Open ghost opened 9 months ago
This also means checking all library compatibilities, modifying all deprecated methods and running tests on numerous devices to check for regression. I upgrade API versions when I can, but it's a lot of work.
Hi! Thank you for your answer. Not having enough time resources is always a good answer when it comes to FOSS project.
On a side note: i noticed the your app min target is android 4.0.1 (SDK 15). My question is: it is worth it ? I’m asking seriously!
Currently, around three people out of 100 are using a phone running Nougat 7.0 / 7.1 (sdk 24 and 25). Lollipop is used by less than two people out of 100. (Source: https://www.gizchina.com/2023/10/30/android-version-distribution/)
I think that try to maintain your code for such old and not used Android versions is not worth the cost in term of workflow and time spent to check for regression etc.
I think the approach of supporting only the last four major releases of Android (in this case Android 11, 12, 13, 14) is too aggressive. Maybe target min SDK to 26 (Android 8) could be a good idea ? Almost nine people out of 100 are using a device with Android Oreo on it. In this way you will have to deal with 8 versions of Android, instead of 19.
Let me know what you think!
Higher SDK means higher security. There are a lot of new restrictions for apps targeting Android 14.