00:36:52.441 --> 00:36:59.356
to install another package to actually
have TCP/IP on Windows at the time. If
00:36:59.356 --> 00:37:03.506
you didn't have Windows, I will get to
that, and I'm talking about the pre-Linux
00:37:03.506 --> 00:37:09.120
days here. So what did you do if you
wanted to do internet on a PC before Linux
00:37:09.120 --> 00:37:14.257
was around? I didn't have a 386 initially,
I had a 286. And on a 286 of course you
00:37:14.257 --> 00:37:18.396
couldn't run any multitasking operating
system because it doesn't have a real
00:37:18.396 --> 00:37:24.444
protected mode. So no Linux, no BSD, but
there was something called KA9Q NOS. And
00:37:24.444 --> 00:37:29.525
now I want to see hands: who has ever
heard of or used KA9Q NOS? Yeah! Ok...
00:37:29.525 --> 00:37:34.169
laughs
Audience member shouts: It is a person's
00:37:34.169 --> 00:37:36.799
callsign.
LaForge: Yes, "It's a person's callsign"
00:37:36.799 --> 00:37:41.966
was the comment from the audience, this is
correct. KA9Q is Phil Karn in the US and
00:37:41.966 --> 00:37:47.135
he wrote a network operating system the
KA9Q NOS, the network operating system.
00:37:47.135 --> 00:37:51.683
And it is an implementation of - he
started actually in the 80s with this on
00:37:51.683 --> 00:37:56.645
CPM and then later ported it to DOS - and
it implements TCP/IP, SLIP, PPP including
00:37:56.645 --> 00:38:02.368
POP3 server, SMTP server + client, IP
routing, telnet, ARP and so on. And you
00:38:02.368 --> 00:38:08.150
could do all this on DOS. I used it quite
a lot at my home. You could do routing and
00:38:08.150 --> 00:38:13.336
you had multiple applications at the same
time all on top of DOS. It was a fantastic
00:38:13.336 --> 00:38:20.132
piece of software. And then you could
build a router to ethernet and you could
00:38:20.132 --> 00:38:24.290
have multiple other machines in your home
and you have more and more cable in your
00:38:24.290 --> 00:38:30.101
home. And more and more connected machines,
yeah, actually, yeah we will get to that, ok.
00:38:30.101 --> 00:38:37.088
PPP superseded that. At some point ISDN
came around, particularly in Germany. ISDN
00:38:37.088 --> 00:38:41.072
is the digital version of telephony
system, so instead of having analog
00:38:41.072 --> 00:38:48.154
circuits you now transfer digital bits.
That could be audio, digitized audio, but
00:38:48.154 --> 00:38:53.351
of course it could be any other
transparent digital data. In Germany ISDN
00:38:53.351 --> 00:39:00.509
was first put in operation in 1989. Until
'93 it used a German protocol standard
00:39:00.509 --> 00:39:06.141
called 1TR6, and from '94 onwards the
European E-DSS1 protocol standard was
00:39:06.141 --> 00:39:13.347
available. It was hugely popularized from
1995 onwards by subsidies. So at the time
00:39:13.347 --> 00:39:20.161
if you actually ordered an ISDN connection
and at the same time you bought a, let's
00:39:20.161 --> 00:39:26.271
say a small PBX or a phone or a modem or
something like that, you could [get]
00:39:26.271 --> 00:39:33.531
subsidies from Deutsche Telekom. So, I
think it went up to 700 marks - not sure
00:39:33.531 --> 00:39:39.562
if somebody remembers the exact figures -
and so you've got quite a bit of money to
00:39:39.562 --> 00:39:44.448
buy equipment to switch to this new
technology. So when ISDN you don't have a
00:39:44.448 --> 00:39:48.815
modem because there's nothing to modulate
or demodulate, it's digital, so it's
00:39:48.815 --> 00:39:56.548
called a terminal adapter, and it adapts
the bitstream, the synchronous serial
00:39:56.548 --> 00:40:04.582
bitstream of the ISDN to your operating
system or your computer and there was
00:40:04.582 --> 00:40:09.060
something called V.110 as a rate
adaptation to do asynchronous serial like
00:40:09.060 --> 00:40:17.130
RS-232, sort of, over a synchronous ISDN.
