At WWDC25 we launched a new type of Lab event for the developer community - Group Labs. A Group Lab is a panel Q&A designed for a large audience of developers. Group Labs are a unique opportunity for the community to submit questions directly to a panel of Apple engineers and designers. Here are the highlights from the WWDC25 Group Lab for UI Frameworks. How would you recommend developers start adopting the new design? Start by focusing on the foundational structural elements of your application, working from the "top down" or "bottom up" based on your application's hierarchy. These structural changes, like edge-to-edge content and updated navigation and controls, often require corresponding code modifications. As a first step, recompile your application with the new SDK to see what updates are automatically applied, especially if you've been using standard controls. Then, carefully analyze where the new design elements can be applied to your UI, paying particular attention to custom controls or UI that could benefit from a refresh. Address the large structural items first then focus on smaller details is recommended. Will we need to migrate our UI code to Swift and SwiftUI to adopt the new design? No, you will not need to migrate your UI code to Swift and SwiftUI to adopt the new design. The UI frameworks fully support the new design, allowing you to migrate your app with as little effort as possible, especially if you've been using standard controls. The goal is to make it easy to adopt the new design, regardless of your current UI framework, to achieve a cohesive look across the operating system. What was the reason for choosing Liquid Glass over frosted glass, as used in visionOS? The choice of Liquid Glass was driven by the desire to bring content to life. The see-through nature of Liquid Glass enhances this effect. The appearance of Liquid Glass adapts based on its size; larger glass elements look more frosted, which aligns with the design of visionOS, where everything feels larger and benefits from the frosted look. What are best practices for apps that use customized navigation bars? The new design emphasizes behavior and transitions as much as static appearance. Consider whether you truly need a custom navigation bar, or if the system-provided controls can meet your needs. Explore new APIs for subtitles and custom views in navigation bars, designed to support common use cases. If you still require a custom solution, ensure you're respecting safe areas using APIs like SwiftUI's safeAreaInset. When working with Liquid Glass, group related buttons in shared containers to maintain design consistency. Finally, mark glass containers as interactive. For branding, instead of coloring the navigation bar directly, consider incorporating branding colors into the content area behind the Liquid Glass controls. This creates a dynamic effect where the color is visible through the glass and moves with the content as the user scrolls. I want to know why new UI Framework APIs aren’t backward compatible, specifically in SwiftUI? It leads to code with lots of if-else statements. Existing APIs have been updated to work with the new design where possible, ensuring that apps using those APIs will adopt the new design and function on both older and newer operating systems. However, new APIs often depend on deep integration across the framework and graphics stack, making backward compatibility impractical. When using these new APIs, it's important to consider how they fit within the context of the latest OS. The use of if-else statements allows you to maintain compatibility with older systems while taking full advantage of the new APIs and design features on newer systems. If you are using new APIs, it likely means you are implementing something very specific to the new design language. Using conditional code allows you to intentionally create different code paths for the new design versus older operating systems. Prefer to use if #available where appropriate to intentionally adopt new design elements. Are there any Liquid Glass materials in iOS or macOS that are only available as part of dedicated components? Or are all those materials available through new UIKit and AppKit views? Yes, some variations of the Liquid Glass material are exclusively available through dedicated components like sliders, segmented controls, and tab bars. However, the "regular" and "clear" glass materials should satisfy most application requirements. If you encounter situations where these options are insufficient, please file feedback. If I were to create an app today, how should I design it to make it future proof using Liquid Glass? The best approach to future-proof your app is to utilize standard system controls and design your UI to align with the standard system look and feel. Using the framework-provided declarative API generally leads to easier adoption of future design changes, as you're expressing intent rather than specifying pixel-perfect visuals. Pay close attention to the design sessions offered this year, which cover the design motivation behind the Liquid Glass material and best practices for its use. Is it possible to implement your own sidebar on macOS without NSSplitViewController, but still provide the Liquid Glass appearance? While technically possible to create a custom sidebar that approximates the Liquid Glass appearance without using NSSplitViewController, it is not recommended. The system implementation of the sidebar involves significant unseen complexity, including interlayering with scroll edge effects and fullscreen behaviors. NSSplitViewController provides the necessary level of abstraction for the framework to handle these details correctly. Regarding the SceneDelagate and scene based life-cycle, I would like to confirm that AppDelegate is not going away. Also if the above is a correct understanding, is there any advice as to what should, and should not, be moved to the SceneDelegate? UIApplicationDelegate is not going away and still serves a purpose for application-level interactions with the system and managing scenes at a higher level. Move code related to your app's scene or UI into the UISceneDelegate. Remember that adopting scenes doesn't necessarily mean supporting multiple scenes; an app can be scene-based but still support only one scene. Refer to the tech note Migrating to the UIKit scene-based life cycle and the Make your UIKit app more flexible WWDC25 session for more information.

