Christian Hoffmann

> I think it would be better to kill the clients first I think I did, will make sure in my next test tomorrow. > Could you repeat the test...

I did not get around to working on it recently. I still think this should be analyzed properly. Keeping it open, but moving it to Backlog for that reason. Feel...

Meanwhile, the exact problem/error message and/or even `adb logcat` output may be helpful (when posting the latter, ensure to sanitize all private information first).

@NickHyHo Thanks! This is a very good hint. Here's our manifest: https://github.com/jamulussoftware/jamulus/blob/master/android/AndroidManifest.xml#L2 The [`android:versionCode` and `android:versionName`](https://stackoverflow.com/questions/9629125/versioncode-vs-versionname-in-android-manifest/10269683#10269683) are [supposed to come from macros which are handled by androiddeployqt](https://doc.qt.io/qt-5/deployment-android.html). [`qmake` seems supposed...

I've pushed a completely untested PoC here: https://github.com/hoffie/jamulus/tree/autobuild/set-android-version Builds should be available at the following URL in about an hour. @NickHyHo If you could check whether this improves things, this...

~~Checked the resulting `apk` via `apktool` and I suspect that this does not fix it. I'm not seeing the relevant headers in AndroidManifest.xml.~~ Will have another look this evening.

Looks like this approach *did* work, but [apktool does not put versionCode and versionName into the decoded AndroidManifest.xml](https://stackoverflow.com/questions/30774833/how-to-increment-versioncode-using-apktool/31951115#31951115). `apkgtool.yml` does show the proper info, as does https://www.sisik.eu/apk-tool Gonna submit a...

The PR (#2398) does not solve this issue fully yet (but I think it's still related and required). As @ann0see suspected, the current builds do not seem to be signed...

Can't say anything specific offhand either, but `--showanalyzerconsole` (a hidden command line flag) might help provide some additional insights?

When considering changing them for solely performance reasons it might also make sense to validate whether the compiler doesn't already perform such optimizations.