Okay and how did we get internet access?
00:40:17.130 --> 00:40:22.699
Well, it was, if you were not in academia
or something like that, there were a few
00:40:22.699 --> 00:40:27.986
commercial ISPs like XLink or EUnet. They
were very expensive and of course you
00:40:27.986 --> 00:40:34.235
didn't have local dial-in in all the
different cities around Germany, but you
00:40:34.235 --> 00:40:38.954
had grassroot groups of enthusiasts that
established themselves in some
00:40:38.954 --> 00:40:46.592
associations to make sure the members can
get internet access. In my region in
00:40:46.592 --> 00:40:52.440
Nuremberg Kommunikationsnetz Franken was
particularly active. They started with
00:40:52.440 --> 00:40:57.664
dial-up UUCP services and later IP for
non-commercial users - and I have to say
00:40:57.664 --> 00:41:04.045
with an extremely high technical standard
which I'm still fascinated by today.
00:41:04.045 --> 00:41:08.126
Kommunikationsnetz Franken had points of
presence in various different cities in
00:41:08.126 --> 00:41:12.390
the region because not everybody could
call to Nuremberg as a local call and
00:41:12.390 --> 00:41:18.080
every user got six static IP addresses,
routed to wherever he dialed in. The use of
00:41:18.080 --> 00:41:23.238
OSPF in the mid-1990s to make sure you
have static IP addresses wherever you dial
00:41:23.238 --> 00:41:28.746
in. Some people still don't have that in
2017 and I'm not even talking about the
00:41:28.746 --> 00:41:35.354
static IP addresses, but anyway. So about
800 users peak at that association at the
00:41:35.354 --> 00:41:41.624
time. And there was an umbrella
organization called "Individual Network
00:41:41.624 --> 00:41:49.148
e.V." (IN). This was established.
Individuals could not become members in
00:41:49.148 --> 00:41:52.380
that association so it's - the name is a
bit interesting - it's called Individual
00:41:52.380 --> 00:41:56.498
Network, because it's about networking for
individuals, but the members were the
00:41:56.498 --> 00:42:00.970
regional associations such as
Kommunikationsnetz Franken, who then
00:42:00.970 --> 00:42:07.480
basically used this umbrella entity to
negotiate decent rates to get internet
00:42:07.480 --> 00:42:12.904
connectivity and so on. And apparently the
IN members served more than three hundred
00:42:12.904 --> 00:42:17.660
thousand users at some point - so it
scaled quite a bit - was dissolved in 2000
00:42:17.660 --> 00:42:22.170
when lots of commercialized ISPs were
around and also when the remaining member
00:42:22.170 --> 00:42:26.740
entities, which many of which still exist
today such as Kommunikationsnetz Franken,
00:42:26.740 --> 00:42:31.640
they didn't need this umbrella entity to
get decent internet rates or tariffs
00:42:31.640 --> 00:42:37.740
again. So, with packets which TCP/IP we
just need one number that we call at some
00:42:37.740 --> 00:42:41.580
point We're not dialing into hundreds of
different BBS's anymore but we're actually
00:42:41.580 --> 00:42:47.347
connecting always to the same number which
is our ISP, and then when we have that
00:42:47.347 --> 00:42:52.349
connection we exchange packet data with
systems worldwide which brought new
00:42:52.349 --> 00:42:57.634
purpose to lease lines. Analog leased
lines were basically telephone lines that
00:42:57.634 --> 00:43:02.810
were permanently switched, or actually
permanently wired at the exchange. So you
00:43:02.810 --> 00:43:06.963
had two wires of copper between one
location and another location and they
00:43:06.963 --> 00:43:10.514
were physically connected you could apply
a DC voltage and the DC voltage would come
00:43:10.514 --> 00:43:16.617
out at the other end. You could get this
from Deutsche Post or Telekom at the time.
