What Is Linux Mobile

January 6, 2024

So, you just heard the term Linux Mobile for the first time, maybe you saw a random post on social media, saw it mentioned at a conference, or maybe something else entirely. Let me explain…

This post is based on a page I wrote for linmob.net. I have a lot to cover and this one might be a bit of a whirlwind.

The technical explanation

If you aren’t super into self-analysis and materialism then the rest of this article probably won’t interest you much, I wanted this post to be accessible, so let’s get this part done.

Linux Mobile is for everyone, it’s a loose collection of groups with a shared goal: running Linux on Phones. For many of us this has strong ideological ties. We care about the environment, sustainability, privacy, and user freedom. For others, it just looks cool and is a fun way to chase clout.

postmarketOS is a big tent kinda project, we have some hard limits on e.g. proprietary software. We allow firmware of course (everything is a brick without firmware), but we focus on engaging with upstream projects - the Linux kernel, the GNOME and KDE desktop communities, as well as being the home for others like SXMO (it’s Sway but for your phone, if you’re into that). We are the home for most efforts to run modern Linux kernels on smartphones.

That is to say, almost all of the well-supported postmarketOS ports involve running a modern kernel on a device that traditionally would use a heavily modified vendor kernel. In the case of the OnePlus 6 for example, we have ~100 patches which aren’t yet in upstream Linux (supporting call audio for example), but otherwise we ship linux mainline.

We also collaborate closely with the Mobian project; where we run a customised Alpine Linux on phones, Mobian run a customised Debian. We have a shared interest in doing “real Linux”, in carving out a space for mobile tech in the otherwise desktop-centric world of FOSS. There are other distros doing similar things, but none that we explicitly collaborate with to the same extent.

On the other side of the coin we have Halium based projects, Ubuntu Touch and Droidian are the main contenders here. The notable difference is that these projects attempt to build on top of the existing Board Support Package (BSP) originally provided by the vendor (that is the kernel and a bunch of proprietary userspace libraries and daemons). This is powered by the aptly named libhybris project, allowing for the proprietary binary-only Android bionic-libc Hardware Abstraction Layer components (HALs) to be used on a glibc or musl Linux system.

This has a lot of benefits, all the fancy platform support and integration can be used, the camera stack, the step counting, audio routing… The power management features can actually work. It all comes for “free” by basing your entire OS on top of the Android interfaces and Qualcomm/Samsung/Mediatek platform quirks. This gets modern phones going faster, with more things working, recent Droidian demos of their camera stack on the OnePlus 6 for example are really awesome.

But (because their has to be a but) there are a few noteworthy addendums here. This model is hard to scale, involves often heavy modifications to parts of the normal Linux userspace stack, and (for better or worse) is generally frowned upon by the upstream community. This approach also relies on running outdated software: the last security update for the OnePlus 6 came out in February 2022, and updated to the November 2021 Android security patch level. The Linux 4.9 kernel used by this device stopped receiving LTS stable patches in early 2023. And regardless, there’s over 5 million lines of code added by Qualcomm to enable support for the platform - plenty of attack surface here which will never be fixed.

This is inevitable with old devices, in fact there was a fairly recent security exploit affecting the modem firmware of Qualcomm devices like the OnePlus 6. Vendors who care provided fixes, but the OnePlus 6 is End of Life, and thanks to Qualcomms security model it is IMPOSSIBLE for us to patch the firmware ourselves without finding an exploit in the chain of trust, or OnePlus releasing their signing keys.

Doomed if you do, doomed if you don’t.

I don’t think these are issues we should be heavily concerned with, unless you have reason to believe you’re going to be directly targeted, in which case you absolutely should not be using Linux Mobile as your daily drive. Full disk encryption and common sense are likely more than adequate protection.

The economy and the environment

It is common knowledge by this point that the INSANE growth in the tech sector has had devestating outcomes for developing countries. Finding sustainable tech is expensive at best, impossible at worst. As individuals the best we can possibly hope to do is to reduce our consumption. The smartphone market is, in 2024, pretty much entirely saturated. From a capitalist perspective there is pretty much no more room for growth. Almost everyone owns at least one phone, and it’s not really feasible in the current landscape to encourage people to replace their phones more often than they already do (at least no explicitly).

