I created a Radar for this FB14766095, but thought I would add it here for extra visibility, or if anyone else had any thoughts on the issue. Basic Information Please provide a descriptive title for your feedback: iOS 18 hit testing functionality differs from iOS 17 What type of feedback are you reporting? Incorrect/Unexpected Behavior Description: Please describe the issue and what steps we can take to reproduce it: We have an issue in iOS 18 Beta 6 where hit testing functionality differs from the expected functionality in iOS 17.5.1 and previous versions of iOS. iOS 17: When a sheet is presented, the hit-testing logic considers subviews of the root view, meaning the rootView itself is rarely the hit view. iOS 18: When a sheet is presented, the hit-testing logic changes, sometimes considering the rootView itself as the hit view. Code: import SwiftUI struct ContentView: View { @State var isPresentingView: Bool = false var body: some View { VStack { Text("View One") Button { isPresentingView.toggle() } label: { Text("Present View Two") } } .padding() .sheet(isPresented: $isPresentingView) { ContentViewTwo() } } } #Preview { ContentView() } struct ContentViewTwo: View { @State var isPresentingView: Bool = false var body: some View { VStack { Text("View Two") } .padding() } } extension UIWindow { public override func hitTest(_ point: CGPoint, with event: UIEvent?) -> UIView? { /// Get view from superclass. guard let hitView = super.hitTest(point, with: event) else { return nil } print("RPTEST rootViewController = ", rootViewController.hashValue) print("RPTEST rootViewController?.view = ", rootViewController?.view.hashValue) print("RPTEST hitView = ", hitView.hashValue) if let rootView = rootViewController?.view { print("RPTEST rootViewController's view memory address: \(Unmanaged.passUnretained(rootView).toOpaque())") print("RPTEST hitView memory address: \(Unmanaged.passUnretained(hitView).toOpaque())") print("RPTEST Are they equal? \(rootView == hitView)") } /// If the returned view is the `UIHostingController`'s view, ignore. print("MTEST: hitTest rootViewController?.view == hitView", rootViewController?.view == hitView) print("MTEST: -") return hitView } } Looking at the print statements from the provided sample project:
 iOS 17 presenting a sheet from a button tap on the ContentView(): RPTEST rootViewController's view memory address: 0x0000000120009200 RPTEST hitView memory address: 0x000000011fd25000 RPTEST Are they equal? false MTEST: hitTest rootViewController?.view == hitView false RPTEST rootViewController's view memory address: 0x0000000120009200 RPTEST hitView memory address: 0x000000011fd25000 RPTEST Are they equal? false MTEST: hitTest rootViewController?.view == hitView false iOS 17 dismiss from presented view: RPTEST rootViewController's view memory address: 0x0000000120009200 RPTEST hitView memory address: 0x000000011fe04080 RPTEST Are they equal? false MTEST: hitTest rootViewController?.view == hitView false RPTEST rootViewController's view memory address: 0x0000000120009200 RPTEST hitView memory address: 0x000000011fe04080 RPTEST Are they equal? false MTEST: hitTest rootViewController?.view == hitView false iOS 18 presenting a sheet from a button tap on the ContentView(): RPTEST rootViewController's view memory address: 0x000000010333e3c0 RPTEST hitView memory address: 0x0000000103342080 RPTEST Are they equal? false MTEST: hitTest rootViewController?.view == hitView false RPTEST rootViewController's view memory address: 0x000000010333e3c0 RPTEST hitView memory address: 0x000000010333e3c0 RPTEST Are they equal? true MTEST: hitTest rootViewController?.view == hitView true You can see here ☝️ that in iOS 18 the views have the same memory address on the second call and are evaluated to be the same. This differs from iOS 17. iOS 18 dismiss RPTEST rootViewController's view memory address: 0x000000010333e3c0 RPTEST hitView memory address: 0x0000000103e80000 RPTEST Are they equal? false MTEST: hitTest rootViewController?.view == hitView false RPTEST rootViewController's view memory address: 0x000000010333e3c0 RPTEST hitView memory address: 0x0000000103e80000 RPTEST Are they equal? false MTEST: hitTest rootViewController?.view == hitView false The question I want to ask: Is this an intended change, meaning the current functionality in iOS 18 is expected? Or is this a bug and it's something that needs to be fixed? As a user, I would expect that the hit testing functionality would remain the same from iOS 17 to iOS 18. Thank you for your time.