I'm trying to achieve a specific UI design in SwiftUI where the bottom section of my List has a different background color than the top section. For example in the Medications portion of the Health app, the "Your Medications" Section has a different background than the top "Log" Section. How do I achieve this?: Here some example code. I wonder if I am supposed to use two Lists instead. If I use two Lists though and nest it in a ScrollView, the height of the lists needs to be specified. I am working with dynamic content, though so I don't think that is ideal. class ProtocolMedication {} // Example model struct HomeView: View { @Query private var protocolMedications: [ProtocolMedication] var body: some View { NavigationStack { List { // Upper sections with default background Section { Text("Content 1") } header: { Text("Log") } // Bottom section that needs different background Section { ForEach(protocolMedications) { medication in Text(medication.name) } } header: { Text("Your Medications") } } .listStyle(.insetGrouped) } } }

I have an app that uses UITextView for some text editing. I have some custom operations I can do on the text that I want to be able to undo, and I'm representing those operations in a way that plugs into NSUndoManager nicely. For example, if I have a button that appends an emoji to the text, it looks something like this: func addEmoji() { let inserting = NSAttributedString(string: "😀") self.textStorage.append(inserting) let len = inserting.length let range = NSRange(location: self.textStorage.length - len, length: len) self.undoManager?.registerUndo(withTarget: self, handler: { view in view.textStorage.deleteCharacters(in: range) } } My goal is something like this: Type some text Press the emoji button to add the emoji Trigger undo (via gesture or keyboard shortcut) and the emoji is removed Trigger undo again and the typing from step 1 is reversed If I just type and then trigger undo, the typing is reversed as you'd expect. And if I just add the emoji and trigger undo, the emoji is removed. But if I do the sequence above, step 3 works but step 4 doesn't. The emoji is removed but the typing isn't reversed. Notably, if step 3 only changes attributes of the text, like applying a strikethrough to a selection, then the full undo chain works. I can type, apply strikethrough, undo strikethrough, and undo typing. It's almost as if changing the text invalidates the undo manager's previous operations? How do I insert my own changes into UITextView's NSUndoManager without invalidating its chain of other operations?

Hi, I see some apps like LinkedIn that doesn't support multi view or split views on iPad, but seems this feature is enabled by default to any new project in Xcode, how to disable it ? Kind Regards

Hi everyone, I’m working on an iOS app using both UITableViewDiffableDataSource and SwiftUI, and I’m facing two separate but puzzling issues: UITableViewDiffableDataSource Not Reusing Cells on first applying after initial Snapshot. After applying first time it is working as expected from second time. SwiftUI View’s inside UITableViewCell onDisappear Not Triggering the on first changes of snapshot after initial snapshot. With normal UITableView it is working fine. Issue causing - it is causing player & cells to retain memory extensively Sample gist code for reproducing with diffable (DiffableTableViewExampleViewController) and working fine without diffable (RegularTableViewExampleViewController) https://gist.github.com/SURYAKANTSHARMA/d83fa9e7e0de309e27485100ba5aed17 Any insights or suggestions for these issues would be greatly appreciated! Thanks in advance!