00:43:16.617 --> 00:43:23.626
When I could finally afford one in '98 for
900 marks installation cost and in my case
00:43:23.626 --> 00:43:31.000
180 marks per month, was sixty marks
per hop. Hop means: telephone exchange. So
00:43:31.000 --> 00:43:35.105
if between the other end where you want to
connect to and where you are, are three
00:43:35.105 --> 00:43:40.000
telephone exchanges, you had three times
sixty marks or 180 marks per month. And
00:43:40.000 --> 00:43:43.929
then I connected to a system that looked
like this, which is called the Hub
00:43:43.929 --> 00:43:48.532
Nuremburg of this Kommunikationsnetz
Franken, which is in the basement of one
00:43:48.532 --> 00:43:55.391
of the members. You have basically a PC
running Linux of FreeBSD, no it was BSD
00:43:55.391 --> 00:44:01.240
actually, with like a 16-port serial card
and various modems stacked on various
00:44:01.240 --> 00:44:05.955
shelves to interconnect all these
different leased lines and which then had
00:44:05.955 --> 00:44:11.987
one ISDN leased line with 128 kilobits to
some internet uplink. Yeah that's the
00:44:11.987 --> 00:44:20.324
obligatory ISDN network termination and
telephone sockets, which brings us to ISDN
00:44:20.324 --> 00:44:27.990
leased lines. There was a product called
SPV "Semi-Permanente Festverbindung",
00:44:27.990 --> 00:44:32.770
which is not really a leased line - it's
semi-permanent - and it's basically a
00:44:32.770 --> 00:44:37.630
flat-rate call to one specific destination
telephone number, which you could get in
00:44:37.630 --> 00:44:42.433
national 1TR6 ISDN and which was rather
inexpensive and what many people used who
00:44:42.433 --> 00:44:47.910
wanted more than the ISDN speeds. Okay I
have to speed up a bit, time is running
00:44:47.910 --> 00:44:52.505
out! The first step of abusing analog
lines, which we did, is by deploying a
00:44:52.505 --> 00:44:58.122
device called an ICU-T, which is the
inverse of an ISDN NTBA. So in ISDN you
00:44:58.122 --> 00:45:01.621
still have the telephone exchange and you
have a network termination, the NTBA, on
00:45:01.621 --> 00:45:09.290
your line. And basically the the ICU-T was
a single line telephone exchange side of
00:45:09.290 --> 00:45:13.040
this protocol. So you could use an analog
line which you normally used for analog
00:45:13.040 --> 00:45:17.222
modems but you remove the two analog
modems you put an NTBA on one end, you put
00:45:17.222 --> 00:45:21.629
the ICU-T on the other end and suddenly we
can get 128 kilobits over that line which
00:45:21.629 --> 00:45:26.438
previously you could only do 33.6 without
having to pay any additional cents or
00:45:26.438 --> 00:45:31.435
money to Deutsche Telekom, of course. And
then there was some special ISDN routers
00:45:31.435 --> 00:45:35.953
which could use the signaling channel, the
16 kbps signaling D-channel on ISDN also
00:45:35.953 --> 00:45:41.944
for data, so you get 128 + 16 kilobytes of
data, because well, there's no signaling,
00:45:41.944 --> 00:45:46.024
you're not dialling anyone so you can as
well use that. Now this is sort of the
00:45:46.024 --> 00:45:50.480
hierarchy of the leased line
infrastructure at this entity. I'm not
00:45:50.480 --> 00:45:54.632
showing every leased line here, but
basically I was at the upper left corner
00:45:54.632 --> 00:46:00.560
here connecting with 33.6 kbps to this hub
Nuremburg, which connects to 128K to a
00:46:00.560 --> 00:46:04.949
machine in a Nuremberg building of the
University of Erlangen, which then
00:46:04.949 --> 00:46:09.250
connects over X21 to the University of
Erlangen, where then all kinds of other
00:46:09.250 --> 00:46:14.003
leased lines come together. That was the
the architecture of what we deployed
00:46:14.003 --> 00:46:18.330
there. Some more pictures: this is in
Fürth, a neighbor city of Nuremberg. The
00:46:18.330 --> 00:46:24.680
collection of telephone outlets and the
collection of modems and the machine - oh
00:46:24.680 --> 00:46:29.488
there was, I'm missing one picture sorry
for that - anyway you can see a pile of
00:46:29.488 --> 00:46:34.220
modems here and some more modems here and
the machine over there. And then we went
00:46:34.220 --> 00:46:38.840
into phase two of abusing analog telephone
lines, when the first DSL modems came out.
