I worked at France Télécom 1994-2000, including bringing up their ISP, Wanadoo. At the time Minitel generated $1B/year in revenue and they had clout in the organization, and risked stifling the Internet (it would have come anyway, but FT would have been left out). I can’t even recount the amount of time I had to spend explaining how web pages could not be priced on a time-spent basis like Minitel.
Hey Fazal, LTNC =) Before you arrived in NL I had similar issues with EuroNet management. I remember you telling me about your Java applet to access Minitel from the web. Good times.
Fond memories of lugging a 4.77MHZ Compaq "laptop" [0] to and from Paris to hook a Prestel game up to Minitel via X25.
Which fits the mindset as you say, since the game was priced per-unit time and BT were coining it in on the UK side (this wasl still a few years pre-eternal September).
[0] Plasma screen, 1 16-bit ISA expansion slot and 1 8-bit, built in PSU.
Interesting I worked on Prestel (Uk Teletext) billing for a while as well as the Telecom Gold (Dialcom) systems and they had super complex and expensive tariffs which stopped wide uptake.
We used to joke that for Prestel you could have a custom tariff for all services almost down to the user level - People whose name is fred and own a goldfish we used to joke.
Prestel Billing use to run on a Unisys mainframe and a super niche language called LINC - and was baroque and crufty to an extreme - there was a BBC micro that sat in a corner and Bonged occasional that was part of it (exactly what I never found out)
I was responible for, ah, certain outages at the time. Not in any malicious sense, but because my game (running on a 4Mhz Z80) managed to overload the whole network.
Ah no worked for Telecom gold near London bridge then network house in Aspley spit after workstyle 2000 when they merged telecom gold and prestels developers.
Did you know Steve Cornelius? my manger later on - you arnt that Hzaeii from Shades BTW?
Recently I worked at the next building over in the same building as the mermaid theatre and used to dink in The cockpit" the "old geezers pub" as a young colleague called it.
Do you happen to have more info on the people who worked on minitel? Maybe it's because most of the source material is in French, but it would be nice if I could flesh out my page on it.
I only joined FT in 1994 and most of them had moved on. My former boss Jean Lebrun worked on the marquee phone directory application, which was at the time probably the world’s highest transaction-rate database. I believe he was also involved in the abortive effort to sell the tech to Qwest in Denver.
Most of the R&D was done by the CCETT lab in Rennes, Brittany. Well into the 90s they were advocating for a graphics-oriented protocol named VEMMI to replace videotex instead of the web. Sort of a cross between X11 and a IBM 3270 terminal conceptually. There were a lot of pitched battles between the Not-invented-here and Web standards camps at FT R&D in those days.
As for the roots of the Minitel, they are a bit more prosaic. France believes in industrial policy. When president Giscard made upgrading the phone network a national priority after years of neglect under de Gaulle, the department became the largest government department by budget. When that started to taper off in the late 70s, they needed to find a new outlet for the manufacturing capacity that had hitherto being making Alcatel E10 TDM switches, and the Minitel was that outlet. One of my 6th-grade classmates was actually part of the Vélizy trial project. The paid external provider model was the result of the press worrying FT would compete with them in information services and insisting on an open model.
Later Minitels were VT100 compatible, albeit with an atrocious keyboard. I used one as terminal to connect to my university’s Vax running Unix over a 9600 baud serial line from my dorm room. Considering it was free when a real VT100 was more expensive than a PC, that was a substantial cost engineering achievement.
Thank you. I was wondering if we could continue the conversation via email? My email should be in my profile. No worries if you don't have time. Mine is limited as well. I just feel that there's a lot history being lost when we don't record any history from engineers who weren't project heads or the original creator. In Minitel's case, it feels worse because it was hard to find information on Bernard Marti, at least in English. I also get this sick feeling knowing that many of the 20th century's engineers have been dying and much of that history is lost forever.
I was 16 years old.
It was 1986 I believe.
With my twin brother we discovered the Minitel in our parents's bedroom.
We spend all evenings on the minitel after class.
Boy it was fun !
We had 'special' codes so we did not pay as much as adults. And we chat with Men (we pretended to a woman). We laught at every kinky sentence they made to us. It was like a writing experience. We chatted with 8 men at th ame time.
Hopefully we never told our parents aout what we were doing..
It was my first dating App experience. 10 years before having a Web connexion.
Imagine :
- All french citizen had a free minitel device (like a mini computer). It was amitious for sure...
- Now the only problem with State driven technological projects : Once it was delivered, Th french telecom did not continue to improve the system. Oterwise, France would have been the center of the Web I supose...
> All french citizen had a free minitel device (like a mini computer)
Nitpicking: AFAIR, the Minitel was not "free" but rented - just like most ISPs in France today rent an ADSL modem-router.
Also, the Minitel was not really a mini-computer but merely a screen, a keyboard and a modem in a single box. People could tell the difference with a computer, because at that time 8-bit micro-computers (Amstrad CPC, Commodore C64, ZX Sinclair, Atari ST, you name it) were not uncommon at home.
> Now the only problem with State driven technological projects : Once it was delivered, Th french telecom did not continue to improve the system
Actually, they did. The French Wikipedia page [1] is much more detailed than its English language counterpart. The latest model could display JPEG black-and-white images, and spoke at 9600bps (versus 1200bps for the first gen.). But it is true it was too little too late; it could not rival the multimedia PC/Modem combo that started to be affordable for the masses at that time.
