“show when run”

Ok, here is my step by step (what I think is) a solution guide... (worked for me) Fix: Xcode Hangs on macOS Tahoe 26.4 (SDKs Status Unavailable) The Problem macOS Tahoe 26.4 introduces a bug where a CoreSimulator subprocess (simdiskimaged) repeatedly exceeds its 20MB memory limit and gets killed by the system. This causes CoreSimulatorService to hang indefinitely, which in turn causes Xcode to freeze, SDKs to show status unavailable, and all simulators to become inaccessible. The root cause is a runaway loop that creates new SimRuntimeBundle files every 11 seconds in /Library/Developer/CoreSimulator/Cryptex/Images/mnt/, bloating the state that simdiskimaged must load on startup beyond its jetsam memory limit. The Fix Step 1 — Grant Terminal Full Disk Access The files that need to be deleted are protected by TCC (Transparency, Consent, and Control). Without this step, sudo rm will return Operation not permitted. Open System Settings → Privacy & Security → Full Disk Access Click the + button Add Te

Here is a summary of my findings so far working with AI: The Bug: macOS Tahoe 26.4 triggers a loop in the CoreSimulator subsystem where a new SimRuntimeBundle is created in /Library/Developer/CoreSimulator/Cryptex/Images/mnt/ every 11 seconds. This accumulates thousands of files and bloats the images.plist manifest. Why Xcode Hangs: When Xcode launches, it starts CoreSimulatorService, which spawns a helper daemon called simdiskimaged via XPC. simdiskimaged is responsible for managing simulator disk images and runtimes. On startup, it tries to load and validate the entire contents of images.plist and the Cryptex directory. Because of the runaway loop, there are now 7,000+ files and a bloated manifest to parse. simdiskimaged exceeds its hardcoded 20MB jetsam memory limit and gets killed by the system (SIGKILL/-9). CoreSimulatorService sent a synchronous XPC message to simdiskimaged and is now waiting for a reply that will never come — it hangs forever on mach_msg2_trap. Everything downstream hangs: xcrun simctl

Thanks for the post. I'm not an expert on audio and less when that involves watchOS. So other watchOS engineers are hopefully jumping into the thread soon. What I know is the behavior you are seeing is due to the strict differences in how watchOS handles background execution and audio compared to iOS. When your app goes into the background on watchOS, custom audio playback like AVAudioPlayer or AVAudioEngine playing a voice or beep is immediately suspended unless specific background modes and sessions are active. The reason the ping sound and haptics continue is likely because you are triggering them via WKInterfaceDevice.current().play(.notification) or similar system-level calls, which are handled by a separate system and can fire as long as your app has any background execution time. Since you want your app to play over music apps without stopping them, you must configure your AVAudioSession to mix with other audio. You should also consider using https://developer.apple.com/documentation/avfaudio/avaudiose

Thanks for the post I believe iOS categorizes it as a secondary audio source, by design, secondary audio sources are not allowed to take over the system's Now Playing deck. The Now Playing screen (and the navigation button to reach it in CarPlay) is strictly reserved for the primary media app. .defaultToSpeaker and .allowBluetoothA2DP are strictly invalid for the .playback category as they are only for .playAndRecord. I believe, you need to use the .playback category without any options that downgrade your app to a secondary audio source or conflict with the category as CarPlay is very strict about when it shows the Now Playing button. CarPlay often refuses to show the Now Playing icon in the navigation bar until the app has actually pushed audio buffers through the active AVAudioSession. Remove .mixWithOthers and the invalid flags, ensure your playback state is set to .playing, and the CarPlay Now Playing deck should behave exactly as expected? https://developer.apple.com/documentation/AVFA

Thanks for the post a the details including the screenshots showing exactly what you are seeing. Yes, there are many changes after Liquid Glass. Your understanding is 100% correct. Rebuilding a UITabBar entirely from scratch is a massive undertaking for a complex app. Because of this maintenance burden and the risk of breaking standard navigation paradigms, replacing the standard UITabBar with a custom view is generally not the recommended path forward. When the Liquid Glass appearance is enabled, the selected capsule is the system-standard visual indicator for the active tab. The framework enforces this specific visual treatment to maintain a consistent design language and accessibility standard across all the different systems. Because Liquid Glass relies on a specific compositing and rendering pipeline to achieve that look, standard legacy overrides are often ignored or overridden by the system to ensure the capsule remains visible. Since you mentioned that disabling Liquid Glass gives you the exa