Throwing in the Towel on Mobile Linux
drewdevault.com
drewdevault.com
Honestly, it's the only Alternative Non-android based phone OS that works. I have been using it for the last 2 years and it works great. The android support works well enough that YouTube, my bank, waze work nicely.
Sailfish has a ton of great apps, the community is very active. Can't find a native app, android support is here.
And yes, it's not "truly" open source, but it's way more than android. Yes, android has an Opensource license, but good luck doing something that google don't want.
Unless I'm missing something, LineageOS and other AOSP-based projects are more open source than Sailfish, can obviously run Android apps better than Sailfish, and can probably do better at pretending to be Linux too (through Termux and its package manager).
It uses all the usual Linux build blocks (rpm,wayland,qt,systemd,pulseaudio ...)
Sailfish is a Linux distribution with a proprietary DE (built in qml, so modification are easy)
How is it "way more" than Android? Doesn't Google's team contribute the most to Android development? Sailfish contains closed source components, so I don't understand how it's inaccurate to say Android is entirely open source, and Sailfish OS is almost entirely open source.
Since when does open source imply that you can submit anything you want and expect it to get merged?
A good 4 hours on YouTube(android app) is easily doable.
I love to tinker with things, but my phone is not one of those things because it needs to hold an extremely high bar of reliability. I can mess around with my PiHole or my PLEX box and if that breaks then it's ok I can fix it. But when it comes to my phone I often find myself in situations where I need to pull up a boarding ticket, or order an Uber, or call a friend when we are trying to find each other. In each of these case, something not working is just not an option.
Specifically my phone has my password manager 1password:
- If I wanted to restore my computer from a disk backup I'd need the backup encryption key.
- If I wanted to recover my computer from a fresh install I'd need my GitHub password, Google Drive password (I sync my Documents folder to drive), email password, etc.
I have paper backups of my 1Password account recovery information but I wouldn't necessarily have those accessible while traveling. And I consider my phone to be the primary and most reliable of those.
(I'd really like a backup of my 1password but they don't make that trivial so it's on my to do list. At least each of their apps on my devices acts as an offline backup. I think that I can copy the SQLite database and the exact version of the app on my computer to a usb drive but that's obviously something I'd want to test)
There are also lots of little annoyances you don't necessarily think of. If you break your computer's network stack it's really annoying to debug without a way to Google, for example. I could borrow my partner's laptop for a quick check, but I'd rather not for an extended session.
My tolerance for tinkering with anything that matters is already low, but with phones it's pretty much zero. I don't want an OS that has less than a billion other people using it.
I feel bad about e-waste, but I don't care nearly as much as I would for other devices. I'll happily upgrade my phone every 4 years, because it means there's a dozen other things I don't need to own in the first place.
Besides, I like android better than Linux by a lot. I use Linux because of the ecosystem of stuff that runs on it, but I'd be jumping for joy if Android for Servers became a thing and they ported the desktop apps I use to it.
A lot of OSS is the best of breed, but the structure of mobile OS development never seemed to attract that. LKML vs xda maybe? The feeling that xda trended very young, the total absence of build systems for most projects...
PMOS is definitely the highlight, and I remember old cyanogen on my nexus one having nightlies, but they were the exception, and I've lost track of how many such projects have imploded. Others just attach zips to xda posts.
I think it's an ecosystem problem, not a hardware one. There probably needs more corporate buy-in, RHEL/Suse style.
I was in the same boat as you were and got started couple years ago with the Pinephone. Tried multiple OS's, loved the experience / freedom but at the end of the day, wasnt satisfied with any of the OS's since there was something usually broken.
Found /e/OS one night and everything that i typically do to harden the phone (degoogle the phone, alternative DNS, alternative app store, etc) is built into the phone. Supports a wide amount of devices with A11 / A12 builds coming out soon. The built-in tracker blocker is awesome to have. Uses MicroG instead of Google Play Services. Most apps works, but some are broken because of this
Check it out - https://e.foundation/
The main software issue has been banking apps overly focused on security. Solution: Just use the web site. The only exclusive the installed app provides is paper check deposit --- and who uses paper checks anymore?
Lots of people don't know that an icon can easily be created to launch a web site just like an app. Most users won't notice the difference.
This is a disappointing assessment but not totally surprising given how the odds are stacked against mobile linux happening at scale.
Having a polished, user-friendly, fully functional mobile linux is maybe the the single most disruptive element one could insert into the modern tech landscape. But it would take some major funding and support to get there.
It a strange time for open source personal computing. A compelling case can be made that with all the advances in ML etc even the Linux desktop is more relevant than ever as the primary platform for "local AI" that empowers users. But the user base for desktops is shrinking and it may take a long time before that is reversed.
In contrast there is no excuse for not having a genuine third option in the mobile space when the user base is literaly billions.
Lets hope that "for the time being" is a really short period.
A third Mobile OS would create more fragmentation, and could drive more things to either be web-first, or make "Everything is an app" less viable.
Plus, Android is batteries-included. Everyone who works on Linux likes modularity, they will fight hard against adding the thing that makes Android great, which is that there's API levels and every devices with a given level is more or less the same.
I'd much rather see a FOSS Android fork that can run AppImages or something, or just one of the tech companies doing a grant program for writing FOSS apps on Android.
But we have it, it’s the biggest mobile OS even.
Also, it is absolutely unreasonable to expect any one company to produce apps for 3 platforms. So anything that is not out of the box compatible with one of the big ones is dead in the water.
The incomplete state of the software isn’t actually a problem for me. If something is missing, I am happy to try to add it myself, and I loved hacking new features on my Nokia N900 a decade ago. But that device at least felt usable hardware-wise.
Otherwise, there's the Librem 5 - it's enough faster to make a difference compared to the PinePhone. As many people receive theirs now (who moved on in the time since they ordered), getting one that's barely used is a possibility.
I would give everything for even an older iphone that could be tinkered with, though.
I know it's a copout, but all I care about is functioning data, and that's not a huge bar to clear.