00:46:38.840 --> 00:46:45.809
So we imported some Ascend DSLpipes in '99
from the US and with some firmwares you
00:46:45.809 --> 00:46:49.602
could operate them back to back without
the DSLAM so basically you operate one DSL
00:46:49.602 --> 00:46:54.370
modem at one end of the leased line and
another DSL modem at the other end, and if
00:46:54.370 --> 00:46:58.260
you are close enough like with a single
hop at the single telephone exchange you
00:46:58.260 --> 00:47:04.300
could get up to 2.3 megabits symmetric
over your analog line. And that in 1999
00:47:04.300 --> 00:47:09.701
was quite a lot of speed, especially if
you're not paying for traffic or anything
00:47:09.701 --> 00:47:14.274
like that. Some less alternative, less
expensive one alternatives came out. Okay!
00:47:14.274 --> 00:47:24.500
Before I wrap up, a short detour or one
thing still to mention. Another phenomenon
00:47:24.500 --> 00:47:29.730
back then - I'm not sure if this happened
in other cities too - and in my area in
00:47:29.730 --> 00:47:35.041
Fürth we had an entity called Falcons
Maze, which was called an online bistro.
00:47:35.041 --> 00:47:41.726
I became a regular there around '94. They
initially had four DOS PCs, each of them
00:47:41.726 --> 00:47:46.884
with a modem and with a dedicated call-
charge meter. And you could basically go
00:47:46.884 --> 00:47:50.564
there, it's a cafe, you can have, you know
you can eat and drink and so on, and you
00:47:50.564 --> 00:47:55.614
can sit at the PC and you can then from
there dial into BBSs and basically do
00:47:55.614 --> 00:47:59.250
things if you didn't have a modem or a PC
at home. But the interesting part of
00:47:59.250 --> 00:48:02.862
course was that there all the other peoples
were hanging out, the other BBS users,
00:48:02.862 --> 00:48:08.810
sysops and so on. At some point the PCs
were networked with 10base2, so people
00:48:08.810 --> 00:48:14.457
could play doom when it came out, I think
in - not sure when it reached us in
00:48:14.457 --> 00:48:20.473
Germany - '94 maybe or so, and yeah. The
internet became more popular. It started
00:48:20.473 --> 00:48:26.320
subsidiaries and we set up ISDN SPVs, the
"semi-permanente Verbindung" as an
00:48:26.320 --> 00:48:33.278
internet uplink from there, so that also,
I mean, you can find some sources that
00:48:33.278 --> 00:48:36.976
this apparently, allegedly was the first
internet cafe. I'm not sure if anyone else
00:48:36.976 --> 00:48:41.350
has contested that. Something like that.
Anyway, after lots of anecdotes I want to
00:48:41.350 --> 00:48:46.720
give you some time for Q&A. To summarize:
the first decades of wide area
00:48:46.720 --> 00:48:51.510
communications were powered by a community
of enthusiasts or rather communities that
00:48:51.510 --> 00:48:56.359
were disjunct and not connected, largely
motivated by non-commercial motives. Of
00:48:56.359 --> 00:49:02.120
course there were commercial BBSs but by
far not without much corporate or
00:49:02.120 --> 00:49:06.914
government influence, right? There was no
Google and there was no ministry that was
00:49:06.914 --> 00:49:12.583
putting censorship or something like that.
