When France gave away terminals, and a dozen email services bloomed

Consider the phone book.

What started as a single cardboard sheet listing 50 names with whom the phone operator could connect you, two years after the first telephone was invented, had ballooned by 1975 into a multi-volume directory in Paris that ran 1,480-pages long for the A to I listings alone. And by 1981, listing every phone number in France required 95 volumes weighing 110kg in total; distributing those numbers required printing 16 million phone books (listing a single district’s numbers, in a more reasonable book or three) every year.

Numbers aside, it was in many ways a modern marvel. For a brief period of human history, it was possible to find any other human’s phone number, as long as you knew their name, from a directory that was mailed, gratuit, to your home. You could meet someone, a Jean-Pierre Durand or Marie Dubois, look them up in the phone book, and if not find them directly at least narrow your search to a few dozen of the most likely suspects.

But the numbers could not be ignored. The phone book, it was estimated in 1979, lost the telephone company 120 million Francs (around $120 million at the time) and used 20,000 metric tons of paper annually in France, volumes that were predicted to grow five times by 1985. A world awash in newsprint, forests raised to find phone numbers, telecoms bankrupted by the bulging directories, that would not do.

So France made a computer terminal, in an attempt to substitute silicon for paper, lent one to French households for free, and inadvertently recreated email, again and again, along the way.

The Télécom Télétel

The idea was similar to one a startup, Micro-soft, was already peddling: A computer in every home. Instead of printed phone books, France could put a computing terminal in every home, where the French search through an online database and find people’s names and numbers. That, and more: They could read the news, check the weather, book plane tickets. The mind boggled with the possibilities.

“The instrument is the computer; the concept, information,” as The Computerization of Society, a 1978 report written for French President Valery Giscard d’Estaing, defined the future.

But it wasn’t exactly a computer that the PTT, for the ‌Postes, Télégraphes et Téléphones body that ran the French postal, telegraph, and telephone services, envisioned replacing phone books. Computers were expensive. The Apple ][ retailed at the time for around $1,300, the Commodore PET for $795. PTT wanted to spend only $50 per machine, affordable enough to offset the cost of printing and shipping phone books within a few years.

Plus, home computers weren’t useful for all that much on their own. They required expensive software and technical know-how to operate, and besides, it was networked computing where the true excitement lay. The earliest tech entrepreneurs cut their teeth on timeshare machines, with a “dumb” terminal in their office connected to the remote server over the telephone lines. The screen and keyboard could do little more than relay characters and process received bits; the software and data lived on the server.

That put far more computing power at one’s fingertips, for the cost of a metered connection fee, than a personal computer at the time offered. It also enabled the earliest of what we’d come to know as online communications. Any one terminal user could see any public files and databases on the shared server, as early web sites. And they could leave mail messages for other users, in a simulacrum of today’s web apps. “It was all of this online interaction,” as Ray Ozzie described the University of Illinois PLATO networked computer. Any terminal-based setup could offer similar advantages, for the low price of a monitor, keyboard, and networking equipment.

It was exactly what the French were after: They wanted to replace paper directories with an online, always up-to-date database, something even the best personal computer could not offer and a terminal could. Tantalizing, the terminal would also boost usage of France’s telephone network, potentially saving printing costs and growing revenues at the same time.

It was a gamble. “Either you target a small audience—in which case computer terminals suffice,” as Le Monde explained it, “or you aim for a mass market (which was the case here), requiring enormous resources to make the devices available at an affordable cost.” The final price of France’s mass-market Télétel—what’d come to be called the Minitel—ballooned to $140 per device.

But the gamble paid off. The payback time shortened from a projected five years to “four years in light of higher than projected revenue flows.”

And, to the average person, the Minitel was free. In an innovation familiar to anyone who’s acquired a “free” iPhone thanks to an expensive monthly carrier deal, the Minitel devices were loaned to French households by PTT. Acquiring it was easy; using it, though, cost 1 French Franc (around $0.30 then or an inflation-adjusted $0.50 today) per minute to use. Thus the low payback time, and the high phone bills French became accustomed to as every minute they were connected to the Minitel network, looking up phone numbers and train tables, was added to their phone bill.

