Using SwiftData and this is the simplest example I could boil down: @Model final class Item { var timestamp: Date var tag: Tag? init(timestamp: Date) { self.timestamp = timestamp } } @Model final class Tag { var timestamp: Date init(timestamp: Date) { self.timestamp = timestamp } } Notice Tag has no reference to Item. So if I create a bunch of items and set their Tag. Later on I add the ability to delete a Tag. Since I haven't added inverse relationship Item now references a tag that no longer exists so so I get these types of errors: SwiftData/BackingData.swift:875: Fatal error: This model instance was invalidated because its backing data could no longer be found the store. PersistentIdentifier(id: SwiftData.PersistentIdentifier.ID(url: x-coredata://EEC1D410-F87E-4F1F-B82D-8F2153A0B23C/Tag/p1), implementation: SwiftData.PersistentIdentifierImplementation) I think I understand now that I just need to add the item reference to Tag and SwiftData will nullify all Item references to that tag when a Tag is deleted. But, the damage is already done. How can I iterate through all Items that referenced a deleted tag and set them to nil or to a placeholder Tag? Or how can I catch that error and fix it when it comes up? The crash doesn't occur when loading an Item, only when accessing item.tag?.timestamp, in fact, item.tag?.id is still ok and doesn't crash since it doesn't have to load the backing data. I've tried things like just looping through all items and setting tag to nil, but saving the model context fails because somewhere in there it still tries to validate the old value. Thanks!

Why did my sandbox test fail due to not being able to obtain the product list?

I use Homepod Mini as the gateway and have bound 9 Matter lights and 7 Matter switches. Each of my switches has 12 buttons, and each button supports three functions: single click, double click, and long press. Therefore, a total of 252 actions can be configured for 7 * 12 * 3. Currently, a total of 71 actions are configured for the 7 Matter switches. When configuring the 72nd action, the app will prompt that the operation cannot be completed. But if you delete a few previously configured actions, such as 68 actions, then you can configure 3 more actions (69, 70, 71). However, as long as you configure the 72nd action, the app will prompt that the operation cannot be completed, as if the available space is occupied. What is the reason for this?

There's plenty of articles out there about programatically grouping push notifications. However I have tried setting the thread-id in the push payload when sending a push, or setting the threadIdentifier for a received push in a notification service extension to be the same for several pushes. But if within the iPhone Settings / Notifications the user selects to display pushes as List and turns off Notification Grouping, then each notification resulting from the push appears on its own separately. Is there something other than thread-id/threadidentifier that is used to programmatically group them? If not then whats the point of these as grouping and display is actually under the control of user.

I keep seeing a red action item to complete a task in settings. IOS beta keeps indicating that I need to enable sensor & usage data. Everything is enabled. I literally enabled everything to get it to go away, but it is still there. Suggestions?

I'm sending push notifications to a notification extension, and within the extension setting the threadIdentifier to be the same. But I'm observing inconsistent grouping behaviour, and behaviour that changes over time. The general iPhone settings are to display notifications as a Stack, and the app settings are to show on lock screen, notification center and banners and the notification grouping is set to by app (changing it to automatic doesn't affect the behaviour below). Pushes are displayed on the lock screen grouped together, then if the device is roused and the screen swiped down to reveal the notification center then they are still grouped. So far so good. If the iphone is active then the notifications appear at the top of the screen, one by one, but in this case if there is a swipe down to reveal the notification center then the notifications are not grouped when displayed, but shown individually. But then if one waits a few minutes and then displays the notification center for a 2nd time, sometimes now they will be grouped, but sometimes not. Why are they not (always) being displayed as grouped in the notification center?

Hi! I use Tips with CloudKit and it works very well, however when a user want to remove their data from CloudKit, how to do that? In CoreData with CloudKit area, NSPersistentCloudKitContainer have purgeObjectsAndRecordsInZone to delete both local managed objects and CloudKit records, however there is no information about the TipKit deletion. Does anyone know ideas?

After updating to ios18.4, 3d scanning function, including AR function in apple clips, cannot be used. Does anyone else have the same problem?