And the BBS community is a distinct
00:49:12.583 --> 00:49:18.370
subculture so it has different norms and
it has different values, different from
00:49:18.370 --> 00:49:22.557
the ham radio guys, different from free
software guys, of course some overlap, but
00:49:22.557 --> 00:49:28.337
still a separate community with separate
norms. What I personally think is the big
00:49:28.337 --> 00:49:37.124
loss, other than the loss of picture on
the screen, is that back then the networks
00:49:37.124 --> 00:49:40.500
were distributed. There was no single
point of failure. The infrastructure was
00:49:40.500 --> 00:49:45.286
owned and operated by its users, by
individuals. The connection speeds were
00:49:45.286 --> 00:49:50.493
symmetric and there was no, like, data
center versus consumer separation that we
00:49:50.493 --> 00:49:55.440
have in the internet day and age of today.
And that's, yes, I really think this
00:49:55.440 --> 00:50:02.590
autonomy and decentralization is a big
loss to society or the community as a
00:50:02.590 --> 00:50:07.854
whole. Ok, some pointers: if you want to
read up more or look at some ANSI artwork
00:50:07.854 --> 00:50:12.940
or log into BBSs, the telnet BBS guide I
can highly recommend that. You can also
00:50:12.940 --> 00:50:19.891
find the BBS I looked into. Ok, good.
Which brings us to the point where we can
00:50:19.891 --> 00:50:24.000
have some questions.
00:50:24.000 --> 00:50:34.659
Applause
00:50:34.659 --> 00:50:40.930
Herald: The microphones here in, 3, 1, 2
and 4, but first we have questions from
00:50:40.930 --> 00:50:43.800
the signal angel. So what's the question for?
00:50:43.800 --> 00:50:47.022
Signal Angel: The internet wants to know,
"What was the highest phone bill you ever
00:50:47.022 --> 00:50:51.978
got back then?"
LaForge: To be honest, I don't remember
00:50:51.978 --> 00:51:00.758
but for sure it was four digits. I'm quite
sure it was. It was quite devastating,
00:51:00.758 --> 00:51:04.590
yes.
Hearld: There is another question from the
00:51:04.590 --> 00:51:06.246
internet.
Signal Angel: And there's another
00:51:06.246 --> 00:51:12.924
question, "You mentioned that there are
very few books around those topics. Which
00:51:12.924 --> 00:51:16.774
ones would you recommend regarding BBS,
Usenet and so on?"
00:51:16.774 --> 00:51:22.263
LaForge: I cannot respond to this directly
I don't remember that. I can put it
00:51:22.263 --> 00:51:26.736
together and people can reach out to me
or I put it in the slides when I submit
00:51:26.736 --> 00:51:33.067
them into the frap system, sorry for that.
Herald: So we have a question from the
00:51:33.067 --> 00:51:38.019
microphone number two please.
Mic 2: Yes, back in the 90s most of the
00:51:38.019 --> 00:51:43.367
voice was uncompressed and actually
direct. Modern technologies usually, I
00:51:43.367 --> 00:51:48.054
think, voice always compressed transferred
over IP. Do you know for any modern
00:51:48.054 --> 00:51:53.390
modulation formats the text can survive
several codecs voice codecs or data
00:51:53.390 --> 00:51:57.199
transmission?
LaForge: I'm not the expert on that
00:51:57.199 --> 00:52:03.148
subject. I know there are some codecs,
yes, but they are extremely slow. So you
00:52:03.148 --> 00:52:10.442
are happy if you get something like 1200
or maybe 2400 bps of data through a modem
00:52:10.442 --> 00:52:14.687
that survives multiple codecs and then of
course always the question of which
00:52:14.687 --> 00:52:20.242
codecs.
Herald: Okay microphone number four
00:52:20.242 --> 00:52:22.570
please.
Mic 4: Okay I don't have a question to
00:52:22.570 --> 00:52:26.319
Herald actually, but thanks for the talk.