I have an NSStatusBar application. This is my first in SwiftUI. And I need to know when the window is closed so that I can disable some of menu commands. I can use NSWindowDelegate with AppDelegate as follows. import SwiftUI @main struct SomeApp: App { @NSApplicationDelegateAdaptor(AppDelegate.self) var appDelegate @StateObject private var menuViewModel = MenuViewModel() var body: some Scene { WindowGroup { ContentView() .environmentObject(menuViewModel) } } } class AppDelegate: NSObject, NSApplicationDelegate, NSWindowDelegate { private var menuViewModel = MenuViewModel() func applicationDidFinishLaunching(_ notification: Notification) { if let window = NSApplication.shared.windows.first { window.setIsVisible(false) window.delegate = self } } func windowWillClose(_ notification: Notification) { menuViewModel.windowClosed = true } } When the window will close, MenuViewModel (ObservableObject) will receive a call, which I want my ContentView to receive. But, so far, it won't. import SwiftUI struct ContentView: View { var body: some View { ZStack { ... ... } .onReceive(statusBarViewModel.$windowClosed) { result in // never called... } } } Can a SwiftUI View receive a call somehow when its window closes? Muchos thankos.

Hi, everyone I have an app already in production that uses SwiftUI's lifecycle (paired with an AppDelegate). Due to some specific behaviour of the app, we decided to migrate the app to use UIKit's lifecycle, adding the corresponding SceneDelegate to the app, as well as modifying the Info.plist file accordingly to accommodate to these new changes. Although everything seems to work when installing the app from zero, when installing it on top of another version, the screen goes black and the user cannot interact with the app at all unless they reinstall it completely. As I've read online, iOS is reusing the window configuration from the previous execution of the app. I know this because the AppDelegate's application(application:connectingSceneSession:options) is not being called when coming from a previous version of the app. I would love to know what can I do to make this work because, as you may understand, we cannot ask our user base to reinstall the application. Thank you very much.

When displaying a view with a Button inside a ScrollView using the sheet modifier, if you try to close the sheet by swiping and your finger is touching the Button, the touch is not canceled. This issue occurs when building with Xcode 16 but does not occur when building with Xcode 15. Here is screen cast. https://drive.google.com/file/d/1GaOjggWxvjDY38My4JEl-URyik928iBT/view?usp=sharing Code struct ContentView: View { @State var isModalPresented: Bool = false var body: some View { ScrollView { Button { debugPrint("Hello") isModalPresented.toggle() } label: { Text("Hello") .frame(height: 44) } Button { debugPrint("World") } label: { Text("World") .frame(height: 44) } Text("Hoge") .frame(height: 44) .contentShape(Rectangle()) .onTapGesture { debugPrint("Hoge") } } .sheet(isPresented: $isModalPresented) { ContentView() } } }

Ok… I'm baffled here… this is very strange. Here is a SwiftUI app: import SwiftUI @main struct StepperDemoApp: App { func onIncrement() { print(#function) } func onDecrement() { print(#function) } var body: some Scene { WindowGroup { Stepper { Text("Stepper") } onIncrement: { self.onIncrement() } onDecrement: { self.onDecrement() } } } } When I run in the app in macOS (Xcode 16.0 (16A242) and macOS 14.6.1 (23G93)), I see some weird behavior from these buttons. My experiment is tapping + + + - - -. Here is what I see printed: onIncrement() onIncrement() onIncrement() onIncrement() onDecrement() What I expected was: onIncrement() onIncrement() onIncrement() onDecrement() onDecrement() onDecrement() Why is an extra onIncrement being called? And why is one onDecrement dropping on the floor? Deploying the app to iPhone Simulator does not repro this behavior (I see the six "correct" logs from iPhone Simulator).

