Nowadays, I just get old laptops with broken screens and take the mainboard out. They are dirt cheap, often even have 16GB of RAM, you can just dump pretty much any linux on them and they are usually way faster than Pis.
The boards are of course bigger, which I guess will be a non-starter for many. They'll also require active cooling unless you downclock them considerably. They often also require more power, although many mainboards can be configured down two a few watts of power consumption, depending on what peripherals and speed you need.
Yeah, it's more work than just getting a Pi, but given the current prices, it makes sense for me.
The goal of raspberry pi's was to have a standard GPIO connected to a moderately powerful and cheap computer for educational purposes. Then it slowly evolved into many use cases because of it's vast community and standardization and availability, not to mention the large number of boards that connect to it. The standardization did erode a little but it still maintains that reputation because of the community and availability and now because its turned into a big brand. Many industrial projects depend on it.
If you use the raspberry pi for homeservers or general compute (which I did/do sometimes) then the value proposition is quite poor regardless of pre AI prices unless you have very strict power requirements (in which case it probably fits the above use case already). The comparison with NUCs or mini PCs or laptop boards feels like it's a wrong comparison.
The chips used in PIs are repurposed Broadcom SOCs that have inferior bus speeds compared to a general purpose processor from intel or amd, even few generations old. For any general purpose computing these chips will always be slower. its like comparing apples and oranges. I never really understood why people did that.
Yeah, I don't disagree, but I would argue the Raspberry was always an applorange, and that's exactly what made it popular (together with the cheap price in the beginning, which made it basically an impulse buy). It brought people into microcontroller/embedded systems territory without having to actually deal with all of the stuff that made this so daunting. As you said, it is basically a normal Linux PC, just that it runs Arm and has GPIO.
People who are really into electronics/embedded systems will use a proper microcontroller nowadays anyway, these have gotten so powerful and easy to use that there's really no reason for a Pi anymore. And on the other side, people who want a little server, a pi-hole, tailscale or octoprint or whatever, can just use an old used laptop/miniPC or similar. Since the Pi has gotten so expensive, I see basically no reason to get one anymore.
I would argue that even if you need "a cheap SBC with GPIOs", at this point you'd be better served by connecting a tiny $5 microcontroller-powered board with a USB PHY + its own GPIO headers, to any cheap "computer" (i.e. a GPIO-less SBC, or an old laptop, or a phone like in the article, etc.) over USB; where the MCU board essentially serves the host a "GPIO gadget."
You can make such a dongle yourself from ~any MCU dev board, but there are also productized "USB GPIO breakout" adapter boards, e.g. the Adafruit MCP2221A.
The "Pi as an SBC with GPIOs" case now only makes sense if you really, really require the Pi's compactness and/or monolithic-ness (as might be the case for e.g. ruggedized embedded systems like robots.)
I always thought the value-prop of a Pi was the monolithic nature. If you needed GPIO and Linux, you didn't have to glue two things together to make a solution, you could just buy one device that met the requirements.
I use a pi (4th? generation) to run homeassistant, calendar sync target and a few other random small services on. It's great. It works, it doesn't make noise, it's small, it plugs into USB-C, the power brick is small, too. The SSD I'm storing everything on because I had a few failed SD cards has a larger footprint than the rest of the computer. What not to like? Nothing I run on it requires faster CPU, and the small footprint is worth something to me.
Sure, if you have a big house with a dedicated server room/hobby space, you might as well have a NUC/old laptop/whatever.
> The comparison with NUCs or mini PCs or laptop boards feels like it's a wrong comparison.
For many (maybe even most) people they do the same thing. Some just want a HTPC, basic desktop, or small server, no GPIO needed. Raspberry Pis were definitely cheaper early on because you could actually buy boards at or near MSRP, but over time the price rose and mini PCs became cheaper.
or maybe the mini pcs started becoming available at all. I think these were a concept since core 2 duo and first gen i3 era. before then the chips were too much wattage for a mini pc. and the first generation mini pcs were also costly for enterprises to adopt in bulk. its only after 4th and 6th gen intel that these were adopted in large quantities in enterprises and once they dumped those in the used market they became comparable in pricing and somewhat in power to a rpi.
Used powerful PCs have always been available for dirt cheap on eBay even when the first pi was launched. just that mini pcs are a more recent thing.
Well, the earlier Mini-PCs were a good bit more anemic, that's why adoption was slow; I had one or two earlier ones; both were AMD Bobcat cores; one was a 'not quite finished' thing where I took a dreamcast and put a small board in it (Where did that go, anyway?) and the other was a small, almost router-sized box. And both of those were, well... good for most of what I wanted (i.e. 8/16 bit console emulation and basic media/browsing) but definitely constrained. (As a side rant, I still wish AMD didn't kneecap non-console Jaguar parts, that single channel limit always struck me as a 'we don't want to show how much we fucked up with bulldozer' sort of thing...)
As far as mini-pcs... The first Intel NUC was a Sandy bridge, i.e. Gen2, FWIW. And we did have the 'Mac Mini' as far back as 2005, buuuuut that wasn't cheap because Apple.
I think SSDs were another big part of mini-PCs though; it's easy to forget for granted the risks of having a hard drive in a poor thermal environment, as well as the general size constraints one imposes. [0]
[0] - Surrrre, you could use a 2.5 inch HDD instead, but 2.5 inch HDDs almost always were slower than a 3.5 inch HDD (excluding the VelociRaptor, but that would only be more demanding as far as thermal...)
> The goal of raspberry pi's was to have a standard GPIO connected to a moderately powerful and cheap computer for educational purposes.
At first it was a blessing that the Raspberry Pi gained popularity as a general purpose computer, not just for education and embedded applications. The additional demand and popularity brought more money and development attention to the ecosystem.
At some point, which I could never really put my finger on, this general purpose computer popularity grew so large that the embedded systems and educational aspects of Raspberry Pi seemed almost forgotten by the mainstream users. At a local in-person meetup group it was a running theme for people to brag about how many Raspberry Pis they owned for different purposes and to boast about using them for applications where they didn’t really fit. In person I’ve heard more about Raspberry Pi NAS systems (where people are always disappointed with the performance and the price of forcing a Pi into this role) or Raspberry Pi Kubernetes clusters (which mostly sat idle and could have been replaced by a single PC with more RAM).
This reached such a critical mass that it almost seemed like the mainstream discussion forgot the original purpose of the Pi and thought it was literally just another computer, but small. Most people who tried to use it for this purpose were left slightly frustrated by how it didn’t perform like they wanted a computer to perform.
So now every discussion about Raspberry Pi attracts comments about how nobody should buy a Pi when you can buy a NUC or laptop or some other used piece of hardware and it’s so much “better”. This is the right advice for a lot of people who wanted a powerful general computer but bought a Pi instead, but it’s also furthering the misunderstanding of what the Pi is for.
The real advice is that everyone should be shopping for the right tool for the job. There are a lot of cool applications for the Pi where a NUC or old laptop or phone doesn’t make sense because you actually need the compact size, the low-level I/O, or some other feature. However anyone who can put a NUC or laptop side by side with a Raspberry Pi and consider them both viable options for their project probably shouldn’t have been looking at a single board computer to begin with.
There is a parallel situation happening with the “just buy an ESP32” arguments. If an ESP32 solves your problem, you also probably shouldn’t have been in the market for a Raspberry Pi.
> There is a parallel situation happening with the “just buy an ESP32” arguments. If an ESP32 solves your problem, you also probably shouldn’t have been in the market for a Raspberry Pi.
I have mixed feelings on that.
For the longest time it was safe to say "just buy a Raspberry Pi" because they were cheap enough and powerful enough. If memory serves me correctly, the base board cost roughly the same as an Arduino Uno and offered considerably more. (Of course, there were cheaper Uno clones, but those depended upon support offered by official Arduino software development tools.)
While there are some projects which would only be possible with something like the Uno, there were far more projects that would only be possible with something like the Pi. So if you wanted to give generic go-to advice to a novice, the Pi it was.
The right advice is to use the right tool for the job, but that is something that a novice (or fly-by-night forum posters) aren't prepared to handle.
Of course, things change. There are more inexpensive microcontroller development boards and the microcontrollers themselves are much more powerful. While that doesn't negate the benefits of a Linux SBC, it does raise questions now that those SBCs are considerably more expensive.
The GPIO in Pi is almost useless. Nobody runs servos off of those. Pis are mostly used as just Linux PCs with USB ports that reliably power itself on.
IMO, the problem with the Pi GPIO is that it's a collection of GPIO pins and low pin count buses for controlling peripherals, rather than a singular bus. The fact that it runs Linux without an RTOS monitor program or a sidecar of sorts adds to the problem.
Much like the goal of the PS3 was to have a powerful gaming console that also let you run Linux. The market quickly deals with any hardware that is cheaper than its value.
For me the value prop is a low-power device that can be passively cooled but has a full Linux and tcp/ip stack, at a price point where I don’t care if I accidentally wreck it.
Fantastic for one-off projects.
Nowadays they require cooling fans and cost absurd amounts considering the performance so there seems little point.
Most of my one off projects these days are now on microcontrollers.
The Lenovo “Tiny” line of ultra small form factor PCs is effectively laptop hardware in a proper case, with ethernet and no battery to deal with. There’s a steady supply of used ones at good prices from enterprise IT liquidators. I’ve had one on my desk as a home server for years.
In a better world we'd all have one in our homes, storing all our personal data and documents. The cloud would only be used for storing encrypted backups.
Many small businesses, kindergartens etc. could also have all their digital needs served this way.
