Summary On Mac Studio systems (no built-in camera), macOS does not initialize camera services after a normal reboot if no physical camera is present. As a result, Continuity Camera does not appear anywhere in the system. Observed behavior System Information → Camera reports “No video capture devices were found.” Continuity Camera (iPhone) is completely absent from camera lists. Plugging in any USB UVC webcam immediately initializes camera services and causes both the USB camera and the iPhone (Continuity Camera) to appear. The USB camera can then be unplugged and Continuity Camera continues working until the next reboot. Reproduction steps Use a Mac Studio (no built-in camera) on recent macOS. Ensure no USB webcam or external camera is connected. Reboot the Mac normally. After login, open System Information → Camera. Expected Camera services should initialize even when no physical camera is present, allowing Continuity Camera to be available as the primary camera. Actual No camera devices are present unless a physical USB camera is connected at least once after boot. This reproduces 100% of the time on Mac Studio and appears to be a camera service bootstrap issue where Continuity Camera cannot be the first camera device. Issue has been filed via Feedback Assistant.

Which HomeKit API serves for the Home application scene (HMActionSet)-related functionality “Remove from Home View” and “Add to Home View”? There must be a public API for that, for at the very least one 3rd party application shows/hides scenes appropriately as they are set up in Home; nevertheless, whatever I try, I can't find the API. Thanks!

My team has developed an app with a biref Matter commissioner feature using the Matter framework on the MatterSupport extension. Our app support iOS and Android. However, we ran into a problem that the control certificate generated by the iOS app could not control the device on the Android side. And the control certificate generated by the Android app could not control the device on the iOS side. The Matter library used by Android is compiled by connectedhomeip. Does anyone have the same problem as us? How to solve this? Thank you

My MacBook Pro M5 running MacOS Tahoe 26.3 beta fails to detect two identical ASUS ROG Swift OLED PG32UCDM monitors simultaneously. Only one display is recognized at a time. One potential root cause might be that both monitors report identical binary EDID serial numbers (0x01010101), and the MacBook Pro M5 appears to use this value exclusively for display identity rather than combining it with other more detailed information (e.g., port, or alphanumeric serial number). I've verified that the monitor EDID binary serial numbers are in fact identical -- however the alphanumerical serial numbers are not identical. NOTE: This behavior is specific to the MacBook Pro M5 — when connecting both monitors via usb-c to a Mac Mini M4 Pro running the same MacOS Tahoe 26.3 beta, the monitors work fine. The OS detects both and assigns different names to them (PG32UCDM (1) and PG32UCDM (2)). NOTE: I could be wrong about this root cause, I don't have a way to disprove it, though the fact the monitors work fine on a Mac Mini is suspicious. What I have tried: Connecting the two monitors using different monitor ports (one on DisplayPort, another on HDMI, etc.), and different MacBook ports (one on HDMI, another on USB-C, etc.) Bumping down the resolution on the monitors to "1920x1080 (low resolution)" and 30Hz to rule out bandwidth issues. Connecting one, or both, monitors to CalDigit TS5 Plus dock. Neither alternate configuration yields the device recognizing both screens. Using BetterDisplay to import a manually-edited EDID for the screen, with a different binary EDID value, manufacturer name, etc. I've also verified that if I plug in my Apple Studio Display as one of the monitors, then the MacBook recognizes both one of the PG32UCDM monitors and the Studio Display at the same time. The issue seems to occur only when both monitors plugged into it are the same PG32UCDM model. When I have both monitors plugged into my MacBook, each time I disconnect the cable to whichever monitor is currently recognized, it immediately recognizes the other monitor. Plugging the cable for the disconnected monitor back in has no effect. I'm at a loss. Has anyone run into this issue and found a successful workaround that is not one of the approaches I've described above?

