Addressing some valid criticism people had on Magisk: there never seems to be a universally "stable" release.
Unfortunately, there are so many moving parts in Magisk that even if unit tests exists, it still cannot cover the full functionality of it.
Unfortunately, there are so many moving parts in Magisk that even if unit tests exists, it still cannot cover the full functionality of it.
For projects like Linux, they branch the source code each release. The community test/report bugs for a specific branch, and bug fixes will be pushed to each release branch.
Linux maintainers will pick a release as "Long-Term Support", and that specific kernel version will...
Linux maintainers will pick a release as "Long-Term Support", and that specific kernel version will...
...receive bug fixes for many years down the road. Basically, users who need a rock solid kernel that wants to perform an upgrade to a specific LTS release will thoroughly test the LTS branch chosen using their own setup, report/fix bugs, then finally switch to the new version.
If Magisk needs to be rock solid, then it will have to follow the Linux development model, since full regression tests is basically impossible; just imagine how many devices/Android version/ROM combination is out there.
Unfortunately, I personally do not have the manpower and time to maintain multiple branches of Magisk and backport bug fixes to older versions. I manage Magisk in rolling releases, and I do in fact perform feature freeze for quite a while before marking a build as "stable".
However, once I decided to add some new feature (either out of necessity e.g. Android 11, or for fun), then things may go off the rails and have high chance of introducing regression bugs in the meantime.
But some part of Magisk does in fact is well regression tested!
But some part of Magisk does in fact is well regression tested!
Thanks to @osm0sis_xda's suite of boot image collections (gitlab.com/osm0sis/boot-i…), for each release, he will run a script to go through all boot images and compare boot image binaries to make sure no regression is introduced when magiskboot's source code is modified.
Anyways, my current plan is to freeze the native portion of Magisk and focus on stabilizing the redesigned Magisk Manager (that is a LOT of code changes!) before introducing any changes for Android 11, because trust me, it will be TONs of refactoring, well, thanks to Google (duh)
