The net, before it was the internet

What you just saw is my own history, played on the machines it happened on. None of it is drawn from memory: the number being dialled is mine, the board was called The Eagle, and the screens are built out of the same eighty by twenty-five characters they arrived in.

The internet is older than people think. ARPANET in 1969, NORSAR outside Oslo in 1973, TCP/IP in 1983. But for most of us it did not begin there. It began with a modem, a telephone number, and somebody who had left a PC switched on.

And for me it did not even begin there. It began with a television set.

Around 1991 I started playing with Scala on the Amiga. Even then you could build genuinely impressive presentations: graphics, text, transitions and sound. Leaving all that on my own screen seemed unnecessarily limiting.

My father had a fairly powerful RF transmitter lying about. It put the video signal out on an ordinary television channel, and people in the nearest blocks of flats could tune in and suddenly receive my own little local station.

This was, for reasons I still consider greatly exaggerated, not equally popular with everybody. The content was apparently not entirely of its time. What nonsense. I prefer to believe the audience simply was not ready for the concept.

What I only found out years later is who was unhappy about it: the estate's own caretaker association. They ran a local channel of their own, you see. On Scala. Whether they could not stand the competition is not for me to say.

The project had a limited life, naturally. But the fascination stayed, and it is really the whole reason you are reading this: making something on a computer and getting it out to other people. It went on to modems, to machines talking to each other, and in the end to a board.

The board was called The Eagle, and it was left switched on. That was the whole idea. A BBS was a PC sitting there with a modem in it, waiting for the telephone to ring.

One line meant one caller at a time. If somebody else was in you got the busy signal, and you rang again in ten minutes. 2400 bits a second: the text arrived about as fast as you could read it, and that was all that was ever coming. There were no pictures. Music was a MIDI file, which is notes rather than sound, and that difference comes back further down this page.

It sounds thin. It was not. You knew the people who were on, you left mail for them, you got into arguments in the conferences, and you knew exactly who was up at night.

Here is what you could actually do in there. Read and write in conferences, which were threads with a subject and a few hundred messages in them. Leave mail for a named person, and know they would read it the next time they rang in. Download files over Zmodem and watch the meter crawl down. Play a door, which was a game running on my machine and drawing itself down the line onto yours. And ring the bell for the sysop, which was me, and get an answer if I was awake.

Here is what you could not. You could not see a picture inside a piece of text. You could not click through to another board; you had to hang up and dial a different number. You could not search across anything at all. You were on one board at a time, and you paid for every minute you were there.

What made the board better than the web was for the next five years was not the features even so. It was that it was the same people. A board's crowd was remarkably stable: the same twenty handles, evening after evening, for years. You knew their machines, you knew how they typed, you knew who was foul before ten and who never went to bed. It was a place, and it had people in it.

I made my first home page in 1995. Exceedingly simple.

But before the page, the line, because the connection is a far bigger part of this story than it looks like now. The internet was not simply there. You had to connect to it.

The machine rang a modem bank at the internet provider, almost exactly the way it rang a board. What differed was where it rang. You did not dial the number of the machine holding the thing you wanted to see; you dialled your provider, and once that connection stood, you had the whole internet. That is the entire difference between a board and a network, and it is one second apart to listen to and a world apart to use.

And while it stood, the telephone was engaged. If somebody was on the internet, nobody else in the house could ring out or in. The meter ran the whole time.

This is easy to underrate now. You generally paid in two places: the subscription with the provider, and the telephone call while you were connected. The internet was not somewhere you were. It was something you did, with a clock running behind it. You connected, did what you came for, and disconnected.

My page could do one thing the board never could: it was simply there. Anybody, anywhere, without dialling a number of mine and without paying for the minutes at my end. That is an enormous thing, and in 1995 I noticed none of it.

Because it did not answer. Nobody left anything. Nobody knew they had been there at the same time as you. There was a guestbook, and the guestbook was empty. A board was a conversation you arrived at. A home page was a poster you put up.

The Eagle MBBS was built because we finally got ISDN at home, which opened up something new entirely. Two channels at 64 kbit a second each, and with that several people could be inside at once.