Are there some undocumented (or, well, documented, but overlooked by me) prerequisites to the readValueWithCompletionHandler: method? The reason I ask is that occasionally I am getting the Read/Write operation failed error in the callback, even in cases where direct, non-deferred reading of the value worked properly. It seems to happen very consistently with some accessories and characteristics, not randomly; thus, it is not likely a temporary quirk in the communication with the device. Probably I am overlooking something of importance, but it does not make a good sense to me. My code (is it right, or can you see anything wrong in there?) // in an HMCharacteristic category if ([self.properties containsObject:HMCharacteristicPropertyReadable]) { id val=self.value, ident=[NSString stringWithFormat:@" [%@] %@ (%@)", self.uniqueIdentifier, self.localizedDescription, self.service.accessory.name]; NSLog(@"nondeferred '%@'%@", val, ident); if (self.service.accessory.reachable) { [self readValueWithCompletionHandler:^(NSError * _Nullable error) { if (error) NSLog(@"deferred ERROR %@ -> %@", ident, error); else NSLog(@"deferred '%@'%@", self.value, ident); }]; } } for most accessories/characteristics works properly, but for some of them I am consistently getting results like nondeferred '70.5' [64998F70-9C11-502F-B8B4-E99DC5C3171B] Current Relative Humidity (Vlhkoměr TH) deferred '70.5' ERROR [64998F70-9C11-502F-B8B4-E99DC5C3171B] Current Relative Humidity (Vlhkoměr TH) -> Error Domain=HMErrorDomain Code=74 "Read/Write operation failed." UserInfo={NSLocalizedDescription=Read/Write operation failed.} Do I do something wrong in my code, or is that normal with some devices? If the latter, is there perhaps a way to know beforehand that I should not use readValueWithCompletionHandler: (for it is bound to fail anyway), and instead I should simply use self.value non-deferred? For some time it seemed to me it happens with bridged accessories, but not really, this hypothesis proved wrong by further testing. Thanks!

I am developing a standard UAC 2.0 device and encountered an issue where the channel names do not update according to the iChannelNames field in the Class Specific AS Interface Descriptor when switching between different channel counts. For example: AS1 (6 channels) is configured with the following channel names: ADAT 1, ADAT 2, ADAT 3, ADAT 4, HP L, HP R AS2 (4 channels) is configured with: ADAT 1, ADAT 2, HP L, HP R However, when switching from AS1 (6 channels) to AS2 (4 channels), the channel names displayed in Audio MIDI Setup do not reflect the change as expected. The actual result is: ADAT 1, ADAT 2, ADAT 3, ADAT 4 The system simply hides the last two channels; the names of the remaining channels are not updated. Initial Topology My original topology was as follows: Later, I discovered that macOS uses the iChannelNames field from the Input Terminal to display channel names. Therefore, I modified the USB device descriptors and updated the topology to the following: To distinguish the channel names for different channel counts, each Input Terminal is assigned a unique iChannelNames value. This method worked perfectly on macOS 15. However, after updating to macOS 26, this topology no longer displays the correct channel names. Question On macOS 26, what is the correct method to ensure that the channel names update dynamically when switching between different audio channel configurations?

I am developing a standard UAC 2.0 device and encountered an issue where the channel names do not update according to the iChannelNames field in the Class Specific AS Interface Descriptor when switching between different channel counts. For example: AS1 (6 channels) is configured with the following channel names: ADAT 1, ADAT 2, ADAT 3, ADAT 4, HP L, HP R AS2 (4 channels) is configured with: ADAT 1, ADAT 2, HP L, HP R However, when switching from AS1 (6 channels) to AS2 (4 channels), the channel names displayed in Audio MIDI Setup do not reflect the change as expected. The actual result is: ADAT 1, ADAT 2, ADAT 3, ADAT 4 The system simply hides the last two channels; the names of the remaining channels are not updated. Initial Topology My original topology was as follows: Later, I discovered that macOS uses the iChannelNames field from the Input Terminal to display channel names. Therefore, I modified the USB device descriptors and updated the topology to the following: To distinguish the channel names for different channel counts, each Input Terminal is assigned a unique iChannelNames value. This method worked perfectly on macOS 15. However, after updating to macOS 26, this topology no longer displays the correct channel names. Question On macOS 26, what is the correct method to ensure that the channel names update dynamically when switching between different audio channel configurations?

Hi everyone, We are currently exploring ways to implement a frictionless Wi-Fi setup for our hardware devices without requiring a dedicated third-party application. We are interested in leveraging Apple's WAC (Wireless Accessory Configuration) to sync Wi-Fi credentials directly from iOS devices. However, we have struggled to find comprehensive technical documentation or specifications regarding the WAC service. Could anyone point us to the official source for these materials? Additionally, we have a couple of technical questions: 1.We are testing WAC provisioning and found that the Home app can discover our device and successfully get it online. However, it always ends with a "Failed to add accessory" message. Does WAC support imply that a device should be addable via the Home app? If not, why is the Home app able to discover and start the setup for a non-HomeKit WAC device? 2. Our device is already Apple AirPlay certified. Does implementing WAC require additional standalone certification, or is it covered under the existing MFi/AirPlay certification umbrella? Any insights or guidance would be greatly appreciated. Thank you!