Yes they would sometimes break down, but with the right setup, the fix would be a matter of booting a replacement from the right thumbdrive.
With unattended-upgrades enabled and nothing open to the internet (I run Tailscale to be able to access it when I’m out of the house) I don’t worry too much about mine.
If you use something like Fedora CoreOS, the system keeps itself and all container images updated. And because system upgrades are also just a new immutable image, very little can go wrong.
I have a Nextcloud server running on it and the maintenance is negligible.
During the recent PHP-vulnerability I simply turned it off. The next day I booted it, and it automatically installed the new Nextcloud image with the vulnerability fixed.
Makes me wonder if I could have it check for CVEs and automatically shut down vulnerable containers until updated images are out.
I was just thinking last night about setting up an agent that would daily check in and do a full maintenance scan on a server or a home lab box. Check for operating system upgrades. Schedule them if needed to move from LTS to LTS. Check for CVEs against daemons you know you're running. Stuff like that. I've used agents and harnesses to set up machines before and do system administration tasks, but it feels like you could sort of set it up in a loop and just let it go forever.
I was treasurer of the kindergarten parents' association when my daughter was in that age. It’s a small org with 10-12 employees, 30-40 customers (children) twice as many stakeholders (parents, relatives) constrained budgets and strict regulations, eg also a privacy officer. We had a static website, everything with PII, so, basically everything, was on a laptop with encrypted backups. Everything was managed with an Excel Sheet and an accounting software. Payroll was handled by a tax advisor. We cloud have used cloud services but too expensive, too many compliance headaches, and too constrained. I think an agent with local inference on a Mac would be a good investment, today.
My first child's kindergarten used Facebook to share photos. You could opt-out and then - if they remembered - they'd leave your child out of the photos. Everything is an upgrade over that I suppose.
After pi’s got expensive I started picking up refurbished Dell micros and Lenovo tinys cheap and haven’t looked back. They are also perfect for the 10” mini rack[0].
Shoutouts as well for the USFF HP Elitedesk and Prodesks. They are amazing. I even got one with a Ryzen CPU that had a dedicated GPU option - that GPU part was on eBay brand new for $60 and replaced the hard drive bracket. Not on par with a desktop GPU but way better than integrated graphics, for things like Minecraft in my case. I use a second one, this one Intel with 32GB of RAM, as my home server. (Jellyfin, Home Assistant, Tailscale, miscellaneous stuff).
Same. I have two Dell SFF 3070s and some MSI micro units in a rack - planning on getting some more for my Proxmox cluster. ServeTheHome has a great resource for the compute nodes [0] and GeekPi makes some awesome 10" racks [1]
A few years back I finally got tired of the quirks with ArchlinuxARM (some things only worked with the forked kernel, some only worked with mainline) and replaced all my persistent RPi usage with refurbished Lenovo mini PCs from Amazon. Even with Amazon’s markup it wasn’t that much more than a beefy Pi4 plus case and adapters.
This let me standardize on amd64 and simplified a bunch of my deploy code.
I also have one and use it as a home server. They’re insanely capable (I’m constantly running a lot of resource heavy things on it), and it cost a lot less than a pi even though the specs of mine are way better
The wattage on labels on PSUs are just safe max power draw, above which the supply has the right to shut itself down. It's not 1:1 representative of idle draw.
There are adapters which take a USB-C connector on one side and offer a barrel connector on the other one. They're available in many sizes for a few €$¥£. Saves you a brick assuming you already have the required USB-PD adapter.
I bought an HP thin-client PC for a few bucks that was basically just an HP Stream system board without battery, keyboard, or display. Delightfully, frustratingly rinky-dink but it seems to run Linux and OpenBSD well.
I've yet to find a laptop mainboard that doesn't work without a battery. As I've written below, I'm getting extremely boring old office laptops with integrated graphics. Usually phones also work fine without a battery. However, I'm not touching phones because I will not fiddle with locked bootloaders and "the Android situation". Running plain Linux is a hard requirement for me.
If only Frameworks were “extremely boring old office laptops” with costs reflecting that, instead of very interesting and desirable hardware for which people will pay a slightly lower than Apple premium.
Yeah, I really just want a refreshed early 2010s Thinkpad with a wide-open BIOS. The modern Chinese mainboards you can get for old Thinkpad chassis are very cool but the prices aren't so cool.
I have a ThinkPad x230 (3rd gen i7) which makes an incredible Windows XP nostalgia device. It’s basically the newest hardware that exists with full XP drivers (other than USB3 which I don’t need) so it’s massively OP, which was the point of that project. Basically, “what if I had access to insanely futuristic hardware in 2002.”
Obviously I wouldn’t sacrifice that hardware but it might be cool to try that on yet another laptop. (Hey, I don’t have a laptop collecting problem!)
Have you guys had any experience or luck adapting Framework motherboards to run on Thinkpads? I’ve thought of doing this side project several times for my old Thinkpad x201.
I'm just looking at an old Dell laptop refusing to boot without manual intervention because it looses its time setting since I removed its main battery. If I can fix that it'll become a very capable server for my home setup. Not sure it'll work though, as the bios can't be configured to ignore the time reset.
All old office laptops I had so far still had the good old CMOS/RTC battery... So if indeed the RTC is solely powered by the main battery, I would guess there's at least a backup capacitor somewhere so that it survives at least a short disconnect? If you can find that one, you can probably connect power there, but that's just me guessing...
Factor in power usage, thermals (esp if they're sharing a living space with you), and time. This is where the modern CPUs on 2nm process nodes really start to pull ahead, especially when you can S3 sleep them or use IPMI type BMCs on stuff like real workstation-class systems and servers. Even an M6 Mac Mini can do a ton of work these days, and single-core speeds on those especially when you look at perf/watt are insane.
My home server is a first-generation Framework 13 laptop that I put into an enclosure that they sell on their website. I sold the battery and the case on their forum. It works great!
Yes, good video, and it covers one of the most important bits: how do you turn that thing on. :-) (EDIT: holy crap he's delaminating a whole old Apple iMac display from the glass, I did this once with a phone(!) screen and I swore I'd never do this again... doing this with a whole iMac screen is hours and hours of absolutely painful work, it looks way too easy in this video...)
How do you power them? Via the battery? I was thinking about turning my old mbp into a build machine, but i am a bit concerned with running it via the battery 24/7
I feel like The Internet has this urban myth that says keeping a charged battery plugged in all the time will eventually cause it to explode and burn your house down, guaranteed.
I get that keeping a battery at 100% charge is not great for it in terms of longevity (ability to hold a charge). And it's obvious that heating the battery up by running the laptop at full tilt inside a backpack or something will turn it into a spicy pillow.
But I struggle to imagine the failure mode that reliably shows up when keeping the laptop powered on 24x7 that is never a problem for 8-16 hours of continuous use while plugged in (a very typical work day for my laptops).
I used a MacBook Pro as a server for years. It was fine for its purpose, however the battery did swell and need replacing. It may have damaged the trackpad, and like the other comment mentioned, idk what would have happened if I had left it swelling for a couple more years. Probably not a battery fire? But the $230 or so that a massively more powerful HP USFF that I replaced it with was worth it, even versus a 0.1% chance it would burn my house down and possibly kill my whole family in the process.
I ended up repairing the Mac, replacing the battery, and installing Chrome OS on it. (chrome sucks though, someday I’ll get it back out and switch to real Linux)
> I feel like The Internet has this urban myth that says keeping a charged battery plugged in all the time will eventually cause it to explode and burn your house down, guaranteed.
I had a MacBook Pro where the laptop worked but the battery became swollen so much that it distorted the case. If I wasn't there to notice it, it would probably burst at some point and Lithium is very flammable.
It's probably rare and overblown, but it can happen.
you're lucky you're running into models that don't panic in some way without a battery.
for example some alienware and XPS models rely on the battery for amperage capacitance when their big onboard GPUs get busy; allows them to run cooler/less-wattage/smaller wall-warts but it also makes for a platform that'll crash when it hits a high GPU usage without a decent way to make up the amps.
'spose you could use a bigger power supply if you get bios/uefi to play nice. hard proposition sometimes with Dell products.
Well, of course I'm not getting old gaming/pro laptops with dedicated gpus, they would consume far too much power, require extensive cooling, and these are way too expensive anyway, even when broken. I usually get old, boring office laptops (lenovo/dell) from bulk resellers, with integrated graphics who are sufficiently used/broken so they are dirt cheap. As long as they still turn on, the mainboard is usually fine (I've yet to get a broken one).
You can just limit maximum charge (using something like aldente), then replace the battery every 4 years if you need extra peace of mind.
Batteries degrade much faster when fully cycled repeatedly, and even then just about any laptop battery can easily make it 6+ years with low-to-zero risk of failure.
Most old PC laptops have external batteries you can detach by pulling out on a springloaded tab. Even MacBooks are possible to run without the battery, but it requires some NVRAM trickery IIRC.
Some older MacBook and MacBook Pro models reduce processor speed and GPU performance roughly by half when running without a battery, to avoid shutting down if the computer needs more power than the AC adapter can provide.
Apple documented the behavior in a support note. I’m not sure if the Apple Silicon models do the throttling as well. I guess the situation doesn't come up as often since the batteries are no longer user-removable.
I’ve got a 2017 MacBook Pro with a dead screen, dead keyboard & trackpad and no battery sitting behind my sofa, closed in a vertical stand and running macOS as a test server. Open Core Legacy Patcher will allow you to boot without the CPU slow down: https://dortania.github.io/OpenCore-Legacy-Patcher/
I have the same device, did you just remove the battery after installing opencore? last I tried, I got omarchy running on it but now it just boots to the startup disk always. Not sure what's wrong.