Before ios18, when two fingers are switched to single fingers, the printing scale value will not change. When switching to two fingers again, the pinch and zoom printing scale will change. The same operation is performed after ios18, and two fingers are switched to single fingers. Switch back to two fingers, and the scale printing will not change. code here: - (void)pinchGesture:(UIPinchGestureRecognizer *)recognizer { NSSet <UIEvent*> *events = [recognizer valueForKey:@"_activeEvents"]; UIEvent *event = [events anyObject]; if (recognizer.state == UIGestureRecognizerStateBegan) { NSLog(@"---- begin finger count: %d scale: %f",event.allTouches.count,recognizer.scale); recognizer.scale = 1.0; } else if (recognizer.state == UIGestureRecognizerStateChanged) { NSLog(@"---- change finger count: %d scale: %f",event.allTouches.count,recognizer.scale); // recognizer.scale = 1.0; } log image for iOS 17.7 log image for ios18.0.2

UITabBarController | | VC_Tab1 --------------------------- VC_Tab2 | | | | VC_Tab1_Child VC_Tab2_Child | (HeaderView) | (MyButton) The structure of the view controllers and views in the project is as described above. <case 1> self.navigationController?.popToRootViewController(animated: false) tabBarController.selectedIndex = 1 When popToRootViewController(animated: false) is called in VC_Tab1_Child, followed by setting the tab controller’s selectedIndex = 1, the following results are observed: viewWillAppear(_:), <VC_Tab2_Child> deinit, <VC_Tab1_Child> viewDidAppear(_:), <VC_Tab2_Child> The originally expected results are as follows viewWillDisappear(_:), <VC_Tab1_Child> viewDidDisappear(_:), <VC_Tab1_Child> deinit, <VC_Tab1_Child> deinit, <HeaderView> deinit, <MyButton> headerView.backButton.rx.tap -> Event completed headerView.backButton.rx.tap -> isDisposed viewWillAppear(_:), <VC_Tab2_Child> viewDidAppear(_:), <VC_Tab2_Child> The HeaderView belonging to VC_Tab1_Child was not deallocated, and the resources associated with that view were also not released. Similarly, VC_Tab1_Child.viewWillDisappear and VC_Tab1_Child.didDisappear were not called. <case 2> self.navigationController?.popToRootViewController(animated: false) DispatchQueue.main.async { tabBarController.selectedIndex = 1 } After performing the pop operation as shown in the code and waiting for a short period before testing, the expected results were generally achieved. (However, rarely, the results were similar to those observed when called without async.)” <case 3> self.navigationController?.popToRootViewController(animated: false) DispatchQueue.main.asyncAfter(deadline: .now() + 0.1) { tabBarController.selectedIndex = 1 } When a sufficient delay was ensured as described above, the expected results were achieved 100% of the time.” The abnormal behavior is more pronounced in iOS versions prior to 18 and varies depending on the iOS version. I couldn’t find any documentation explaining the unexpected behavior shown in the results above. What could be the cause? The simulation code is provided below. https://github.com/linusix/UITabBarController_Test2

Hi, I am in need of a solution that takes data from SwiftData classes and generates a PDF with the given data. Any guidance would be greatly appreciated.

Description I'm developing a tvOS application where I utilize a UISearchController embedded within a UISearchContainerViewController for search functionality. In a particular flow, a custom view controller contains a TVDigitEntryViewController as a child, with its modalPresentationStyle set to .blurOverFullScreen. The issue arises when a user initiates the PIN entry but decides to cancel and return to the search interface without entering a PIN. Upon returning, the search keyboard is no longer visible, and attempts to focus or interact with it are unsuccessful. Steps to Reproduce Initialize and present a UISearchContainerViewController that contains a UISearchController with a results view controller. Within the search results, present a custom view controller containing TVDigitEntryViewController as a child, setting its modalPresentationStyle to .blurOverFullScreen. Dismiss the custom view controller without entering a PIN (e.g., by pressing the Menu button on the remote). Observe that upon returning to the search interface, the keyboard is missing, and focus interactions are unresponsive. Observed Behavior After dismissing the custom view controller with TVDigitEntryViewController, the search keyboard does not reappear, and the focus system seems to lose track of the search input field. Expected Behavior The search keyboard should remain visible and functional after dismissing the custom view controller, allowing users to continue their search without interruption. Additional Context I have reviewed the TVDigitEntryViewController documentation (developer.apple.com) and related discussions on the Apple Developer Forums but have not found a solution to this specific issue. Questions Has anyone encountered a similar issue or have insights into why the search keyboard becomes unresponsive after dismissing a .blurOverFullScreen modal with a child TVDigitEntryViewController? Are there recommended practices to ensure the search keyboard remains active and focusable after such modal presentations? Any guidance or suggestions would be greatly appreciated. Thank you!