This is, convenient timing with the right to repair movement and recent EU mandates on software support, but nevertheless, it’s good this things are starting to level out, as continued growth would come with even more continued devestation.

Phones are a status symbol, the iPhone is PROLIFIC in the US, and it’s no surprise that the quintisential smartphone is a slab of glass with a flashy camera bump. Some companies try to differentiate themselves, like the Nothing phone with it’s LED stripes and otherwise extremely meagre offerings.

The smartphone market is going to be undergoing some huge changes, with AI being the investor buzzword, flexible displays, and in-screen cameras. But at the end of the day, smartphones are basically solved. For sure we could do with nicer screens, better haptics, connectivity leaves a lot to be desired (bluetooth is a huge mess, like wtf).

But really, a smartphone from 2015 is still pretty damn good in many cases. And the idea that these devices are destined to run outdated software forever is absolutely onconsciouble. We do not live in a world where we can consume infinitely. Entropy will come for us all eventually, and we should make the most of what we have.

Don’t forget about society

We (the Linux Mobile zealots) are sick of our needs not being served, of our phones being hostile towards us, and of the eternal profit motive driving planned obselence and insecurity.

We believe that your phone should work for you; it shouldn’t sap your attention with dark patterns, nor should it stop receiving security updates or replacement parts while still more than capable of serving your every day needs.

But let’s be clear, the ideological drive should not be a necessary prerequisite to get involved. I’m not purist, I spend way more time than I’d like to admit sucked in by social media, YouTube, and other evils that no doubt many of my fellow developers would scoff at.

Gatekeeping and purity testing are something many online progressives out there are all too familiar with, it’s something we need to be careful of. We can push out those who don’t share enough of our views, and risk killing the movement. Or we can let everyone in and risk the resulting infighting and power struggles as we compete for relevance within the community.

I guess this is nothing new or unique to our movement. But regardless I feel like there are those who are blissfully unaware of their role within this dynamic. For those in positions of power (such as project leaders and open source maintainers) it’s especially important to consider the influence you have on the rest of the community, and take care where possible.

The software

These stream of consciousness style blogposts are great for me to actually get something out there (get that engagement girl). But there are some things I want to get to more specifically here.

Whatever you want to do with your phone, it should still be your phone. You shouldn’t be arbitrarily limited in what software you can run on it, you should be able to customise the interface to suit your needs. Obviously not everyone has the time to go and build and this custom software, but some folks do, and you should be able to make use of it with ease.

UX is really frickin hard. Android frankly SUCKS to use, in ways we don’t even understand. You know that feature where you tap and hold on the spacebar and it turns into a touchpad - letting you move the cursor around? Well on Android it only goes left and right, it simulates left and right key presses. But on iOS… Holy shit, it just actually gives you a 2D touchpad, you can move the cursor wherever you want. It’s way better because mapping the location of your thumb directly to the location of the cursor within a particular area is much MUCH more intuitive than trying to press the left and right arrow keys in sync with your movements…

Interacting with text is AWFUL in general, selecting and copy/pasting sucks. It doesn’t need too - surely we can do better. But hey, good luck actually doing this in the real world and offering something people can just use.

The software duopoly on mobile has absolutely killed creativity, and frankly we deserve better.

Linux Mobile is for everyone

We’re all just trying to get along in this crazy world, and for those of us trying to build a better smartphone ecosystem, we should take the time to do things properly.

Some people are probably not gonna be happen to hear this, but I really do think it’s imperative that we target the Rational Consumer here. It’s clear that most people care about progressive causes, and the motivations behind Linux Mobile are clear. But without a competitive UX we’ll always fall short.

I’m certainly not hopeless here, there’s a lot of really awesome work being done, and I’m extremely excited to see things continue to evolve.

And hey, dear reader, the thing we need the most right now is confidence. The more people donate to projects like postmarketOS, the more they talk about the work we’re doing, and the more folks are able to get paid to do this work, the more we can continue to grow. We’re still oh so small, and not yet at the point of self sustainability. But no doubt we’ll get there.