Has WKExtendedRuntimeSession changed the way timers work?

84 views Asked by At

I have a stand-alone watchOS app that starts a WKExtendedRuntimeSession and a 1-second repeating timer when a button is pressed. No CPU-intensive stuff happens during this extended (background) time.

On watchOS 9, this worked perfectly until I stopped it again (or after the hour I get out of a Mindfulness Session Type has passed, and the extended runtime session expires). The screen could go dark, and the timer would keep running.

On watchOS 10, I get about 50 seconds when the screen goes dark, and then the timer just stops. No delegate methods are called (i.e. no "extendedRuntimeSession...willExpire()" or "...didInvalidateWith()" is called), though. Also, the scenePhase does not change at this point, it remains at inactive. The execution just stops. When I tap the watch to turn the screen back on, the timer resumes as if nothing happened.

Does anybody know what's going on here? Has something changed in WKExtendedRuntimeSessions in watchOS 10, or is it a bug?

I tried removing the timer and changing it to a recursive DispatchAsync.main.asyncAfter block, but it yielded the same result.

1

There are 1 answers

0
eternalstorms On BEST ANSWER

it was my doing after all. CPU usage of my watchOS app skyrocketed when put in the background for some reason or another, and that's why watchOS suspended my app. Deservedly so. Anyway, it's all good now.