I can't shake the "I don't think I did this correctly" feeling about a change I'm making for Image Playground support. When you create an image via an Image Playground sheet it returns a URL pointing to where the image is temporarily stored. Just like the Image Playground app I want the user to be able to decide to edit that image more. The Image Playground sheet lets you pass in a source URL for an image to start with, which is perfect because I could pass in the URL of that temp image. But the URL is NOT optional. So what do I populate it with when the user is starting from scratch? A friendly AI told me to use URL(string: "")! but that crashes when it gets forced unwrapped. URL(string: "about:blank")! seems to work in that it is ignored (and doesn't crash) when I have the user create the initial image (that shouldn't have a source image). This feels super clunky to me. Am I overlooking something?

Some of our users keep reporting that occasionally some widgets (which initially were working perfectly) are now broken and show only skeletons. Note that the app has multiple widgets, and this happens only to some of them (that is when some widgets get broken, some other widgets are still working normally). While developing the app, I came across this symptom as well, and when I read the device console, I found the following errors in the console (which the system was repeatedly reporting to the console): Unable to get connection interface: Error Domain=LNConnectionErrorDomain Code=1100 "Unable to locate com.example.app.MyAppntents for the com.apple.intents-service extension point" UserInfo={NSLocalizedDescription=Unable to locate com.example.app.MyAppIntents for the com.apple.intents-service extension point} The affected widgets use the 'UserDefaults with groups' to read the data that is stored by the host app. And I think that sometimes the widget fails (or even crashes) while reading the data from the UserDefaults and it ends up showing the skeleton. The only remedy I found was Uninstall the app Reboot the device Install the app again. After this, the widget starts to work again.. So, what can be the reason behind this? Is this an iOS bug or what and how to fight it?

