I'm in the process of refactoring a humongous monolithic app with multiple extensions into a setup where we have a service library which is linked to each of the targets. However, this causes a significant issue when the service library is being referenced from the notification extension, as that extension has a hard memory limit of 12MB on an iPhone 5s, which we are still supporting, and our memory usage is at about 11MB which is way too close for comfort. When the memory usage surpasses the limit, the notification extension is automatically killed by the system. Profiling with Instruments/Allocations has shown that a significant chunk of that memory is allocated (wasted, really) by the internal AVSecondScreenController.load() class initializer from the AVKit library. My approach is now to try to figure out what causes AVKit to be loaded in the first palce. I have no explicit imports for AVKit in the notification extension and attempts to figure out what triggers it from the service library failed as well. I did add a symbolic breakpoint for AVSecondScreenController.load() which is hit, but the backtrace is useless. Does anyone have any idea how I can pinpoint what causes AVKit to be loaded in the first place? Thanks!

I've been trying to get something to work with the Screen Time API, but almost everything related to it is completely and truly broken/unimplemented. I've tried all of this in the iOS 15 Beta 2 and 3 and Xcode 13 Beta 2 and 3. These are just a handful of the bugs I came across: Screen Time APIs do not work on the simulator. Calling AuthorizationCenter.shared.requestAuthorization on a simulator with a logged in child account always results in "FamilyControls.FamilyControlsError Code=3" FamilyActivityPicker does not list installed apps on either of the guardian's or child's devices, it only lists the categories. It's not clear at this point whether FamilyActivityPicker needs to be called on the parent or the child device. Assuming restrictions can actually be set, there is no API to allow setting up of restrictions for individual kids. After successfully authorizing a child device via AuthorizationCenter.shared.requestAuthorization, DeviceActivityCenter.startMonitoring always results in an MonitoringError.unauthorized error. AuthorizationCenter.requestAuthorization error in callback cannot be mapped back to FamilyControlsError because it is a plain NSError with an error code and FamilyControlsError does not expose an initialiser that takes a rawValue. Even though an app is authorized to managed ScreenTime on a child device, the child can always just delete the app. (I would try to submit a bug with device diagnostics about this using Feedback Assistant, but to top it all off at the moment I am receiving an error in FA that says "Unable to gather diagnostics" / "A problem was encountered when trying to gather diagnostics. Try gathering diagnostics again"). There's no device activity / shield configuration or shield action extension templates in Xcode. For those wondering, I created a CallDirectory extension and then changed the NSExtensionPointIdentifier in the plist to the appropriate value. What is the appropriate value? It's not mentioned in the documentation :), so I had to dig around in the Xcode support files to find out the following identifiers: com.apple.ManagedSettings.shield-action-service com.apple.ManagedSettingsUI.shield-configuration-service com.apple.deviceactivity.monitor-extension And this leads me to my last item: the documentation which is absolutely lacking. There is no programming guide for the frameworks and the Homework app we saw developed during WWDC is not available. It's not clear how the settings from the guardian app get applied to the child app. Is there a built-in mechanism or do we need to use our own implementation to transfer those changes (via push or whatever)? Of course I reported almost all of this via Feedback Assistant, but considering I had previous feedback for other APIs sit there untouched for years in the past, it's really frustrating and it feels like I'm just talking to myself with those bug reports. So, did I do something wrong and completely misinterpreted the capabilities and functionality of the ScreenTime API or is it truly broken and non-functional in iOS 15 Beta 2/3? There's absolutely no mention of these issues in the release notes so I am starting to question my sanity here.

I've been going over the new ScreenTime APIs and I haven't been able to identify any API that would allow the parent to receive the "allow more time" request from the child's device. Is it even possible to handle that request or is that handled only by the system ScreenTime app?