Hacker News new | past | comments | ask | show | jobs | submit login
Minitel, the Open Network Before the Internet (theatlantic.com)
86 points by ForHackernews on June 17, 2017 | hide | past | favorite | 53 comments

Minitel had one huge drawback: it was too good. It put France a solid 4 years behind the rest of the world when the web hit. Everything in France was done via Minitel, and in the beginning the web was still so small and capable of much less that the French (rightly) laughed at it as an inferior version of what they already had. But that changed rapidly and before long the head start turned into a lag and a ton of inertia. The world had nothing to compare the web with so that's what it moved to at an incredible clip. Roughly around 1996/1997 France saw the light and then they moved quite rapidly.

Even so it took until 2012(!) before Minitel was finally decommissioned.

> the French (rightly) laughed at it as an inferior version of what they already had

What they didn't see though was that Internet was a decentralized, truly democratic network unlike Minitel and the likes (e.g. CompuServe, AOL, Bitnet I think?). For its time Minitel was quite good aesthetically and in terms of UI, but as a centralized consumer-oriented network it had no chance to compete with the Internet.

It was not totally centralized technically. You could host your own full service, using your own modem banks and servers. The drawback was that you would not be discoverable in the minitel directory and you would not get the billing system etc. People needed to know your phone number and dial it to connect.

Minitel took a long time to die because it was a cash cow, both for France Telecom and service providers. The economic model was very simple: users pay the time spent online, and the revenue is shared between the telco and the service provider. No ads, no tracking ;)

It was also successful because every household got a Minitel for free in the mid-80's, when personal computers were very uncommon. And if you had a computer, you could use the Minitel as an external model with a serial port adapter and reach BBSes! Because of that I once had to explain to my parents why I was spending hours calling a number in Germany, which was not a cheap thing :P

When I was in middle school, I was participating in a meteorological club, and we had such a minitel server.

People from around could phone in with their minitel, and get the information we'd put there for them. The few kids in the club collected data from a small weather station, compiled it, and put it there. We'd also sometimes redesign the presentation of the minitel "site".

We were coached by a maths teacher, and that was part of what got me into computers (I had an Oric Atmos at home, the club was the first time I had access to PCs (with Windows 2! But we'd most usually run DOS); we had access to some Thomson MO5 in primary school before that).

IIRC some variants of the Minitel had a RS232 port. I've seen it used to program alarm systems on site.

30 year ago, I hooked my mother's Minitel to my Atari 800 XL's joystick port, using (IIRC) a couple of transistors on a breadboard, and a software serial interface that probably consisted of a few tens of lines of assembly.

This was the summit of my hardware career ;)

That's more than most programmers.

IIRC, all the minitels above and including the 1B had a serial port, but it was through a DIN connector. We were using that to test our "site" by feeding it directly from a PC.

I'm not sure the Minitel 1 had one.

Even the Minitel 1 had a rs232 port, on a non standard connector. Super slow though, 75/1200 bauds :)

> every household got a Minitel for free

People got minitels (available for free) because phone books were eliminated, so even people not interested by the online world (shopping, train/flight reservations, dating "sites", high school diploma results, weather forecast, etc...), would get one in order to be able to look up phone numbers.

Phone books were not eliminated. Minitel gave you access to the phone numbers in the whole country, whereas phone books were limited to your department (for foreigners: France is divided in about 100 departments).

Seems correct - weird, I was sure this was the case... They are eliminating the white pages only now:


So it was only hosehold with a minitel that stopped receiving white pages:

"les foyers équipés de Minitel ne recevaient plus que les pages jaunes" (Homes with Minitel received the yellow pages only)


You ll find more about the minitel history here: http://www.inaglobal.fr/telecoms/article/du-minitel-linterne

Phonebooks ARE not eliminated. I got one last year (it was a pile outside and one could grab one).

You got it wrong. To get a minitel without paying for it you had to opt out of receiving the local white pages phone book. Minitel costs was marginally recouped by saving on printing and recycling phone books and mostly by paid services.

The white pages phone book was still around last year.

Also with 6.5 millions minitel owners at peak in 1993 we were far from every household having one.

The minitel was centralized technically as in dumb terminals connecting to a smart server[1]. Everyone wanting to connect to your own hosted full service would have to connect to a central point: your own hosted full service. But there was one service, 3618 or service MIAMI, that allowed direct connection minitel to minitel but this was using a different network and stayed available 2 years after the minitel network was shut down.