I have a NavigationSplitView with all three columns, and NavigationLinks in the sidebar to present different Lists for the content column. When a list item is selected in the content view, I want to present a view in the detail view. Since different detail views should be presented for different content views, I represent state like this (some details omitted for space): // Which detail view to show in the third column enum DetailViewKind: Equatable { case blank case localEnv(RegistryEntry) case package(SearchResult) } // Which link in the sidebar was tapped enum SidebarMenuItem: Int, Hashable, CaseIterable { case home case localEnvs case remoteEnvs case packageSearch } // Binds to a DetailViewKind, defines the activate SidebarMenuItem struct SidebarView: View { @State private var selectedMenuItem: SidebarMenuItem? @Binding var selectedDetailView: DetailViewKind var body: some View { VStack(alignment: .leading, spacing: 32) { SidebarHeaderView() Divider() // Creates the navigationLinks with a SidebarMenuRowView as labels SidebarMenuView(selectedMenuItem: $selectedMenuItem, selectedDetailView: $selectedDetailView) Spacer() Divider() SidebarFooterView() } .padding() .frame(alignment: .leading) } } struct SidebarMenuRowView: View { var menuItem: SidebarMenuItem @Binding var selectedMenuItem: SidebarMenuItem? @Binding var selectedDetailView: DetailViewKind private var isSelected: Bool { return menuItem == selectedMenuItem } var body: some View { HStack { Image(systemName: menuItem.systemImageName).imageScale(.small) Text(menuItem.title) Spacer() } .padding(.leading) .frame(height: 24) .foregroundStyle(isSelected ? Color.primaryAccent : Color.primary) .background(isSelected ? Color.menuSelection : Color.clear) .clipShape(RoundedRectangle(cornerRadius: 10)) .navigationDestination(for: SidebarMenuItem.self) { item in navigationDestinationFor(menuItem: item, detailView: $selectedDetailView) } .onTapGesture { if menuItem != selectedMenuItem { selectedMenuItem = menuItem } } } } // Determines which detail view to present struct DetailView: View { @Binding var selectedDetailView: DetailViewKind var innerView: some View { switch selectedDetailView { case .blank: AnyView(Text("Make a selection") .font(.subheadline) .foregroundStyle(.secondary) .navigationSplitViewColumnWidth(min: 200, ideal: 350)) case .localEnv(let regEntry): AnyView(EnvironmentDetailView(regEntry: regEntry)) case .package(let searchResult): AnyView(PackageDetailView(searchResult: searchResult)) } } var body: some View { innerView } } struct ContentView: View { @State private var detailView: DetailViewKind = .blank var body: some View { NavigationSplitView { SidebarView(selectedDetailView: $detailView) .navigationSplitViewColumnWidth(175) } content: { HomeView() .navigationSplitViewColumnWidth(min: 300, ideal: 450) } detail: { DetailView(selectedDetailView: $detailView) } } } My issue is that the detail view is not updated when the ContentView's detailView property is updated. I've verified that the value itself is changing, but the view is not. I searched around to see why this would be (this is my first Swift/SwiftUI application) and from what I gather a @Binding alone will not cause a view to be updated, that binding either needs to be used in the view hierarchy, or it needs to be stored as a @State property that gets updated when the binding value changes. I added a dummy @State property to DetailView and that still doesn't work, so I'm a little confused: struct DetailView: View { @Binding var selectedDetailView: DetailViewKind @State private var dummyProp: DetailViewKind? var innerView: some View { switch selectedDetailView { case .blank: AnyView(Text("Make a selection") .font(.subheadline) .foregroundStyle(.secondary) .navigationSplitViewColumnWidth(min: 200, ideal: 350)) case .localEnv(let regEntry): AnyView(EnvironmentDetailView(regEntry: regEntry)) case .package(let searchResult): AnyView(PackageDetailView(searchResult: searchResult)) } } var body: some View { innerView.onChange(of: selectedDetailView) { dummyProp = selectedDetailView } } } Any ideas?

Hello, I am having the following issue: I am using the packages DSWaveformImage and DSWaveformImageViews if that matters. My code's goal is to generate waveform images from files embedded in the application. That works great on a simulator device, but when I run the code on a physical device (iPhone 15 Pro, iOS 18.2), I get this error (shown in the attached screenshot below):

I’m trying to record videos with AvAssetWriter but sometimes my videos exempt audio buffers recorded and when audio is included, the stream of video buffers stops. the gist below is my code. https://gist.github.com/kwameaj67/70a3409c84d48cf758b3734c08a46244

I'm adding Admob ads to my app, and Admob needs to know the width of the view, so I'm using GeometryReader for that. To prevent GeometryReader from grabbing screen space, I've wrapped the main view in GeometryReader { }. I then use geometry.size.width in my call to the adView. This all works fine. I have two main screens where I show ads, and they both work, until I rotate the device. Then the app crashes! If I comment out the GeometryReader code and pass a fixed value to the ad view, I can rotate the device with no fear of a crash. My question is: Do I have to accept that GeometryReader will crash the app when it's rotated, or is there another, stable way to get view dimensions?

While using my xmas present of a new iPhone and iOS 18.2, I figured I'd try the Eye Tracker app. I've been working with clients successfully using Tobii and other existing eye trackers. In my limited tests, Apple has room for improvement. My main issue is with the camera app which cannot be used at the same time while using the Eye Tracker app. I get an error popup from Apple: Camera is use by another app The image below is from my app showing the popup message "Camera in use by another app", but the same error occurs on the installed camera app. This error is from Apple, not my app. For terminology: 'front' camera is the one pointing at the user (the selfi camera) while 'back' camera is the main one with multiple lenses. Eye tracking needs the 'front' camera. It seems when an app uses the camera, it takes over both the front and back facing cameras (since you might swap them). Thus another app, especially Eye Tracking, cannot use just the front facing camera at the same time. That limits use of Eye Tracking, in particular one cannot take pictures or click any buttons on an app that uses the camera. Anyone know of a way for an app to not take over both front and back cameras at the same time? If I can separate them, the Eye Tracker could use the front camera while the camera uses the back camera.