The thin clients that I've seen were all Linux/unix based machines. A Windows box sounds like a fat client to me and I'd be concerned about security updates and how locked down they are.
They're marketed as "thin" clients for whatever definition of it, probably for ease of management. What matters to surplus seekers is that they are great as Pi substitutes.
How do you configure them after pulling out the mainboard ? For example if you need to ssh into them, how does it work, if it is not configured correctly before removing the display ?
You can put it in a custom case. Releasing the space constraints enforced by a laptop design means you can easily have a better cooling arrangement so experience less thermal throttling if you ask it to perform longer compute-heavy tasks.
If you are building custom cases for laptop boards, just buy an old tiny PC at that point. They are like $50-100 on eBay if you don't need killer specs. Unless you're just doing it for the love of the game or happen to have a lot of laptop boards sitting around you can sell, it's not worth it.
If you already have a laptop with a board you can extract, and a 3D printer you can make a custom case with for a couple of quid, you'd rather reuse parts of the old laptop than let it go to waste, and you'd enjoy the whole project, why instead spend $50-100 and have to deal with the [censored]s on eBay?
> Unless you're just doing it for the love of the game
Isn't that way a lot of people do the home-lab thing and related projects what would need the Pi these ideas are talking about replacing? It is why I do it [or used to be before working in tech for decades killed a lot of the enjoyment I had for playing with tech].
> or happen to have a lot of laptop boards sitting around you can sell
You are the first in this thread to bring up what any of this is worth on a larger scale, selling the results. Everyone else is talking about putting together one simple computer or jumped-up microcontroller for a task of their own, and maybe more later for other projects. Some of us find this sort of thing fun, and in the shit world we are all living in right now having a bit of fun and achieving something [sort of] practical out of it, can be priceless.
Well, this is about replacing a Pi, so I guess people want a raw board. Indeed, if you just want to run a pi-hole, take any old laptop and just wedge it someplace...
Wanted to add that LLMs have become incredibly good at reverse engineering such devices, even when they don't have postmarketOS (now "Nura") or upstream kernel support yet. Been able to merge kernel changes and backport driver support to upstream kernels pretty easily by just giving them access to UART. They'll just go ahead and figure out all the partitioning and firmware quirks by themselves, and e.g. OpenWRT instructions seem to be well documented in the wiki so they'll understand the process on how to get there almost completely by themselves.
For my malware home lab, I decided to just reuse old phones and routers that nobody wanted anymore (quite literally got them from the trash bins). Now they're running an ARMv8 kernel as the host kernel and the rest is emulated or containerized, depending on the use case.
Works like a charm for most tasks I throw at it, and many old cheap Android phones have at least quad or octa cores in them, making them pretty awesome for podman/docker usage, too.
The only thing you have to look out for is to remove the battery, and that you use a separate power circuit with a step down converter to power it. But other than that, old smartphones have become amazing for selfhosting stuff now.
> The open question is supply. Finding two good used phones is easy; finding a few hundred matching ones, with healthy batteries and predictable camera behaviour, would take more work. I feel increasingly confident that it’s possible.
RaspberryPies and such are commodity hardware. You can buy one, two, ten, twenty, a hundred, and it just scales along. All of those devices will, barring perhaps a few dead-on-arrival ones covered under warranty, just work out of the box.
A used smartphone will have issues which can vary from device to device, so scaling that would take a very dedicated refurbishing setup where the quirks are ironed out (and the unsalvageable ones discarded).
I would also expect that you are starting out at the other end of the bathtub curve: with new devices you will have a few that fail soon after unboxing and using them a bit, and then they will run fine for years until you start hitting the other end of the curve and wear-related faults pop up. With old cheap phones you would start out at that side of the curve, so if you buy one hundred of those, a bunch will fail at various points soon after deploying them.
The cost of that unpredictable behaviour seems high to me. So much so that I wonder if it wouldn't be cheaper to buy a lot of unsold stock of a certain type of budget smartphone from two or three years back. Those devices could run fine as long as they aren't connected to the internet directly.
> RaspberryPies and such are commodity hardware. You can buy one, two, ten, twenty, a hundred, and it just scales along.
Not my experience at all. Pies are hard to find, get more expensive each week. Some are impossible to find (Pi zero).
The most fragile part of a phone is the screen. If you don't care about breaking it, then access to the battery becomes pretty easy, and therefore easy to replace.
Where are you based? I just bought 25 from CPC without any issues. Delivered in two days. They aren't as cheap as they used to be, but I understand the reasons why, given how expensive RAM is. Yes, the zero is trickier, but the 2W is in stock on PiHut at the moment.
Thank you so much for the heads up! They're as 'rare as rocking horse ...' as an old teacher of mine would say. And I've needed one for a project for ages.
Unfortunately thanks to the EU's new tariff regime it would cost over 30eur to ship it here, but luckily I can smuggle one back from my next trip to the UK...
That said for 4 and 5s - I've not had a problem here (Finland) in literally getting them off the shelf of a local retailer (Partco) for the going rate anywhere else.
I think battery replacement is only part of the problem, though. Even if you can swap batteries cheaply, you're still dealing with different firmware versions, component revisions
Phones are pretty robust though, meant to handle everyday use and abuse. In terms of failure, the battery is really the only thing to be concerned about. Most phones have some battery limiting option, and bypass the battery when the charger is connected. And if you're determined enough you can run without a battery.
Just plug all the battery slots in parallel to a 4.1V power source (that's a nearly full battery voltage). You probably have to solder them to make a reliable connection.
I'm pretty sure the third/fourth pins are used for communication to the battery. I doubt the phone would be happy it wasn't responding. I vaguely remember some phones can just use a resistor for that pin? Either way it's not quite that simple.
Maybe. Try it and see. If it doesn't work it's probably just a temperature sensor (thermistor) and a resistor indeed simulates a certain temperature the same way your power supply simulates a charge state.
You could probably open the wrap of a battery to show the individual components including the cells and other guts. Once you identify what the cables lead to then simulate it or replicate it.
Or we could look at how it communicates in-circuit, since that matters more than what it looks like to the eyeballs.
Most/all phones have one cell and therefore no balancing magic. Two connections do the job of connecting to the battery cell to the rest of the phone. These two are easily identified with a DMM.
If the phone works with only these two connections, then we're done; the phone will run from a bench power supply and there's nothing left to pursue.
If it does not work, then the functions of any additional connections be further explored while still never unwrapping the battery. Serial data? NTC thermistor? Something else? Fine; record that and emulate it. NTCs are cheap and so are MCU dev kits. (And if we don't have a decent-ish oscilloscope yet, then this is a good excuse to get one.)
1 point by leptons 16 minutes ago | root | parent | next | edit | delete [–]
If you don't have any real training in electronics, please refrain from giving advice like this. It can start fires, especially when lithium batteries are concerned.
That is not what I said at all, but thanks for taking it completely of context for your shitty troll comment.
Do whatever the fuck you want and suffer the consequences yourself for doing stupid things you don't understand, but don't encourage other people who also may lack understanding to follow you over a cliff.
If fucking around with lithium batteries doesn't make you pause a bit, then you're just ignorant.
The type of advice that 99% of HN commenters would be able to do properly…but that last 1% … I hope they don’t cause the battery to blow up when they try to open it up in just the wrong way!
Still though, could be a lot of screwing around needed to get this to work. Likely the extra wires go straight to a chip which expects to be able to monitor the voltage of the battery, and which might be alarmed by a power supply that acts unnaturally for a battery.
Hobbyist don't usually need to scale beyond one device. Warranty is not that a big issue, if you buy a used hardware you know what are getting into, that's why they are cheaper than new devices.
we are talking about embedded devices here. you dont need a second one because your web server cant serve enough requests, you need a second one because you want separate projects to have separate hardware.
I've chatted to the guy running Citronics, they provided old but working Fairphone 3 to a workshop we had at Goatmire Elixir. They try to tackle repurposing and certifying the quality of old phones. Interesting effort and well positioned for the current climate.
Their board currently on sale is this. Based on Fairphone 2. Targeting commercial or industrial more than hobbyist though. I've been trying to convince them to sell the old phones as is, as well.
https://shop.citronics.eu/
Thanks to Marc Lainez and his token spend we have a great Nerves system for running the Fairphone 3 by just programming Elixir. And people are flipping new (old) phones over to be put to work every week in the Nerves Discord.
We have a good UI framework in Emerge. We have media handling with Membrane. And Nerves handles the OS integration for stuff like networking. I thoroughly enjoy making up my own mobile OS for my own purposes.
You lose the GPIO/I2C/SPI/UART as easy extension points compared to the Pi. But the stuff that is there is already well integrated.
The main issue with old phones are NAND Flash degradation and battery problems. As long as you fix those two issues everything else is simple. These days with AI it's almost trivial to build and flash your own ROM and even the most hardcore of binary blob drivers can be reversed so long you have enough GPUs.
It's not the dark days of android where something like CyanogenMod takes a whole team of experienced system and embedded engineers to port and you don't have to hunt for information on XDA anymore.
Oh and cooling. That's something that the Pi has an edge on.
Early Android hackers might remember something called the Android Arduino Mega ADK. I suspect nowadays if you need real time control you can probably just have smartphone control the peripherals via the ESP32 using either USB C or Bluetooth. Developing over the wire protocols is trivial since LLMs can hammer out protobuf definitions like there's no tomorrow.
for applications like that your should remove the battery and put small cheap circuit that emulates it. then you can run such server very very safe, unlike with 100x batteries always harging.