Free, though, made Minitel tantalizingly hard to ignore. “Anyone curious about the new system being promoted on TV could simply go to the post office and return home with a shiny new Minitel box,” relayed the IEEE of Minitel. Two years after launch, “by 1984, six million citizens were reading newspapers, sending emails, buying train tickets, paying taxes and engaging in scandalous chat on the Minitel system,” says the Computer History Museum. And a decade or so later, “the French owned about 9m Minitel devices, with 25m users connecting to more than 23,000 services,” reported The Guardian. It faced little of the chicken and egg problem of not enough people versus not enough to do on the network that other computing networks would encounter, with over half of the French population online at Minitel’s peak.

It was also easy to use. “All you had to do was plug the terminal into the wall, dial the local gateway, et voilà, you were transported into cyberspace,” explained the IEEE of Minitel’s URL-type addresses. The address book, Annuaire Électronique, was free to access by dialing the short code 3611, while other codes required a number and name and cost per minute, like 3615 SNCF to book a train ticket.

It wasn’t a closed system, limited to the Télécom, either. Anyone with a personal computer or server could start offering services over the Minitel’s X.25 network (a sister set of networking protocols to the failed X.400 email standard) and register a shortcode with the Télécom. Better yet, the phone company shared over 60% of revenue with service providers. The longer someone used your service, the more you’d make once they paid their phone bill at the end of the month, without needing to build out a new payment system. There were even multiple rates: “3614 was cheaper, and 3616 more expensive,” explained @wazoox. “Some services were accessible through several prices, for instance 3614 AIRFRANCE and 3615 AIRFRANCE: you dialed the 3614 first, and if there was no line available, tried the more expensive (but otherwise identical) 3615.” That turned the network into a highly-profitable ecosystem, where publishers and service providers rushed to offer any computerized service people wanted.

As it turned out, what people wanted most was to talk.

Let the people talk

The Télécom envisioned people checking databases. And that they did, looking up phone numbers and checking the price of pork. But also, along the way, they learned to use the Minitel to communicate.

Each Minitel service was an island unto itself, reminiscent of web pages. You’d dial into one service to book flights, another to book train tickets, another to check the weather. Each service lived on a single server, making communication possible with anyone else logged in to that service, and notifications of new data only possible when you reconnected to that service.

The earliest network started out with a few services built to test out the system. One, M3V, was built to send feedback to the team with electronic mail, mail that you could also send to any other user of the M3V service. And that quickly became one of the most popular Minitel services, one users lobbied to keep alive long past its intended expiration date at the end of the original testing period. “Five service providers alone accounted for over half the calls,” said the telecom of its 2,500 first users; one of those was the burgeoning electronic mail service.

Other mail services popped up from the most unexpected places. The Les Dernieres Nouvelles d'Alsace newspaper decided their Minitel service, Gretel, should also offer a way for readers to get in touch. “We realized we needed to be able to explain to people who had lost their way how to find their bearings again,” as developer Michel Landaret was quoted in The Minitel Saga about their system that let users exchange three lines of text. “We devised an in-house electronic mail service; they discovered it; they used it; and then the flood gates opened.”

Messaging worked in unique but similar ways on the official 3615 TELNET service and the cheaper 3612 MINICOM from France Télécom, in experiments like M3V and as unexpected extras in services like Gretel, and in dedicated popular messaging services including GEN4 and RTEL. Typically you’d claim an anonymous username, and could send a private message to anyone else on the service. Some offered additional features; “A key feature of M3V is the ability to quickly send the same message to a group of subscribers,” noted one guide, in an early version of the email newsletter. TELNET let you send SMS messages. RTEL and the official MINICOM service, by 1997, would offer gateways to real, internet email—but inside Minitel, every message was kept inside its own service. You had no central mailbox, no offline list of mail on your Minitel device. Each message was stored on its respective service, and you had to check each one to see if there was anything new.

You might message one friend on one service, log into another to send work messages to colleagues, and then dial up a new service in pursuit of new friends and interest groups, in a bifurcated messaging space similar to DMs inside of social media apps today. Hitchcock, the pseudonym of one Minitel user interviewed by Le Monde, mentioned having “become a regular user of the two or three messaging services he now frequents regularly, after having tried almost all of them.”

