I don't think this is a fight that's sustainable - a giant graveyard of Android privacy forks shows that. In the end, the major contributor to this OS is and will remain Google.
In that respect, Purism's PureOS seems like a more sustainable effort.
I don't think this is a fight that's sustainable - a giant graveyard of Android privacy forks shows that. In the end, the major contributor to this OS is and will remain Google.
In that respect, Purism's PureOS seems like a more sustainable effort.
I have a very high respect for all the work that's been done on the various GNU/Linux mobile distributions, but it's still so far away from Android wrt usability...
I agree there is something weird about fighting Google with Android, but still it makes perfect sense.
From a business or engineering PoV, you want best features for the fastest MVP. Android gives you just that. Actually it goes way beyond, /e/ didn't do much of engineering, they mostly re-used what's done by the Android FLOSS community. (Honestly I don't really like /e/ because they basically don't give anything back to the community but well).
If /e/ is ever successful, they can hire many engineers to actually fork Android to have their own. That's what Android did for Linux in the first place anyway! Most OEMs have a lot of changes to their Android, even though they keep using Google apps, and they are just fine. I'm pretty sure that Samsung have an order of magnitude more changes in Android than LineageOS (which /e/ is a pretty direct fork of), and they maintain it just fine.
Micay, of GrapheneOS and original creator of CopperheadOS, is not some random. They have pushed AOSP security ahead and have made many upstream contributions over the years.
Micay originated the Android Hardening Project, I believe it was called, and is the brains behind what was CopperheadOS, and now the superior GrapheneOS.
He was also a pretty instrumental Rust developer.
This is where I ended up after countless hours of research looking for security/privacy.
Has anyone found sources for this?
That was indeed a losing battle because Google has far more resources to add features ans beat CyanogenMod.
On the other hand, a de googlized and privacy focused platform provides something that Google can never provide.
There’s also a change in the environment surrounding all Of this. Thanks to the weaponization of Android by the American government, every country and company in the world is looking for a De GoogliZed alternative.
CopperheadOS did not die. https://copperhead.co/android
/e/ mostly uses LineageOS code without providing attribution. They have a fair fight ahead of them and the Android FLOSS community benefits from this internal competition.
Android is not designed for the community to be hacking on it. While these projects occasionally appear they don't last because of the incredible amount of work required to maintain them. You're much better off finding a normal Linux distro/phone pair that can make phone calls etc.
What makes you interested in it?
It has become a lot easier thanks to Project Treble.
Doesn't the DRM stop working once the bootloader is unlocked? That'd be a big concern since most streaming apps would refuse to install or work in that case.
What other such changes is Google pushing into Android that discourages unlocking the bootloader / rooting?
So first point about the DRM itself: - Google here is "benevolent". Devices are allowed to keep working Widevine L1 working after bootloader unlock (without needing relock), and it does work just fine on Pixels. - Some devices clears their key after bootloader unlock, so unlocking means you'll never get Widevine L1 - Some devices have the key tied to being locked, to relocking will work - Some devices doesn't have restrictions, but have poorly written drivers that need to be hacked around to allow to work on unlocked bootloader
And then, there is the extra issue of what streaming apps decide to give you. For instance Netflix will not use Widevine L1 even if it works unless your device is whitelisted. So if you flash a "simple" alternative ROM on a Pixel 5, you still would get only 480p. The ROM needs to lie, and tell it is the original Pixel's ROM. (Also Netflix won't be listed from PlayStore unless the ROM lies and says the app that yes the bootloader is locked)
In my opinion, the biggest vendor lock-in about bootloader unlocking is that you'll loose all your data when unlocking, with no way to properly backup your data on stock ROM.
As far as I could see in your GitHub project wiki, most devices tend to have a few glitches (probably due to issues in the hardware abstraction layer from the manufacturers side?): https://github.com/phhusson/treble_experimentations/wiki
I could see some that look mostly flawless, like Samsung Galaxy Tab 8.0 2019. It'd be good to have a curated list of preferred devices. Buying Android devices with ROMs in mind is quite frequent, and your generic ROMs are a huge leap ahead!
PureOS has to try and reimplement an entire mobile app ecosystem just to get to parity with existing competitors.
I don't know if all of these would work in e.foundation, but presumably at least some would run.
The telemedecine app is a big one I think about. I don't know what I would have done if I hadn't had a modern iOS or Android device when the pandemic hit in March, and my doctor appointments suddenly moved to this remote system.
Custom AOSP apps can be supported via Anbox on upstream linux.
SailfishOS has this, and doesn't do tracking.
PureOS and Purism One implement some FOSS techniques and rebrand them. I believe Nextcloud does this as well. Not sure why projects don't clearly mention what they're based upon, especially when its a lot like original.
What I seek in a phone is control over my privacy, not necessarily a vegan FOSS system.
What that means is things like:
- Fine-grained control of permissions to apps (e.g. access to rear camera only, access to only city-level accuracy of locations, access to read from only directories I specify)
- For apps that insist on having permissions to things like location and wi-fi scans to use them, the ability to make the app think it got said permissions, but receives fake data. And no, Android's mock location feature doesn't work, because apps can check if the feature is enabled or not
- The ability to fake IMEI, phone number, contact list, installed apps, and other identifying data
- Ability to generate fake IMU, proximity, temperature, barometer, and all other sensor data that could conceivably be used for fingerprinting
I think the effort is very good. The result is excellent and it's really addressing a much bigger audience that PureOS.
I think they legally can't call it "Android" in this case so it's not really their fault that they aren't mentioning this.
[0] https://docs.google.com/spreadsheets/u/0/d/1bx6RvTCEGn5zA06l...
The officially stated reason (I'm paraphrasing) was because it allows for fudging the signatures of apps and services on the OS, and this breaks the security model.