I'm currently working on implementing a character limit for Korean text input using UITextField, but I've encountered two key issues. 1. How can I determine if Korean input is complete? I understand that markedTextRange represents provisional (composing) text during multistage text input systems (such as Korean, Japanese, Chinese). While testing with Korean input, I expected markedTextRange to reflect the composing state. However, it seems that markedTextRange remains nil throughout the composition process. 2. Problems limiting character count for Korean input I’ve tried two methods to enforce a character limit. Both lead to incorrect behavior due to how Korean characters are composed. Method 1 – Before replacement: func textField(_ textField: UITextField, shouldChangeCharactersIn range: NSRange, replacementString string: String) -> Bool { guard let text = textField.text else { return true } return text.count <= 5 } This checks the text length before applying the replacementString. The issue is that when the user enters a character that is meant to combine with the previous one to form a composed character, the input should result in a single, combined character. However, because the character limit check is based on the state before the replacement is applied, the second character does not get composed as expected. Method 2 – After change: textField.addTarget(self, action: #selector(editingChanged), for: .editingChanged) @objc private func editingChanged(_ sender: UITextField) { guard var text = sender.text else { return } if text.count > limitCount { text.removeLast() sender.text = text } } This removes the last character if the count exceeds the limit after the change. But when a user keeps typing past the limit, the last character is overwritten by new input. I suspect this happens because the .editingChanged event occurs before the multistage input is finalized, and the final composed character is applied after that event. My understanding of the input flow: Standard input: shouldChangeCharactersIn is called replacementString is applied .editingChanged is triggered With multistage input (Korean, etc.): shouldChangeCharactersIn is called replacementString is applied .editingChanged is triggered Final composed character is inserted (after all the above) Conclusion Because both approaches lead to incorrect character count behavior with Korean input, I believe I need a new strategy. Is there an officially recommended way to handle multistage input properly with UITextField in this context? Any advice or clarification would be greatly appreciated. MacOS 15.5(24F74) Xcode 16.4 (16F6)

News link: https://developer.apple.com/news/?id=12m75xbj If your app offers Sign in with Apple, you’ll need to use the Sign in with Apple REST API to revoke user tokens when deleting an account. I'm not good English. I'm confused about the above sentence Do I have to use REST API unconditionally or can I just delete to the account data?

This is an issue that occurred while using SwiftUI. Cannot find '$state' in scope The other view finds properties normally. May I know why the error is occurring? The following code is the full text of the code that causes problems. import SwiftUI @Observable class HomeState { var title: String = "Home" } struct HomeView: View { @Binding var state: HomeState var body: some View { Text(state.title) } } #Preview { @Previewable @State var state: HomeState = .init() HomeView(state: $state) /// Error: Cannot find '$state' in scope } The same error occurs when using the String type rather than the object. What did I do wrong?

I am currently studying the Accelerate library by referring to Apple documentation. Here is the link to the referenced document: https://developer.apple.com/documentation/accelerate/veclib/vforce When I executed the sample code provided at the bottom of the document, I found a case where the results were different. let n = 10_000 let x = (0..<n).map { _ in Float.random(in: 1 ... 10_000) } let y = x.map { return sqrt($0) } and let y = [Float](unsafeUninitializedCapacity: n) { buffer, initializedCount in vForce.sqrt(x, result: &buffer) initializedCount = n } The code below is provided to observe the issue described above. import Accelerate Task { let n = 1//10_000 let x = (0..<n).map { _ in Float(6737.015)//Float.random(in: 1 ... 10_000) } let y = x.map { return sqrt($0) } try? await Task.sleep(nanoseconds: 1_000_000_000) let z = [Float](unsafeUninitializedCapacity: n) { buffer, initializedCount in vForce.sqrt(x, result: &buffer) initializedCount = n } } For a value of 6737.015 when calculating the square root: Using the sqrt(_:) function gives the result 82.07932, While using the vForce.sqrt(_:result:) function gives the result 82.07933. Using a calculator, the value comes out as 82.07932139, which shows that the result from vForce is incorrect. Could you explain the reason behind this difference?