RE: https://androiddev.social/@MishaalRahman/115849772521760098

It’s worth repeating. Android is not a viable base for an independent or even just collaborative operating system. Android is Google and only Google’s project.

If you want to see an actually transparent, international, and collaborative system on phones, support @postmarketOS

#android #opensource

@thibaultamartin @postmarketOS It's the reason I bought a @WeAreFairphone for me and my wife. If Android becomes fully enshittified, there is a high chance postmarketOS will be compatible enough at that time to be a real alternative. I support pmOS with 5€ a month as well. It's nice. It feels like I directly funded the recent call audio work.

@opensourceit @thibaultamartin

forking AOSP would be a ridiculously huge effort, since Google has evolved to scale through increasing abstractions you need to have teams of people working maintenance like security fixes across bespoke compilers, apis and code, responsible for having compatible expensive test farms, and establishing governance for changing UI features, updating security APIs, and interacting with vendors.

To support that you would need funding that really can only be obtained by having lots of vendors using your OS, but vendors aren't gonna do that without the governence in place.

basically this is limitied to corporations and states.

@opensourceit @thibaultamartin more or less yeah that's my take.

I believe this is one of the reasons postmarketOS is better. We intentionally work towards decentralisation in a many2many style, so that the number of centralised components that could bottleneck the project are kinimal.

we aim to make it easier to set up your own paclage repositories, CI, and automated testing, and do t force people to pick a side by ensuring that our automated hardware testing allows one device to be tested from multiple gitlab instances.

the binary repo and tooling like pmbootstrap let you easily replace on the fly any part of the OS (same way you would replace a package on your desktop linux pc) since it's literally Linux. The immutable version will also make this possible

@opensourceit @thibaultamartin @postmarketOS the scale of a project like this means, it can't meaningfully be forked. You'd need a team of experienced developers just to keep up, let alone make some headway.

So
"fork": yes
Fork and be viable long-term: extremely hard
😩

@opensourceit @thibaultamartin @postmarketOS I can't speak about PostmarketOS because I have no idea how this project is structured or how it keeps running. From what I could gather quickly, it seems to be based on pretty standard HUGE projects (linux kernel, GNOME, Waydroid, Alpine packaging). So the "damage" here would be the device-specific customisations they seem to be in charge of. 1/

@opensourceit @thibaultamartin @postmarketOS 2/ Even with this *much* smaller problem to solve, this project is not ready:

> The goal is to make postmarketOS usable for non-technical people too, but we are not there yet. Usability and most importantly stability issues need to be worked out first.

from: https://postmarketos.org/state/

So, I would say: no, postmarketOS *currently* is not an alternative. And I personally doubt it will be any time soon.

@opensourceit @thibaultamartin @postmarketOS 3/ I would love to see a viable linux phone. I have looked at this space on and off all the way back to the neo Freerunner (2008). I met someone who had one. And while it's of course a great effort and achievement to get a 90% working phone, a 90% working phone is still *not working*.

@opensourceit @thibaultamartin @postmarketOS 4/4 Without backing of a steering entity and MASSIVE effort (human labour, infrastructure, money) I don't see this happening. To reach something that can even begin to compete with current Android or iOS, you'd have to be a pretty large company. Not even Samsung does its own operating system.

It would basically take a nation state or the EU to start from scratch and be competitive in 5 years.

@claudius @opensourceit @thibaultamartin @postmarketOS you're largely spot on that the "damage" is largely device specific support at least from a hardware enablement standpoint.

We have a very upstream-first community (it's something we take extremely seriously for the obvious reasons in this thread but also because it is just the right thing to do) and as a result upstream projects also increasingly consider mobile usecases, for example GNOME's libadwaita graphics toolkit treats mobile as a first-class citizen, and all GNOME core apps works great on a phone through convergence.

Despite this, we're still operating at a very small scale when it comes to working on things for our own sake and we are still a while away from being able to fully fund even core maintenance of the project.

That being said, 2025 has imo been a year of preparation for us, we have improved our governance with the aim of making it easier to grow the team and scale more horizontally, we also made a lot of progress towards automated testing which is gonna be huge!

With immutable pmOS also on the horizon I'm really excited for 2026 when it comes to reliability!

@claudius @opensourceit @thibaultamartin @postmarketOS

In the long term I think we'll grow to fill a niche and ideally be self-sustaining along with our community of contributors. We are always trying to bring users in and turn them into developers, return to when you could just tinker with your devices! Our next biggest hurdle is probably going to be figuring out how to collaborate with vendors and shipping pmOS out of the box.

There is also the growing inevitability of climate and economic collapse and an increasing necessity for longer lasting hardware, if there comes a time when Google/Android is no longer a viable option for the EU, I want postmarketOS to be there when that time comes

@cas @opensourceit @thibaultamartin @postmarketOS thank you for the added details. I hope you were not offended by my post. I really like what you are doing. But it is a HUMONGOUS task that you picked. I would love nothing more than this to get funding to a degree where everyone could work on this as much as they wanted.

Basically, I wish the EU had given you the millions of euros it chose to burn up in AI subsidies.

@cas @opensourceit @thibaultamartin @postmarketOS that said, longer lasting hardware will absolutely make your task a little easier. One of the huge problems in this space was "dang they are supporting a two year old phone" used to be a deal breaker. This is no longer the case, a phone today can easily productively work for many years. This really honestly was not the case ten years ago, when even flagship phones were struggling to keep up at three or four years old.