The software was Worldgroup, which had been called Major BBS before that. Now people could sit in teleconference and write to each other while it was happening, rather than only leaving messages for one another. A board with four nodes is a room. A board with one line is a letterbox.

ISDN made life easier, and the internet was still something you rang. One channel gave you 64 kbit/s and left the telephone alone. Both channels gave you 128, which was very fast in 1997, and two channels are two calls to pay for. Speed could be bought, in other words, and the price showed up on the telephone bill.

The 56k modems arrived and the web became something everybody had. Pictures loaded from the top down, and you knew how much was left because the status bar told you.

You may be wondering how one streamed. One streamed in RealAudio, at between 8 and 20 kbit a second, in mono. It sounded like a telephone. Video was a postage stamp of 160 by 120 pixels at a few frames a second, and it stopped to buffer about as often as it played. Nobody really watched anything. You waited for something, and then you got to see a bit of it.

You do not have to take my word for it. The modem is in the museum, and it puts the whole site on 33,600 while you carry on reading.

The site got its first domain in 1999. Until then it had lodged with other people, at an address with a tilde in it.

The enormous leap web pages took between 1998 and 2000 opened the door to the Flash era, DHTML, higher resolution and real interactivity.

And it went fast. In 1998 a web page was a document with pictures in it: text from the top down, a table holding it in place, and a couple of animated GIFs. Two years later it ran things. Flash played animation, sound and whole games inside the window. DHTML let something happen without the page loading again. Resolution went from 800 × 600 to 1024 × 768, colour from 256 to all of them, and the line finally kept up with what people wanted to make.

That is where the web stopped being a library and became somewhere things happened. And the moment it did, it had everything the board had without a single one of the limits: pictures inside the text, sound, movement, links out of the house, search across all of it, and nobody paying by the minute to be there.

So the BBS disappeared into eternity as far as I was concerned. Nobody switched it off. The people simply left, one at a time, for wherever everybody else was.

It was no longer only about machines talking to each other. This was the era where you turned code into experience. That is the era I am still standing in, and this site is roughly what it looks like when nobody has asked for it.

Technology moves at a tremendous pace today, and the paradox is that we notice each single leap less and less. We get a little more resolution, slightly better light, a few more frames a second, graphics that keep getting more convincing. It is impressive, and the improvements arrive so evenly that yesterday's impossible has time to become ordinary before anybody stops to look at it.

There was a time when one upgrade could change the experience outright. Going from CGA to VGA, hearing proper digital sound out of a Sound Blaster for the first time, starting a game off a hard disk instead of a stack of diskettes, or watching Quake run on a 3dfx Voodoo card, was not simply better. It felt as though somebody had swapped the whole computer for something out of the future.

That feeling is the hard one to reproduce now. The technology has become unimaginably better, and we mostly move towards the future in small continuous steps. In the eighties and nineties the future occasionally arrived in a cardboard box, and once you had fitted it, the world was simply not the same any more.

Then the barriers started coming down, one at a time, and none of them was about speed.

The price first. Competition between the telephone companies grew, network traffic could be priced below an ordinary call, and towards the end of the nineties “free internet” turned up. Free came with substantial modifications: you were spared the subscription with the provider, and the call to get there cost exactly what it had cost before.

Then flat rate. CyberCity was early, and in 2000 you could use an ordinary modem or ISDN and pay the same every month with no minutes hanging over you. It was sold as unlimited use. It did not take long to establish that “unlimited use” and a Norwegian network enthusiast's reading of unlimited use were not the same thing: some left the line up around the clock, capacity came under strain, and the limits followed. Others had hours included from the start, like Telenor's Online 150 and its hundred and fifty a month.

The difference was enormous even so. You still had to dial. But you could stop working out what the next twenty minutes were going to do to the bill.

And then came the thing that actually changed everything, and it was not the speed. Cable modems first, in some places, then ADSL and SDSL over the telephone network, and eventually fibre. What mattered was not the megabits. It was that you stopped ringing the internet. It was simply there.

That is a much larger shift than the numbers suggest. Before, you sat down and decided to go on the internet: connect, use it, disconnect. With broadband it gradually stopped being something that started and ended. It became infrastructure.