We are currently testing the implementation of our Location Push Service Extension (LPSE) in both Ad Hoc and Release environments. We have encountered an issue where LPSE notifications, which were previously working correctly, suddenly fail to be delivered on some devices. After a period of several hours, the notifications resume arriving, but the issue remains intermittent. Notably, during these periods of suspected delivery restriction, regular push notifications (e.g., those using apns-push-type: alert) are delivered and displayed without any problem. [Detailed Situation] Test Environment and Scope We are testing LPSE after obtaining the necessary entitlements, in both Ad Hoc and Release environments. The issue is not observed on all test devices; only certain devices are affected. Observed Behavior Under normal circumstances, LPSE notifications are received and the extension is activated; however, on some devices the notifications suddenly stop arriving. During these periods, even when sending notifications with apns-push-type: location directly via the CloudKit Push Notification Console, no response is observed on the affected devices. The APNs server (api.push.apple.com) always returns a 200 OK response via HTTP/2, and our server-side logs and configurations (DNS resolution performed on every request, using the same JWT token for 59 minutes per session, communication via HTTP/2 with ALPN Protocol: h2) show no issues. Other app functionalities (network communication, UI responsiveness, etc.) work normally. Sending content When sending notifications from our server to APNs (api.push.apple.com), we use the following configuration (over HTTP/2): const payload = { aps: { 'content-available': 1 } }; const headers = { ':method': 'POST', ':path': /3/device/${apnsToken}, 'Authorization': bearer ${jwtToken}, 'apns-topic': 'ot.Here.location-query', 'apns-priority': '10', 'apns-push-type': 'location', 'Content-Type': 'application/json' }; We perform DNS resolution for every request, use the same JWT token for a 59-minute period per session, and communicate via HTTP/2 with ALPN Protocol: h2. Hypothesis on the Cause We suspect that due to an implementation issue, silent push notifications (using content-available: 1) were being sent every few minutes concurrently, which may have triggered an APNs delivery restriction (rate limiting). As a countermeasure, we have completely stopped sending silent pushes and any other background notifications aside from LPSE; however, the issue persists. Additionally, even after resetting affected devices, the delivery problem continues to occur. [Questions for Diagnosis] Given the above situation, is it reasonable to suspect that excessive silent push notifications have triggered an APNs delivery restriction? Does such a silent push restriction affect LPSE notifications (i.e., those sent with apns-push-type: location)? Do APNs delivery restrictions persist even after a device has been reset? Can a high volume of LPSE notifications alone (without silent pushes) also trigger a delivery restriction? → This is our primary concern since it poses a significant implementation challenge. Please let us know if any additional information is required for diagnosis.

Sorry but I'm not good at English. Is the originalTransactionId received for the same subscription by the same user the same in Store Kit v1 and Store Kit 2? Also, will the originalTransactionId be the same even if the subscription price is changed?

Hi, I'm developing a bluetooth peripheral. During factory reset I generate a new IRK and drop any bond information it has. When I then try to bond again with the device iOS returns bonding failed with 0x08 (unspecified reason). I assume that iOS somehow still thinks my reset device is the same as the device it has already bonded with. What information about the device is it using to draw this conclusion? Bonding works if I remove the "pre-reset" peripheral from the iOS list of bluetooth devices. Regards

I’ve created several shortcuts that tell me the stock price of a given company. The shortcut queries Yahoo Finance using Get Contents of URL, with the URL https://finance.yahoo.com/quote/TICKER SYMBOL/, for example https://finance.yahoo.com/quote/PLTR/ for Palantir or https://finance.yahoo.com/quote/AAPL/ for Apple, etc. Then it uses RegEx to parse out the numbers which it then formats and displays in a notification. Simple. It works great for several stocks, but for some reason, it does not work correctly for Palantir. It shows an older “previous close” price. Oddly, when I go to the website myself, it shows me the current stock price. So for today Mar 21 https://finance.yahoo.com/quote/PLTR/ shows me $90.96 (correct) but the shortcut, via Get Contents of URL, shows $87.39 (incorrect). This $87.39 price is listed further down in the page as a "previous close” price. I don’t get it. Here is a link to my Palantir shortcut: https://www.icloud.com/shortcuts/edea6ee0261245f49b078efc74d632dd Here is a link to my Apple shortcut: https://www.icloud.com/shortcuts/54a416393203432aa356fe76373e3f8b So the question is, why does Get Contents of URL show an old stock price but when I go to the site myself, it shows the correct stock price … and only for Palantir? I have about six shortcuts running correctly. Palantir is the only one that does not work. Been banging my head on this one for weeks. Any advice would be much appreciated. Thank you, Rob

I've successfully started the Live Caller ID Lookup example and initialized the PIRService. I added several identities to the input.txtpb file, some with block: true and others with block: false. Here is the file but modified phone digits: identities { key: "+40790123123" value { name: "Blocking 1" cache_expiry_minutes: 7 block: true } } identities { key: "+972526111111" value { name: "Blocking 2" cache_expiry_minutes: 7 block: true } } identities { key: "+123" value { name: "Adam" cache_expiry_minutes: 8 block: false category: IDENTITY_CATEGORY_PERSON } } identities { key: "+972526111112" value { name: "Identified Business Name 1" cache_expiry_minutes: 1 block: false category: IDENTITY_CATEGORY_BUSINESS } } identities { key: "+972526111113" value { name: "Identified Business Name 2" cache_expiry_minutes: 1 block: false category: IDENTITY_CATEGORY_BUSINESS } } The main issue is that only the number marked as +40790123123 was actually blocked, while "Blocking 2" appeared as identified contacts with their assigned name displayed. Notably, the only blocked number was a foreign number with a different country code than the number being called. The other numbers belonged to the same country. Can someone clarify whether this is a bug in the example project or an issue with the data file?