I would like to ask the audience because
00:52:26.319 --> 00:52:30.490
many, I think, users and operators of BBSs
are here. Who wants to meet this evening,
00:52:30.490 --> 00:52:34.944
at I would say nine o'clock, in one of the
seminar rooms for talk about the back old
00:52:34.944 --> 00:52:41.619
times? Yeah, so I will try to lock a self-
organized session at the seminar room
00:52:41.619 --> 00:52:46.333
1415, I think it's called, at 9 o'clock.
LaForge: Ok, thank you very much.
00:52:46.333 --> 00:52:49.472
Mic 4: So, see you there and talk about
the good days of and some more stories I
00:52:49.472 --> 00:52:52.228
think.
Herald: There are still more people
00:52:52.228 --> 00:52:59.525
queuing up. Microphone number 4, please.
Mic 4: I've got a question about the
00:52:59.525 --> 00:53:05.187
political bulletin board systems. Could
you tell us a bit about the CL-Net and the
00:53:05.187 --> 00:53:11.116
fascist clone the Thule-Net? What was the
dynamics back then and the fights? What
00:53:11.116 --> 00:53:16.682
were the conflicts in those boxes?
LaForge: I have to admit I cannot say too
00:53:16.682 --> 00:53:22.189
much about it. I know, of course, CL-Netz
was a network mainly for left-wing
00:53:22.189 --> 00:53:26.590
political activists and groups and yes
there was Thule-Netz, a right-wing
00:53:26.590 --> 00:53:29.780
Network, and I knew there was discussions
and so on and there were people trying to
00:53:29.780 --> 00:53:36.462
hack each other's mailboxes and so on,
but I was not participating or involved
00:53:36.462 --> 00:53:43.094
in these discussions to an extent that I
can really comment on it sorry.
00:53:43.094 --> 00:53:46.150
Herald: Microphone number one, please.
Mic 1: Hi Harald. I still remember when I
00:53:46.150 --> 00:53:50.750
started with an acoustic coupler. I did
that because there was a severe threat of
00:53:50.750 --> 00:53:55.486
punishment if you used an illegal modem at
the time from the Deutsche Bundespost. So
00:53:55.486 --> 00:54:00.010
I was actually never aware that a little
bit later you could actually do an end,
00:54:00.010 --> 00:54:05.552
back to back DSL modem connection over an
analogue exchange. So at that time you did
00:54:05.552 --> 00:54:09.950
that, what was the the punishment
situation from the Bundespost or whatever
00:54:09.950 --> 00:54:14.271
it was called at the time if they would
have ever caught you doing that? Do you
00:54:14.271 --> 00:54:17.010
remember?
LaForge: I have no clue. Yes, it sort of,
00:54:17.010 --> 00:54:25.363
and I mean the... How can I say? The the
criminal offense, I think, stopped in '92
00:54:25.363 --> 00:54:30.331
when Deutsche Post was privatized. So
until '92 it was a criminal offence to
00:54:30.331 --> 00:54:34.730
operate a non-approved modem at the German
telephone network, because was government
00:54:34.730 --> 00:54:39.825
owned. It was a crime, not a minor
offence. But afterwards I don't really
00:54:39.825 --> 00:54:44.850
know to be honest. I don't think anyone
bothered at the time and nobody, I mean
00:54:44.850 --> 00:54:49.994
the, we never had any trouble with these
DSL things and so on, that we did over
00:54:49.994 --> 00:54:53.823
analog circuits.
Herald: Microphone number two, please.
00:54:53.823 --> 00:54:58.725
Mic 2: Okay, hello I'm from Taiwan and I
just want to share something interesting
00:54:58.725 --> 00:55:05.200
for everyone. In Taiwan is a small country
in Asia. We are still using BBS. The
00:55:05.200 --> 00:55:12.090
largest is named PTT and exported to use
SSH or WebSocket you can edit, and the
00:55:12.090 --> 00:55:16.580
source code is open available on GitHub.
