If the device can run PostmarketOS with a mainstream kernel, then it can run any Linux distribution. (I put Arch ARM on such devices, since I like that distro.)
That's the big hurdle though - mainstream kernel support.
For most devices, even if they can be rooted and jailbroken, you're stuck with the kernel they come with. Doesn't have a new feature you need? A horrible security flaw in the network stack? You're out of luck. Most "repurpose your old phone" approaches have this problem. You can make it a server but you wouldn't want to expose it to the public Internet.
But yes, this is definitely an issue. I've been playing with a 2013-era Samsung device that came with a 3.0 kernel. It can run pmos with said kernel but there are multiple root LPE vulns. I've been looking into getting it to run a mainline kernel just for fun, but it's not going to be easy.
I note that Linux mainline has a device tree for the "Samsung Galaxy S1 (GT-I9000) based on S5PV210", not sure how complete it is though. Lots of others too:
This is the kind of task I've found tools like Codex to be pretty good at. You just have to be able to give it good enough access to test and debug its work.
x86_64 is the only official Arch Linux. All other ports are unofficial. They are community projects where many of the members are the same as the main Arch Linux.
I think it's basically for the same reason as why they dropped 32-bit x86 support about 8 years ago. Not enough users. (That resulted in the unofficial Arch Linux 32 to maintain support.)
I think the reason is they don't want to become debian where deciding anything takes foverever. Another architecture is a liability, so it lives in another "project" that official arch is not committed to.
I write this from arch on arm (orange pi) thingy, btw
The thing that holds me back from this is always the battery. I want to have my battery removed so that it doesn't eventually become a time bomb, but it's a pain on modern phones and I'm not even sure if they boot without. The mobile hardware reuse space can suck for hobbyists.
Most phones can have battery removed somewhat destructively, but without affecting the rest of the phone.
Generally, as long as you keep the phone plugged in, the battery should be safe virtually indefinitely - the battery management on board will keep it in a state where its a constant charge which means the chemistry will be stable.
There were several generalizations in that statement that align with my similar fears to the OP. Most firmware should minimize the charge cycling, most batteries should be stable at constant charge... most isn't great for something that I want to sit in the corner undisturbed for a decade just chugging along - I have a few old desktops I use whenever I need a stand alone server or to host something web-live for a while. They'll eventually have hardware failures, but I have a lot more confidence that when they fail it won't be dramatic or destructive - ditto with old laptops, the serviceability expectations are much higher than phones so I have yet to meet a laptop I can't pop open and just pull the battery out of to run on AC alone - in the case of a power failure the UPS can't cover I'd rather the machine just power off rather than needing to deal with the possibility of dramatic failure.
I think if you're considering re-harvesting old devices to use for hosting and get far enough down your list to get to phones then you've likely got enough constant maintenance costs in overseeing things that the additional worry of fire risk just isn't worth it.
I'm not educated enough in this area to have any expertise, however, in my personal experience leaving a lithium-ion battery plugged all the time results in scary semi-exploded batteries that also stop working.
Would you say this is a chemistry/QA problem? Have there been advances in battery / controller technology that achieves the above?
Of the six old Android phones I have around, two of them I don't dare turn on due to swollen batteries. I guess it depends how old the devices are whether this was a real risk, but I won't leave devices plugged in anymore for this reason.
Depends on your phone. Just has to replace the battery on a generally always-plugged in Moto (at least after a certain age). Battery had pillowed out. It's acting as our "landline" with a link2cell on some old DECT handsets.
So the phone effectively becomes a 4U rack server that's probably not much of a fire hazard. We'll tuck it away behind some wood for extra safety. Never liked sleeping with my eyes shut anyway!
In theory, you can replace the battery with a chunky enough capacitor (to get past the power-on surge) and a power source at the right voltage attached right where the battery would go. The soldering points are way too tiny for my amateur soldering skills, though.
> I grabbed a few power point timer switches, and set them to only over up the charger for a hour a day. Never had another battery puffing failure - at last not in the next 2 or 3 years before I left.
hopefully "bypass charging" becomes more of a thing in the future. a few of the latest pixel phones use it but the only other time ive seen it is on tablets aimed at gaming
I tried this once a few years ago.. had half a dozen Samsung Android phones running an SSH daemon with some functionality that could be remotely accessed. However, what I learned is that phones generally don't like to run 24/7 as servers. They start giving you trouble after a while, never figured out why.
But I suspect it's just the "always on" nature and the battery. The usage pattern is just entirely different than having a phone in your pocket and using when you need it.
You're welcome to try though, maybe phones got more reliable.
All my old phones used to become BOINC nodes doing WorldCommunityGrid or seti@home, at least until we got to the point where you couldn't run the phone without a battery anymore. Came home to one too many spicy pillow'd phones even keeping them in a cool area with a rigged up fan blowing on them.