Continuing with my investigations of several issues that we have been noticing in our testing of the JDK with macosx 15.x, I have now narrowed down at least 2 separate problems for which I need help. For a quick background, starting with macosx 15.x several networking related tests within the JDK have started failing in very odd and hard to debug ways in our internal lab. Reading through the macos docs and with help from others in these forums, I have come to understand that a lot of these failures are to do with the new restrictions that have been placed for "Local Network" operations. I have read through https://developer.apple.com/documentation/technotes/tn3179-understanding-local-network-privacy and I think I understand the necessary background about these restrictions. There's more than one issue in this area that I will need help with, so I'll split them out into separate topics in this forum. That above doc states: macOS 15.1 fixed a number of local network privacy bugs. If you encounter local network privacy problems on macOS 15.0, retest on macOS 15.1 or later. We did have (and continue to have) 15.0 and 15.1 macos instances within our lab which are impacted by these changes. They too show several networking related failures. However, I have decided not to look into those systems and instead focus only on 15.3.1. People might see unexpected behavior in System Settings > Privacy & Security if they have multiple versions of the same app installed (FB15568200). This feedback assistant issue and several others linked in these documentations are inaccessible (even when I login with my existing account). I think it would be good to have some facility in the feedback assistant tool/site to make such issues visible (even if read-only) to be able to watch for updates to those issues. So now coming to the issue. Several of the networking tests in the JDK do mulicasting testing (through BSD sockets API) in order to test the Java SE multicasting socket API implementations. One repeated failure we have been seeing in our labs is an exception with the message "No route to host". It shows up as: Process id: 58700 ... java.net.NoRouteToHostException: No route to host at java.base/sun.nio.ch.DatagramChannelImpl.send0(Native Method) at java.base/sun.nio.ch.DatagramChannelImpl.sendFromNativeBuffer(DatagramChannelImpl.java:914) at java.base/sun.nio.ch.DatagramChannelImpl.send(DatagramChannelImpl.java:871) at java.base/sun.nio.ch.DatagramChannelImpl.send(DatagramChannelImpl.java:798) at java.base/sun.nio.ch.DatagramChannelImpl.blockingSend(DatagramChannelImpl.java:857) at java.base/sun.nio.ch.DatagramSocketAdaptor.send(DatagramSocketAdaptor.java:178) at java.base/java.net.DatagramSocket.send(DatagramSocket.java:593) (this is just one example stacktrace from java program) That "send0" is implemented by the JDK by invoking the sendto() system call. In this case, the sendto() is returning a EHOSTUNREACH error which is what is then propagated to the application. The forum text editor doesn't allow me to post long text, so I'm going to post the rest of this investigation and logs as a reply.

Hello! I'm working with universal links in my app and have configured the /.well-known/apple-app-site-association file. Currently, I use the paths array in this file to define URL routing rules. However, I’m struggling to find up-to-date documentation on the pattern syntax supported by the paths field. "paths": [ "/page/*", "NOT /page/*/subpage" ] Could someone clarify: Is the paths array still officially supported, or is it deprecated in favor of the newer components dictionary (as referenced here https://developer.apple.com/documentation/bundleresources/applinks/details-swift.dictionary/components-swift.dictionary)? If paths is still valid, where can I find documentation for its pattern-matching capabilities? I want to ensure my implementation aligns with Apple’s current best practices. Thank you!

What is the version policy for the Apple Pay SDK Javascript ? The documentation refers to this link : https://applepay.cdn-apple.com/jsapi/1.latest/apple-pay-sdk.js The future updates will overrride the file on that link ? Is there a way to be notified of any changes ? We are using a previous version named v1 : https://applepay.cdn-apple.com/jsapi/v1/apple-pay-sdk.js What are the risks not using changing to the lastesdt link ? Thank you for your help.