Making the call as: UIApplication.shared.open(websiteURL) doesn't open the browser and gives the following in the console: BUG IN CLIENT OF UIKIT: The caller of UIApplication.openURL(:) needs to migrate to the non-deprecated UIApplication.open(:options:completionHandler:). Force returning false (NO). However, making the call as: UIApplication.shared.open(websiteURL) { _ in } opens the browser and there is nothing in the cosole. Does someone understand why is this or if it's Apple's iOS 18 bug? In the iOS & iPadOS 18 RC Release Notes | Apple Developer Documentation there is a section around resolving this or something similar, unsure.

I'm using NSTextTable to format panels of stand-out text within body text. Paragraphs within the panel are handled as individual NSTextTableBlocks within the table. Each block is added to the NSMutableParagraphStyle that is part of the attributed string within the block's attributes. That's all fine and it works. But... Occasionally I see undrawn lines within the panel. These disappear (or sometimes appear) when the parent window (and thus the NSTextView holding the rendered attributed string) is resized. Lines do not always appear, and when they do they are not always in the same place. The height of the gap varies. I see this behaviour with these panels and with tables. What's common to both cases is not only the use of NSTextTable and NSTextTableBlock etc., but crucially (I think) the use of block margins. If I disable margins (which looks OK for the panels, but isn't right for tables), the problem disappears. So, a bug or (more likely) I'm missing a key part of view rendering or margin set up. But what? Code here: https://github.com/smittytone/PreviewMarkdown/blob/930f5f32aa0b3b77ec3f4f53436a79e10bb26f18/Markdown%20Previewer/Styler.swift#L882 Running 14.6.1 on an M3. I'm using TextKit 1 because I'm using an NSLayoutManager subclass to override certain text underlines (not used in panels as outlined above, or tables).

When I copy and paste example code in apple developer documentation, LongPressGesture does not work as expected in Xcode Version 16.0 (16A242) and iOS 18. It seems updating(_:body:) method does not work when used with LongPressGesture. When I make a breakpoint in updating(_:body:) method and long press the blue circle on the screen of simulator(or device), it is expected to be caught in breakpoint or it is expected that color of circle turns from blue to red to green. However, it is not caught in breakpoint and never turns to red. Question of Stackoverflow is about same issue and I can not use onLongPressGesture method to implement required feature of my app. Development environment: Xcode Version 16.0 (16A242), macOS 14.5 Run-time configuration: iOS 18.0

Based on this TextKit 2 demo project I thought that I could implement syntax highlighting by parsing syntax block tokens (e.g. comments like <!-- --> or /* */) in processEditing and storing their locations, and then actually applying the rendering with NSTextContentStorageDelegate in textContentStorage(_:textParagraphWith:) by checking the location of each paragraph against the store of syntax tokens. This sort of works except that the rendering is only updated for paragraphs which are changed. Is there a way to trigger NSTextContentStorage to re-fetch paragraphs in a given range? Or is this a totally misguided approach to the problem?