More specifically, if I have 2 pinned tabs that are displayed in different order for compact and regular and switch between compact and regular, when it gets to regular mode, the images for both tabs will not be rendered. The tabs can still be pressed, just the icon is missing.

Hi, I'm working on an app that will mostly live in the menu bar. I'm trying to make a menu item that looks similar to the Tailscale app's menu: Note: I'm inspired by how Tailscale's menu is rendered: I have made a View that shows my avatar, name, and optionally the company I work for: import SwiftUI struct MenuWhoAmI: View { var username: String var binding: String? var body: some View { HStack { AsyncImage(url: URL(string: "https://avatars.githubusercontent.com/u/76716")!){ image in image.resizable().scaledToFit() } placeholder: { ProgressView() } .clipShape(Circle()) VStack(alignment: .leading) { Text(username) if let binding = binding { Text("\(binding)").foregroundStyle(.secondary) } } } } } #Preview { VStack(alignment: .leading) { MenuWhoAmI(username: "grahamc").padding() Divider() MenuWhoAmI(username: "grahamc", binding: "DeterminateSystems").padding() }.padding() } I tried using it in my menu bar: import SwiftUI @main struct DeterminateApp: App { var body: some Scene { MenuBarExtra("Determinate", image: "MenuIcon") { MenuWhoAmI(username: "grahamc") Button("Two") {} Button("Three") {} Divider() Button("Quit") { NSApplication.shared.terminate(nil) }.keyboardShortcut("q") }.menuBarExtraStyle(.menu) } } and it renders differently: After reading the forums and documentation, I understood the MenuBarExtra only renders certain elements. I then tried to use an NSStatusBar with an AppDelegate: import AppKit import SwiftUI @main struct DeterminateApp: App { @NSApplicationDelegateAdaptor private var appDelegate: AppDelegate var body: some Scene { Window("Authentication", id: "login") {} } } class AppDelegate: NSObject, NSApplicationDelegate, ObservableObject { private var statusItem: NSStatusItem! func applicationDidFinishLaunching(_ notification: Notification) { statusItem = NSStatusBar.system.statusItem(withLength: NSStatusItem.variableLength) if let button = statusItem.button { button.image = NSImage(named: NSImage.Name("MenuIcon")) } statusItem.menu = NSHostingMenu(rootView: Group { Button(action: { print("hi") }) { MenuWhoAmI(username: "grahamc") } }) } } and still, the avatar/name doesn't render like I'd expect, missing the circle clipping: ...and I'm a bit mystified. How can I make this menu render the way I'm trying for? Thank you!

Using gesture recognizers it is easy to implement a long-press gesture to open a menu, show a preview or something else on the iOS platform. And you can provide the duration the user must hold down the finger until the gesture recognizer fires. But I could not yet find out how to determine the default duration for a long-press gesture that is configured in the system settings within the "accessibility" settings under "Haptic Touch" (the available options are fast, standard and slow here). Is it possible to read out this setting, so my App can adapt to this system setting as well?

