This is discussed here: https://developer.apple.com/forums/thread/801455 Essentially my users are finding there can be inconsistencies in the configurations offered for my app complications depending on if they are doing so on their Apple Watch directly or via the Watch Face Gallery in the iPhone Apple Watch app. Are there any new changes here or any best practices to follow when creating WidgetConfigurationIntents?

Does PaperKit expose the underlying PKDrawing / PKStroke / PKStrokePath data in a way that preserves stroke order, point sequence, and per-point properties, or is it primarily intended for markup persistence and rendering?

One screen of our app has a calculator interface that takes over the toolbar area to afford sufficient tap targets for buttons. (Just like Apple's Calculator app without interference from the clock.) watchOS 11: works watchOS 27 b1: works Clock avoidance does not work in watchOS 26. Is there a specific API call we can use for this one screen in 26?

Hi! I've seen during the Keynote that watchOS will support Siri AI (when your iPhone is nearby), which sounds really exciting. Out of curiousity, I also noticed that both Core AI, and now the Foundation Models framework, are available on watchOS 27. This sounds amazing - bringing intelligence features to the watch could enable some great tailored experiences on the go. Here are some of my questions: Do these require a nearby iPhone? Do they work offline (on-device)? Lastly, are there any device limitations? Also, if there are any additional details or fun facts about using Foundation Models or Core AI on watchOS (or perhaps iOS, if making use of the companion app to run those models on the iPhone), I’d love to hear them!

Moving the minimum OS version from 4 to 9, I see a message "WatchKit Settings Bundles Deprecated". Settings used to be handled by providing a plist which the Watch app on the iPhone would then use to display settings and allow the user to change them. How, nowadays, should we make settings available to the user to be viewed and changed?

Xcode 27 refuses to build common Swift Packages for watchOS, such as swift-algorithms or swift-collections. ... 'WATCHOS_DEPLOYMENT_TARGET' is set to 8.0, but the range of supported deployment target versions is 9.0 to 27.0.x. The error disappears by explicitly declaring in the SPM manifest "platforms: [.watchOS(.v9)]". (Patching the "SwiftStdlib" availabilities in swift-collections is insufficient.) Is there a better workaround than mirroring our whole SPM graph to declare the platform minimum? Our projects' minimums are correctly set at 11 or 26. (Also happens in the hello world Xcode 27 watchOS only project.) FB23037701 ref: https://developer.apple.com/forums/thread/829540

Our app displays a calendar of the week ahead. On tapping a date, it offers three pieces of text for that date. These pieces of text can be quite long. What is the best overall architecture? Reading the documentation and watching the videos, it seems to be a NavigationSplitView whose items are the dates, and whose detail is a TabView which in turn contains three TextViews. Is this right?

Many Swift Packages, including those from Apple, like swift-collections, don't compile with Xcode 27. It shows these error messages: error: The watchOS deployment target 'WATCHOS_DEPLOYMENT_TARGET' is set to 8.0, but the range of supported deployment target versions is 9.0 to 27.0.x. (in target 'DequeModule' from project 'swift-collections') error: The armv7k architecture is deprecated. You should update your ARCHS build setting to remove the armv7k architecture. (in target 'DequeModule' from project 'swift-collections') Is this an expected behavior? If so, how can we fix it? If not, when will this be fixed?

Let's take an example where a user is on an iPhone 12 running iOS 27. They have an Apple Watch Series 10 paired running WatchOS 27. Are they able to access SiriAI queries from the Apple Watch? If not is there a technical limitation on why this isn't possible? On Apple Watch models where SiriAI is fully supported, does it still have any on device capability or is it all now requiring Internet access? And can this be Internet over cellular without the paired iPhone nearby?

What is the recommended way to analyze PencilKit strokes in near real time without interfering with Apple Pencil latency? Should analysis occur only after canvasViewDrawingDidChange, after each stroke ends, or through lower-level touch handling?

Will the sample code of the "Unwrap PaperKit" session be made available?

On the new iOS 27 Beta, after updating- the Transactions messages are moved to trash and cannot be recovered

I wrote a python script called MXSymbolicate (https://github.com/OliveTreeBible/MXSymbolicate) to symbolicate the crash reports produced by the old version of MetricKit. The new MetricKit in the '27 releases appears to be basically the same data but in type-safe form, with the addition of state tracking. Notably, the new CrashReportExtension API seems to include the ability to symbolicate a crash stack on-device, but it looks like the CallStackTree type is still just binary names and frame offsets and such - no symbol names. Should I keep symbolicating crash reports from the new MetricKit APIs the same way I did the old ones? Or is the expectation that I should use the CrashReportExtension system for symbolicated crashes, and use the MetricKit crash reports for other purposes, like state tracking?

I am writing a QuickLook extension using Swift (specifically QLPreviewingController). Extensions can be opened from both the previewer (pressing Space with a file selected) and Get Info. How do I reliably detect which mode my extension is being started from? The problem is that when it's in the Get Info window, it forces a fixed width (but not height). If I set the size statically based on that width, then when it launches in its own window (via Space), you can't resize the preview window. What am I missing? It seems like I should be getting more context about where I'm being inserted so I can lay out accordingly? Does anyone have any suggestions or have faced this problem before?

Hi there, we deploy upwards of 12-15 accessories (containing BLE) at a time, in a single system instal. Can AccessorySetupKit be used to streamline the pairing process for all of these accessories at once, so that the user isn't required to step through the process for each individual accessory?

Hi there, we deploy upwards of 12-15 hardware accessories containing BLE at a time, in a single system instal. Can AccessorySetupKit be used to streamline the pairing process for all of these accessories at once, so that the user isn't required to step through the process of pairing with each individual accessory?

I forget which extension type it is, but there is one that can discover the devices in a sandbox. Xcode 26 shows me Media Device Discovery, Xcode 27 doesn't. It could be a different one. Which ever one it is, do you know off hand if the new StateReporting framework would report performance and usage privately to MetricKit? I know that some frameworks report data to Apple's analytics reports, I'd potentially want to capture some performance metrics and app state changes that are happening in supporting extensions.

Hello, Currently we have the sample project from WWDC2024 as an example on how to do it as a live example but only as a simulated project on the Accessory end. The example is good but being implemented a rich platform like iOS or iPadOS, the rich stack of APIs are provided. Embedded devices such as the ESP32/RaspberryPI do not have this rich API stack, such as CoreBluetooth and so on, so a mirror stack of API/functions must be implemented to give the equal experience. Are there any examples or design/technical guidelines for the Accessory end to implement AccessorySetupKit on embedded devices. (Like a list of technical requirements or checklist of functions/implementations/services we can go through on a ESP32/RapberryPI) to ensure that the Accessory has all the needed code/technical implementations with AccessorySetupKit for both Bluetooth and WiFi support, especially on the Bluetooth end. Best to have an ESP32 project, implemented in Embedded Swift but at least a checklist (of items/situations/error handling) to confirm that it works.

Like in the case that two products are advertising with the same company identifier, both matching a single display item passed in when displaying the accessory picker. Would it be expected for two items to show in the accessory picker, one for each advertising product?

What would you recommend to teams that want to act as both a central and peripheral role? I want to use the ASK permission model for Central mode, but doing so I can't build connections to Apple Watch and Apple Vision Pro that don't support peripheral mode forcing the iPhone to do that.