I was recently working on porting PMOS to Xiaomi Poco X3 Pro, though switched to NixOS due to much faster update cycle. I got much further than I thought I would, fixed everything including cameras except mic working in calls. Still bummered about that because I really wanted to try daily driving linux phone but alas not lucky and alternatives like Fairphone have way too shitty hardware and too high price.
This is awesome stuff. Fully mainstream Linux with a normal userland running on a modern handset with all features will unlock a tremendous amount of change once it's there.
Things like running a distributed database for a community on top of the devices they own, hopefully also running true mesh networking.
The part about getting authorized to test the emergency calling capabilities is cool! I should've expected that would be a thing but I didn't ever consider it.
Apologies for self-linking here, but it's not just cool things, in my opinion it could transform humanitarian aid work. I wrote a use-case/short-story about it here:
Like using your phone as a desktop computer, when it's connected to a screen and keyboard and runs a full desktop OS. Having a full access to all desktop apps, development environments. Also this, https://news.ycombinator.com/item?id=49201496.
Also, choosing the OS means freedom to choose the interface you like or develop your own (or to pay somebody to do that for you). Freedom to not be in walled gardens.
Calling our friends, maybe? No idea; I've tried desktop Linux on phones years prior and besides maybe being a potential portable cyber security rig it was only a cool oddity.
Does ubuntu touch or postmarket os work reliably on the fairphone 6 now, or is this still just components working? Is it possible to use it as a daily driver?
What is that bit about BedrijfsaansprakelijkheidsÂverzekering at the end? I've never heard of business liability insurance before. Does Germany not have an equivalent of an LLC?
Thatâs Dutch, and the same sentence mentions Netherlands. Equivalent of LLC exists everywhere, but business insurance still makes sense if you want to stay in business after paying damages.
My advice would be to not take the frustrating and useless road of a non-profit structure for some things that you do for pleasure.
The for-profit will be simpler and give you more liberties, and if, in the end, you have the good mindset that you want your creations to be shared freely, you can do it anyway. It is a state of mind.
I m quite sure that even in the Netherlands, you will have a lot of obligation and reporting that will be imposed on you because of that. To ensure that the non profit is not used as a for profit.
Even in the day to day, like for him to pay himself a salary or do expenses, things could be tricky as administration could easily consider as personal abuse of funds of a non profit.
PostmarketOS can't partner with any OEM/ODM and volunteers have to use hacks to unofficially implement basic Android 1.0 functionality years after a device was released? That's sad to see. This is just lame, drivers should be implemented by those who work on them..
We had Librem 5 which had real OS support that didn't have to rely on third-party volunteers, I wish they released something more modern..
Is there any precedent to believe that such a thing is in fact possible?
> and volunteers have to use hacks
I have terrible news for you about how the official firmware works.
> to unofficially implement basic Android 1.0 functionality years after a device was released?
Conversely, in other places it's far better than Android 1.0. But yes, if you can only start developing against the device after it's released, then your software will be released after the hardware ships plus development time.
> That's sad to see. This is just lame, drivers should be implemented by those who work on them..
My general understanding of the Android ecosystem is that they do write drivers, and you don't want those drivers. I like to imagine that somewhere that out there somebody must be making phones with high quality software, but I've never heard of them.
Murena could, Jolla could, GrapheneOS could do that too (possibly Nexphone also did but we'll find out; only GrapheneOS will give us flagship experience though). Of course these (except Jolla?) are still based on the Android model, but this itself doesn't preclude creating a non-user-hostile OS (of course Google would never certify it), proprietary black-box OEM code only assumes some Android elements, not the end user experience..
> I have terrible news for you about how the official firmware works.
I've seen some e-waste ODM (Tinno) code and their freestyle philosophy on modifying the kernel code.. however they at least have to fix this so that basic functionality works as there's QA and serious bugs will affect certification and they'd not get paid. But here if you report a bug to PostmarketOS (but well, even LineageOS), nothing will happen unless you fix it (and if you submit a patch, it will sit there forever, welcome to FOSS).
> My general understanding of the Android ecosystem is that they do write drivers, and you don't want those drivers
What's wrong with pixel drivers for instance? I'm of course not talking about those written by e-waste ODMs like Tinno who can't even release GPL kernel sources properly.
> What's wrong with pixel drivers for instance? I'm of course not talking about those written by e-waste ODMs like Tinno who can't even release GPL kernel sources properly
A couple of issues , back from my sailfish os porting days (haven't touched it in half a decade so things might have changed):
1. You would still need to run a lot of Android user space to make use of these drivers. We had a separate second init system that runs a stripped down version of Android init, starting enough services to make use of these drivers.
2. These binary drivers are tied to the kernel and Android versions that were released with the phone. So if you wanted some of the newer kernel features (back then it was like better container support), you're out of luck.
Usually, all this means you rely on a lot of hacks to fix some bugs... And porting takes a lot of time and effort.
I was recently working on porting PMOS to Xiaomi Poco X3 Pro, though switched to NixOS due to much faster update cycle. I got much further than I thought I would, fixed everything including cameras except mic working in calls. Still bummered about that because I really wanted to try daily driving linux phone but alas not lucky and alternatives like Fairphone have way too shitty hardware and too high price.
This is awesome stuff. Fully mainstream Linux with a normal userland running on a modern handset with all features will unlock a tremendous amount of change once it's there.
Things like running a distributed database for a community on top of the devices they own, hopefully also running true mesh networking.
