| ▲ | d3Xt3r 4 hours ago |
| There's already an alternative, it's called Linux. |
|
| ▲ | DaSHacka 4 hours ago | parent | next [-] |
| Though, as the narcs are quick to point out, Android 'is' Linux. So really, the alternative is proper GNU/Linux (or as I've recently taken to calling it...) |
| |
| ▲ | pjmlp 3 hours ago | parent | next [-] | | It really isn't, it more like Inferno would have replaced its kernel with the Linux kernel. It is as close to GNU/Linux as Inferno was to UNIX System V. Which is why Termux has to work around what is actually available as public APIs on the NDK. | | |
| ▲ | jeroenhd 3 hours ago | parent [-] | | It's still Linux, though, the same way Ubuntu is still Linux despite carrying custom patches and security frameworks. Termux has to work around security restrictions in a way not dissimilar to how you would need to work around an extremely securely configured SELinux setup on RHEL. | | |
| ▲ | pjmlp 2 hours ago | parent [-] | | Ubuntu is GNU/Linux, you don't run Afterstep on Android via NDK, unless you rewrite X Windows or Wayland to run on top of OpenGL ES or Vulkan surface that is then hosted inside a AndroidView or JetpackCompose display surface. Which is what Termux has to do, just like they have to call into specific Android services via JNI, as the NDK only exposes official NDK APIs, alongside the C and C++ standard libraries. |
|
| |
| ▲ | fsflover 3 hours ago | parent | prev [-] | | Android uses Linux but it has no consequences for users or developers. So this is practically irrelevant. See also: https://news.ycombinator.com/item?id=50057296 https://news.ycombinator.com/item?id=50060611 https://news.ycombinator.com/item?id=50060215 |
|
|
| ▲ | fsflover 4 hours ago | parent | prev | next [-] |
| Related recent discussion: https://news.ycombinator.com/item?id=50051530 |
|
| ▲ | microtonal 4 hours ago | parent | prev [-] |
| Yes, let's throw away a fully mature smartphone OS (AOSP) and replace it by something that does not even have proper app sandboxing. /s The alternative are other AOSP-based OSes like GrapheneOS and Lineage. Once they are used widely enough and funded enough, they could fork AOSP when needed. In some sense they have already soft-forked in that they use major and QPR2 releases, but put a lot of patches on top to fix issues, improve security, etc. GrapheneOS has an estimated 500,000 active installs now, apparently LineageOS has 1.4M active installs. Probably around two orders of magnitude more than traditional Linux desktop-based phones, which have been trying since, checking notes, Maemo in 2005? |
| |
| ▲ | flaburgan 4 hours ago | parent [-] | | Btw /e/ OS people have already shared that they were getting ready for a hard fork of AOSP looking at how Google is acting.
But to be fair the challenge is more around the hardware support than the OS itself, so in the end it will be the chip makers which will decide which OS can run or not. If they go the AMD way then great, if they go the old nvidia way where even basic reclocking is impossible then we are screwed. | | |
| ▲ | microtonal 3 hours ago | parent [-] | | I would be surprised if they have the expertise and manpower to do that. They are mostly relying on LineageOS now and putting their own cloud, microG, and App Store app and branding on top of Lineage. Even now they are not able to roll out Android releases and WebView updates timely. Them maintaining an AOSP fork would probably end up in an outdated, broken mess. My money is on GrapheneOS and maybe LineageOS to do it when needed. |
|
|