Took me hours to replace the battery in a pixel. They make it almost impossible to open a phone without breaking it. I can't imagine doing that for more than 2 phones
Pixel 7a. To be fair, it was a screen + battery replacement. The problem was the adhesive that they use. I warned it up, tried to open it and the screen shattered into tiny pieces when I tried to open it (because it was cracked).
I also don't have a heat gun, so I had to use a Ziploc bag with warm water.
Did the 7 and 8 have that much worse of a battery? I'm currently replacing the two Pixel 6 in my household (end of security updates), and there is nothing wrong with those. The hardware, battery included, is as good as new (or at least, not impeding our use).
I still use a pixel 6a. I don't really care about security updates. This was my wife's phone. I would say battery life is a bit worse in the 7a, but if you are already opening it (to change the screen), then you may as well replace the battery
I have recently taken to gathering old phones and putting Nerves on them (Nerves packages up the Erlang VM into a mainline kernel to run on a variety of devices). I got inspired by a talk by Marc Lainez at Goatmire who had put it on a variety of exotic devices (e.g., a Wii).
I am by no means an expert at this, but starting of with a device on the Postmarket OS list and Claude, you can get a lot of these old devices working with mainline Linux and nearly all drivers. The newer devices are a bit harder, but if you go back nearly a decade, they're very easy to get working.
So far I have repurposed a Fairphone 3, Nexus 5, Galaxy S20, and Galaxy Tab 3.0 7". All of them (except the s20) with all drivers working compared to stock. Others have done a lot too: https://github.com/mlainez/awesome-nerves-systems
For DIY projects you can use host USB, or get one of the Adafruit FT232H type boards for GPIO pins.
Neat. I've heard good things about Erlang, though I haven't taken a proper look at it. What kind of things can you do with it? Or rather, where does it shine?
So Nerves folks typically run Elixir on the Erlang VM. So you have the Phoenix web framework, Membrane media framework, Nx (nx_eigen) for basic inference, Emerge does well for embedded UI.
I've played around with clustering a bunch of these devices over Tailscale across my home. Opening cameras from one device and streaming to anothers display, willy-nilly. All experimental for me.
I've put Homex and Espex on them to have them integrate with Home Assistant. Feeding sensor data up or whatever I feel like.
Elixir already has a production embedded ecosystem through Nerves and web services have been the bread and butter for a long time. So you can do almost anything.
To a degree, I see the author's point. But there is a catch for me: I have a few older phones and the reason why I generally prefer Pi over an old phone is the OS. If I could get a full-fledged linux on a phone, I don't have a big reason to look at Pis: I don't need the gpio in almost all cases. For most phones however, this is not only not straightforward but in 99% of cases not possible at all.
Yes, the terms "kernel" or "security updates" don't exist in the article at all.
Everything without upstream kernel support is inherently vulnerable.
This is why most phones are unfortunately not general purpose computers. They don't run the software you want, but only work on closed binary blobs hardcoded against specific kernel versions.
If only we had some magical software that could compile the vanilla Linux kernel, diff it with the vendor kernel, reverse engineer the differences and patch the original sources for you while you sleep...
I’m out of the loop on this sort of workflow. I guess this is referencing the LLMs super reverse engineering capabilities. They are impressive but how much would that sort of task cost? There isn’t a ton of headroom if we’re going to beat a Pi, right? I mean they aren’t super cheap anymore, but you also have to factor in the phone.
You can reverse quite a large program on a Claude $100 sub if you already have one lying around. I have to assume it's cheaper if you have the sources of a kernel that's 99.99% the same.
In the long run I also imagine this sort of thing being something someone does once per phone model and uploads it.
Often enough for those android devices you already have the sources for most components but only compatible with some age old android kernel and it's "just" porting it and poking around. LLMs are quite good at this.
i'm not them, but i'm doing it right now. it's going well. i'll probably write and post something about it soon (in one or two weeks, because the phone's still being ported (the phone is already very functional, but i want to port more things)).
Yeah like its a weird article idea where the pi is expensive compared to an old phone which you have for free. But that doesn't make the used phone cheaper. You can just as well say why use an expensive PI when I have an old computer with a 1680ti and 12 gigs of ram. Like yeah okay but what does that have to do with Pi? Entirely different device classes imo
Why get an expensive Pi when I have an old free Pi? My issue is with the expensive qualifier. Anything is expensive compared to something you have lying around for free.
If you want to compare the expensive factor, at least use some kind of cost to compare like the price of a used phone etc or the cost of a basic server to see how they differ.
> Yeah like its a weird article idea where the pi is expensive compared to an old phone which you have for free. But that doesn't make the used phone cheaper.
Perhaps not in the first world reality you live in, where you can buy a rpi 4 for a couple of bucks. But in third world countries they aren't readily available, and usually the imported prices exceed that of a decent PC.
You can find an old used phone for almost half the price that is still good enough for many daily processing tasks.
> You can just as well say why use an expensive PI when I have an old computer with a 1680ti and 12 gigs of ram.
That is a logical fallacy. A high end PC and an old used phone that you can buy in Ebay are in entirely different price ranges.
> Like yeah okay but what does that have to do with Pi? Entirely different device classes imo
Entirely different device classes, yeah, but both can execute the very same functions. The difference is one is much cheaper and you get 2x the processing power than the "more specialized" other. You can use both as local llm, serv
ers, home automation... and even GPIO if you expose some wires from the phone.
1. I don't live in the "first world". I own a pi 4 and have a couple of cellphones.
2. Again, "cheaper" is a valid qualifier if you compare costs. Phones vrs pi is not what I am talking about, free vrs second hand (not free) is what I am talking about. Something cannot be cheaper than free so comparing costs is kind of weird.
Sorry for any offense this semantic may have caused.
You're correct but they did lay out the table oddly because the first line has the price for the Pi but not the phone, which made it look like it was assuming they already had the phone.
Reasons to get a Pi: GPIO support, software support, OS updates, free root & bootloader, more flexibility and customizability, extensive documentation and support, consistency, simplicity, more ports, battery optional.
Used phones are good for tinkering, especially for "display with sensors and battery attached" usecases, but there are still plenty of downsides. And unfortunately Linux on phones is just not ready for primetime, last I checked.
Cool, didn't know that. But like are they physically accessible? I don't think you can use the USB port as GPIO. Maybe some unused pins on the SOC itself, but that means opening up the phone and probably removing the shielding/heatsink...
I think you can attach a keyboard to a phone over USB, and in theory it can provide both input and output (you can control the LEDs from the host computer). That would be a more convenient route, I suppose.
Atom x5 Z-8350, 2GB of RAM, 8-16GB of flash, some USB, and there's a board on that page you can have done up to do a slapdash SD adapter inside the case for more flash.
Phones are one of the biggest offenders in this too. Like the article mentioned, Xiaomi actually make rooting/flashing their phones reasonably easy, but other manufacturers actively try to block it. Also if you do flash it you'll be without most Play Integrity checks which will hamper its usage as a primary phone, so this sort of reuse is very good.
You could consider getting an older flagship phone perhaps 2022-2023 and have a strong portable device with a decent amount of compute power and not to mention ram. I'm starting to believe that these devices are under utilised and we will see some interesting applications using either with AI or without.
I particuarly think the economics of this are interesting, if you take the one plus 11 you can get one for around £200-250. The CPU is very capable and the alternatives are a PI 5 or N100. You get 16b of ram, a screen, gps, bluetooth a camera... the list goes on.
These devices could really shine as edge computing devices and I believe they are overlooked.
The problem with modern smartphones is that they run on battery, even when plugged into external power. You cannot use it without the batteries without some more hardware hacks. As the batteries aren't designed to be always in a fully charged state, if you decide to use the phone with the battery (with 24x7 power supply) then you also have to carefully monitor the batteries too. Batteries tend to heat up when charged / discharged, old batteries can also bulge or, worst case, even explode. Another thing you need to consider is cooling - Smartphones are designed with the idea that their CPUs won't run at 100% for a very long duration. A smartphones CPU, Camera, battery etc. tends to get hot on prolonged use. So if you are, for example, planning to use it as a mini webserver or media server or as a 24x7 streaming cam, you need ensure it remains adequately cool. And ofcourse, the best smartphones for these kind of projects is one capable of running Linux.
Some OnePlus devices have a "bypass charging" system. My OP8 Pro is running since 2y now as a Kubernetes node, and nothing bad has happened to its battery (yet).
I'm honestly curious to see if bypass charging is on, but it's been a while since I touched the device itself and it has been fine do far. YMMV
>>The problem with modern smartphones is that they run on battery, even when plugged into external power.
That's not universally true - a lot of android phones have a "bypass charging" mode where you run directly from the power supply, and indeed you can run them with the battery removed.
I think sometimes playing games on the phone while plugged in would actually drain battery because the CPU just used too much power (obviously a while back when charging was lower power).
Not sure if those phones would work without a battery and throttle the CPU, or if they would crash.
A lot more mobile devices will require removable batteries after 2027 due to EU regs.
There are exceptions, so not all phones will, but more will. I hope there'll be more time/effort put into hacking removable battery phones to accept hardwiring a PSU onto the battery terminals over the next 5 years.
I'm a big fan of the esp32. I needed a sensor to beep when a board came down a conveyor, and an esp32 with "VL53L0X Time-of-Flight Ranging Sensor Flight Distance Measurement Module" and a buzzer works perfect for about $10 in parts. Minus the 3d printed box and articulated arm to hold it the right distance away. Call it $10.25.