Hello there! I wanted to give a native scrolling mechanism for the Swift Charts Graph a try and experiment a bit if the scenario that we try to achieve might be possible, but it seems that the Swift Charts scrolling performance is very poor. The graph was created as follows: X-axis is created based on a date range, Y-axis is created based on an integer values between moreless 0-320 value. the graph is scrollable horizontally only (x-axis), The time range (x-axis) for the scrolling content was set to one year from now date (so the user can scroll one year into the past as a minimum visible date (.chartXScale). The X-axis shows 3 hours of data per screen width (.chartXVisibleDomain). The data points for the graph are generated once when screen is about to appear so that the Charts engine can use it (no lazy loading implemented yet). The line data points (LineMark views) consist of 2880 data points distributed every 5 minutes which simulates - two days of continuous data stream that we want to present. The rest of the graph displays no data at all. The performance result: The graph on the initial loading phase is frozen for about 10-15 seconds until the data appears on the graph. Scrolling is very laggy - the CPU usage is 100% and is unacceptable for the end users. If we show no data at all on the graph (so no LineMark views are created at all) - the result is similar - the empty graph scrolling is also very laggy. Below I am sharing a test code: @main struct ChartsTestApp: App { var body: some Scene { WindowGroup { ContentView() Spacer() } } } struct LineDataPoint: Identifiable, Equatable { var id: Int let date: Date let value: Int } actor TestData { func generate(startDate: Date) async -> [LineDataPoint] { var values: [LineDataPoint] = [] for i in 0..<(1440 * 2) { values.append( LineDataPoint( id: i, date: startDate.addingTimeInterval( TimeInterval(60 * 5 * i) // Every 5 minutes ), value: Int.random(in: 1...100) ) ) } return values } } struct ContentView: View { var startDate: Date { return endDate.addingTimeInterval(-3600*24*30*12) // one year into the past from now } let endDate = Date() @State var dataPoints: [LineDataPoint] = [] var body: some View { Chart { ForEach(dataPoints) { item in LineMark( x: .value("Date", item.date), y: .value("Value", item.value), series: .value("Series", "Test") ) } } .frame(height: 200) .chartScrollableAxes(.horizontal) .chartYAxis(.hidden) .chartXScale(domain: startDate...endDate) // one year possibility to scroll back .chartXVisibleDomain(length: 3600 * 3) // 3 hours visible on screen .onAppear { Task { dataPoints = await TestData().generate(startDate: startDate) } } } } I would be grateful for any insights or suggestions on how to improve it or if it's planned to be improved in the future. Currently, I use UIKit CollectionView where we split the graph into smaller chunks of the graph and we present the SwiftUI Chart content in the cells, so we use the scrolling offered there. I wonder if it's possible to use native SwiftUI for such a scenario so that later on we could also implement some kind of lazy loading of the data as the user scrolls into the past.

Hi! When building my app using Xcode 16, fast scrolling (using scrollViewProxy.scrollTo) a list would result in items not appearing even when scrolling stopped. This does not happen when the app is built with Xcode 15, even on iOS 18. I'm also getting this error in the logs: List failed to visit cell content, returning an empty cell. - SwiftUICore/Logging.swift:84 - please file a bug report.

Hi everyone. I'm trying to use the new ControlWidget API introduced on iOS 18 to open a sheet that contains a form when the user taps on the button on the control center. This is my current code. It opens the app, but I haven't found how to do an action inside the app when the app is opened. @available(iOS 18, *) struct AddButtonWidgetControl: ControlWidget { var body: some ControlWidgetConfiguration { StaticControlConfiguration(kind: "com.example.myapp.ButtonWidget") { ControlWidgetButton(action: LaunchAppIntent()) { Label("Add a link", systemImage: "plus") } } .displayName("Add a link") .description("Creates a link.") } } @available(iOS 18, *) struct LaunchAppIntent: AppIntent { static var title: LocalizedStringResource { "Launch app" } static var openAppWhenRun: Bool = true func perform() async throws -> some IntentResult { return .result() } }

Using desaturated mode on an image in a widget will break any links or buttons that use the image as their 'label'. Using the following will just open the app as if there was no link at all - therefore just using the fallback userActivity handler, or any .widgetURL() urls provided. Link(destination: URL(string: "bug://never-works")!) { Image("puppy") .widgetAccentedRenderingMode(.desaturated) } The same goes for buttons: Button(intent: MyDemoIntent()) { Image("puppy") .widgetAccentedRenderingMode(.desaturated) } I've tried hacky solutions like putting the link behind the image using a ZStack, and disabling hit testing on the image, but they don't work. Anything else to try? Logged as Feedback #15152620.

struct ContentView: View { var body: some View { NavigationSplitView { List { Text("row 1") Text("row 2") Text("row 3") } .toolbar(content: { ToolbarItem { Button("aa", action: onToolbar) } }) } detail: { HSplitView { VStack { Image(systemName: "globe") .imageScale(.large) .foregroundStyle(.tint) Text("Hello, world!") } .toolbar(id: "toolbar", content: { ToolbarItem(id: "toolbar-1") { Button("bb", action: onToolbar) } }) .padding() Text("right") } }.navigationTitle("") } func onToolbar() {} } Run & Crash NSToolbar 0x6000005665b0 already contains an item with the identifier com.apple.SwiftUI.splitViewSeparator-0. Duplicate items of this type are not allowed.

