The signs were in plain sight:

✅ having a dev funded to work on the Client
✅ our support of #NGI
✅ #FDroid being a key partner in #Mobifree
✅ splitting client functionality into libs
✅ wrestling with #targetsdk
✅ dipping our toes into #MaterialDesign

What do all these compound to? All revealed in https://f-droid.org/2025/04/29/mobifree-client-funding.html

I believe I have something like the final #rc for the next update of #TinyWeatherForecastGermany after an other week of testing.

This version moves from #targetSDK 30 to 33, comes with a new & better data update logic, improved home screen #widgets, amongst other improvements.

Also happy news for everybody using a #legacy #device saving the environment: the app launch time is now remarkably faster compared to previous versions!

Also fixes some #bugs regarding #gadgetbridge support.

Lets see if this is a neat or stupid idea to move the #targetsdk of #TinyWeatherForecastGermany from 30 directly to 33.

At least, I installed the self-compiled #sdk successfully and the first compilation runs.

However, I suspect that the automatic fetching of weather warnings and weather data via a foreground service will fail and I will need to re-write some stuff to make it work again....

When you do NOT publish your #android #app in the Play Store (e.g. #gadgetbridge, #fdroidapp , #imagepipe), is there a good reason to increase the #targetSdk ?

Does it matter at all?

The last sdk updates mainly impose restrictions without much benefit for developers and make development harder.

Why upgrade targetSdk when you can completely ignore G***le's rules about minimum targetSdk?

What are your thoughts?