I received several dozen comments to my previous article, "Will Linux phones become a viable alternative to the Android/iOS duopoly?" (https://news.ycombinator.com/item?id=50051530), asking why bother developing mobile Linux when we already have a mobile Linux operating system available with the Android Open Source Project (AOSP). They argued that there is no reason to develop another mobile Linux OS, when we already can use the most popular one available, and it gives us access to millions of apps.
In this blog article, I go through the reasons why the community would struggle to fork AOSP or even add changes to each new AOSP release from Google that empowers users and maintains their freedom like mobile Linux. Here is the central argument:
> Given how Maemo Leste and UBports have struggled to maintain the large codebases developed by Nokia and Canonical, respectively, it is not realistic to expect that the community can fork an enormous project like AOSP with 309 million lines of code. Even if the AOSP fork can be rebased every year on each new AOSP release by Google, it will still be a huge amount of work to keep refactoring the code each year so the AOSP fork respects and empowers the user like mobile Linux does. If Google ever stops releasing AOSP, there is virtually no way that the community will be able to replicate all the code changes that Google is making with each new version of Android, so the community fork will diverge over time from Google’s version. As more and more Android apps become incompatible with the community’s AOSP fork, fewer people will be willing to use it.
robtherobber•56m ago
That was a good read, and my respect for Purism increased a bit. (I say "a bit" because I knew very little about it before.)
amosbatto•1h ago
In this blog article, I go through the reasons why the community would struggle to fork AOSP or even add changes to each new AOSP release from Google that empowers users and maintains their freedom like mobile Linux. Here is the central argument:
> Given how Maemo Leste and UBports have struggled to maintain the large codebases developed by Nokia and Canonical, respectively, it is not realistic to expect that the community can fork an enormous project like AOSP with 309 million lines of code. Even if the AOSP fork can be rebased every year on each new AOSP release by Google, it will still be a huge amount of work to keep refactoring the code each year so the AOSP fork respects and empowers the user like mobile Linux does. If Google ever stops releasing AOSP, there is virtually no way that the community will be able to replicate all the code changes that Google is making with each new version of Android, so the community fork will diverge over time from Google’s version. As more and more Android apps become incompatible with the community’s AOSP fork, fewer people will be willing to use it.