With the same firmware, OTA testing on the DCL test network was successful in September 2025, and the Home app was able to deliver software update notifications. Since the beginning of 2026, however, the Home app no longer delivers software update notifications. This is bug number: FB21922369

Hello, I am an individual developer working on a macOS application using SwiftUI and RealityKit. I would like to understand the feasibility of face-related tracking on macOS when using an external USB camera, compared to iOS/iPadOS. Specifically: • Does macOS provide an ARKit Face Tracking–equivalent API (e.g., real-time facial expressions, gaze direction, depth)? • If not, is it common to rely on Vision / AVFoundation as alternatives for: • Facial expression coefficients • Gaze estimation • Depth approximation • In an environment without dedicated sensors such as TrueDepth, is it correct to assume that accurate depth data and high-fidelity blend shape extraction are realistically difficult? Any clarification on official limitations, recommended alternatives, or relevant documentation would be greatly appreciated. Thank you.

Environment: iPhone 17 / iPhone 17 Pro (Apple N1 chip) iOS 26.x Xcode 26 Framework: Flutter app with native iOS BLE library (CoreBluetooth) We have a production IoT app that communicates with BLE nodes (Nordic, PIC, EnOcean peripherals) using an advertising/scanning-based protocol — not GATT connections. The app broadcasts commands via CBPeripheralManager (advertising service UUIDs) and receives responses by scanning with CBCentralManager (reading manufacturer data and service UUIDs from advertisement packets). This workflow has been reliable across all iPhone models from iPhone 8 through iPhone 16 Pro Max. On iPhone 17 devices, we are experiencing multiple failures in this workflow. Architecture: Sending commands: We use CBPeripheralManager.startAdvertising() with CBAdvertisementDataServiceUUIDsKey to broadcast a UUID-encoded command to nearby nodes. Receiving responses: We use CBCentralManager.scanForPeripherals(withServices: nil, options: [CBCentralManagerScanOptionAllowDuplicatesKey: true]) and filter responses in centralManager(_:didDiscover:advertisementData:rssi:) by matching CBAdvertisementDataServiceUUIDsKey or CBAdvertisementDataManufacturerDataKey against expected UUID masks. Communication pattern: Advertise a command → stop advertiser → start scanner → wait for matching response → process result. Typical timeout is 1.5 seconds per exchange. Issues observed on iPhone 17: peripheralManagerDidStartAdvertising behaviour change After calling CBPeripheralManager.startAdvertising(:), the delegate callback peripheralManagerDidStartAdvertising(:error:) either fires with errors that did not occur on previous hardware, or advertising does not appear to reach the peripheral nodes at all. The same advertising payload works immediately when tested on iPhone 15/16. Is the N1 chip's Bluetooth 6 stack handling CBAdvertisementDataServiceUUIDsKey advertising differently? Are there new constraints on advertising payload size or format? Scanner returning fewer/no results with withServices: nil Our scanner uses scanForPeripherals(withServices: nil) because we need to read manufacturer data from advertisement packets and filter using a custom UUID mask. On iPhone 17, we observe significantly fewer didDiscover callbacks compared to iPhone 15/16 in the same physical environment, with the same nodes advertising. We understand that passing service UUIDs in withServices: is recommended, but our protocol requires reading raw manufacturer data bytes that aren't associated with a single service UUID — we use mask-based matching (e.g., filter mask 11110000-0000-0000-0000-000000000000 against scan results). Has the N1 chip changed the rate or filtering behaviour of unfiltered BLE scans? Is there a new throttling mechanism? Background scanning stops immediately When the app moves to background, scanning appears to stop entirely on iPhone 17 — even with bluetooth-central in UIBackgroundModes. On iPhone 16, background scanning continued (at reduced intervals) and delivered results for peripherals advertising filtered service UUIDs. Aggressive session termination on app backgrounding Our advertise-then-scan sequences (typically 1.5s round-trip) are being interrupted when the user briefly switches apps. The CBPeripheralManager stops advertising and the CBCentralManager stops scanning, causing timeout errors. This was not observed on previous iPhone models with the same iOS background mode configuration. Questions for Apple: Are there documented changes to CoreBluetooth behaviour on the N1 Bluetooth 6 chip that affect advertising-based (non-GATT) communication patterns? Has the scan response rate for scanForPeripherals(withServices: nil) been intentionally reduced on iPhone 17? Is CBCentralManagerOptionRestoreIdentifierKey now required for reliable background scanning on iPhone 17, or is this a known regression? Are there new advertising payload constraints (size, format, interval) that we should be aware of for the N1 chip? What we've tried: Added NSBluetoothAlwaysUsageDescription and NSBluetoothWhileInUseUsageDescription to Info.plist Confirmed Bluetooth permissions are granted Tested with identical BLE nodes that work on iPhone 15/16 Verified CBManagerState.poweredOn before all operations Any guidance or known workarounds would be greatly appreciated. Happy to provide sysdiagnose logs or a minimal reproducible sample project.