The (maybe?) too slow evolution of the tech probably had to more to do with the huge userbase and backwards compatibility issues than with "State driven tech projects" (BTW French civil nuclear power and high speed trains were for a long time "State driven tech" and are still alive and more-or-less kicking). Just like IPv4 seems to not want to die today.
You're right. And to be more precise, French government at that time pushed its citizen to use the minitel:common example that come at the top of mind: when I was younger minitel was mandatory to register yourself at school or university which explains the quite important development within the country.there were even store offering minitel service.
However this monopoly position were disrupted like any other monopoly by better and more advanced technology.
Speaking for myself, my friends and I would do this as teens in '99 with a open protocol chat program called MIRC. This was chatting with strangers, and back then everyone opened conversations by asking "ASL?" which meant "Age, Sex, Location".
We figured out pretty early on that by saying "Female", you opened yourself up to a torrent of creepy come-ons, which to us was absolutely hilarious. Here was this complete stranger who did not have so much as a profile picture to go off of, either professing his undying love to you after a few messsages, or straight up using abusive and degrading language.
As an old millenial now, it's been pretty shocking to me how open Gen Z is about revealing numerous identifying details about themselves on social networks. I'm under 40 and we were raised to protect our privacy online, and this was when surveillance tech was in the stone ages.
I believe it is a mistake to think that. I have some notes from an undergraduate course in digital signal processing, dated 1979. There is a section written by someone from "Joint Speech Research Unit, Cheltenham" there can only be one organisation in that town interested in researching speech, good friends of No Such Agency.
But it's precisely that. The prevailing attitude is that, well, advertisers are going to track that you googled "ob/gyn prenatal care XYZ area" and hostiles are going to be able to doxx you anyway -- you have to take proactive measures to ensure privacy so fuck it.
I think it's a pretty broad hobby for early teenagers and children. For me personally, I just thought it was funny. It's also interesting, because you find out very fast that there's some guys just a little too into it when you say you're a 13yo girl. That I think informs your behaviour online also. It's a consequence-free way to learn about loads of impulses other people have, without the shadow of adult oversight.
When I was a young teen my friends and I picked female avatars on Runescape because we'd get free items from presumably male players. Sometimes you'd have to 'flirt back' but inevitably you'd get some good items, gold, and some creepy messages. As a 14 year old kid, it was hilarious to think about some sad older man flirting via Runescape gifts.
I had this happen as recently as probably 2007. I was in my 20s at the time, and it also happened in an online game (WoW). Albeit, my experience certainly wasn't deliberate.
For whatever reason, I've almost always preferred playing female characters, but it quickly became an eye-opener to me how creepy other dudes are online. One of the more abusive of these sorts (funny at the time until I realized it was a behavior he probably continued to actual woman) was when I'd gone in to help some RL friends who were stuck in a dungeon, and they had this random chap in the group. As soon as I joined up, he started hitting on me. I ignored it for a while until it started getting really strange, and I told him rather flatly:
"I'm a dude, dude."
He went ballistic and called me all manner of names, inferring that I must clearly be homosexual because I was playing a woman (not sure how that works). After launching into his insults, he quickly left the group. I think he tried harassing me via private messages, but I blocked him by that point.
It occurred to me sometime later that the fact he realized he was hitting on another guy must have challenged his sexuality and made him incredibly uncomfortable doubtless leading to his anger and lashing out.
I noticed this behavior dropped off dramatically in the years following, either because more dudes (like me) were playing female characters and the creepy sorts realized it was a minefield to take the risk against random other players who may or may not troll them. Or perhaps harassment of the sort started to be taken seriously by the moderation team as the number of women playing increased.
I have reasons to doubt the latter and suspect it's a combination of the former. I also suspect most of the harassment was probably concentrated in guilds by that point where I wasn't as likely to see it firsthand as I played largely with people I personally knew. Leastwise, the weird random encounters dropped off precipitously within about a year.
It's true: The experience was eye-opening to me, and that was what--13 years ago? Actually longer, now that I think about it, since I started playing at release in 2004 (also on female characters).
And you're right, I think it's a valuable experience that has to be seen to be believed. It's easy for me to post about it here, but it's difficult to convey how annoying it was (at best). Not simply from borderline sexual harassment either. There was, at the time, an undercurrent of harassing accusation that any dude playing a female character clearly had to be "gay." Looking back on it, this whole idiotic and immature belief was laughable, because all the gay guys we had in our guild played male characters. They also gave me a hard time for being hit on with some regularity, so at least there was some levity among our group to make light of an otherwise sad and disgraceful situation.
Unfortunately, that's not the only circumstance I can remember. There's easily half a dozen others varying from odd to outright creeper material that stand out, though not as much as the one I shared above. I figure if it happened to me that frequently, it's absolutely happening to women whom other players know are women MUCH more often. It wasn't just limited to long term encounters, either. Just standing around waiting for something as a female character in the early days of the game sometimes painted me as a potential target.
A friend of mine who also played female characters--and who moonlighted as a stand-up comedian--had similar experiences, but his dialog was absolutely hilarious to watch.
I am pretty certain it’s especially fun for kids to discover what men are into so that when they finally grow up they could be better in touch with their own needs
Poster said he was a teenager. For adolescences, everything around sexuality is interesting. Some of those men he was tricking were probably other teenagers pretending to be adults!