The economic model was definitely not what you said, there was a wide range of possibilities: fully free (3623), free with a time limitation (3611), paid for by the service provider (3613), paid for by the user with no money for the service provider (3614), paid by the word (3656 to send a telegram), paid by the user with money split with the service provider a.k.a. kiosk debuting in 1984 (3615), overcharged services (3617, 3628, 3629). Some had a fixed costs at connection.

Ads were a thing minitel services, from a simple banner to a full page interstitial graphics that you had to wait a little while for it to load. Tracking also though much less elaborated and certainly not exploited the way it is today.

Minitel was not given to all for free, it was offered for a limited time to people in exchange for opting out of receiving the free local white pages phone book, this economic model of giving away the terminal and recouping the costs through services prices was much criticized but has now become common (printer, smartphones, ...). Then you could either rent or buy a minitel, usually a more evolved version.

The use of the minitel was mostly the phone directory, buying train tickets and erotic messaging. The kiosk where the service provider get its share of the money made the minitel successful. Among services were 3615 internet to connect to the internet with your minitel which paid for altern.org and ultimately lead to gandi.fr, also notable is Xavier Niels who debuted his career to becoming the billionaire behind free.fr and proxad.net as a high schooler with a minitel erotic messaging service.

To get deeper see this INA article[2].

[1]: This is the basis of the famous talk "internet or minitel 2.0" from 10 years ago about the evolution of internet https://www.fdn.fr/actions/confs/internet-libre-ou-minitel-2...

[2]: http://www.inaglobal.fr/telecoms/article/du-minitel-linterne

> Internet was a decentralized

That's the key word: was.

Now that the web has mostly eaten the internet, the network has been centred around a few nodes. And we still have the asymmetric bandwidth that is the primary driver of centralisation.

Make no mistake, the internet has become just a centralised and consumer-oriented than the minitel was. Even I don't host my own web site, for 2 reasons: HTTP doesn't scale, and I don't have enough upload.

If your website is small enough, not getting many visits, does it need to scale?

CDNs are becoming a cheap commodity.

> CDNs are becoming a cheap commodity.

And then all sites are centralised again at Cloudflare and a handful of others.

There are more and more cheap CDNs out there, not just Cloudflare! A lot of them will cost you the price of a VPN or the cost of the electricity to run your own server.

Most CDNs, and especially the newer ones can work with things like the Cache-Control header, which means that for a basic small site you don't even need to get into their console or know much about how one particular CDN, and ti means that switching to a different one is as simple as a one DNS record change, hence my saying that they have become a commodity, completely the opposite of your assertion that "all site are centralised".

Note that I do think that the internet is becoming more centralised, but in my opinion the issue is long term maintenance. Small site run by a single person will disappear when that person stops maintaining it, or dies... I have no solution to this issue, but it is one of the reason I give to archive.org.


Forget about a small technical blogs with almost no multimedia content (just text and code). They're not the problem. Or maybe they are… One reason my web site is light is because of the scale issue.

Think a mildly popular vlog, or basically any YouTube video you've ever watched. I expect the costs so soar pretty quickly, if one wants to sustain the slightest Reddit Hug.

Thing is, it doesn't have to go through the web. Bittorent for instance handles massive distribution in a way that doesn't DDOS the sender. It doesn't work as well as it could however, because our collective upload is severely limited.

Also, the fact one has to rely on an external provider at all it already a problem.


The internet is more centralised than you realise. To take just one example, think about how much Alphabet actively spies on for profit: most web searches, analytics for most popular web sites (found this with NoScript), all the content of Gmail, including when one is not on Gmail (even I send email to Gmail users).

Sure, we can still escape it. But there will come a time where our choice will be much simpler: either we use some big provider that will spy the hell out of us, or we stay out of the network altogether, effectively ceasing to participate in society at all.

I have a Facebook story. I first subscribed under pressure from my peers from my music club (ostensibly because it's more convenient). I got flooded by notifications I did not want. I cut the notifications, but still didn't get into the habit of checking out Facebook every day. Then I started missing things that were talked on Facebook (decisions, invitations…). And then I got told "you didn't know? But I sent the email…" (he didn't).

I have since shut down that account, and vowed never to use Facebook ever again. Next time a group forces me to use Facebook, I'll consider not being part of this group.

