A very long time ago (~1985), I dropped my Minitel and the screen was dead.
I cabled an Ascii keyboard on the Mother board and made an interface for a Black& White TV with only one transistor (2N2222 probably) and a few resistors.
The motherboard was inserted in a "Traoumad" metal box for the famous biscuits.
It's still in a garage in Rennes, I will take a picture next time I go there, but it's in a dirty state.
I was very lucky to find a mechanical keyboard with a matrix that was very similar to the one used by the Minitel. And as I worked in an in-house teaching unit of France Telecom near Lannion's CNET R&D unit, I had access to Minitel's schematics. If I remember correctly, there was no mapping for the "connection" key.
And at that time Traou Mad was commercialized in steel boxes.
One of the things that blog post makes me realize is that we don't have instruction manuals that are intended for a technical audience anymore. The odds of seeing an instruction manual with pinouts for connectors and voltage levels are pretty much nil nowadays.
Instruction manuals, interface documentation… yeah, modern devices are treated like black boxes.
Say you go buy a Fitbit, is the manufacturer going to tell you how the protocol works to communicate with it? No, they’re going to give you an app, that may or may not track usage data, and that they can decide to stop supporting whenever they want. If you want to write your own, you’ll have to reverse-engineer it.
> Say you go buy a Fitbit, is the manufacturer going to tell you how the protocol works to communicate with it? No, they’re going to give you an app, that may or may not track usage data, and that they can decide to stop supporting whenever they want. If you want to write your own, you’ll have to reverse-engineer it.
I agree 100%.
Wouldn't it be nice if protocol documentation was mandatory? I understand why it's done this way (because 99% of consumers wouldn't care and it allows them to change the protocol later), but it's really annoying to have to reverse engineer everything.
It's funny too, I bought a fitness device a few years back (Lumo lift by Lumobody tech) and the app doesn't work since the company got acquired by another. If there was documentation, someone might have stepped up and written a replacement app or something to connect to the device.
At this point, this kind of puts me off from buying any kind of hardware tech from a startup that might not be around for a few years, since it might just turn up to be another tech paperweight.
> At this point, this kind of puts me off from buying any kind of hardware tech from a startup that might not be around for a few years, since it might just turn up to be another tech paperweight.
Same. The vast majority of peripherals meant to be used with mobile phones I just wont buy at this point, because they’re almost all like this.
At least I know I’ll be able to use my Pebble or PineTime indefinitely; the former because it’s been reverse-engineered, and the latter because it’s open from the start. As appealing as the newer Fitbits are, I can’t say the same for them.
> Wouldn't it be nice if protocol documentation was mandatory?
Yeah - I’d love to see this, just like we’ve seen a push for repair documentation with recent right to repair legislation.
Indeed. Consumers in general may not care - but imagine if every device that's currently a black box could instead have competitors and an ecosystem build up around it.
Odds are that the communications protocol is part of a DRM system, too. In the U.S. that will make it illegal to reverse engineer without an exemption.
It’s wonderful how Raspberry PI has restored this old tradition. I love that I can download the schematics for any board and that their site has a ready made guide to interface with the header pins.
There is probably a market for x64 PCs that would let you build Linux from scratch and run it on hardware that is hackable. I’m not sure if licensing would permit this.
I've never encountered middle managers who think this way. They'd be overjoyed if the engineers want to write documentation. Getting engineers to write docs on the other hand... sigh. Most engineers I've worked with don't even want to comment their code beyond the bare minimum and throw a temper tantrum when having to write more than a few sentences on a service they've written. Heck I know engineers who see another non-updated comment in the codebase and use that as justification to not write docs themselves.
In my university days I used this setup, via the school's then brand-new PBX and the built-in serial port in the phone in my dorm room, to connect to the university VAX. The 9600bps speed was, shall we put it, leisurely, but still a major improvement over the original Minitel's 1200/75. Ethernet was far from widespread in the early 90s, certainly not all the way to the dorms.
Thats really cool! I set up one as well (the Italian version) - if someone is interested in some more pictures and Getty/tmux config I published everything here: https://pojntfx.github.io/minitel/main.html
When I first saw minitel in the early 90s what surprised me was not just the technology, but the sense from the denizens that this was just a completely normal service to have.
I’m sure they had a honeymoon period when it felt magic, but thereafter it was such a useful part of life that it just became a commodity.
I wish I could objectively see how the true internet affected me in that same light. Being immersed in it, I had no idea how it went from a toy to part of everyday life.
By far the best feature of the minitel 1B for a kid like me with no budget was that it was free, as long as you could get your hands on one.
I have fond memories of endlessly annoying my parents and the local post office in my quest to get that model, the only free one with a plug i could connect to my PC. Good times.
Maybe it's a silly question, but why is it needed to have a USB converter? Specifically, in the article, they mention connecting to a raspberry Pi. Couldn't they use the UART serial port on the GPIO of the Rpi? I see the baud rate seems limited to 4800, would that be the issue? (I don't know how this stuff work, I often use serial port to connect to my SBCs - RPi or similar - and I would love to use a minitel for that after reading the article ^^)
You don't if you are going into a serial port on the other side. I chose to go USB because it gives me flexibility to connect into a Raspberry Pi or my Mac or ...
Hi Alduin32,
May I ask you a related question?
I bought recently a VaxStation 3100 and tried to connect to it with a USB FTDI (I didn't try hard), but I was unable to get anything.
https://en.wikipedia.org/wiki/Traou_Mad