Its important to note that this same app did not have this issue in iOS 17. Ever since iOS 18 I have noticed that when application written in SwiftUI uses Label with the default color (which auto changes between light and dark appearance), from time to time when resuming an application that has been in the background, the color of those labels (only the Label elements) switches from the opposite to the correct one. Here is an example: Steps to reproduce Phone is in dark appearance Open app Labels text color is white and labels background is black Go to home so that app is on background Wait random time (does not happen all the time), some times 1 min some times 10 Reopen the application. During the opening transition the Label text color was changed while the app was in suspended mode to the light appearance variant (black) Once the app opening transition finishes the Label text color switches back to the correct color for dark appearance (white) Same issue happens if you start from light appearance. I cannot reproduce this on Xcode simulators, I have tried to force memory warning to check if that has anything to do with it but that also does not cause the issue to appear on simulators. For now I can only reproduce this on real device. Screenshots Here is screenshots of the above example: During transition After transition

My App always encounter with CoreAutoLayout invade My SnapKit layout constraint as follow: popBgView.snp.makeConstraints { make in make.centerY.equalToSuperview() make.leading.equalTo(assistantTeacherView.snp.trailing).offset(.isiPad ? -50 : -40) if TTLGlobalConstants.isCompactScreen320 { make.width.lessThanOrEqualTo(300) } else { let widthRatio = .isiPad ? 494.0 / 1024.0 : 434.0 / 812.0 make.width.lessThanOrEqualTo(TTLGlobalConstants.screenWidth * widthRatio) } bubbleViewRightConstraint = make.trailing.equalToSuperview().constraint } ..... popBgView.addSubview(functionView) msgLabel.snp.remakeConstraints { make in make.leading.equalToSuperview().inset(Metric.msgLabelHorizantalInset) make.centerY.equalToSuperview() make.trailing.lessThanOrEqualToSuperview().inset(Metric.msgLabelHorizantalInset) make.top.equalTo(Metric.msgLabelVerticalInset) } functionView.snp.makeConstraints { make in make.leading.equalTo(msgLabel.snp.trailing).offset(Metric.msgLabelFunctionSpacing) make.centerY.equalToSuperview() make.trailing.equalToSuperview().offset(-Metric.msgLabelHorizantalInset) } msgLabel and functionView superview is popBgView However, when I try remove from superview for functionView, There is low probability crash: OS Version: iOS 16.1.1 (20B101) Report Version: 104 Exception Type: EXC_BAD_ACCESS (SIGSEGV) Exception Codes: SEGV_NOOP Crashed Thread: 0 Application Specific Information: Exception 1, Code 1, Subcode 14967683541490370463 > KERN_INVALID_ADDRESS at 0xcfb7e4e0f8fe879f. Thread 0 Crashed: 0 CoreAutoLayout 0x382555f44 -[NSISEngine positiveErrorVarForBrokenConstraintWithMarker:errorVar:] 1 CoreAutoLayout 0x382555e9c -[NSISEngine positiveErrorVarForBrokenConstraintWithMarker:errorVar:] 2 CoreAutoLayout 0x3825557e4 -[NSISEngine removeConstraintWithMarker:] 3 CoreAutoLayout 0x382555198 -[NSLayoutConstraint _removeFromEngine:] 4 UIKitCore 0x34d87961c __57-[UIView _switchToLayoutEngine:]_block_invoke 5 CoreAutoLayout 0x382556e8c -[NSISEngine withBehaviors:performModifications:] 6 UIKitCore 0x34d8a1c38 -[UIView(AdditionalLayoutSupport) _switchToLayoutEngine:] 7 UIKitCore 0x34d7f01b0 __57-[UIView _switchToLayoutEngine:]_block_invoke_2 8 UIKitCore 0x34d879770 __57-[UIView _switchToLayoutEngine:]_block_invoke 9 CoreAutoLayout 0x382556e8c -[NSISEngine withBehaviors:performModifications:] 10 UIKitCore 0x34d8a1c38 -[UIView(AdditionalLayoutSupport) _switchToLayoutEngine:] 11 UIKitCore 0x34d8a1848 __45-[UIView _postMovedFromSuperview:]_block_invoke 12 UIKitCore 0x34e7ff8d0 -[UIView _postMovedFromSuperview:] 13 UIKitCore 0x34d85e3c8 __UIViewWasRemovedFromSuperview 14 UIKitCore 0x34d85b1a4 -[UIView(Hierarchy) removeFromSuperview] 15 Collie-iPad 0x203001550 [inlined] InClassAssistantView.functionView.didset (InClassAssistantView.swift:105)