xcode github

I spent the entire day debugging a network issue on my Apple Watch app, only to realize the problem isn't my code—it's Apple's inflexible design. The Context: I am building a generic MCP (Model Context Protocol) client for watchOS. The nature of this app is to allow users to input their own server URLs (e.g., a self-hosted endpoint, or public services like GitHub's MCP server) to interact with LLMs and tools. The Problem: When using standard URLSession to connect to widely trusted, public HTTPS endpoints (specifically GitHub's official MCP server at https://mcp.github.com), the connection is forcefully terminated by the OS with NSURLErrorDomain Code=-1200 (TLS handshake failed). The Analysis: This is caused by App Transport Security (ATS). ATS is enforcing a draconian set of security standards (specific ciphers, forward secrecy requirements, etc.) that many perfectly valid, secure, and globally accepted servers do not strictly meet 100%. The Absurdity: We cannot whitelist domains: Since this

Hello, Is there any plan to integrate realtime sanitier (RTSan) in Apple Clang? This tool has been introduced in Clang 20.1.0 back in March 2025. This tool is incredibly useful, and having it in AppleClang (and ideally integrated in Xcode and xcodebuild) would be great. Thank you, Best, Thomas

Where are the instructions for this process

I have a ScrollView with several Buttons and a .refreshable modifier. As soon as I pull to refresh and the refresh indicator appears, the tap targets no longer match the visible button positions. For example, tapping button A triggers button B’s action, as if the hit-testing region didn’t move along with the content. This only happens while the refresh indicator is shown. Before pulling to refresh, everything is correct and afterwards as well. Tested on iOS 18 and 26 (Xcode 16.4, Xcode 26). Here is a minimal reproducible example: import SwiftUI struct ContentView: View { var body: some View { ScrollView { VStack { Button(Button A) { print(A) } .buttonStyle(.borderedProminent) Button(Button B) { print(B) } .buttonStyle(.borderedProminent) Button(Button C) { print(C) } .buttonStyle(.borderedProminent) Button(Button D) { print(D) } .buttonStyle(.borderedProminent) Button(Button E) { print(E) } .buttonStyle(.borderedProminent) } .frame(maxWidth: .infinity) } .refreshable { try? await Task.sleep(

I’m developing a share extension for iOS 26 with Xcode 26. When the extension’s sheet appears, it always shows a full white background, even though iOS 26 introduces a new “Liquid Glass” effect for partial sheets. Expected: The sheet background should use the iOS 26 glassmorphism effect as seen in full apps. Actual behavior: Custom sheets in my app get the glass effect, but the native system sheet in the share extension always opens as plain white. Steps to reproduce: Create a share extension using UIKit Present any UIViewController as the main view Set modalPresentationStyle = .pageSheet (or leave as default) Observe solid white background, not glassmorphism Sample code: swift override func viewDidLoad() { super.viewDidLoad() view.backgroundColor = .clear preferredContentSize = CGSize(width: UIScreen.main.bounds.width, height: 300) } Troubleshooting attempted: Tried adding UIVisualEffectView with system blur/materials Removed all custom backgrounds Set modalPresentationStyle explicitly Questions: Is

Hi everyone, In the previous Xcode version, I was using Intelligence with ChatGPT to refactor and fix code. It worked great: it applied changes across files in a few seconds and the code preview updated quickly. In the new Xcode version, something strange is happening: • With Claude, responses are almost instant. • With ChatGPT, it starts fine but then it shows the entire file instead of just the relevant section and, after a few lines, it starts streaming character by character, becoming extremely slow. This only started after updating Xcode. Is there any setting (streaming, token limits, formatting, etc.) that I can tweak to get ChatGPT back to the previous performance? Is anyone else seeing this issue with ChatGPT in Intelligence, or is it just my setup? Thanks in advance for any help!

I'm reading a bit outdated book about Cocoa/Objective-C applications development and I'm trying to code some examples. And so I ended up with an almost empty NSDocument-based app with runtime errors like that: Internal inconsistency in menus - menu Title: Window Supermenu: 0xbc726d080 (Main Menu), autoenable: YES Previous menu: 0x0 (None) Next menu: 0x0 (None) Items: ( , , , ) believes it has Title: Main Menu Supermenu: 0x0 (None), autoenable: YES Previous menu: 0x0 (None) Next menu: 0x0 (None) Items: ( ) as a supermenu, but the supermenu does not seem to have any item with that submenu It looks like as if the menu tries to include itself as a submenu. Am I right? I have no ideas what led to this. If I'm not mistaken this has started since macOS Tahoe. The code that is not a boilerplate one I have: // Document.m #import Document.h @implementation Document @synthesize text; - (NSString *)windowNibName { return @SampleDocument; } - (void)windowControllerDidLoadNib:(NSWindowController *)aController { [super w