Hello. In my app, I use RegisterEventHotkey to implement global keyboard shortcuts to trigger actions. Up until macOS Sequoia, I was able to use a keyboard shortcut with option and shift as the modifiers, like option shift 2 (⌥ ⇧ 2). Now, on macOS Sequoia, using RegisterEventHotkey to register a hotkey with those exact modifiers (option and shift), regardless of the key, fails with the error -9868 (eventInternalErr). Is this a documented and wanted change, or is this a bug? Other modifier keys (just command, command option, command shift, command control, control shift, etc), all work. Any insight into this would be appreciated. (Feedback filed: FB15163561) Thank you, Matthias

Okay I know, fill a bug... but here is a super simple app that will demostrate that the navigation bar chnages from Dark scheme to light scheme when you tap back after it goes to the second view. import SwiftUI struct ContentView: View { var body: some View { NavigationStack { NavigationLink { Text("Tap back and notice the navigation title changes to black text instead of white") .toolbarBackground(.visible, for: .navigationBar) .toolbarBackground(Color.orange, for: .navigationBar) .toolbarColorScheme(.dark, for: .navigationBar) } label: { VStack { Text("First page is the sweetest") } .padding() } .navigationTitle("First Page") .toolbarBackground(.visible, for: .navigationBar) .toolbarBackground(Color.orange, for: .navigationBar) .toolbarColorScheme(.dark, for: .navigationBar) } } } #Preview { ContentView() }

I have a NSWindow subclass. The window has a custom shape, and thus has a custom contentView which overrides drawRect to draw . Since the window has a custom shape it cannot use the system provided titlebar. The problem I'm having is when there are multiple screens, if my window is on the inactive screen (not mainScreen with menu bar) and I move the mouse over to the second monitor and click the window....the menu bar doesn't travel to the screen my app is on after the window is clicked. This does not happen with any other window. In all other windows, the menu bar moves to the screen once you click a window on that screen. So it appears this is because my window is not using NSWindowStyleMaskTitled. As far as I know, I can't use the system title bar and draw my custom window shape. Abandoning the custom window shape is not an option. Without going into too many details as to why I care, the menu bar really should travel with first click on my window like other apps.. Is there a way to tell the system (other than using the NSWindowStyleMaskTitled) that clicking on my window should make that screen the "main screen" (bring the menu bar over? I tried programmatically activating the application, ordering the window to the front, etc. but none of this works. This forces the user to click outside my app window, say on the desktop, to move the menu bar over, which feels wrong. Thanks in advance if anyone has any suggestions.