There was a minitel service "3615 internet" that offered internet access to minitel users, the guy behind this service valentin lacambre used the money to finance altern.org.

Here's an interesting snippet of history from "une contre-histoire des internets": https://www.youtube.com/watch?v=A2hA3Bvm1n4

>Even so it took until 2012(!) before Minitel was finally decommissioned.

As a point of interest, there are still buildings that use (local) minitel systems to manage security, elevators, electricity, etc. It's not uncommon to see contractors with a minitel terminal in the back of the truck.

This seems similar to the situation the US is in with broadband and mobile. We already had this big decades-long investment in copper (cable and phone) and old cell towers, so it was easier to just keep using, maintaining, and expanding it while the rest of the world got (or is getting) affordable high cap (or no cap) 4G and fiber to the home.

Comcast is in the middle of rolling out gigabit everywhere after Google lit a fire under their butts, so I guess we'll catch up with the rest of the world eventually in our own stumbling, expensive way.

The France Telecom monopoly on telephone line probably did more to slow down broadband adoption that the minitel.

Both actually, France Telecom (now Orange) was not ready for broadband, their infrastructure needed an overhaul. At the same time they had the minitel where they were the middle man with a 45% share of everything for them, and almost all internet was through PSTN which made them money too. All in all: a multi billions incentive not to develop broadband internet.

At the time Orange mostly worked to block competitors while overcharging their own services. Ultimately they got sued, lost after years and paid a record fine amounting to a fraction of their profits.

So they continued engaging in this practice to this day with now quite a record of record fines for abuse of dominant position and such. Recently there was a year they made an unexpected bump in yearly profit because they had secured money to pay the next big fine which didn't happen.

Right now Orange is in the process of securing a monopoly on french fiber on public funds.

I don't think so. There was a huge raise in prices when the market opened.

I don't remember that. I mostly remember that I could finally use by 8Mbit modem at 8Mbit instead of the 1Mbit or 500kbit (can't remember which) to which France Telecom was artificially capping it.

Are you talking about local-loop unbundling ?

As stated by Jacquesm, Minitel slowed down Internet deployment in France. Every bank, mail ordering sellers or alike already had their Minitel "site" (in which they invested money) so jumping in the new train took time.

At the end of the 90's, I had to develop a Minitel Emulator for a French bank who wanted to put PC based kiosks where people could have many informations (static coded texts) but also connect to their accounts. Not a web kiosk (we were specialized in that at the time). A minitel one.

Just for the record: We had a similar experience with banks in Germany, it was called "Bildschirmtext" or "BTX". It took a long time and a lot of emulators until they finally adopted the Internet. Besides this, BTX was not as omnipresent as Minitel was in France.

  "Bildschirmtext" (BTX) is almost as old as Minitel and 
  technically very similar, but it was largely unsuccessful 
  because consumers had to buy expensive decoders to use it. 
  The German postal service held a monopoly on the decoders 
  that prevented competition and lower prices. Few people 
  bought the boxes, so there was little incentive for 
  companies to post content, which in turn did nothing to 
  further box sales. When the monopoly was loosened, it was 
  too late because PC-based online services had started to 

Surprisingly, the Internet is becoming centralized like the Minitel was. Is history repeating itself?

I remember discussions circa 1995 on how the web becoming commercial would ultimately destroy the democratic potential of the internet to mirror the issues of capitalism such as centralizing power and wealth.

In 2003 (was it 2001?) there's the 2nd edition TAZ preface by Hakim Bey[3], in 2007 there's this infamous talk "free internet or minitel 2.0"[1] at the Libre Software Meeting[2]. Closer to now there's RIPE publishing "The death of transit"[3] and recently Peter Sunde told us we have lost the internet and it's all about damage control now.[4]

I would not say it's a surprise, it's simply the consequences of applying the same model that's making our civilisation fail, removing the conditions favorable to life on the planet, causing the current mass extinction, etc.

[1]: https://www.fdn.fr/actions/confs/internet-libre-ou-minitel-2...

[2]: https://en.wikipedia.org/wiki/Libre_Software_Meeting

[3]: https://labs.ripe.net/Members/gih/the-death-of-transit

[4]: https://thenextweb.com/eu/2017/06/09/pirate-bay-founder-weve...