Yes, I was 16 as well in 1986 and was also interested in technology, especially when it was called 3615 ULLA (https://www.youtube.com/watch?v=9fmsF6PDXwc, probably NSFW because it is horribly politically incorrect today, like most of what we had in the 80's)
On the serious side - Minitel was such an incredible experience! The only feeling similar to that was the first email I sent in 1992.
I feel it is a horribly lost opportunity to lead the web, teen like us were driving the idea a lot (I was trying to explain my parents to not take the phone book anymore becauseit was all there, a search away...)
By 1988 three million terminals were installed, with 100,000 new units installed monthly. The telephone directory received 23 million calls monthly, with 40,000 updates daily. About 6,000 other services were available, with 250 added monthly. France Télécom estimated that almost 9 million terminals—including web-enabled personal computers (Windows, Mac OS, and Linux)—had access to the network at the end of 1999, and that it was used by 25 million people (of a total population of 60 million). Developed by 10,000 companies, in 1996, almost 26,000 different services were available. [0]
A similar (yet not nearly as successful as Minitel) service existed in Germany (BTX [1]), and I remember my uncle still using it for online banking in 2005 (it was officially shut down in 2001, but online banking continued to work until around 10 years ago). He used a special software for Windows 98 which acted as a kind of BTX browser, so for around 10 years he used BTX and the Internet in parallel, on the same machine.
The failure lies in the ownership model: a telco company trying to build a walled garden. Other walled gardens failed as well of course including some that were basically much more competitive in terms of networking. E.g. AOL failed a few years later because the world outside their walled garden was more interesting. It never really stood a chance of surviving despite being much more similar to the internet at that time.
Operators trying to build walled gardens is a pattern that repeated itself until it became such a low margin business that the only operators able to make a profit were those merely offering full internet over 3G and 4G and not even bothering with the black hole of R&D needed to keep up with what was happening elsewhere. There were attempts in both standards to grandfather in a lot of stuff that is now exclusively done over IP but none of it ended up being popular. The history of mobile phone operators is just a long succession of attempts to copy things from the internet at the time and trying to turn them into walled garden things. Each of those things failed. We're not watching TV via operator provided services. We're not doing video calls via operator provided services, their game portals have long died and faded away. We're not browsing any WAP sites. Instagram wiped out what little adoption MMS ever had. None of it stood a chance.
The incentives and ownership were just wrong and the role as operators as essential middlemen just never panned out in the real world as they hoped.
Many telcos of course managed to convert themselves into successful mobile operators ultimately but their role was diminished to providing network access. My android phone has a phone stack. The only stuff that comes in via my phone number is basically recruiter cold calls. It's a small internet device with some quaint legacy features that I rarely use. Really, numeric identifiers that you have to punch in? In 2020? How is that still a thing? I remember memorizing shit like that a few decades ago.
So, minitel is merely an early and probably expensive incarnation of that. France telecom subsidized the devices and had to earn it back via regular network charges. I doubt that generated anywhere near what they hoped in revenue.
Yeah but their business model was different: they started with a compelling garden and then users convinced themselves they needed to be in it.
Basically the telcos came from another direction: "hey our users are not paying us for that stuff they are doing with some other company on our network! Lets fix it by copying that stuff in some half-assed way and then charge for it."
That's a bad idea because now you are charging for stuff that users are already getting elsewhere that they don't really need you for. So, you basically end up with an empty room problem. Worse, due to the network effect some of these things don't make sense if they are operator specific. E.g. An O2 only social network or video call service is just not a thing. So, now we are talking standardization and design by committee. This is exactly why most of the stuff in 3G flopped (other than IP networking): they were years late to market with stuff that barely worked with basically noone caring about it.
> My android phone has a phone stack. The only stuff that comes in via my phone number is basically recruiter cold calls. It's a small internet device with some quaint legacy features that I rarely use.
Funny how people live in totally different bubbles. I see you are based in Berlin (I'm in Denmark) and I have no idea how it is there but here I -as a geek with very geeky friends- don't know anyone who use their phone as you describe. Sure a few calls are via Facebook with video and some messages too (though many are starting to refuse to use social media) but the overwhelming majority of use is normal old phone-calls, SMS and MMS. Inside my bubble things like you describe (internet based stuff) is where "old people" have moved and younger people are moving away from. Especially those who hate all the data mining done by Apple, Facebook, Google, etc. I don't have a single app on my phone with those capabilities you use and I'm not missing out. Of course I'm sure other very different bubbles are right next door :-)
>Operators trying to build walled gardens is a pattern that repeated itself
Phone calls or text messages are rare for me. I can't remember the last time I sent an SMS message. I've not received an MMS message in over a decade and even then that was relatively rare (less than 10 times in the years that that was a thing at all). So, that stuff is completely dead as far as I'm concerned. I don't think most young people even know stuff like MMS exists. Did Apple ever bother to implement that at all on IOS? As I recall they were not in a hurry to implement that when the iphone launched originally.
The only numbers I call are people like my Doctor to make an appointment because they don't do email for whatever reason (which would make a lot of business sense for them). But mostly my phone is a mini tablet for me. Young people here mainly distinguish themselves by maybe using different networks.
People who hate Facebook are common. People who understand Facebook actually also runs Instagram and Whatsapp are less common. Last time I checked they are still getting a lot of daily use, also from Denmark.
> I don't think most young people even know stuff like MMS exists. Did Apple ever bother to implement that at all on IOS?
iOS supports SMS/MMS in iMessage. These conversations have green bubbles instead of blue bubbles[0].
This occasionally becomes a status symbol, where showing up as green on someone's iMessage indicates that you don't have an iPhone, and thus are poor / low status.
Absolutely all true, these notions actually share the same model: centrally controlled, no control at terminal / no open access for new entrants, consumption devices are locked to consume only, hard to “become center” or provider, etc. The networking model was a logical consequence (and practical implementation) of the the centrally controlled walled garden you describe.
Many hackers were trying to escape it by becoming their own providers and also connecting with other international x25 servers and foreign hackers. For example, Assange was actually from these times.
Nevertheless, it did generate much more revenue than what they initially envisioned.
I think its more correct to say x.25 was virtual circuit based - over a packet based system.
x.25 had a MFT (Metric Fuck Ton) of error checking (at each layer). I used to work in OSI based networking - debugging OSI issues was an art - no wireshark back then
However You could do some cool things with osi based networking binary attachments natively supported, years before the internet had it.
Interworking between legacy system like Telex, I used to have my own telex number and answerback and could have had an x.400 email address at the country level.
Ps id take site you linked with pinch of salt no mention of PAD's for example.
It’s true that most applications were stream oriented, and Minitel was, as the users’ sessions were only bytes (in videotex encoding) rendered by the Minitel as a serial terminal integrating a modem (1200 bauds reception, 75 bauds emission!).
There was a smart gateway node that translated complex x25 from the Minitel service providers nodes into simple bidirectional stream of bytes to the Minitels, it was a specialized type of packet asssembler disassembled (PAD) called a PAD Videotex (PAVI).
Now x25 did have packets, Minitel terminals did not. So the PAVI and x25 servers were working very differently and with more complexity (hence smart) than Minitel serial terminals (hence dumb).
Haha. I've been "working" on an X.25 implementation on a PC and it took me an embarrassingly long time to figure out what the hell a PAD was ever needed for.
My understanding 30 years ago was a "PAD" was the TCP/IP equivalent of a "terminal server", translating serial streams to network connections. This may be an overly simplistic view, I don't know. It's been a very long time!
Are there any public x.25 networks still operating?
> Are there any public x.25 networks still operating?
Not that I know of. My plan if to use old line simulators and other tools i can find on ebay and telcom equipment reseller sites to create a "fake" private one. This is one of those projects that I'm working on purely out of novelty interest :P.
You are confusing connection-oriented phase of x.25 with “streams”.
X.25 is a “packet-switched” “connection-oriented” protocol.
[which p.s. makes sense for traditional telecom’s view of the world, which as you know is circuit-switched. One of the main beefs of telcos in regards to IP was traffic management. Compare X.25 congestion management with IP. ]
Some kind soul has delved into the details of modalities of networking protocols:
I'm not sure it can be described as a failure, it made a lot of money and provided many jobs in France (I think my father worked on it during the 80s as a telco engineer).
The last time I had to use it must have been in 1999 as the only choice to enter your college preferences. And there was "Minitel over the internet" for some time. Apparently it only closed in 2012!
Transitioning the country to the Internet was botched in a typical French fashion though.
> The failure of Minitel lies mainly in the underlying networking model, and resulting capability to become a service provider.
Not sure it's that simple. While you couldn't easily host your own services on the telco mainframes, there was nothing stopping companies from offering services over the phone. We had a Minitel service where I grew up in Brazil and numerous BBSs supported the 1200/75 dial in speeds and had specific UIs (or, at least, "ASCII" art) for the terminals.
Supporting the cheap and easily available modems was a no-brainer for the BBS owners, even though they needed to do bit-banging to make the asymmetric speed work.
Think of the terminal as a primitive but ubiquitous browser and the phone system as a circuit switched network and the dial in services as websites and you have the 80's web.
AX.25 is to a large extent completely different beast than what people mean when they talk about "X.25 network". X.25 as a specification is to a large extent an "API" and not particular network implementation. AX.25 is complete protocol stack that implements network with superficially similar and heavily simplified API.
I lived in France for a few years around the turn of the millennium and picked up a Minitel 1 cheap at a local flea market. I set it up in the living room as a GUI for a sound system controller connected to a Linux box, playing MP3s. I don't remember the details but there was something odd about the serial port... funky voltage or baud rate or something. It seemed very retro at the time... the design exuded 1970s style in some way (it had some style in common with the Poly-1 computers from my high school in New Zealand, originally released around the same year, 1981). Sadly I left it behind! The keyboard was frankly terrible. Vive l'azerty.
I don't remember much but I see there are many articles from people who've hooked a Minitel up to a Raspberry Pi or similar in recent times. I think it had a non-standard DIN connector, and you could make up a special cable to get RS232, possibly with a voltage change, not sure about that, and then I think you just had to talk to it very slowly; maybe 1200/75 baud, something like that. There were terminfo definitions for minitel (various models) and I see they are still there in Debian, which means you can write programs using libcurses to do primitive, slow 80s style user interfaces!
If I had space in my life for more retro-junk I'd definitely love to have a working DEC VT100 or Lear Siegler ADM-3A if I could find one with the tube in good sharp condition so you could actually really use it to hack code on... but they probably go for a fortune by now, and ... well I just don't have space. For a slightly less rare and valuable terminal, a Wyse 60 would also be a good option (unless I'm misremembering, the keyboards were very nice and you could absolutely work on one of those today). When I was a kid all the libraries had Wyse 60s as catalogue search terminals, and I also learned how to use Unix through one of those beautiful things.
I had a couple of Wyse 60s that looked basically new when I got them, and they both broke (refused to display anything) quickly. It is rare in general for computer equipment to break when in my possession.
What a missed opportunity. I remember when we got a free minitel from the post office when I was a kid.
It was so expensive to use it that I was barred from even touching it. The only time I used it was many many years later, to check the result of the ceremonious baccalaureat high school exam.
It was slow and had ugly graphics compared to the VGA of my 386 but it still felt magical.
I note that the anglophone vorticist magazine Blast[1], No 1 (1914) p.27, blesses france for its mastery of "rose" services.
An irony of the Minitel:
> "... à l’époque, les femmes à qui je racontais que j’étais donc payé pour faire fantasmer des hommes ... étaient plutôt séduites par ma façon de le faire ... paradoxalement, parler de mon boulot d’animatrice minitel rose était une super technique de drague."[2]
Back in the 1980s when I was a l33t hax0r, I used to bounce around the world on X.25 networks, many of which I suspect were operated by Minitel. There was a chat system/early BBS/something called QSD in France, where a lot of hackers would congregate to chat and trade credentials for systems. I can't find much on the net about QSD now, except a few old Phrack articles. Does anyone know what the relationship between Minitel and QSD was?
Minitel is the terminal system, with the CPU, keyboard & screen, but by extension, it includes the protocol to display & download the data.
Actually, i was used to connect to "minitel" servers, RTC or BBS-like that were dedicated to the developement of HP48 calculator applications. It was a very rich scene in France, with RTC servers named Pulsar, RTC-one, etc. Basic BBS boards with discussion boards (like HN!) and where you could download HP 48 programs.
Some history (in french) about one of these boards http://wiz0u.free.fr/pulsar/historique.php
It was back in 1997 with a pre-internet scene, the first IRL parties, internet friends, etc
Basicly it was some smart guy hosting an atari st in the garage, with a dedicated telecom line (so 1 people connected at the same time!).
Yeah, lutzifer was another X.25 hacker chatroom. I think it was based in Germany. I used to reach it via Telenet [sic]. In retrospect, I think it may have been run by a US government intelligence agency. Someone on there doxxed me despite hopping through a lot of relays that should have been nearly impossible to trace through.
you must be as old a me. i was regular on QSD via X25. Great Days. Big big phone bills.
QSD was a 3rd party chat service, accessed Via minitel but with a per minute surchage. I met a lot of good friends their, swapped some postcards with some french girls. Great days.
I dont know how it came to be accessible via x.25 though.
Yes, QSD was hosted on a Minitel serving node/system but was mainly used for international chats. So we can say that QSD “abused” the x25 interconnection of a Minitel server in order to provide chats mainly to hackers.
Common people there: BayernPower, Gandalf, ... long time ago.
I remember some dude abusing a VAX that had x.25 access to connect out to all that stuff. "SET HOST /X25", anyone? Word spread like wildfire. At one point every elite dude and his mother was on the system. I think it took a little over a month to figure out something was awry. Probably the Sprint / Telenet bill was a clue...
Minitel terminals always were considered to be 'dumb' machines, as mandated by how the service worked. No storage, no graphics capabilities...
However, they were running their little 8 bit microcontroler at 14 Mhz, had 32 kB of ROM... (Ok, just 128 bytes of RAM)
This guy made it do things it had no business doing:
http://hxc2001.free.fr/minitel/https://youtu.be/a2HD6OzNoEo
In 1992, with a minitel 1B (80 columns), I have used the 3621 to connect to a service (129040134) that connected me to a vax station at school. From there I was connected to sun station where I edited my thesis using latex and vi. My teacher never knew I was 700km away from him when he received my daily updates.
.The minitel was free (we had to choose between having a minitel or having a paper annuary). The communication on 3615 was expensive (9.15€ per hour). On 3614 or 3621, it was around 3.05€ per hour (sometimes half).
The internet clearly predated Minitel. The big difference is that it was not initially used for commerce. It was to link the military with universities.
I joined Xerox in 1985. Two years earlier the network we were on converted from Arpanet to Internet. There might have been some group somewhere in Xerox that was still on the Arpanet but the military were worried about security so they created the Internet. You can call the exact date of the split between arpanet and internet as being the birth of the internet.
Sometime in 1980 I was also involved in hooking up the internet to my college.
Urls also predated the web because they were used on the internet.
The web made the internet easier to use - that is true. But before then we still had email and discussion boards and distribution groups. It was all text based and somewhat ugly.
There was an experimental network in France before the Transpaq network of the minitel that was designed for inter-connectivity.
It strongly inspired TCP/IP which is, in my opinion, what actually created the Internet of today (having a world-wide network of networks).
https://en.m.wikipedia.org/wiki/CYCLADES
I don't think the vision of the minitel was to build an international network though, so there might not have been much interest in inter-connectivity I guess.
> "Not only were links within North America difficult to establish, but Dik Winter, from the Netherlands, describes how the first cross Atlantic Usenet link was delayed until 1982-1983 because of the difficulty of acquiring an autodialer modem that conformed to European standards. The mail link ... connected the site decvax at Digital Equipment Corporation (DEC) in the U. S. with mcvax at Mathematisch Centrum (MC) in the Netherlands."
> "CWI established the first connection between Europe and the Internet on November 17, 1988”
Not at all true; I used to connect from the Centre Mondial in Paris via transpac to MIT-Multics to read my email in 1982. And that only worked because those links already existed.
suggests all may be true: TRANSPAC was X.25 (presumably terminated similarly at MIT? I know nothing about ITU.), USENET piggybacked on voice networks, and 1988 was presumably when regular IP packets started being routed.
URLs were invented by Tim Berners-Lee. They do not precede the web. There were separate namespaces before the web; there was Gopher and FTP, and various shorthands to indicate the combination of a site and a path, but no standard, universal resource locator (see what I did there?).
The split you talk about, did the networks become MILNet and NSFNet? I always thought Arpanet was decommissioned, not transferred to the military. Or was it just the wording that made it seem Arpanet was shut down?
Genuinely curious about Internet history details. Thank you!
Amusingly, I learned about Minitel just a few hours ago when I stumbled upon this video from John Stossel's show (obligatory thanks to YouTube algorithm) where he mentions Minitel as an exhibit of government/public sector's failure to innovate and a way for the French government to 'control the growth' of this new technology before private players. Here's the video (with the exact time stamp) where Mr. Stossel talks about Minitel -- https://youtu.be/rOkkqoDJl-s?t=208
I then searched the web (through Google) for information on Minitel and this IEEE article showed up.
The Minitel wasn’t the Internet, though, and was in its time quite innovative. There are much better examples of strategic errors from various governments (and private companies as well, but it does not fit the same agenda).
[edit] in any case, thanks for the link. That was an interesting read.
> John Stossel's show (obligatory thanks to YouTube algorithm) where he mentions Minitel as an exhibit of government/public sector's failure to innovate and a way for the French government to 'control the growth' of this new technology before private players.
I wonder what he thinks ARPANET is an example of. (I don't)
I don't call myself a libertarian, but I am sympathetic to a lot of their arguments. Stossel comes across as a whiny spoiled brat who can't handle being told 'no' to doing things like lighting up a cigarette in a public school, because it's bad enough that it's a public school but I can't smoke either?
The selective arguments he makes can usually be picked apart very easily, too.
I don't consent with the gist of that show. Rather, I think, the Minitel approach was much like todays app stores: still allowing for private entrepreneurs and innovation, but with a regulating entity, which is also claiming ownership of the transactions, on top of it.
Ah, thanks for that article, it might come in handy :)
I thought about doing it this way, but was rather planning to do it slightly differently, by building the right modulator (at, what was it? 2600 bauds? it should be pretty doable to bit-bang using an arduino, or the RPI's GPIO. Maybe even with a digital output, though I'm thinking of using a DAC. Or hell, if using a raspberrypi, why not the audio or video output?
Don't see why not, most European equipment is built to handle a range of 220-240V at least, whatever the marking says. At worst the screen might display slightly smaller than intended.
I bought one last year and it's honestly pretty fun to get them to work and display a Linux tty. I'll try to get some home automation stuff running on it if I can.
Also in the early days of the internet, most dialup modems sold in France would also allow you to connect to the Minitel (rather than the internet) from your computer.
Minitel was great for it's time, but its huge success in France resulted in the country later being slower than other European countries to adopt the WWW.
Bill Gates writes about this in one of his books. Not about Minitel, but other technological first movers. His example was how the US first got television but then were stuck with fewer lines per frame than we had in Europe.
First mover isn't always an advantageous position to be in.
I think the primary problem with NTSC was more colour-related than lines-related, though the argument still basically stands.
Interestingly, this one went back and forth, AIUI. The initial British (monochrome) standards, which was very early, was quite flawed; later US standards improved on that, then NTSC improved on those, then PAL and SECAM improved on that (and bounced off each other a bit; the version of PAL that was eventually released used concepts from SECAM).
France, being France, did its own thing on the colour transition; SECAM wasn't used anywhere else in Western Europe (SECAM was better than both PAL and NTSC for the end user, but was hard to work with and not commonly used).
British and Irish TV channels, incidentally, continued to broadcast the pre-PAL standard on separate frequencies until surprisingly late. I believe the Irish broadcaster only stopped in the _90s_, when their conversion equipment became unreliable.
Similarly, the French first program ("première chaine de l'ORTF" before 1975, TF1 after) was transmitted in 625 lines SÉCAM on UHF band and simultaneously on 819 lines black & white on VHF.
Old black & white 819 lines was the more highly defined television system before HDTV.
It's true. Though it must be noted that infrastructure wise France has caught up pretty nicely [0]. I think in terms of the country adopting the web for it's everyday life, like administrative procedures, it's still far behind other countries. It would be nice if there were some quantitative values out there to be able to compare objectively.
Fiber in France is very patchy, I had fiber in my previous tiny apartment in a building that was almost falling down, moved to new construction building aaaaand... back to ADSL, 10mbit, less on rainy days.
They missed the occasion to have built something global that would probably have carried European values. Probably because of a lack of communication outside french language. If I remember correctly all the parts where build in France. Later they could also have competed against YouTube, with one of the first video sharing websites, too bad they totally gave up against the US brands.