I want a user to be able to save a URL of a folder on a cloud share using the standard FileManager APIs. I'm testing with Dropbox in particular. The initial interaction is working (I can select a folder, save it as a bookmark, and scan the files in that folder). However, no changes made externally to the folder will be reflected in the app when I refresh the contents. Launching the Files app and browsing to the folder DOES show the updated contents, and once that step is complete, then my app will again show up-to-date contents. Is there perhaps some API I should be calling to trigger the Dropbox File Provider extension to update it's cache? Sample project demonstrating issue: https://github.com/dhennessy/FolderScan STEPS TO REPRODUCE Launch the app on a physical device Tap Choose, browse to folder on a Dropbox share, tap Open to select The app will show the contents of the folder (the 'test' folder) Switch to the Dropbox app and create a new subfolder of the test folder Return to the test app and tap Refresh. Notice that the changes do not appear Re-launching the app also does not show the changes Workaround Launch the Files app (or re-open the UIDocumentPickerViewController by tapping choose and then dismiss it) Tap Refresh and the changes will appear in the app Note: None of the other 'cloud file providers' (google drive, one drive, box) even allow the user to even select a folder.

Hi, I want to offer an auto-renewable subscription (e.g., $1/month) that grants users (10 document analyses per month), with the count resetting at the start of each billing cycle. -Unused analyses will not roll over to the next month- Additionally, any analyses generated while the subscription is active will remain accessible to the user permanently, even if they cancel the subscription. The paywall, app description, and metadata will clearly state that the subscription grants (10 document analyses per month with no rollover) We want this to be implemented as an auto-renewable subscription model, not as a consumable service or a token/credit system (which we want to avoid). Is this model acceptable under Apple’s guidelines, or would it be considered a token/credit system? Any insights or alternative suggestions would be appreciated.. Thanks

Hi, We are trying to make the PKAddPaymentPassViewController to show the correct list of devices to where the pass can be added. We have analysed the documentation and we are using the PrimaryAccountIdentifier field which is the field that supposedly controls this behavior but the list of devices presented in the view controller always include one iPhone and one Apple Watch, regardless of where the card has been already added. We are initializing the PKAddPaymentPassRequestConfiguration object with: PKEncryptionScheme PrimaryAccountIdentifier CardholderName PrimaryAccountSuffix LocalizedDescription PaymentNetwork PrimaryAccountIdentifier CardholderName PrimaryAccountSuffix LocalizedDescription We have also verified the configuration in our payment pass processor and everything should be ok. We would like to have some help on achieving the desired flow for Apple Pay, which is to present the PKAddPaymentPassViewController with the correct list of available devices and not the full list. Thank you.

After copying and inserting instances I am getting strange duplicate values in arrays before saving. My models: @Model class Car: Identifiable { @Attribute(.unique) var name: String var carData: CarData func copy() -> Car { Car( name: "temporaryNewName", carData: carData ) } } @Model class CarData: Identifiable { var id: UUID = UUID() var featuresA: [Feature] var featuresB: [Feature] func copy() -> CarData { CarData( id: UUID(), featuresA: featuresA, featuresB: featuresB ) } } @Model class Feature: Identifiable { @Attribute(.unique) var id: Int @Attribute(.unique) var name: String @Relationship( deleteRule:.cascade, inverse: \CarData.featuresA ) private(set) var carDatasA: [CarData]? @Relationship( deleteRule:.cascade, inverse: \CarData.featuresB ) private(set) var carDatasB: [CarData]? } The Car instances are created and saved to SwiftData, after that in code: var fetchDescriptor = FetchDescriptor<Car>( predicate: #Predicate<Car> { car in car.name == name } ) let cars = try! modelContext.fetch( fetchDescriptor ) let car = cars.first! print("car featuresA:", car.featuresA.map{$0.name}) //prints ["green"] - expected let newCar = car.copy() newCar.name = "Another car" newcar.carData = car.carData.copy() print("newCar featuresA:", newCar.featuresA.map{$0.name}) //prints ["green"] - expected modelContext.insert(newCar) print("newCar featuresA:", newCar.featuresA.map{$0.name}) //prints ["green", "green"] - UNEXPECTED! /*some code planned here modifying newCar.featuresA, but they are wrong here causing issues, for example finding first expected green value and removing it will still keep the unexpected duplicate (unless iterating over all arrays to delete all unexpected duplicates - not optimal and sloooooow).*/ try! modelContext.save() print("newCar featuresA:", newCar.featuresA.map{$0.name}) //prints ["green"] - self-auto-healed??? Tested on iOS 18.2 simulator and iOS 18.3.1 device. Minimum deployment target: iOS 17.4 The business logic is that new instances need to be created by copying and modifying previously created ones, but I would like to avoid saving before all instances are created, because saving after creating each instance separately takes too much time overall. (In real life scenario there are more than 10K objects with much more properties, updating just ~10 instances with saving takes around 1 minute on iPhone 16 Pro.) Is this a bug, or how can I modify the code (without workarounds like deleting duplicate values) to not get duplicate values between insert() and save()?