There are a few issues with Android devices. I have 2 Android phones lying around in my rack, as well as an Android tablet that I use as a dashboard.
1. Batteries swell over time. Replacing them is easy, but it's still harder than changing one inside a UPS hat. Most phones won't even start without a battery.
2. Software support will be awful. Maybe you'll get lucky and your device will be able to run postmarketOS.
3. There's no modularity, you're stuck with the full package. Don't feel comfortable having a camera or a microphone on a device with no security updates? Too bad, there's not much you can do about it.
Don't get me wrong, if you understand these risks, are fine with them, and Android devices cover your use cases — that's great, go for it! I just find SBCs with decent software support to be a better option for general projects.
#1 is not so simple - most phones (Android at least) know how big their battery is supposed to be and will shut themselves off if they think they've consumed the whole thing. You usually need to root them and mess around with some undocumented firmware settings. (It's also a pain in the ass to crack open a phone to get access to the battery.)
I once tried powering an android tablet with straight DC, turns it is EXTREMELY sensitive to slow power supplies, ie. it required a very beefy one. I had to add capacitors even to get it to work, it was really weird. But considering a battery can dish out some serious amps....
Lost in translation, I think. I mean you reuse the connector and controller and sensor from the original battery. Then you can connect any voltage 3.7-4.2 Volts to instead of Li-On cells.
There is of course the issue that a phone uses momentarily high current when transmitting. So there needs to be beefy capacitor too.
I have done this many times, of course. Especially on Ebooks, for which you cannot find replacement superthin batteries. You use original connector and any small lithiums in parallel. Discarded e-cigarettes are good source for small lithiums.
Rooting or flashing a modern phone is pain in the ass. Ideally the battery needs to be replaced with a 3.7V PSU (and BMS must be salvaged without cutting the battery accidentally). And for old phones it either old linux kernel, partially unsupported hardware or fork maintained by handful of hackers for free.
I guess laptops are way easier: more upgradeable and repairable, don't require unlocking mostly, battery can just be thrown away and it works with its native PSU. Just need to make sure power won't go out. And it is bulkier, though maybe less susceptible to overheating.
The main reason is that they are one of the few ARM computers sold with a pre-installed officially supported GNU/Linux distribution, at affordable prices.
Even with the prices increases that the PI has suffered, there no systems of vendors like Tuxedo, or System 76, at similar price point, with ARM support.
Well that's because for the Pi you have GPIO. If you just want to run homelabs or software only stuff, why not go for ITX motherboard with just in time components
I recently used Termux on a 7 year old low-end android to run a WordPress site with sqlite db and a golang sse pubsub service at a conference, via cloudflare tunnel, with dozens of people connected and doing some realtime collaborative stuff. Worked perfectly.
It was vastly more powerful than the old RPi 3 that i have, and significantly easier to work with given the screen, keyboard, wifi etc...
Nice thing about RPI is gpio, that can be used to configure the rpi with jumpers. All you need is a script that checks gpio state on boot and run scripts depending on jumpers placement. So you can get desired services started, configs applied etc without even connecting to the rpi.
Recently did this with a few old Android phones I had lying around, now they are three k3s nodes added to my existing homelab, 24 cores and 30ish GB of RAM, battery limited to 80%, running vanilla debian apart from the custom kernel build
This isnt new and has a major flaw. Phones are not made to be continously charged. You have to find or tinker your own solution to power it without battery which is more cumbersome than just getting off the shelf used hardware in the raspberry range.
Is this an actual issue though? Even if it accelerates battery degradation, that doesn't matter in this situation. Won't modern battery controls stop the battery overheating?
Depending on what you are trying to do one of the biggest problems with repurposing old mobile stuff (including laptops) is that when they are repowered they need manual intervention to get back to business.
In other words if the power goes out you need to physically be there to get the device running your always on stuff.
I guess in general, the development experience on a Pi is much better, because it’s made to be tinkered with, right? In the other hand, Android phones are also nicely hackable. I’d just guess that for beginners, it’s much easier to install the Raspberry IDE and plug in the Pi (as opposed to learning how to connect and develop for an Android phone). But yeah, for pros or vibe coders, an Android phone might be the right or even better choice. And it’s great that Android phones can therefore be used a little longer before landing in a landfill.
The top comment should be that without opening the case and wiring in the power directly (not easy) for such old phones THIS IS DANGEROUS. The battery, when under load and plugged in the whole time, can and will swell and explode somewhere in the future.
Pis are too expensive.
I just picked up a new Bosgame AMD 5300U 4/8 core 8 GB 256GB SATA SSD, 2.5 Gbx2 NICs, 2x SSD slots, for $199 on Amazon. Great OpenWRT/Proxmix/Mini-NAS. 8 watts idle.
Phones are cheap for what you get, but problems arise when you try to add something. Driving LEDs is the start. So how do you do that? Via a cloud service? How do you add an air monitor or servos? The pi and esp32s have an ecology of add-ons and a community to support you. Sure that could be built around a rooted phone, but it would take years.
You use an ESP32 or RP2040 or STM32 for that and connect it to the phone with USB. This is sometimes done with a Pi to begin with - if you need to run a 3D printer, a Pi's GPIO will not be enough. Or not, these chips might be able to handle what you need to do all on their own.
>Why get an expensive Pi when I have an old phone?
>I’ve unlocked and rooted the Redmi
Herein lies the issue. I don't know fuck all about phones. I don't want to. They are engineered to keep me out, and that is a pain in the ass I don't want to deal with when doing hardware work. I want to own something that works the second I get it and doesn't fight me the entire way. Sure, I could let an LLM guide me through it or even do it for me, but why bother when the whole point of the hardware is for me to fuck off and use it as a subscriber to the OEM and carrier? This is tedium, removed by further spending. Not to mention documentation - what bullshit is the OEM hiding in their board or firmware thats going to ruin my day? I'm sure that some of these devices are great for reuse as a self contained server with media and modem, but is my old phone going to do this?
I'm not gonna spend $50 on ewaste to try to do something unless I want to use that ewaste (old keyboards, old toys? :3). Old phones are a non-sequitur because, again, they're engineered to keep me out.
>>Herein lies the issue. I don't know fuck all about phones. I don't want to. They are engineered to keep me out, and that is a pain in the ass I don't want to deal with when doing hardware work.
I mean, you do know we're on a forum called hacker news, right?
In most cases it's not some magic voodoo process that you need an LLM for, it's the first result on Google and one command to run or in some cases you just hold two buttons together when booting up and click the unlock button.
Like, I completely understand your perspective and desire to have things which "just work"(I'm in the same boat a lot of the time lately) but like, let's be realistic - rooting an old phone and running some software on it is really not a big deal, I'd hope especially for people on here.
I tried this once - with a little "fleet" of phones running 24/7 serving things. But it turned out that phones do not really like running 24/7 doing something. Eventually they started failing in various ways.
This is valid for one off projects but if you want something that can replicate/scale and easily integrate into a product then RPi or Jetson boards are great
Old phones are fine if you want to tinker but are a nightmare to try and actually 'do' something with. It sucks and shouldn't be the case but here we are.
Even better are old smartphones with Sailfish OS, Ubuntu Touch or PineOS - they are based on proper Linux with full control, not crippled like stock Android, unless you flash it (if you can).
Depends on what you want to do. In terms of ram and CPU I understand but if you want HDMI, I/O... And lot of documentation about existing projects, troubleshooting....
Agreed, dealing with deprecated android devices can be frustrating even when rooted. There was a project to bring a standard deb linux distro build to native mobile hardware, but most features like cell data access become a brick.
It is the drawback of locked proprietary hardware builds no one makes anymore.
An recent laptop with a broken LCD is cheap e-waste, and works just fine to setup on an external monitor. Mini PC are also cheaper than most pi5 now, so if you don't need the gpio... save the money for a real SSD. =3
Not as a knock to the author but I feel like this is a wildly optimistic / simplified take. The type of audience pis serve is either industrial as part of a product or process or tinkerers that often want to follow some set process like installing retroarch.
Not to mention not all phones are hackable, they might not have the gpio or software you want or ecosystem (I’m quite fond of the Poe hats for example) and frankly if you’re buying a pi you’re obviously not the person who has the perfect hackable phone lying around or otherwise can easily source one
Thats a great read. I explored similar things with my rooted phone. Samsung galaxy m30s. Wrote something on it as well on what I did, may be you will also like this type of experimentation?
On a similar note, old smartphones (depending how old of course) are way more powerful than a Raspberry Pi. You can turn them into a Kubernetes cluster [1].
Hey, I can answer that! Because me and my family have always had Verizon and when I recently went through alllllll the devices I have ever owned, I would need to go back to circa 2010's Galaxy S7 to be able to get a rooted device, which as Claude keeps reminding me, "isn't what I want, but might be good enough"....
Raspberry PI's don't give me near the struggle and I never dumbly signed a contract that I didn't really own the device I was purchasing (that I am completely aware of)
I was thinking about using my old phones with something similar, but I'm worried about battery. When permanently connected to charger, won't they swell, start fire or something like that after longer(er) period of time?
Yeah, people mostly just buy things to satisfy their wants. the PI sells because its the more reliable, low-friction option for a lot of use cases,even if an old phone is technically stronger and cheaper
For me if I don't have a need for the GPIO pins, it's "why get an expensive pi when I can get a used ex-corporate small form factor quad core, intel core i7-something desktop PC with 16GB of RAM and a real SSD"... For like 150 bucks on eBay. RAM apocalypse and price of equipment has affected this a lot (these used to be like 90 bucks depending on era of core i7), but they're vastly more useful when a general purpose x86-64 linux is installed on them. Or even just put proxmox on it and then various things inside that.