@claudius @opensourceit @thibaultamartin @postmarketOS thanks for the response! I'm definitely with you when it comes to funding heh, and no your description was pretty realistic imo, no offence taken.

the upside is that even if we don't succeed in taking over the world, the journey is absolutely one for a lifetime, it's amazing to be part of this project and work on something that brings joy to peoples lives today and now

@claudius

In fully agreement with @cas, I think your assessment is fair, even if it can sometimes hurt seeing it written like that. Personally, one of the things that motivates me the most is that we all know that the task is humongous... But honestly, so was the effort needed to get from https://news.ycombinator.com/item?id=14427232 to where we are now.

We (postmarketOS but also the whole mainlineLinux Mobile community: Mobian, Phosh, Plasma Mobile, etc.) have already succeeded in something that is a humongous task, that most considered completely impossible, and that very few even dared to dream. Yet, here we are, and we're not stopping any day soon. Personally, this looking back and seeing what once was thought as impossible as accomplished, is one of the things that motivates me the most to keep going.

@opensourceit @thibaultamartin @postmarketOS

@thibaultamartin @postmarketOS It is worth repeating. Android was never good and it is getting worse. Move away...

@pabloyoyoista @claudius @cas @opensourceit @thibaultamartin @postmarketOS Hopefully you don't mind me chipping in here.

I used to maintain my own fork back in high-school (when LineageOS was still called CyanogenMod... ouch, my back). Back in those days smartphones were a wild west. People carrying computers full of personal data in their pockets without any encryption, banking apps and other essential applications didn't care whether you modified your device or not. It was trivial to root your device and relatively easy to maintain your fork.

That started to change around 2016 - 2018. Around Android 8 codebase started increasing, maintaining your forks/rebasing became a pain in the butt. That's why many "Custom ROMs" started disappearing, it simply consumed too much time and required a lot of resources to build the bloody thing.

Around that time prices of smartphones started to drop, and people in developing countries could afford a smartphone (where they couldn't afford a computer). This was the time when we observed a general population shifting their daily computing needs from desktop onto mobile.

With people shifting their habits, companies decided to track users and sell their data in exchange for "free" utilities and forcing everyone to use their stupid apps (where most people couldn't block ads).
Mobile market became a much more lucrative target for spyware and malware, because now everyone had a computer connected to the internet in their pocket at all times. That's why banks started to detect/block people using modified Android builds for "accountability". Soon after, it essentially became a war.

For that reason, I bought an iPhone with my first real paycheck in 2016. Not having to deal with Android's atrocious security, potential malware in custom builds (it happened many times).
It just worked... until 2022, when Apple started adding ads to App Store and so on. It made me angry that I spent a small fortune on a device and manufacturer dared to serve me ads, so I bought a Google Pixel 6 and installed GrapheneOS on it.

To put it bluntly: I despise that phone. I still have it kicking around, but I'm working on gs101-mainline. Heavy, bulky, no expandable storage, no headphone jack, no physical keyboard.
Using GrapheneOS meant I couldn't use contactless payments, notifications didn't work, and in ~2025 my bank added Play Integrity API, which prevented me from using my banking applications whatsoever.

Some people will say it's about keeping users safe, I call it "control". Device manufacturers are trying to take away device ownership from you by removing the ability to "unlock bootloader" (in reality just flipping a flag whether ABL should check signatures) because they want you to stay within their ecosystems.
Google Services that track your every move, $VENDOR shipping uninstallable applications and harvesting your data/serving you ads. I've looked at my brother's Samsung S25 (or whatever it's called) running stock firmware and I was absolutely horrified.
If my brother and I would be talking about sensitive topics, I would ask him and his girlfriend to put their phones into a Faraday cage. That's how bad modern smartphones are.

AOSP is done for, Google doesn't care about opensource anymore (unless it benefits them).
Releases twice a year, delayed security updates. Not publishing trees for Pixels anymore (their own freakin' devices!).
They laid off the entire ChromeOS team, cancelled AMD Chromebooks, and now there are rumors of ditching coreboot in the future (where they could easily lock-down the boot chain and fuse devices with their own signing keys).

One vendor I've worked with (and whom I'll likely work again this year) almost went bankrupt because of Google and amount of control they have on device manufacturers.
Out-of-the-blue Google cancelled Android XR partnership (even though said vendor helped Google get that project off the ground) and left them with AOSP scraps provided by silicon vendor (Qualcomm).
Said company has ~15 employees whom can't possibly write replacements for what Android XR offers nor maintain it down the road. Hell, you need a system with ~64GB of RAM and ~480GB of disk space to build those sources. It's absolutely ridiculous.

While I haven't done much in postmarketOS, I've been tracking the project since 2018 and messing with/maintaining few devices, as well as suggesting ideas myself.
It's the only way we can have computers in our pockets that don't track our every move and that can easily be pwned by 3-letter agencies if your suspected of any (even insignificant) crime.

It's been a wild ride and I'm extremely proud of everyone who's involved in this project. Tremendous amount of work done and plenty still to do. Slowly but surely it will become a viable replacement for Android.

@opensourceit @thibaultamartin @claudius @pabloyoyoista @postmarketOS @elly dude read what she wrote and stop reading into what she didn't

do you think any of the android derivatives you mentioned have the ability to take on a full fork of AOSP?