[5]: I think perhaps the least useful part of the book is its section on the Internet. I envisioned the net as an adjunct to the TAZ, a technology in service to the TAZ, a means of potentiating its emergences. I proposed the term "Web" for this function of the Net. What a joke. Time magazine identified me as a cyber-guru and "explained" that the TAZ exists in cyberspace. "Web" became the official term for the commercial/surveillane function of the Net, and by 1995 it had succeeded in burying the anarchic potential of the Net (if any really existed) under a mass of advertising and dot-com scams. What's left of the Left now seems to inhabit a ghost-world where a few thousands "hits" pass for political action and "virtual community" takes the place of human presence. The Web has become a perfect mirror of Global Capital: bordeless, triumphalist, evanescent, aesthically bankrupt, monocultural, violent -- a force for atomization and isolation, for the disappearance of knowledge, of sexuality, and of all subtle senses.

Specifically, the Apple app eco system comes to mind.

Apple's apps have nothing to do with an open or closed Internet.

Of course they have.

When you have to use an app and a specific hardware to access content instead of the browser it's an open unless its closed internet (see tapas.io for one such example).

The walled garden model is causing the closed internet.

But people seem to think that and they have a reason. When apps replace web content, they quite clearly have something to do with the Internet. Web is part of the Internet. Isn't this obvious?

People like the guy I responded to think that because they have an axe to grind against Apple. Notice how he said "Apple's app ecosystem" and not "Android app ecosystem", when Google's platform is just as robust? And didn't Microsoft start it with ActiveX and IE? Or how about "Works best with Netscape Navigator"?

But no, it's Apple's fault. Everything is Apple's fault.

Maybe I'm wrong but my interpretation was quite different.

To me he's just pointing out a fact, to me along the 20+ years since I've been aware of the issue many things happened or didn't happen but a few specific things come out of the pack, including the one he cited.

To me you came out as the apple fanboy who cannot stand that apple is part of the issue and that it is undeniably apple who popularized the "app store" model, emblematic of the problem at hands. It's only after apple's success that the others followed suit and copied the model to recreate their own distinct impossible to interconnect ecosystem. Even Mozilla has one now.

Yes Apple is among the GAFAM who are to blame among many others for the current state of things, but ultimately the failure is in us who chose voluntary submission to new masters, who failed to use our technical knowledge to provide to other less technically inclined, who failed to force ISP to offer the means to self host, and so on

I guess this is quite an illustration of "don't argue with fools, from a distance people may not be able to tell the difference".

I've always wondered how much Tim Berners-Lee was inspired by Minitel given the fact that he must have frequently seen or even used them in the area around CERN.

The UK had Prestel [1], it wasn't nearly as widely used as Minitel though.

[1] https://en.wikipedia.org/wiki/Prestel

Funny anecdote : the founder of "free" internet provider, xavier niel, made his first bucks running a minitel service ( an erotic service iirc, what people called "minitel rose"), then later completely disrupted french internet dsl access by offering unlimited high speed access for a few bucks a month. I worked at france telecom at the time ( the old monopolistic company that owned all the coper lines), and i remember employees not believing this offer to be real.

Interestingly, the minitel solved the monetization problem as visitors would automatically be billed by the phone company at the rate chosen by the service they visited.

Somehow this does not add up with the PTSN era of the internet and with the current options to be billed on your ISP bill such as internet+[1].

[1]: https://fr.wikipedia.org/wiki/Internet+ http://internetplus.fr/

Rate and time spent on the service.

I never got to use a Minitel. :(

Don't forget Plato. That was my first experience with a networked computer. In about 1985 in Escanaba Michigan no less.

I remember chatting with someone on the Dungeons and dragons type Wizardry game & learning they were in California. Whoa! ( I was about 10 or 11 )


People who find this interesting may like to read more about the AlohaNET, PRNET, SATNET and AMPRNet.



Minitel was also used a lot for chatrooms (most of them for sexual purposes). Of course most of the women in those chatrooms where men working for the service. A few guys has made a lot of money with it (the creator from "Free" french internet provider). It is extremly funny because on the picture you can see a list of nicknames, it looks like some bdsm chatroom :)

Also Marc Simoncini (creator of meetic), and there's a few others.

The real minitel was a public domain modem comms program for MS-DOS/PC-DOS created by the brilliant Tom Jennings in 1983 (the FidoNet guy).

From https://en.wikipedia.org/wiki/Minitel :

The service was rolled out experimentally in 1978 in Brittany and throughout France in 1982

You might even say he invented minitel

Applications are open for YC Winter 2022

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact