|Android Development - Constant Churn|
|Written by Mike James|
|Wednesday, 14 November 2018|
With the Android Dev Summit just over it is a good time to reconsider where we are in the light of what was, and was not, announced.
Android is not young any more and it is unreasonable to expect "one last thing" moments at the Dev Summit. It is more a matter of sniffing out the direction that things are moving in.
On the almost trivial front we have the revelation that dark mode saves battery life - well I never would have guessed that turning down the brightness used less electricity, but there you are! After encouraging us to go for bright white UIs, does this qualify as a revelation? Personally I like bright white and I think our users should get bigger batteries to accommodate them.
More seriously there is news of the API being extended to accommodate folding phones with multiple screens. We only just came to terms with notches and suddenly we have screens front and back to work with as well.
That's about all the "real" news from Dev Summit - the rest is mostly updates on things we already knew about. Kotlin 1.3 is on its way to Jetpack and the main IDE. Talking of Jetpack, this repackaging of the libraries, the Support Libraries, we all knew and had to use, is also getting some new features - the Navigation and Work Manager are moving to beta soon. However, there is still no clear distinction between what is core Android and what is part of a support library. Can anyone tell me the recommended way to use a fragment? Native or support library/JetPack?
Developing on Android feels increasingly like taking aim at a target that moves every time you think you have it in your sights.
This is all part of the constant revolution that the Android Dev platform seems to be. Just as you think you know what best practices are, they change and you have an app whose code is outmoded, if not actually deprecated, and so you have to work on it. Even things that are generally good, like Kotlin, are causing us to have to rework large parts of our code to take avanatage of them.
The Android team seems to be unaware of all of the tutorials and programming help they are invalidating by their constant changes.
Churn is not good.
Then we have Android Studio, another part of the ecosystem in constant revolution. All we want is something that is fast, robust and easy to use. What we get is something that adds features that are good, but of minority concern and increasing confusion about how anything works. So we now have Android Studio 3.3 in beta and the good news is that the team seem to be promising fewer crashes and hangs, better memory use, more bug fixes and so on. Even so, there is still no news on the missing widgets from Android Studio's toolbox and no way to add widgets of any sort. I hope 3.3 really is better, but I'm not holding my breath.
Here is a sign of things to come. Android Studio is an officially supported IDE on Chrome OS - starting early next year. There are all sorts of rumors that Android tablets are dead - certainly up-market Android tablets are not a thing of the future. The replacement is likely to be Chrome OS based Chromebooks and Chrometablets. These run Chrome apps, Android apps and Linux apps. Interestingly you will be able to run your Android app on the Chromebook you developed it on as, of course, Android apps run on Chrome OS. This is perhaps the first time we have been able to develop and run on the same platform.
So which are you going to work with? Do we really need features in Android to allow it to work with large-screen tablets any more? Perhaps Android could just become a phone-based app development system - or perhaps not.
There is still the threat of Oracle hanging over Android and, with the move to charge manufacturers in the EU for using the Google parts of Android, things are decidedly uncertain.
or email your comment to: firstname.lastname@example.org
|Last Updated ( Wednesday, 14 November 2018 )|