Skip to content

Turbulence

Today someone answered my Android question on StackOverflow. I checked when I asked it and it turns out it was December 2010, like over 12 years ago. Just in time, thanks.

So, during this time before my eyes, the technology stack has changed completely about 3-4 times. Completely. At the same time, it seems to me that Android development is now in the most unstable and complex state it has ever been. Never have so many gears "become obsolete" simultaneously.

Compose, coroutines, endless Jetpack libraries that are forced from every iron, KMM, Gradle that changes at a very fast pace. And they're all rushing in parallel, with some forcing you to use others literally like dominoes. Everything is experimental, with issues filed literally a couple of hours ago. Bleeding edge.

We in Android are now at that moment where it's too late to use something proven, and too early to switch to new and trendy. Too late - because soon you'll have to rewrite everything anyway, too early - because the community doesn't have best practices and there's a high chance of doing something stupid being a pioneer.

The golden rule "if it works - don't touch it" - doesn't apply, because what works today doesn't mean it will tomorrow. Even the damn "back" button (gesture) in new Androids is not the same as before. And you need to handle it differently too. I couldn't even imagine that.