Hello! After upgrading to Xcode 16 & Swift 6 & iOS 18 I starting receiveing strange crashes. Happens randomly in different view and pointing to onGeometryChange action block. I added DispatchQueue.main.async { in hopes it will help but it didn't. HStack { ... } .onGeometryChange(for: CGSize.self, of: \.size) { value in DispatchQueue.main.async { self.width = value.width self.height = value.height } } As far as I understand, onGeometryChange is defined as nonisolated and Swift 6 enforce thread checking for the closures, SwiftUI views are always run on the main thread. Does it mean we can not use onGeometryChange safely in swiftui? BUG IN CLIENT OF LIBDISPATCH: Assertion failed: Block was expected to execute on queue [com.apple.main-thread (0x1eacdce40)] Crashed: com.apple.SwiftUI.AsyncRenderer 0 libdispatch.dylib 0x64d8 _dispatch_assert_queue_fail + 120 1 libdispatch.dylib 0x6460 _dispatch_assert_queue_fail + 194 2 libswift_Concurrency.dylib 0x62b58 <redacted> + 284 3 Grit 0x3a57cc specialized implicit closure #1 in closure #1 in PurchaseModalOld.body.getter + 4377696204 (<compiler-generated>:4377696204) 4 SwiftUI 0x5841e0 <redacted> + 60 5 SwiftUI 0x5837f8 <redacted> + 20 6 SwiftUI 0x586b5c <redacted> + 84 7 SwiftUICore 0x68846c <redacted> + 48 8 SwiftUICore 0x686dd4 <redacted> + 16 9 SwiftUICore 0x6ecc74 <redacted> + 160 10 SwiftUICore 0x686224 <redacted> + 872 11 SwiftUICore 0x685e24 $s14AttributeGraph12StatefulRuleP7SwiftUIE15withObservation2doqd__qd__yKXE_tKlF + 72 12 SwiftUI 0x95450 <redacted> + 1392 13 SwiftUI 0x7e438 <redacted> + 32 14 AttributeGraph 0x952c AG::Graph::UpdateStack::update() + 540 15 AttributeGraph 0x90f0 AG::Graph::update_attribute(AG::data::ptr<AG::Node>, unsigned int) + 424 16 AttributeGraph 0x8cc4 AG::Subgraph::update(unsigned int) + 848 17 SwiftUICore 0x9eda58 <redacted> + 348 18 SwiftUICore 0x9edf70 <redacted> + 36 19 AttributeGraph 0x148c0 AGGraphWithMainThreadHandler + 60 20 SwiftUICore 0x9e7834 $s7SwiftUI9ViewGraphC18updateOutputsAsync2atAA11DisplayListV4list_AG7VersionV7versiontSgAA4TimeV_tF + 560 21 SwiftUICore 0x9e0fc0 $s7SwiftUI16ViewRendererHostPAAE11renderAsync8interval15targetTimestampAA4TimeVSgSd_AItF + 524 22 SwiftUI 0xecfdfc <redacted> + 220 23 SwiftUI 0x55c84 <redacted> + 312 24 SwiftUI 0x55b20 <redacted> + 60 25 QuartzCore 0xc7078 <redacted> + 48 26 QuartzCore 0xc52b4 <redacted> + 884 27 QuartzCore 0xc5cb4 <redacted> + 456 28 CoreFoundation 0x555dc <redacted> + 176 29 CoreFoundation 0x55518 <redacted> + 60 30 CoreFoundation 0x55438 <redacted> + 524 31 CoreFoundation 0x54284 <redacted> + 2248 32 CoreFoundation 0x535b8 CFRunLoopRunSpecific + 572 33 Foundation 0xb6f00 <redacted> + 212 34 Foundation 0xb6dd4 <redacted> + 64 35 SwiftUI 0x38bc80 <redacted> + 792 36 SwiftUI 0x1395d0 <redacted> + 72 37 Foundation 0xc8058 <redacted> + 724 38 libsystem_pthread.dylib 0x637c _pthread_start + 136 39 libsystem_pthread.dylib 0x1494 thread_start + 8

After updating to Sonoma, the following is logged in the Xcode console when an editable text field becomes key. This doesn't occur on any text field, but it seems to happen when the text field is within an NSPopover or an NSSavePanel. ViewBridge to RemoteViewService Terminated: Error Domain=com.apple.ViewBridge Code=18 "(null)" UserInfo={com.apple.ViewBridge.error.hint=this process disconnected remote view controller -- benign unless unexpected, com.apple.ViewBridge.error.description=NSViewBridgeErrorCanceled} What does this mean?

I am trying out the new TextField selection ability on macOS but it crashes in various different ways with extremely large stack traces. Looks like it is getting into re-entrant function calls. A similar problem is described on the SwiftUI forums with no responses yet. Here is my simple example struct ContentView: View { @State private var text: String = "" @State private var selection: TextSelection? var body: some View { TextField("Message", text: $text, selection: $selection) .padding() } } Setting text to a value like "Hallo World" causes an instant crash as soon as you start typing in the TextField. Setting text empty (as in example above) lets you edit the text but as it crashes as soon as you commit it (press enter). Any workarounds or fixes?