Everybody can search it. Thank you.
00:55:16.580 --> 00:55:23.593
LaForge: Thank you very much. It's
actually not just for Taiwan, but you can
00:55:23.593 --> 00:55:27.870
find many, I mean maybe it's more popular
there still, but you can find many BBSs
00:55:27.870 --> 00:55:33.049
that are still in operation today in many
different countries even also with BBS
00:55:33.049 --> 00:55:38.240
software that's free software that's
maintained now on GitHub or on other
00:55:38.240 --> 00:55:43.240
repositories with contributors and so on.
So the community still lives, but I think
00:55:43.240 --> 00:55:48.230
at least internationally it's very small
and I'm happy to hear if it's larger in
00:55:48.230 --> 00:55:52.480
some countries.
Herald: You have still time for questions.
00:55:52.480 --> 00:55:57.820
Microphone number four, please.
Mic 4: So you talked about restoring
00:55:57.820 --> 00:56:05.181
decentralization. So, what old systems
would you like to see coming back?
00:56:05.181 --> 00:56:09.030
Something like the Usenet? I mean it's
still there, but you can't access it
00:56:09.030 --> 00:56:13.290
without paying a lot of money to some big
gateway. So, which technologies would you
00:56:13.290 --> 00:56:17.374
like to revive or do you think are
realistic to revive to have
00:56:17.374 --> 00:56:21.640
decentralization again?
LaForge: I don't think the technologies
00:56:21.640 --> 00:56:26.248
necessarily need to be revived because
they are, to a large extent, old and
00:56:26.248 --> 00:56:32.902
people are smarter and the, how can I say,
the capacity and the computational
00:56:32.902 --> 00:56:37.141
complexity of what you can do today and so
on is much better. So we can have much
00:56:37.141 --> 00:56:42.576
better technology. But the thing that I
would like to see revived is more
00:56:42.576 --> 00:56:47.855
decentralization and more people operating
their own technology and that's just, I
00:56:47.855 --> 00:56:53.529
think, I don't really have a plan and I'm
not saying I have a vision I'm just saying
00:56:53.529 --> 00:56:58.620
it has a problem, this development, that
basically it's a consumer / producer model
00:56:58.620 --> 00:57:03.510
and especially with content delivery
networks and with attacks on network
00:57:03.510 --> 00:57:08.356
neutrality and and all these topics, it's
always moving in one direction. It's
00:57:08.356 --> 00:57:12.796
basically turning the user into a stupid
consumer and and making sure all the
00:57:12.796 --> 00:57:20.214
control and all the content, and so on, is
in the hand of large corporations.
00:57:20.214 --> 00:57:28.543
Applause
By the way, one interesting anecdote about
00:57:28.543 --> 00:57:33.644
the... I talked about the asymmetry of the
speed, right? And with DSL at this ADSL
00:57:33.644 --> 00:57:38.207
and the popular technology is always the
downlink is bigger than the uplink. I know
00:57:38.207 --> 00:57:45.110
in Brazil a lot of people, basically in
small, like small size ISPs, they did it
00:57:45.110 --> 00:57:50.290
the opposite way around! So they did one
modem with basically a large downstream
00:57:50.290 --> 00:57:55.380
and small upstream and then they, on
another line next to it, they inverted it
00:57:55.380 --> 00:57:59.130
by using a master modem on one side and a
slave modem on the other so then again he
00:57:59.130 --> 00:58:04.289
had symmetric speed. So, some people had
creative ideas to work around some of the
00:58:04.289 --> 00:58:09.902
technological restrictions.
Herald: So microphone number two, please.
00:58:09.902 --> 00:58:16.090
Mic 2: I also from Taiwan and I want to
add something for my friend. Like, there
00:58:16.090 --> 00:58:25.636
are still like half million people come
here to BBS called PTT, yeah, today. And
00:58:25.636 --> 00:58:34.345
like, there's a, there are 100,000 people
online now, yeah. So, I think the
00:58:34.345 --> 00:58:39.255
community is now like...