I do Wigle wardriving with a dedicated cheap phone these days. (Moto G Stylus 2023)
In order to prevent issues this time around I've preemptively removed the back of the phone, and the camera modules so I can have a nice flat phone. Then I bought a heatsink nearly the same size as the phone itself. I've got thermal pads on the SoC which sits lower than the battery and the heatsink itself had thermal adhesive on it pre-applied which is sticking to the battery/phone frame holding it to the phone. No more phone overheating worries and if the battery goes pillowy it should just pop the heatsink up instead of warping the whole phone.
For some reason, I never buy phones that work with postmarketOS :( And I find phone naming confusing, it's difficult to find a used one locally to play with. Is it a Moto Play 2018 or a Play 2020? Trying to get that information from someone on Facebook marketplace is like pulling teeth.
That makes sense. Most phone users aren't technical. Like -- at all.
If you can think about how deep into technicalities the most average person you know gets, then you can also understand that ~half of everyone is even less technical than that.
There's nothing wrong with this. That's just the way that it is. (We can accept this or be frustrated. Acceptance is more useful.)
As a workaround, I find that searching by part number provides a good filter.
Maybe I want a very particular Moto G Power to use for whatever. I don't search for any permutation of "Motorola G Power" at all, because that description doesn't help me.
Instead, I just find the part number (maybe something like "XT2041-7") and search for that instead.
This excludes a lot of listings straight away, and that's fine: I don't want to stumble through listings from people who don't know what they have. I only want to buy what I want to buy, and what I want is an XT2041-7.
Inspired by this, went to look into how much performance I can squeeze and turns out Qualcomm software practices are so bad that I can’t do much but accept old software.
It sounds like Qualcomm has to do everything from scratch on their hidden Linux software for every new chip.
Readit News
That's the big hurdle though - mainstream kernel support.
For most devices, even if they can be rooted and jailbroken, you're stuck with the kernel they come with. Doesn't have a new feature you need? A horrible security flaw in the network stack? You're out of luck. Most "repurpose your old phone" approaches have this problem. You can make it a server but you wouldn't want to expose it to the public Internet.
But yes, this is definitely an issue. I've been playing with a 2013-era Samsung device that came with a 3.0 kernel. It can run pmos with said kernel but there are multiple root LPE vulns. I've been looking into getting it to run a mainline kernel just for fun, but it's not going to be easy.
I think it's basically for the same reason as why they dropped 32-bit x86 support about 8 years ago. Not enough users. (That resulted in the unofficial Arch Linux 32 to maintain support.)
I write this from arch on arm (orange pi) thingy, btw
Generally, as long as you keep the phone plugged in, the battery should be safe virtually indefinitely - the battery management on board will keep it in a state where its a constant charge which means the chemistry will be stable.
I think if you're considering re-harvesting old devices to use for hosting and get far enough down your list to get to phones then you've likely got enough constant maintenance costs in overseeing things that the additional worry of fire risk just isn't worth it.
What is your source on this?
I've replaced the battery in always-plugged-in iPhone 3 times over 10 years because it was expanding into a spicy pillow.
I too want a way to run phones directly off of USB power, without a battery present.
Would you say this is a chemistry/QA problem? Have there been advances in battery / controller technology that achieves the above?
(This is for a removable battery, but should be close for built-in ones too, I suppose)
[1] https://www.youtube.com/watch?v=7f8SliNGeDM&pp=ygUYZ3JlYXRzY...
[0] https://news.ycombinator.com/item?id=45021233
But I suspect it's just the "always on" nature and the battery. The usage pattern is just entirely different than having a phone in your pocket and using when you need it.
You're welcome to try though, maybe phones got more reliable.
Just thermal paste to the battery hah
In order to prevent issues this time around I've preemptively removed the back of the phone, and the camera modules so I can have a nice flat phone. Then I bought a heatsink nearly the same size as the phone itself. I've got thermal pads on the SoC which sits lower than the battery and the heatsink itself had thermal adhesive on it pre-applied which is sticking to the battery/phone frame holding it to the phone. No more phone overheating worries and if the battery goes pillowy it should just pop the heatsink up instead of warping the whole phone.
If you can think about how deep into technicalities the most average person you know gets, then you can also understand that ~half of everyone is even less technical than that.
There's nothing wrong with this. That's just the way that it is. (We can accept this or be frustrated. Acceptance is more useful.)
As a workaround, I find that searching by part number provides a good filter.
Maybe I want a very particular Moto G Power to use for whatever. I don't search for any permutation of "Motorola G Power" at all, because that description doesn't help me.
Instead, I just find the part number (maybe something like "XT2041-7") and search for that instead.
This excludes a lot of listings straight away, and that's fine: I don't want to stumble through listings from people who don't know what they have. I only want to buy what I want to buy, and what I want is an XT2041-7.
It sounds like Qualcomm has to do everything from scratch on their hidden Linux software for every new chip.