You are asking a meaningless question in order to promote, brag, justify and make public your own know-how having already accepted and understood whatever limitations in order to justify your own choice.
The point of a rasperri PI is to allow general, easy, and intentional bespoke availability of a fully programmable platform, where it also allows you access to low level, but practical means of acheiving home automation via GPIO etc.
I have 5 old phones in my cupboard, the most recent 3 of them have locked bootloaders and could potentially be rooted simply because of them not being region locked, while not gauranteeing me any real and complete support such as a PI would.
Nevermind even thinking as far ahead as GPIO pins etc etc.
At best, 1 out of the 5 phones might be able to eventually, and with effort/time and possible additional cost for extras give me parity to a PI, even exceeed the PI in compute.
But that is not the point.
You know it.
I know it.
Your post is not in search of an answer. You know the answer, you are the minority, and you know this.
You would not get an expensive PI if you had what you need from your old phone and skillz.
So why ask us where you already knew the answer...
It's because you are not asking anything, you are trying to show off and be dishonest by posing it as a question while assuming generousity by 'helping' people not 'waste money'
You are transparent buddy.
Yep, the typos are understood and left as is. Does nothing to take away my meaning.
I've always found it a bit absurd that an old phone can have more sensors and connectivity than a fairly expensive hobbyist computer, yet we rarely consider using it as one
As long as it's not an iPhone. Once Apple has bricked your old iPhone, you can't even run your own software on it anymore. Despite all their lip service to being environmentally friendly, that thing is nothing but an expensive paperweight and e-waste now.
Yes, USB to GPIO adapters are available. Note that you won't be able to implement anything that has timing requirements - you'll need a microcontroller for that.
I think in practice the timing jitter and the low cost difference to something like an Arduino Nano or Pi Pico means people use one of those even if they could technically use a GPIO adapter.
If you mean the 13C then you are no longer receiving security updates as of four days ago. You can go down the ROM route but you'll likely lose your banking and Google Pay.
A lot of phones these days are offering longer update periods. For instance that 13C only offered 3 years from release, but the 15C is offering 6 years. The Pixel phones are getting 7 years.
The boards are of course bigger, which I guess will be a non-starter for many. They'll also require active cooling unless you downclock them considerably. They often also require more power, although many mainboards can be configured down two a few watts of power consumption, depending on what peripherals and speed you need.
Yeah, it's more work than just getting a Pi, but given the current prices, it makes sense for me.
If you use the raspberry pi for homeservers or general compute (which I did/do sometimes) then the value proposition is quite poor regardless of pre AI prices unless you have very strict power requirements (in which case it probably fits the above use case already). The comparison with NUCs or mini PCs or laptop boards feels like it's a wrong comparison.
The chips used in PIs are repurposed Broadcom SOCs that have inferior bus speeds compared to a general purpose processor from intel or amd, even few generations old. For any general purpose computing these chips will always be slower. its like comparing apples and oranges. I never really understood why people did that.
Yeah, I don't disagree, but I would argue the Raspberry was always an applorange, and that's exactly what made it popular (together with the cheap price in the beginning, which made it basically an impulse buy). It brought people into microcontroller/embedded systems territory without having to actually deal with all of the stuff that made this so daunting. As you said, it is basically a normal Linux PC, just that it runs Arm and has GPIO. People who are really into electronics/embedded systems will use a proper microcontroller nowadays anyway, these have gotten so powerful and easy to use that there's really no reason for a Pi anymore. And on the other side, people who want a little server, a pi-hole, tailscale or octoprint or whatever, can just use an old used laptop/miniPC or similar. Since the Pi has gotten so expensive, I see basically no reason to get one anymore.
By the way, I got a zero 2W a few months ago and it costed me 47€
You can make such a dongle yourself from ~any MCU dev board, but there are also productized "USB GPIO breakout" adapter boards, e.g. the Adafruit MCP2221A.
The "Pi as an SBC with GPIOs" case now only makes sense if you really, really require the Pi's compactness and/or monolithic-ness (as might be the case for e.g. ruggedized embedded systems like robots.)
For many (maybe even most) people they do the same thing. Some just want a HTPC, basic desktop, or small server, no GPIO needed. Raspberry Pis were definitely cheaper early on because you could actually buy boards at or near MSRP, but over time the price rose and mini PCs became cheaper.
or maybe the mini pcs started becoming available at all. I think these were a concept since core 2 duo and first gen i3 era. before then the chips were too much wattage for a mini pc. and the first generation mini pcs were also costly for enterprises to adopt in bulk. its only after 4th and 6th gen intel that these were adopted in large quantities in enterprises and once they dumped those in the used market they became comparable in pricing and somewhat in power to a rpi.
Used powerful PCs have always been available for dirt cheap on eBay even when the first pi was launched. just that mini pcs are a more recent thing.
As far as mini-pcs... The first Intel NUC was a Sandy bridge, i.e. Gen2, FWIW. And we did have the 'Mac Mini' as far back as 2005, buuuuut that wasn't cheap because Apple.
I think SSDs were another big part of mini-PCs though; it's easy to forget for granted the risks of having a hard drive in a poor thermal environment, as well as the general size constraints one imposes. [0]
[0] - Surrrre, you could use a 2.5 inch HDD instead, but 2.5 inch HDDs almost always were slower than a 3.5 inch HDD (excluding the VelociRaptor, but that would only be more demanding as far as thermal...)
At first it was a blessing that the Raspberry Pi gained popularity as a general purpose computer, not just for education and embedded applications. The additional demand and popularity brought more money and development attention to the ecosystem.
At some point, which I could never really put my finger on, this general purpose computer popularity grew so large that the embedded systems and educational aspects of Raspberry Pi seemed almost forgotten by the mainstream users. At a local in-person meetup group it was a running theme for people to brag about how many Raspberry Pis they owned for different purposes and to boast about using them for applications where they didn’t really fit. In person I’ve heard more about Raspberry Pi NAS systems (where people are always disappointed with the performance and the price of forcing a Pi into this role) or Raspberry Pi Kubernetes clusters (which mostly sat idle and could have been replaced by a single PC with more RAM).
This reached such a critical mass that it almost seemed like the mainstream discussion forgot the original purpose of the Pi and thought it was literally just another computer, but small. Most people who tried to use it for this purpose were left slightly frustrated by how it didn’t perform like they wanted a computer to perform.
So now every discussion about Raspberry Pi attracts comments about how nobody should buy a Pi when you can buy a NUC or laptop or some other used piece of hardware and it’s so much “better”. This is the right advice for a lot of people who wanted a powerful general computer but bought a Pi instead, but it’s also furthering the misunderstanding of what the Pi is for.
The real advice is that everyone should be shopping for the right tool for the job. There are a lot of cool applications for the Pi where a NUC or old laptop or phone doesn’t make sense because you actually need the compact size, the low-level I/O, or some other feature. However anyone who can put a NUC or laptop side by side with a Raspberry Pi and consider them both viable options for their project probably shouldn’t have been looking at a single board computer to begin with.
There is a parallel situation happening with the “just buy an ESP32” arguments. If an ESP32 solves your problem, you also probably shouldn’t have been in the market for a Raspberry Pi.
I have mixed feelings on that.
For the longest time it was safe to say "just buy a Raspberry Pi" because they were cheap enough and powerful enough. If memory serves me correctly, the base board cost roughly the same as an Arduino Uno and offered considerably more. (Of course, there were cheaper Uno clones, but those depended upon support offered by official Arduino software development tools.)
While there are some projects which would only be possible with something like the Uno, there were far more projects that would only be possible with something like the Pi. So if you wanted to give generic go-to advice to a novice, the Pi it was.
The right advice is to use the right tool for the job, but that is something that a novice (or fly-by-night forum posters) aren't prepared to handle.
Of course, things change. There are more inexpensive microcontroller development boards and the microcontrollers themselves are much more powerful. While that doesn't negate the benefits of a Linux SBC, it does raise questions now that those SBCs are considerably more expensive.
IMO, the problem with the Pi GPIO is that it's a collection of GPIO pins and low pin count buses for controlling peripherals, rather than a singular bus. The fact that it runs Linux without an RTOS monitor program or a sidecar of sorts adds to the problem.
Suddenly the Kodi and the GPIO crowd (Arduino, ESP8266, ESP32, etc.) could work with the same board.
A new era, like the IBM PC or the iPhone. Changed computing forever.
EDIT: like other commenters noted, it seems like it made the intel NUC a reality.
Fantastic for one-off projects.
Nowadays they require cooling fans and cost absurd amounts considering the performance so there seems little point.
Most of my one off projects these days are now on microcontrollers.
In a better world we'd all have one in our homes, storing all our personal data and documents. The cloud would only be used for storing encrypted backups.
Many small businesses, kindergartens etc. could also have all their digital needs served this way.
Yes they would sometimes break down, but with the right setup, the fix would be a matter of booting a replacement from the right thumbdrive.
I have a Nextcloud server running on it and the maintenance is negligible.
During the recent PHP-vulnerability I simply turned it off. The next day I booted it, and it automatically installed the new Nextcloud image with the vulnerability fixed.
Makes me wonder if I could have it check for CVEs and automatically shut down vulnerable containers until updated images are out.
May I ask, how do you think kindergartens would benefit from a home server?
0: https://github.com/geerlingguy/mini-rack#build-showcase
[0] https://www.servethehome.com/introducing-project-tinyminimic...