For a long time it was still a question of where you lived, and being willing to pay did not help. For ADSL the distance to the exchange and the state of the copper decided it: two houses in the same municipality could get entirely different answers, and one of them was told the line was too long. Cable modems needed a cable television network built for two-way traffic. Fibre was more exclusive still. Some places got radio links, which worked well where they existed and did not exist many places. And for the genuinely remote, satellite was a real option with a price of its own: the signal has a long journey, so downloads were fine while anything happening in real time gave away how far it had been.

In Norway, ADSL was still something special in 2001. Then it went quickly. Half of all households had broadband by the second quarter of 2005, and by 2010 it was around four in five. There is no date on which Norway got broadband. For some the revolution came early in the 2000s, and others sat with a modem for years afterwards because nothing better existed where they lived.

The last barrier was the cable on the floor. Broadband had made the internet permanent in the house; wireless made it independent of where in the house the machine stood. With a router and a laptop, the internet was no longer something that lived on the family PC by the telephone socket.

Flat rate removed the clock. Broadband removed the dialling. Wireless removed the cable. Only once all three were gone could the internet become the continuous part of the day it is now.

All of this belongs to the story of home pages, because the line did not only decide how we got onto the network. It decided what it was possible to make.

On a modem a heavy page was not merely irritating, it cost the visitor money. Large pictures came down piece by piece. Sound was heavy. Video was a curiosity rather than a medium. Pages had to be small, and anybody who knew the craft counted kilobytes.

The early network was not small because people lacked imagination. It was small because the audience was smaller, every minute cost, every kilobyte took time, and most people were not there at all. When those limits went, it exploded. More people could be online for longer, more could publish, pictures could be large, games became practical, and radio, video, blogs, forums, shops and services that had been unthinkable could suddenly stand side by side.

And here is a small paradox. The variety exploded once the lines were good, and then most of the activity gathered into a handful of large places. On the old network you had to find your own way: to a private person, a club, a local history society, a university, or a complete stranger who had spent ten years documenting precisely the thing they cared about. It was messy, inefficient and often ugly, and there was no guarantee any of it interested anybody but the person who made it. That is exactly why it could be magnificent.

So to understand why the first home pages looked the way they did, HTML and graphics are not enough. You have to remember the line.

Before broadband we went on the internet. After broadband we were on the internet.

Fibre delivers the picture of the day before you can see it arrive.

Everything that was impossible in 1998 is free now. A page can play film at full resolution, draw in three dimensions, calculate while you type and change itself without loading again. This one fetches the weather from a mast, works out how high the sun stands over Melhus at this minute, and asks itself every hour whether it still answers. None of that was within reach of one person in 1998, and none of it costs anything in 2026.

Something has gone missing on the way even so. At the high-water mark everybody made their own page. Almost all of that has moved into five or six houses now, and a personal page has become a profile: the same template, the same width, the same buttons, and what you write sits somewhere you neither own nor can take with you. It is not worse in every way. But it has become remarkably alike.

Two things stood on every home page back then and stand on almost none today. Both of them are down at the bottom of this one.

The visitor counter. It was mostly a lie. People set it to 10,000 and let it run. But it was the only trace a page had that anybody had been by, and it stood in the open where everyone could see it. What made it go was the number moving into somebody else's dashboard, where it is far more accurate and not yours, and where it is no longer visits being counted but people. A number in the open is also a claim you can be held to. Five hundred looks thin. No counter is safer.

The guestbook. A stranger could leave a trace there without holding an account anywhere. What killed it was spam. An open writing surface on an open network lasts about two weeks. And once it was gone, the conversation about your page had moved out anyway: it is written in a thread somewhere else now, by people who are not on your page, and you rarely get to see it.

The counter at the bottom here is real and counts what it says it counts. The guestbook exists only in 1998 mode, where it is a link to the contact form. That is not a cheat. It is exactly what a guestbook was.

And what you are standing on now is exactly the same urge as the RF transmitter on Romsås: make something on a machine and get it out to people. The difference is that the machines from back then are in here, and they still start.