I have a BLE device which my app connects to and can detect button presses. On a button press, I want my app to start recording using the AudioRecordingIntent. But my app doesn't work and throws a background error. Is there any reliable way I can get the app to start recording audio in the background?

Hello, I am a Bluetooth Engineer at Google investigating an interoperability bug between an Android device and AirPods 4. When requesting an L2CAP connection (with PSM = AVDTP) to the AirPods during SDP service discovery, The AirPods L2CAP layer incorrectly responds with a "refused - no resources available" status followed by a Pending status and a Success status. This violates the specification, which says that the request has been fully rejected after the refused status and should not receive followup responses. I suspect the "no resources available" response is a bug. This prevents A2DP from working with the AirPods. This bug does not exist with AirPods 2 firmware. Here is a packet capture: 1602 1969-12-31 16:07:04.805261 0.062473 localhost () Apple_6b:db:09 (AirPods) L2CAP 17 Sent Connection Request (AVDTP, SCID: 0x22c6) 1603 1969-12-31 16:07:04.810953 0.005692 controller host HCI_EVT 8 Rcvd Number of Completed Packets 1604 1969-12-31 16:07:04.811078 0.000125 Apple_6b:db:09 (AirPods) localhost () SDP 27 Rcvd Service Search Attribute Request : Device Information: [Bluetooth Profile Descriptor List 0x0009] 1605 1969-12-31 16:07:04.821249 0.010171 localhost () Apple_6b:db:09 (AirPods) SDP 19 Sent Service Search Attribute Response 1606 1969-12-31 16:07:04.876396 0.055147 controller host HCI_EVT 8 Rcvd Number of Completed Packets 1607 1969-12-31 16:07:04.876464 0.000068 Apple_6b:db:09 (AirPods) localhost () L2CAP 21 Rcvd Connection Response - Refused - no resources available (SCID: 0x22c6) 1608 1969-12-31 16:07:04.942539 0.066075 Apple_6b:db:09 (AirPods) localhost () SDP 41 Rcvd Service Search Attribute Request : Unknown: [Bluetooth Profile Descriptor List 0x0009] 1609 1969-12-31 16:07:04.951052 0.008513 localhost () Apple_6b:db:09 (AirPods) SDP 19 Sent Service Search Attribute Response 1610 1969-12-31 16:07:05.010605 0.059553 controller host HCI_EVT 8 Rcvd Number of Completed Packets 1611 1969-12-31 16:07:05.080593 0.069988 Apple_6b:db:09 (AirPods) localhost () SDP 27 Rcvd Service Search Attribute Request : GATT: [Bluetooth Profile Descriptor List 0x0009] 1612 1969-12-31 16:07:05.087636 0.007043 localhost () Apple_6b:db:09 (AirPods) SDP 19 Sent Service Search Attribute Response 1613 1969-12-31 16:07:05.209417 0.121781 controller host HCI_EVT 8 Rcvd Number of Completed Packets 1614 1969-12-31 16:07:05.279491 0.070074 Apple_6b:db:09 (AirPods) localhost () L2CAP 21 Rcvd Connection Response - Pending (SCID: 0x22c6) 1615 1969-12-31 16:07:05.280731 0.001240 Apple_6b:db:09 (AirPods) localhost () L2CAP 21 Rcvd Connection Response - Success (SCID: 0x22c6, DCID: 0x0406) Please file this bug with the AirPods Bluetooth team.

Right now, I am scanning for specific BLE peripherals with my iPad app, using this: [self.cbCentralManager scanForPeripheralsWithServices:serviceUUIDsToScanFor options:@{CBCentralManagerScanOptionAllowDuplicatesKey:@YES}]; I have the "CBCentralManagerScanOptionAllowDuplicatesKey" set true because I need to be able to detect when a peripheral is no longer advertising, so I capture each "didDiscoverPeripheral" callback and set a 3-second timer that notifies the user that that peripheral is no longer in range if another didDiscoverPeripheral hasn't been received in that time. The peripherals all advertise at 100ms intervals. What's weird is that if I leave the scan on for a long time, the advertising packets slow down, and eventually one of those timers times out, around about one or two minutes for the first instance, and then every 10-20 seconds after that. I've checked with ATS for all the BLE traffic, and there are indeed > 3-second gaps in the advertising packets that the iPad sees, so it's not my code introducing the gap. Is there some reason long-running scans should not be done on iPadOS (both 18 and 26.1 used)? I've tested out switching my scan to "stopScan" and restart it every 10 seconds, and that seems to have resolved the issue, but it's unclear why that would matter (and that does not seem like an appropriate use of the stop and start scans). Thanks!