Herald: What ist your question? Can you
00:58:39.255 --> 00:58:42.302
please phrase the question?
Mic 2: I just want to add something for my
00:58:42.302 --> 00:58:46.642
friend, yeah.
LaForge: Okay, thank you.
00:58:46.642 --> 00:58:54.641
Herald: Microphone number one, please.
Mic 1: cough You talked about content of
00:58:54.641 --> 00:59:01.551
these mailboxes. Isn't it that the
Freifunk community today is a possible way
00:59:01.551 --> 00:59:11.900
to get this freedom back from what you had
in your mailboxes? The services they were
00:59:11.900 --> 00:59:19.290
offered there, the Freifunk could do the
same today with user own structures and so
00:59:19.290 --> 00:59:21.571
on.
LaForge: That's very correct yes. Freifunk
00:59:21.571 --> 00:59:26.267
definitely is much more in the spirit of
the community owned and community run
00:59:26.267 --> 00:59:31.440
systems, and I see lots of similarities
between the BBS community and what
00:59:31.440 --> 00:59:34.690
Freifunk is doing today. It's correct.
Mic 1: Are you are you doing something
00:59:34.690 --> 00:59:37.630
with Freifunk?
LaForge: Me personally? No, I'm not
00:59:37.630 --> 00:59:40.480
involved.
Mic 1: Okay.
00:59:40.480 --> 00:59:47.203
Herald: I think microphone number two is
waiting way too long.
00:59:47.203 --> 00:59:53.270
Mic 2: Hello, thanks for the talk. You
mentioned that most people didn't have a
00:59:53.270 --> 00:59:59.668
TCP/IP capable operating system at this
time and I started to read recently about
00:59:59.668 --> 01:00:06.313
an operating system called Xenix, X-E-N-
I-X, that was actually developed by
01:00:06.313 --> 01:00:14.609
Microsoft and published in 1983 that could
run on IBM PC compatible machines on the
01:00:14.609 --> 01:00:20.854
x86 processors, and I hear that in the
Russian BBS systems at least it was very
01:00:20.854 --> 01:00:26.480
popular. Did you encounter any Xenix
operating systems at that time?
01:00:26.480 --> 01:00:30.886
LaForge: No I personally did not encounter
Xenix. I read about it, yes, and I know it
01:00:30.886 --> 01:00:35.860
I could have possibly run it on my 286
machine, but I mean, I don't think it was
01:00:35.860 --> 01:00:40.097
something that was readily available for
affordable price to individuals, but maybe
01:00:40.097 --> 01:00:44.280
I'm wrong. No, certainly not, okay, some
people are heavily shaking their heads.
01:00:44.280 --> 01:00:46.580
Mic 2: I think this is why it was popular
in Russia...
01:00:46.580 --> 01:00:49.150
Laughs
LaForge: Possibly. I do not want to
01:00:49.150 --> 01:00:52.440
comment on that...
Herald: We have time for one more
01:00:52.440 --> 01:00:56.596
question. Microphone number 4.
Mic 4: I just wanted to note, in the wiki
01:00:56.596 --> 01:01:00.976
the meeting is up. Search for BBS and this
evening at 9 o'clock I think we can talk
01:01:00.976 --> 01:01:05.482
about all the details of running DSL on
modem lines. I've also got some more
01:01:05.482 --> 01:01:10.284
details on that and a lot of these modems
left if you need some. But I think, so see
01:01:10.284 --> 01:01:13.710
you Harold at 9 o'clock
LaForge: Yeah definitely! Thanks!
01:01:13.710 --> 01:01:16.090
Mic 4: Ok, everybody welcome.
LaForge: Thank you!
01:01:16.090 --> 01:01:17.480
Applause
01:01:17.480 --> 01:01:20.222
Herald: Thank you very much for the talk.
01:01:20.222 --> 01:01:25.425
34C3 Music
01:01:25.425 --> 01:01:43.000
subtitles created by c3subtitles.de
in the year 2020. Join, and help us!