French language might have actually got you further ahead in 40 years ago Europe than any other lang. Minitel arrived in some form or another in most European countries surrounding France. Spanish version was called Ibertexto.
They missed the opportunity as well, a lot did. I am not saying they would have succeeded for any specific reasons, but back then all the possibilities were there. Internet has become a nightmare, we lost everything that made it special when it started. Nowadays probably 90% of the non-porn traffic is going to Google, Amazon or Facebook and it becomes a problem for freedom of speech and "democracy". Concentration of such power or money is a problem.
Minitel was run by the French state, it doesn't get much more centralised than that! And of course far-right or Nazi messageboards would have got censored off it immediately.
(I have no idea whether there was much politics content on minitel, but I've not heard of it, and would be interested in references)
On the french Minitel a lot of discussion forums were private owned and not really censored. I think the censorship is more heavy on facebook than it was on minitel services.
Sorry, I find that statement quite imprecise: you could connect to a single server at a time, just like a browser tab connects you to only one service at a time (even if it uses many libraries from many other http servers), so if you mean peer to peer this way, then yes.
But then to do that, it connected to a massive infrastructure that was there and running with very high availability constraints, ... for the time ;-)
I remember my stepdad talking about “miditel“, which would have been the name of Minitel in Belgium. Don't know if France and Belgium were supposed to be connected, if it was the same network and operating entity or some kind of white mark thing. Or just speculation.
I really doubt it would have won against the internet, but there's definitely a parallel universe in which the internet was preceded by a global (or at least pseudo-global) Minitel.
In Italy we had Videotel which was (IIRC) a localised version of Minitel built for the Italian market. I remember a pub in my neighbourhood had Videotel terminals, I thought it was exciting that you would be able to communicate with someone you didn’t know in a chat room.
I do think that in term of accessibility and ease of use for some remote services the minitel was quite better than the web is.
I worked for a company that ran a service until the very end of the network. This service was used to track mandatory farming informations. Farmers could use the phone, the minitel or the web to do so. There was a whole generation of farmers who used it until the very end and ended up having to use the phone as a replacement. Having to use a computer with the operating system and browser layer made the web version of the service much more complex by nature.
My father used this kind of service on the minitel (maybe the one your company was running), and indeed it was quite confusing to him to switch to the web version. The later also came with much more information to process.
One story: They got more business use than you might otherwise expect because they didn't list what services you paid for on the bill. So businessmen were racking up larger charges for sex chatrooms and what have you without having to explain them to their company.
I used one of those. There was a narrow time window prior to the dial-up age when multple telcos considered something similar, but the investment required (especially terminals) was massive. It was a pretty neat universe. I last played with Minitel data when looking at (of all things) _dial-up_ set-top-boxes with a basic browser (it was an Alcatel thing, I think, that talked to a Minitel gateway to render content).
"Minitel use peaked in 1993". The closest technology at that time outside France would have been bulletin board services (BBS). Each was run as a separate walled garden but many were linked by Fidonet. So not too dissimilar to Minitel, just more diverse and decentralized.
There was a similar system for minitel: these were called RTC servers. Nothing to do with WebRTC.
It was short for Réseau téléphonique commuté. Basically a simple user would run the equivalent of bbs software on his computer, but with specific tweak that it could talk in videotex encoding.
And then of course he would use a modem (even in hackish ways the modem of an existing Minitel) to answer calls on a normal phone line (no x25 here).
Before knowing the Web, I spent afternoons after the school chatting with strangers on the Minitel, instant messaging was the big thing. Seeing the popularity of the chat feature, a lot of entrepreneurs decided to launch erotic services well-known as "Minitel Rose". People essentially spent their time (and paid a fortune) to flirt with chatbots.
When I first learned about the Minitel, most things I found online just said it was done to reduce the cost of printing and shipping phone books. That never sounded right-- very interesting to hear more of the actual story.
The story seems to echo a bit of why Gopher did not become the WWW (with the added protocol scalability issues that tcp/ip catered for) - basically, Gopher developers set a prohibitive pricing model (I think it was U of Minn)
I recall it was the mere fact that they had a pricing model at all when the competition made it clear that they would not attempt to derive any revenue from anything related to the web (e.g., running a web server, distributing a web client program)
The French insisted on doing it all in French while the language of the Internet is English. They isolated themselves and could never have been the center of the web.
Back in the day, not so far from the Minitel days, French was the lingua franca of diplomacy, and spoken all over the world.
English was the language of commerce and technology, but French being the language of diplomacy and intercommunication it would be a game changer on teleco.
But Minitel was a bit like an enclosed shell, like Gopher.
Now Gopher is the opposite, nobody needs a license to host a server.
The switch was somewhere in the 70s. For every two English words I hear on TV now there was one in French and one in my language (Italian). Anyway the exposition to foreign languages was much lower because there was no Internet. English got an advantage that French never had in the 20th century. And Hollywood, UK/USA music, etc. All of them larger than their French counterparts.
The use of French in diplomacy was primarily an outgrowth of it having been the court language of continental Europe. You also had dominant languages for other purposes--for certain fields of science, you sort of needed to know German.
But as the world increasingly globalized, there were a lot of forces to have a fairly universal language. And between the once huge British Empire and the industrial dominance of the US post-WWII, that language naturally became English.