“The Minitel is at once a workplace, a bar, a Monopoly game, and a poker game,” said Hitchcock. And an addictive one at that.

Messaging, one email-of-a-sort at a time, was useful enough. Gretel and others, though, would morph into real-time chat over time, displaying messages from other users as soon as they came through. And that kept people glued to their screens. The longest uninterrupted conversation went on for 72 hours, reported The Minitel Saga, with one user logging on for 200 hours a month in total and another with the longest letter at 630 pages long, propelling Gretel by 1988 to account for “40% of mass-market Minitel traffic.”

There were friends and family reconnecting, businesses corresponding over the early network. That, and the notorious “pink” services, adult chats hosted on Minitel and advertised across France, accounting for “revenues of about $400,000 a month and that as much as half of France Telecom’s Minitel profits,” as the LA Times reported in 1989. That brought up questions of censorship and common carriers, long before the rest of the world worried about such things. “The difficulty is that we consider ourselves the carrier,” said Minitel legal advisor Claire Mialaret. “We don’t want to be considered as a censor. As a public service, we don’t want to be in the position of dictating morality.” Yet it didn’t sit easy with some that the telecom service and leading newspapers were profiting from something they’d rather sweep under the rug.

“We are on the threshold of a far-reaching phenomenon whose significance is equal to that of the advent of the railroad or the airplane,” said French engineer and Minitel leader Gérard Théry, as the network took off.

But it was enough to give Gretel creator Landaret pause.

What hath France wrought?

“He who owns the wire is powerful. He who owns the wire and the screen is very powerful. He who owns the wire, the screen, and the computer has the power of God.” ~ French journalist George Suffert

Minitel was clearly a success, in France at least, enough that Gretel alone in 1986 brought in “13 million francs in annual revenue, with 27% profit before provisions and taxes,” purely from their profit-share from the Telecom. “The economic success is undeniable,” said Landaret. “But the human impact is debatable.”

It was a force for good, in many ways. Farmers called it a revolution, connecting their remote regions with authorities to report animal deaths or labs to check milk tests. Clearly half of French households found it useful enough to keep around, enough that the network was kept around until 2012, long after the internet rendered it obsolete.

Yet it also surfaced the same human problems that the internet would, years later, in its early anonymous spaces. “Gretel and its numerous user-friendly imitators are responsible for about fifty divorces and half a dozen suicides,” worried Michel Landaret. “Anonymity allows people to get away with many things verbally, and sometimes even with actions,” he said. “Everything happened too fast, without study, without reflection, on the phenomenon. And this acceleration is causing damage.” Damage, enough, that the record-breaking chatter who talked for 72 hours had to be “committed to a psychiatric hospital.”

One day the phone company decides to create a terminal to replace the phone book. Another, users turn an internal messaging tool into proto-email. Then come the hordes, and it's hard to trace back what went wrong.

Minitel indeed may have been a victim of its own success. It had all of the data you needed, and tools to message your friends. It even had a business model that paid service providers, long before App Stores and paid subscriptions. What else would you need from the internet? Yet it was a Galápagos Island to itself: Minitel owned France and failed to catch on elsewhere. By the time bridges between Minitel’s early messaging services and email popped up, the rest of the world was rushing headlong into the internet and the dot com bubble, leaving the Minitel network increasingly isolated.

“Telecommunications will, in time, deprive the postal service of its most profitable activity,” concluded The Computerization of Society’s report. And so France raced ahead, replacing paper with screens. For a few decades, as the Concorde ruled the skies and the TGV made high speed rail a reality, Minitel digitized French communications, funneling revenue into the French postal system, telecom, and service providers alike years before other countries would go online. It was built as a digitized phone book; the French turned it into a destination to hang out online. Today, its descendants are in large part why we stopped memorizing friends’ phone numbers decades ago, and the first manifestation of a future where phones would, instead, be internet communications devices, for better or worse.

People like communicating, and the phone book was just the first step towards always-on connection.