You know what would be nice?
An "open firmware" project, where people produce and share FOSS firmware for common household devices, replacing the manufacturer's firmware with a decrapified version.
A smart TV? Reflash it with FOSS firmware, which removes tracking, calls to the manufacturer's servers, and any Internet-connected services that you don't want?
A printer? It's now yours. No ink subscription, or need for an app to print remotely.
And so on.
A man can dream, eh.
@larsmb @neil
fully agree, nice dream, but not happening at the rate at which new devices are brought onto the market and old ones disappear; inherent to the nature of the component market and consumer device price pressure.
So, not really something volunteers writing software solves for more than 6 months.
You see that in libreboot and similar projects: coreboot's successful for servers, LB works for 26 laptop models worldwide; rough guess: 10–30 laptops models enter the market, every day
@larsmb @neil I mean, yes, but no.
That whitelabelling has long happened; for example, in the least-cost home DSL router market; it feels like many routers you get from ISPs are just "intern changing the vendor strings and replacing the placeholder logos from the router OEM SDK, which honestly is no SDK but a GPL violation and five ash scripts in a trenchcoat".
Not the kind of devices that someone who hopes to get lasting or repairable hardware would want to use, usually.
And sure, with like
@larsmb @neil androidTV, things are a lot more interchangeable. But the fact it's android highlights the exact same problems you see in the phone market with alternative firmwares:
a somewhat random selection of flagship phones (which often come with by-far-not-the-worst-on-market android distribution to begin with) are supported, at great maintainence cost, and even these often miss features compared to their stock firmware
Then, 0.5a passes, and there's new flagship phones, or new android.
@larsmb @neil generally, the whitelabeling isn't going to make things more centralized. The advantage of just branding a whitelabel product with your company's logo is that economies of scale apply to you, while you have no R&D effort to keep up with the things that will be cheaply available for mass manufacturers in 2 months.
Thus, white label -> larger volume for same hardware -> quicker amortization of R&D cost vs spot component pricing -> higher model turnover is what I think we see in the…
@larsmb @neil … "nearly no-name TV business", for example.
In the end, whether it's 100 actual design companies holding on to a TV model for 1 year on average, or 10 companies doing 10 products a year because that makes them cheaper on panels, doesn't really matter. The number of different devices per year remains the same.
Starting such a project is relatively straightforward. Reverse engineering firmware is hard, though 😭
I love writing code but I don't have the skill required or the time to acquire it. (Talk to me in ten years, when my kids are older.)
If I win the lottery, I will fund this. I'll let you know.
@neil Well actually for DVB-receiver decoders there is FOSS in "Enigma 2". For TV-sets it's probably not worth, just don't connect them to the network. As for printers, decent printers still exist, and you won't get an inkjet to be useful, no matter what software you put on it.
As for TVs there is a bit of a chance, as the hardware is not terribly diverse. There's only a comparatively small number of TV chipsets around, and displays are, except for calibration, rather similar.
@spacerog @neil I would consider that a bad thing.
It's not that open source is bad.
Rather it's that people can get hurt due to software flaws (e.g. 737 Super Max & Tesla crashes). And without the legal obligation to build quality code people writing that code will, as I have seen time and time again, be sloppy or have not done adequate testing. (My business is building tools to test network protocol software for robust behavior in the face of non-laboratory conditions.)
In automobiles the person who is hurt is as likely as not to be a person other than the one who chose the open source firmware, so this is not a situation where the risk and harm is confined to the person who made the choice.
I've seen open source advocated also for voting systems. But in my experience quality in those kinds of systems, once again, comes from inspection and testing, not that the code is open to use and adaption by others. (I do agree that open code is usually easier to inspect and test.)
@neil well, you can root your smart vacuum cleaner* and install free software on it! what a time to be alive.
https://valetudo.cloud/
*if it's on the list of supported devices, of course
Dream hell, I DEMAND IT.
Sorta like we did in the 00's & very late 90's with just about anything that could be made to hold Linux in firmware? dd-wrt would be another example. Also, the Audrey boards were created around the idea (Audrey was a 3Com all-in-one for email & light browsing; the boards quickly expanded to all sorts of other devices). Also the early wardriving boards. All of this pretty much started declining as smartphones and widespread broadband became a thing.
Usually, the biggest thing stopping alternate firmware is a lack of documentation. It takes a *massive* amount of work to reverse engineer devices and write device drivers if non-standard chips are used. And the entire process has to be re-started from scratch when the original branded product becomes hardware v2, or the manufacturer uses multiple factories, and so on.
@neil
It is *always* worth searching around to see if someone, somewhere has already done just what you describe - especially after a device goes EOL.
I think we're going to see a lot more citizen-level reverse engineering projects bare fruit, as the #enshittification epidemic spreads like #covid.
We're fed up, and those of us who have the skills and time to do these kinds of projects, *are*.
I know I am.
I just saw a guy #reveng his Phillips electric toothbrush. Another replaced the controller board of an otherwise unusable video router with his own microcontroller and code. Tip of the iceberg.
@neil i gladly bumped into tasmota recently which is open firmware for esp devices, so that is already something, but indeed it would be great to have a do-not-phone-home firmware for other devices
No need to dream, you're referring to open source hardware and open source firmware, pretty much.
John "maddog" Hall, chairman of the Linux Foundation, pushes for Open Hardware, the foundation heads projects like RISC-V International, Sound Open Firmware and KiCad.
The Open Source Firmware Foundation manages the Open Source Firmware Conference. Those two are good starting points.
@neil even better, vendors ought to open their firmware.
At the very least when the device is not supported anymore.
That's actually an @fsfe campaign:
https://fsfe.org/activities/upcyclingandroid/
Where do I sign up?
Andrew Pettitt
Neil Brown
timothy
KI7CFO
Pedro Cambra
Emanuele Panz
VolcWolf
Lars Marowsky-Brée 😷
Marcus Müller
Pierre Prinetti
William Leech
Joe Oswald
Götz Hoffart
Christian M. Grube 🐧
My camera shoots fascists
Acesabe
lopta
Rhizomatic Arcade
Christian Berger DECT 2763
Space Rogue
Karl Auerbach
Daniel Durrans
Democracy Dies in Dumbass
noodlejetski 
Raúl Redondo
Adzie 
coolbean (account has been reverted to before we moved back to this instance)
Kevin Russell
Frances Larina
Alan Langford 🇨🇦🧤🧊摏
Sean Boyer 🇵🇸 FREE PALESTINE
BinaryUnit
Ted Johnson
Herz++
Ash
warkolm
Eli the Bearded
Jiří Fiala Total Landscaping
werdahias (tired)
Eckes 
Rogan Dixon
[Yaseenist] CauseOfBSOD 
mmu_man
Ángela Stella Matutina
Popolon 🇵🇸🇳🇨☮️🌳🎋 ᠫᠣᠫᠣᠯᠣᠨ🐎抱抱龙🐉 ⏚φ
Mike Sturm 
Tim B
Michael Bacon
trisweb
enachtpfoetchen6s0
Alex Vengerovsky
danwwilson 