[1] https://www.amazon.com/GeeekPi-Cabinet-Equipment-RackMate-Ra...
This let me standardize on amd64 and simplified a bunch of my deploy code.
Of course, there are a wealth of hardware options out there, so it varies. I've heard as low as 4 watts.
[1] https://www.youtube.com/watch?v=veRja-F4ZMQ
btop says CPU takes 1W when "idle" (seeding some torrents), no idea what the rest of the board and NVMe could take.
Can you explain this a bit? What is the application/use-case where the power adapters are a problem?
Old mobile phones or laptops are riskier because of batteries. Some laptops can work without batteries as well, but do old phones?
Intel MacBook Pro from 2014 works but throttles the CPU to 800 Hz making it barely usable.
I have a ThinkPad x230 (3rd gen i7) which makes an incredible Windows XP nostalgia device. It’s basically the newest hardware that exists with full XP drivers (other than USB3 which I don’t need) so it’s massively OP, which was the point of that project. Basically, “what if I had access to insanely futuristic hardware in 2002.”
Obviously I wouldn’t sacrifice that hardware but it might be cool to try that on yet another laptop. (Hey, I don’t have a laptop collecting problem!)
they did, with modifications
I feel like The Internet has this urban myth that says keeping a charged battery plugged in all the time will eventually cause it to explode and burn your house down, guaranteed.
I get that keeping a battery at 100% charge is not great for it in terms of longevity (ability to hold a charge). And it's obvious that heating the battery up by running the laptop at full tilt inside a backpack or something will turn it into a spicy pillow.
But I struggle to imagine the failure mode that reliably shows up when keeping the laptop powered on 24x7 that is never a problem for 8-16 hours of continuous use while plugged in (a very typical work day for my laptops).
I ended up repairing the Mac, replacing the battery, and installing Chrome OS on it. (chrome sucks though, someday I’ll get it back out and switch to real Linux)
I had a MacBook Pro where the laptop worked but the battery became swollen so much that it distorted the case. If I wasn't there to notice it, it would probably burst at some point and Lithium is very flammable.
It's probably rare and overblown, but it can happen.
If that had been in a closet, it would have been months or years before I realized.
for example some alienware and XPS models rely on the battery for amperage capacitance when their big onboard GPUs get busy; allows them to run cooler/less-wattage/smaller wall-warts but it also makes for a platform that'll crash when it hits a high GPU usage without a decent way to make up the amps.
'spose you could use a bigger power supply if you get bios/uefi to play nice. hard proposition sometimes with Dell products.
Batteries degrade much faster when fully cycled repeatedly, and even then just about any laptop battery can easily make it 6+ years with low-to-zero risk of failure.
Apple documented the behavior in a support note. I’m not sure if the Apple Silicon models do the throttling as well. I guess the situation doesn't come up as often since the batteries are no longer user-removable.
> Unless you're just doing it for the love of the game
Isn't that way a lot of people do the home-lab thing and related projects what would need the Pi these ideas are talking about replacing? It is why I do it [or used to be before working in tech for decades killed a lot of the enjoyment I had for playing with tech].
> or happen to have a lot of laptop boards sitting around you can sell
You are the first in this thread to bring up what any of this is worth on a larger scale, selling the results. Everyone else is talking about putting together one simple computer or jumped-up microcontroller for a task of their own, and maybe more later for other projects. Some of us find this sort of thing fun, and in the shit world we are all living in right now having a bit of fun and achieving something [sort of] practical out of it, can be priceless.
For my malware home lab, I decided to just reuse old phones and routers that nobody wanted anymore (quite literally got them from the trash bins). Now they're running an ARMv8 kernel as the host kernel and the rest is emulated or containerized, depending on the use case.
Works like a charm for most tasks I throw at it, and many old cheap Android phones have at least quad or octa cores in them, making them pretty awesome for podman/docker usage, too.
The only thing you have to look out for is to remove the battery, and that you use a separate power circuit with a step down converter to power it. But other than that, old smartphones have become amazing for selfhosting stuff now.
> The open question is supply. Finding two good used phones is easy; finding a few hundred matching ones, with healthy batteries and predictable camera behaviour, would take more work. I feel increasingly confident that it’s possible.
RaspberryPies and such are commodity hardware. You can buy one, two, ten, twenty, a hundred, and it just scales along. All of those devices will, barring perhaps a few dead-on-arrival ones covered under warranty, just work out of the box.
A used smartphone will have issues which can vary from device to device, so scaling that would take a very dedicated refurbishing setup where the quirks are ironed out (and the unsalvageable ones discarded).
I would also expect that you are starting out at the other end of the bathtub curve: with new devices you will have a few that fail soon after unboxing and using them a bit, and then they will run fine for years until you start hitting the other end of the curve and wear-related faults pop up. With old cheap phones you would start out at that side of the curve, so if you buy one hundred of those, a bunch will fail at various points soon after deploying them.
The cost of that unpredictable behaviour seems high to me. So much so that I wonder if it wouldn't be cheaper to buy a lot of unsold stock of a certain type of budget smartphone from two or three years back. Those devices could run fine as long as they aren't connected to the internet directly.
Not my experience at all. Pies are hard to find, get more expensive each week. Some are impossible to find (Pi zero).
The most fragile part of a phone is the screen. If you don't care about breaking it, then access to the battery becomes pretty easy, and therefore easy to replace.
Unfortunately thanks to the EU's new tariff regime it would cost over 30eur to ship it here, but luckily I can smuggle one back from my next trip to the UK...
That said for 4 and 5s - I've not had a problem here (Finland) in literally getting them off the shelf of a local retailer (Partco) for the going rate anywhere else.
Most/all phones have one cell and therefore no balancing magic. Two connections do the job of connecting to the battery cell to the rest of the phone. These two are easily identified with a DMM.
If the phone works with only these two connections, then we're done; the phone will run from a bench power supply and there's nothing left to pursue.
If it does not work, then the functions of any additional connections be further explored while still never unwrapping the battery. Serial data? NTC thermistor? Something else? Fine; record that and emulate it. NTCs are cheap and so are MCU dev kits. (And if we don't have a decent-ish oscilloscope yet, then this is a good excuse to get one.)
If you don't have any real training in electronics, please refrain from giving advice like this. It can start fires, especially when lithium batteries are concerned.
Do whatever the fuck you want and suffer the consequences yourself for doing stupid things you don't understand, but don't encourage other people who also may lack understanding to follow you over a cliff.
If fucking around with lithium batteries doesn't make you pause a bit, then you're just ignorant.
Still though, could be a lot of screwing around needed to get this to work. Likely the extra wires go straight to a chip which expects to be able to monitor the voltage of the battery, and which might be alarmed by a power supply that acts unnaturally for a battery.
It makes more sense with high availability though. Their current situation seems rough.
Their board currently on sale is this. Based on Fairphone 2. Targeting commercial or industrial more than hobbyist though. I've been trying to convince them to sell the old phones as is, as well. https://shop.citronics.eu/
Thanks to Marc Lainez and his token spend we have a great Nerves system for running the Fairphone 3 by just programming Elixir. And people are flipping new (old) phones over to be put to work every week in the Nerves Discord.
We have a good UI framework in Emerge. We have media handling with Membrane. And Nerves handles the OS integration for stuff like networking. I thoroughly enjoy making up my own mobile OS for my own purposes.
You lose the GPIO/I2C/SPI/UART as easy extension points compared to the Pi. But the stuff that is there is already well integrated.
It's not the dark days of android where something like CyanogenMod takes a whole team of experienced system and embedded engineers to port and you don't have to hunt for information on XDA anymore.
Oh and cooling. That's something that the Pi has an edge on.
Early Android hackers might remember something called the Android Arduino Mega ADK. I suspect nowadays if you need real time control you can probably just have smartphone control the peripherals via the ESP32 using either USB C or Bluetooth. Developing over the wire protocols is trivial since LLMs can hammer out protobuf definitions like there's no tomorrow.
Bot farms works that way.
I also don't have a heat gun, so I had to use a Ziploc bag with warm water.
Still though, they make it really hard
I don't have experience with the ziploc method, but heat guns can melt solder so I'm wondering if temp was not high enough with the ziploc.
Could you say more about this? What kind of setups work well? Is this really straightforward or are there any subtleties?
I am by no means an expert at this, but starting of with a device on the Postmarket OS list and Claude, you can get a lot of these old devices working with mainline Linux and nearly all drivers. The newer devices are a bit harder, but if you go back nearly a decade, they're very easy to get working.
So far I have repurposed a Fairphone 3, Nexus 5, Galaxy S20, and Galaxy Tab 3.0 7". All of them (except the s20) with all drivers working compared to stock. Others have done a lot too: https://github.com/mlainez/awesome-nerves-systems
For DIY projects you can use host USB, or get one of the Adafruit FT232H type boards for GPIO pins.
So Nerves folks typically run Elixir on the Erlang VM. So you have the Phoenix web framework, Membrane media framework, Nx (nx_eigen) for basic inference, Emerge does well for embedded UI.
I've played around with clustering a bunch of these devices over Tailscale across my home. Opening cameras from one device and streaming to anothers display, willy-nilly. All experimental for me.
I've put Homex and Espex on them to have them integrate with Home Assistant. Feeding sensor data up or whatever I feel like.
Elixir already has a production embedded ecosystem through Nerves and web services have been the bread and butter for a long time. So you can do almost anything.
Because the pi has a maintained kernel and a userspace I don't have to fight.
Everything without upstream kernel support is inherently vulnerable.