It's already unlocked on other more open phones. Why do you think this time it will be different?
The part about getting authorized to test the emergency calling capabilities is cool! I should've expected that would be a thing but I didn't ever consider it.
Found the pixel modem engineer.
Ouch.
Awesome! There are so many cool things we could be doing with our phones if they weren't locked to a single OS
Genuinely curious, like what cool things?
Apologies for self-linking here, but it's not just cool things, in my opinion it could transform humanitarian aid work. I wrote a use-case/short-story about it here:
https://ianso.blogspot.com/2025/11/what-we-dont-have.html
Like using your phone as a desktop computer, when it's connected to a screen and keyboard and runs a full desktop OS. Having a full access to all desktop apps, development environments. Also this, https://news.ycombinator.com/item?id=49201496.
Also, choosing the OS means freedom to choose the interface you like or develop your own (or to pay somebody to do that for you). Freedom to not be in walled gardens.
like running K8s on your old phones?
Calling our friends, maybe? No idea; I've tried desktop Linux on phones years prior and besides maybe being a potential portable cyber security rig it was only a cool oddity.
Does ubuntu touch or postmarket os work reliably on the fairphone 6 now, or is this still just components working? Is it possible to use it as a daily driver?
What is that bit about BedrijfsaansprakelijkheidsÂverzekering at the end? I've never heard of business liability insurance before. Does Germany not have an equivalent of an LLC?
He didn't create an LLC he's a 'stichting' which is basically a non profit with a specific goal to achieve.
He doesn't want to create an LLC because he doesn't want anyone to question his motives.
Also this is the Netherlands not Germany :)
Thatâs Dutch, and the same sentence mentions Netherlands. Equivalent of LLC exists everywhere, but business insurance still makes sense if you want to stay in business after paying damages.
I see AF algo has some trouble with defocussing a perfectly sharp scene for no reason. :)
Does this HW have access to PDAF pixel data?
This aged well
Awesome
wow amazing
Great job.
My advice would be to not take the frustrating and useless road of a non-profit structure for some things that you do for pleasure.
The for-profit will be simpler and give you more liberties, and if, in the end, you have the good mindset that you want your creations to be shared freely, you can do it anyway. It is a state of mind.
Terrible advice
Why? Nonprofits involve a specific governance structure that may not be optimal, especially for small operations.
That is very region specific, in the Netherlands it's quite easy to operate a non-profit
I m quite sure that even in the Netherlands, you will have a lot of obligation and reporting that will be imposed on you because of that. To ensure that the non profit is not used as a for profit.
Even in the day to day, like for him to pay himself a salary or do expenses, things could be tricky as administration could easily consider as personal abuse of funds of a non profit.
PostmarketOS can't partner with any OEM/ODM and volunteers have to use hacks to unofficially implement basic Android 1.0 functionality years after a device was released? That's sad to see. This is just lame, drivers should be implemented by those who work on them..
We had Librem 5 which had real OS support that didn't have to rely on third-party volunteers, I wish they released something more modern..
> PostmarketOS can't partner with any OEM/ODM
Is there any precedent to believe that such a thing is in fact possible?
> and volunteers have to use hacks
I have terrible news for you about how the official firmware works.
> to unofficially implement basic Android 1.0 functionality years after a device was released?
Conversely, in other places it's far better than Android 1.0. But yes, if you can only start developing against the device after it's released, then your software will be released after the hardware ships plus development time.
> That's sad to see. This is just lame, drivers should be implemented by those who work on them..
My general understanding of the Android ecosystem is that they do write drivers, and you don't want those drivers. I like to imagine that somewhere that out there somebody must be making phones with high quality software, but I've never heard of them.
Murena could, Jolla could, GrapheneOS could do that too (possibly Nexphone also did but we'll find out; only GrapheneOS will give us flagship experience though). Of course these (except Jolla?) are still based on the Android model, but this itself doesn't preclude creating a non-user-hostile OS (of course Google would never certify it), proprietary black-box OEM code only assumes some Android elements, not the end user experience..
> I have terrible news for you about how the official firmware works.
I've seen some e-waste ODM (Tinno) code and their freestyle philosophy on modifying the kernel code.. however they at least have to fix this so that basic functionality works as there's QA and serious bugs will affect certification and they'd not get paid. But here if you report a bug to PostmarketOS (but well, even LineageOS), nothing will happen unless you fix it (and if you submit a patch, it will sit there forever, welcome to FOSS).
> My general understanding of the Android ecosystem is that they do write drivers, and you don't want those drivers
What's wrong with pixel drivers for instance? I'm of course not talking about those written by e-waste ODMs like Tinno who can't even release GPL kernel sources properly.
> What's wrong with pixel drivers for instance? I'm of course not talking about those written by e-waste ODMs like Tinno who can't even release GPL kernel sources properly
A couple of issues , back from my sailfish os porting days (haven't touched it in half a decade so things might have changed):
1. You would still need to run a lot of Android user space to make use of these drivers. We had a separate second init system that runs a stripped down version of Android init, starting enough services to make use of these drivers.
2. These binary drivers are tied to the kernel and Android versions that were released with the phone. So if you wanted some of the newer kernel features (back then it was like better container support), you're out of luck.
Usually, all this means you rely on a lot of hacks to fix some bugs... And porting takes a lot of time and effort.
> PostmarketOS can't partner with any OEM/ODM
Where did you get this idea from?