Hi all, I’m facing a device-specific issue in a live production iOS app distributed privately via the App Store . The app crashes immediately after login on one client’s iPhone, while the same account works fine on other devices. There’s no crash log generated in Analytics, and the app just pops to the home screen. Environment: App: Production app on App Store iOS version: 26.3 Devices: Only one device exhibits the crash; other iPhones work fine Login flow: App calls an API and writes the response to a local SQLite database immediately after login Distribution: App Store (Privately). The user is install via the redemption codes. Observations: All users on the problematic device crash immediately after login. The crash does not occur on any other devices, including the same iOS version. The client had already uninstalled and reinstalled the app via App Store cloud download, but the crash persisted. No crash log appears in Analytics or Xcode (process just terminates). Device restart had not been attempted before reinstall. App does not use Keychain tokens; local DB is only SQLite in the app sandbox. Hypotheses so far: Corrupted binary or cached app installation on that device SQLite database corruption or write failure Device-specific OS/environment issue (temp files, file locks, provisioning) iOS watchdog silently terminating the app during post-login DB write Language / region differences unlikely Questions: Is it possible for a device to retain a corrupted app binary or cached installation even after uninstall + cloud download reinstall from the App Store? Can uninstalling, restarting the device, and reinstalling guarantee a fresh binary and sandbox? Are there any known iOS behaviors where a local SQLite write could trigger an instant crash on one device only, without generating crash logs? Any other suggestions for diagnosing this device-specific post-login crash in a live production environment? Thanks in advance for any guidance — this issue is affecting a client’s live usage, and we’d like to understand the root cause and best way to resolve it safely.

Hello Apple Developer Technical Support Team, I’m working on an iOS banking/security SDK and we’re trying to match an Android feature that reads payment cards via NFC (EMV). On Android, this is implemented using an NFC scanning screen (e.g., “NfcScanActivity”) that can read EMV data from contactless credit/debit cards. Could you please clarify the current iOS capabilities and App Store policy around this? On iOS, is it currently possible for a third-party App Store app to read contactless credit/debit cards using Core NFC (i.e., accessing EMV application data/AIDs from payment cards)? If this is possible, what are the supported APIs/frameworks and any entitlement requirements (if applicable)? If this is not possible for App Store apps, could you recommend the closest acceptable alternatives for achieving a similar user outcome? For example: Using Apple Pay / PassKit flows for payment-related experiences Card scanning alternatives (camera-based OCR) for capturing card details (if allowed) Using an external certified card reader accessory (MFi) and required approach/entitlements Any other Apple-recommended approach for “card verification / identification” without reading EMV NFC data Our goal is not to bypass security restrictions, but to provide a compliant solution on iOS comparable to Android’s NFC-based card reading, or to adopt an Apple-approved alternative if direct EMV reading is not supported. If helpful, I can share a brief technical summary of the Android behavior and the exact data we need to obtain (e.g., whether it’s card presence verification vs. reading specific EMV tags). Thank you for your guidance. Best regards, Imran

Hello Apple Developer Technical Support Team, I’m working on an iOS banking/security SDK and we’re trying to match an Android feature that reads payment cards via NFC (EMV). On Android, this is implemented using an NFC scanning screen (e.g., “NfcScanActivity”) that can read EMV data from contactless credit/debit cards. Could you please clarify the current iOS capabilities and App Store policy around this? On iOS, is it currently possible for a third-party App Store app to read contactless credit/debit cards using Core NFC (i.e., accessing EMV application data/AIDs from payment cards)? If this is possible, what are the supported APIs/frameworks and any entitlement requirements (if applicable)? If this is not possible for App Store apps, could you recommend the closest acceptable alternatives for achieving a similar user outcome? For example: Using Apple Pay / PassKit flows for payment-related experiences Card scanning alternatives (camera-based OCR) for capturing card details (if allowed) Using an external certified card reader accessory (MFi) and required approach/entitlements Any other Apple-recommended approach for “card verification / identification” without reading EMV NFC data Our goal is not to bypass security restrictions, but to provide a compliant solution on iOS comparable to Android’s NFC-based card reading, or to adopt an Apple-approved alternative if direct EMV reading is not supported. If helpful, I can share a brief technical summary of the Android behavior and the exact data we need to obtain (e.g., whether it’s card presence verification vs. reading specific EMV tags). Thank you for your guidance. Best regards, Anis