This is why most phones are unfortunately not general purpose computers. They don't run the software you want, but only work on closed binary blobs hardcoded against specific kernel versions.
In the long run I also imagine this sort of thing being something someone does once per phone model and uploads it.
If you want to compare the expensive factor, at least use some kind of cost to compare like the price of a used phone etc or the cost of a basic server to see how they differ.
Perhaps not in the first world reality you live in, where you can buy a rpi 4 for a couple of bucks. But in third world countries they aren't readily available, and usually the imported prices exceed that of a decent PC.
You can find an old used phone for almost half the price that is still good enough for many daily processing tasks.
> You can just as well say why use an expensive PI when I have an old computer with a 1680ti and 12 gigs of ram.
That is a logical fallacy. A high end PC and an old used phone that you can buy in Ebay are in entirely different price ranges.
> Like yeah okay but what does that have to do with Pi? Entirely different device classes imo
Entirely different device classes, yeah, but both can execute the very same functions. The difference is one is much cheaper and you get 2x the processing power than the "more specialized" other. You can use both as local llm, serv ers, home automation... and even GPIO if you expose some wires from the phone.
2. Again, "cheaper" is a valid qualifier if you compare costs. Phones vrs pi is not what I am talking about, free vrs second hand (not free) is what I am talking about. Something cannot be cheaper than free so comparing costs is kind of weird.
Sorry for any offense this semantic may have caused.
Used phones are good for tinkering, especially for "display with sensors and battery attached" usecases, but there are still plenty of downsides. And unfortunately Linux on phones is just not ready for primetime, last I checked.
I do use termux daily to ssh into other boxes, that's incredibly useful
If you can access it from within the sandbox, you don't need to be limited to keyboards - the HID framework allows for arbitrary I/O.
https://www.ebay.com/itm/147663241534
https://www.parkytowers.me.uk/thin/wyse/3040/
Atom x5 Z-8350, 2GB of RAM, 8-16GB of flash, some USB, and there's a board on that page you can have done up to do a slapdash SD adapter inside the case for more flash.
I particuarly think the economics of this are interesting, if you take the one plus 11 you can get one for around £200-250. The CPU is very capable and the alternatives are a PI 5 or N100. You get 16b of ram, a screen, gps, bluetooth a camera... the list goes on.
These devices could really shine as edge computing devices and I believe they are overlooked.
I'm honestly curious to see if bypass charging is on, but it's been a while since I touched the device itself and it has been fine do far. YMMV
See: https://blog.denv.it/posts/pmos-k3s-cluster/
That's not universally true - a lot of android phones have a "bypass charging" mode where you run directly from the power supply, and indeed you can run them with the battery removed.
Not sure if those phones would work without a battery and throttle the CPU, or if they would crash.
There are exceptions, so not all phones will, but more will. I hope there'll be more time/effort put into hacking removable battery phones to accept hardwiring a PSU onto the battery terminals over the next 5 years.
1. Batteries swell over time. Replacing them is easy, but it's still harder than changing one inside a UPS hat. Most phones won't even start without a battery.
2. Software support will be awful. Maybe you'll get lucky and your device will be able to run postmarketOS.
3. There's no modularity, you're stuck with the full package. Don't feel comfortable having a camera or a microphone on a device with no security updates? Too bad, there's not much you can do about it.
Don't get me wrong, if you understand these risks, are fine with them, and Android devices cover your use cases — that's great, go for it! I just find SBCs with decent software support to be a better option for general projects.
Then you can use any 4 volt power source.
#2 and #3: You install Termux.
I once tried powering an android tablet with straight DC, turns it is EXTREMELY sensitive to slow power supplies, ie. it required a very beefy one. I had to add capacitors even to get it to work, it was really weird. But considering a battery can dish out some serious amps....
Also 4V sources are not common.
There is of course the issue that a phone uses momentarily high current when transmitting. So there needs to be beefy capacitor too.
I have done this many times, of course. Especially on Ebooks, for which you cannot find replacement superthin batteries. You use original connector and any small lithiums in parallel. Discarded e-cigarettes are good source for small lithiums.
I guess laptops are way easier: more upgradeable and repairable, don't require unlocking mostly, battery can just be thrown away and it works with its native PSU. Just need to make sure power won't go out. And it is bulkier, though maybe less susceptible to overheating.
Even with the prices increases that the PI has suffered, there no systems of vendors like Tuxedo, or System 76, at similar price point, with ARM support.
It was vastly more powerful than the old RPi 3 that i have, and significantly easier to work with given the screen, keyboard, wifi etc...
https://github.com/crazynucfan/phone-nodes
Even using a timer to charge only every odd day still leads to the same thing.
In other words if the power goes out you need to physically be there to get the device running your always on stuff.
>I’ve unlocked and rooted the Redmi
Herein lies the issue. I don't know fuck all about phones. I don't want to. They are engineered to keep me out, and that is a pain in the ass I don't want to deal with when doing hardware work. I want to own something that works the second I get it and doesn't fight me the entire way. Sure, I could let an LLM guide me through it or even do it for me, but why bother when the whole point of the hardware is for me to fuck off and use it as a subscriber to the OEM and carrier? This is tedium, removed by further spending. Not to mention documentation - what bullshit is the OEM hiding in their board or firmware thats going to ruin my day? I'm sure that some of these devices are great for reuse as a self contained server with media and modem, but is my old phone going to do this?
I'm not gonna spend $50 on ewaste to try to do something unless I want to use that ewaste (old keyboards, old toys? :3). Old phones are a non-sequitur because, again, they're engineered to keep me out.
I mean, you do know we're on a forum called hacker news, right?
In most cases it's not some magic voodoo process that you need an LLM for, it's the first result on Google and one command to run or in some cases you just hold two buttons together when booting up and click the unlock button.
Like, I completely understand your perspective and desire to have things which "just work"(I'm in the same boat a lot of the time lately) but like, let's be realistic - rooting an old phone and running some software on it is really not a big deal, I'd hope especially for people on here.
It is the drawback of locked proprietary hardware builds no one makes anymore.
An recent laptop with a broken LCD is cheap e-waste, and works just fine to setup on an external monitor. Mini PC are also cheaper than most pi5 now, so if you don't need the gpio... save the money for a real SSD. =3
Nowadays I dumpster dive for old computers and parts to Frankenstein together what I can with Debian as headless servers.
One man's trash is another man's treasure, repurposed to sail the high seas.
Arrrr!
Not to mention not all phones are hackable, they might not have the gpio or software you want or ecosystem (I’m quite fond of the Poe hats for example) and frankly if you’re buying a pi you’re obviously not the person who has the perfect hackable phone lying around or otherwise can easily source one
https://nimitbhardwaj.github.io/posts/hermes-agent-rooted-an...
[1]: https://blog.denv.it/posts/pmos-k3s-cluster/
Raspberry PI's don't give me near the struggle and I never dumbly signed a contract that I didn't really own the device I was purchasing (that I am completely aware of)
https://f-droid.org/en/packages/mattecarra.accapp/
You can also "just" take the battery out, it might still function without it or with something that pretends to be a battery but isn't
Some are bought with US carrier locks - never activated or unlocked. Any chance they can be unlocked with today's tools and not dumped into ewaste.
There are a few simple projects for home automation i have in mind but the rooting felt like a daunting task.
The pi has pihole and unbound installed. You aren't supposed to use these programs with DHCP but I'm a cowboy.
All clients connected to this AP have less ads.
Built it for some seniors who got sick of ads. Works great for them.
But a new pi is just a bad deal for anything right now.
My blog article about this: https://louwrentius.com/the-raspberry-pi-5-is-no-match-for-a...
Sure, CPU wise an old phone can do the same job, but not the GPIO.
But hey, thanks AI delusion , even a fucking raspberry pi is now a purchase I gotta think 3 times about before pulling the trigger.
You can get a cheap Tracfone for like $20, never pay for the subscription, and run termux on it.
Need to fuck about pretty much per phone model,rather than install Debian, unattended-upgrades and forget about device for 2 years
* Lithium batteries are a problem
* No ethernet
* No prebuilt images like say Talos
* Can't readily attach an SSD
The point of a rasperri PI is to allow general, easy, and intentional bespoke availability of a fully programmable platform, where it also allows you access to low level, but practical means of acheiving home automation via GPIO etc.
I have 5 old phones in my cupboard, the most recent 3 of them have locked bootloaders and could potentially be rooted simply because of them not being region locked, while not gauranteeing me any real and complete support such as a PI would.
Nevermind even thinking as far ahead as GPIO pins etc etc.
At best, 1 out of the 5 phones might be able to eventually, and with effort/time and possible additional cost for extras give me parity to a PI, even exceeed the PI in compute.
But that is not the point. You know it. I know it.
Your post is not in search of an answer. You know the answer, you are the minority, and you know this.
You would not get an expensive PI if you had what you need from your old phone and skillz.
So why ask us where you already knew the answer...
It's because you are not asking anything, you are trying to show off and be dishonest by posing it as a question while assuming generousity by 'helping' people not 'waste money'
You are transparent buddy.
Yep, the typos are understood and left as is. Does nothing to take away my meaning.
https://www.parkytowers.me.uk/thin/
No one cares if you have overpriced white earbuds or not for Pi tasks.
I think in practice the timing jitter and the low cost difference to something like an Arduino Nano or Pi Pico means people use one of those even if they could technically use a GPIO adapter.
A lot of phones these days are offering longer update periods. For instance that 13C only offered 3 years from release, but the 15C is offering 6 years. The Pixel phones are getting 7 years.