Before WWII French was certainly considered the most important foreign language in continental Europe and English eventually became prevalent only after. (I vividly remember my grandma being somewhat disappointed that we were taught English as the first foreign language in school, instead of French.)
It may not be that far fetched to imagine a 1930s internet ("The League of Thoughts") coming somewhat naturally in French.
That's an excellent name! I think about William Gibson's novels about Babbage's machine being hooked up to a telegraph resulting in an earlier "internet" revolution.
My "ligue des pensées" (LDP), AKA "Gedankenbund", AKA "League of Thoughts" (LoT) is certainly based on storing analog signals on refreshing electrostatic drums and projecting the image (images!) on CRTs. There's a nice railway print of the futuristic main hub of the British part of the LoT at New Castle, exposing the stylized cylindrical shape of the storage drums in its main tower (nicely lit in electric blue by high the voltage discharge of the drums inside). A delicate chapter in the history of the LDP is the niche of the "valses roses", which inevitably evolved into a major business. In many countries some opposition of mostly moral objection arose, especially in Germany, where a rising NSDAP feared for the health of the populus and for the sanity of German men who were exposed to the finer details of the anatomy of foreign women, which in turn lead to its smashing defeat in the 1933 election… :-)
----
Edit, on the technology of the LoT: This was some of a joint European effort. A crucial part was the storage of the analogue image signal on a continuously refreshed electrostatic drum, invented by Dr Ludwig Chevenmüller in the early 1920s. A major breakthrough was then the contribution of a young Tommy Flowers, who came up with a way to combine discrete signals suitable for relay-based automatic dialling with the analog signals that provided the content. In a somewhat parallel strain Thierry du Pont du Lac invented his "lecteur discret", which did much the same for character based signals in Baudot code, which was eventually combined in the vertical blank of the analog image signal and stored on the terminal side on a second drum, to be superimposed by a tiny typesetting machine and an array of electrostatic lenses onto the analog Chevenmüller image. (Alternatively, the combined image could be redirected onto a third drum, from which it could be printed using a technique similar to stencils. Seasoned users of the LoT may remember the typical smell quite well.) But the LoT only really took of, when BASF came up with a way to compress the Chevenmüller signal into rapid bursts, which enabled the rather astounding speed of the network.
> not so far from the Minitel days, French was the lingua franca of diplomacy, and spoken all over the world.
What a ridiculous claim. The language spoken for Diplomacy is spoken by diplomats and has virtually nothing to do with what the population speaks. Just like Latin was the language of Religion and Science in Middle-Ages, but certainly not spoken by everyone else.
A classic problem with english translations of War and Peace is what to do about the languages? The main characters start out the novel speaking french with each other and russian with the servants (similar to the role of english for elite households in 2020 india?).
The costume drama https://www.youtube.com/watch?v=J2-fkZ-0Blw claims that in the seventeenth century, the aristocracy would've been familiar with enough with their tenants' dances (indeed, preferring them to french court customs) to perform them themselves, but no matter what the historical accuracy of that portrayal, Tolstoy's take on the nineteenth century situation was different: https://en.wikipedia.org/wiki/War_and_Peace#Language .
Not ridiculous. In the late 70's, everyone's second language in Spain (except OFC the ones speaking Iberian latin dialects or Basque having Spanish itself as a second one) for example wasn't English, but French.
If you wanted a better future (lots of people headed to France to work), you would learn French to make a good chunk of money and cheap goods, period.
> In the late 70's, everyone's second language in Spain (except OFC the ones speaking Iberian latin dialects or Basque having Spanish itself as a second one) for example wasn't English, but French.
1- Spain is a neighbor to France, so it's hardly surprising that French is used a lot in the first place.
2- Spain represents less than 1% of the World population, so it's hardly a proof that French was used all over the world in the 70s.
French is used across the world _today_, so it's not that much of a stretch to imagine that it was also used across the world in the 1970s.
By the Wikipedia numbers [1], French is the fifth most spoken language in the world by total number of speakers, yet only fifteenth by native speakers.
The period prior to the internet existing (before the 60s) ends far before the period before Minitel existing (before the 80s).
While you can have discussions about when it became ‘the internet’ ARPAnet was already an international network far before minitel launched and of course the language was English.
Sure, but the internet did not become a consumer phenomenon till 1993. There were very few rules on the internet of 1992, but one of those rules was that commerce was not allowed. So for example there was in 1992 a newsgroup named ba.jobs which functioned very much like HN's monthly 'who is hiring?' and 'who wants to be hired?' comment sections, but there was no analog to HN's monthly 'Freelancer? Seeking freelancer?' because although it was considered OK to seek W-2 work (i.e., seek to become a proper employee), seeking 1099 work (i.e., seeking a contract to do some programming work) was considered too much like commerce and consequently not allowed on the internet (or more precisely not allowed on the US internet backbone -- and it would've been impractical to run a newsgroup while avoiding the US backbone).
In 1992 visiting a friend's apartment in San Francisco, I had to explain to someone what the Internet was, whereas my friend who was on Compuserv didn't have to explain what Compuserv was.
Note that this thread has evolved / mutated so much that it no longer has any bearing on the question of whether "The French insisted on doing it all in French while the language of the Internet is English. They isolated themselves and could never have been the center of the web" (a question I have no interest in).
The point is that France was ahead of the US in conferring the benefits of computer-mediated communication and computer-mediated commerce to the average consumer.