Overview

Hi, After the release of macOS Tahoe 26.2. We are seeing memory leaks if our Network Protection Extension is used alongside the Apple Built In Firewall, a second Security Solution that does Network Protection and a VPN. Our NEXT, socketfilterfw and the other security solution consume instead of a few MB of Memory now multiple Gigabytes of Memory. This issue started with the public release of macOS Tahoe 26.2, this issue was not present in earlier versions of macOS and the same set of Software. Just testing our solution by itself will not show this behavior. I unfortunately can't try to reproduce the issue on my test device that runs the latest 26.3 beta as I do not have the third party software installed there and I can't get it. Our Network extension implements depending on the license and enabled features: NEFilterDataProvider NEDNSProxyProvider NETransparentProxyProvider For all man in the middle Use Cases we are using Network Framework, to communicate with the peers. And leaks suggest that the there is a memory leak within internals of the Network Framework. Here is a shortened sample of the leaks output of our Network extension. However, the third party NEXT does show the same leaks. More details can be found on the Feedback with the ID FB21649104 snippet is blocking post? sensitive language Does anyone see similar issues or has an idea what could cause this issue, except a regression of the Network.framework introduced with macOS Tahoe 26.2? Best Regards, Timo

Hardware: • Mac Studio (M1 Ultra) • Apple Silicon • Sufficient free disk space (~35GB+ available) • SIP enabled • 4 local Time Machine snapshots present Current System: • Already running macOS Tahoe beta • Attempting upgrade to macOS Tahoe 26.4 Beta (Build 25E5207k) Primary Issue: Software Update downloads successfully, then moves to “Preparing…”. It stalls at “About 5 minutes remaining” for ~10–15 minutes and fails with: “Failed to prepare the software update. Please try again. An error occurred while downloading the selected updates.” The error appears network-related, but download always completes. Observed Behaviour: • “Checking for Updates” completes in <3 minutes. • softwareupdated and mobileassetd CPU usage <1%. • softwareupdate --history shows no record of the failed beta install. • No excessive system load or obvious disk pressure. Steps Already Attempted: 1. Standard Software Update retries • Multiple download/retry cycles from System Settings. • Same Preparing failure every time. 2. Permissions / Disk Access • Granted Terminal Full Disk Access. • SIP remains enabled. • Some system-protected locations still return “Operation not permitted”. 3. Cache / State Cleanup Attempts • Removed /Library/Updates/*. • Cleared temporary softwareupdate folders under /private/var/tmp. • Attempted MobileAsset cache cleanup where permitted. • Killed softwareupdated and mobileassetd processes (auto-restarted by launchd). 4. Beta Channel Reset • Turned Beta Updates OFF in System Settings. • Rebooted. • Re-enabled Beta Updates. • Forced catalog refresh via softwareupdate -l. 5. Snapshot Check • tmutil listlocalsnapshots / shows only 4 snapshots (normal range). 6. Full Installer Workaround softwareupdate --list-full-installers shows: • macOS Tahoe Beta 26.4 (25E5207k) • Tahoe 26.3 / 26.2 / 26.1 • Sequoia / Sonoma / Ventura / Monterey installers Downloaded installer via: softwareupdate –fetch-full-installer –full-installer-version 26.4 Installer downloaded correctly: /Applications/Install macOS Tahoe Beta.app 7. startosinstall Method (Apple Silicon optimized) Attempted install with: sudo “/Applications/Install macOS Tahoe Beta.app/Contents/Resources/startosinstall” –user –agreetolicense –nointeraction –forcequitapps –rebootdelay 30 –passprompt Authorization succeeds, installer begins preparing, but installation still ultimately fails. Additional Notes: • CPU usage remains low during failure. • No excessive snapshot count. • Error messaging consistently references download/network even though download completes. • Issue appears to occur during cryptex/asset preparation phase rather than initial download. Question for Apple / other developers: Has anyone else seen Tahoe 26.4 beta fail specifically during the Preparing phase on Apple Silicon (especially Ultra-class chips)? Looking for confirmation whether this is a known MobileAsset or cryptex staging issue, or if logs indicate a deeper APFS / volume-owner authorization problem.

Hello All, I’m working on a computer-vision–heavy iOS application that uses the camera, LiDAR depth maps, and semantic segmentation to reason about the environment (object identification, localization and measurement - not just visualization). Current architecture I initially built the image pipeline around CIImage as a unifying abstraction. It seemed like a good idea because: CIImage integrates cleanly with Vision, ARKit, AVFoundation, Metal, Core Graphics, etc. It provides a rich set of out-of-the-box transforms and filters. It is immutable and thread-safe, which significantly simplified concurrency in a multi-queue pipeline. The LiDAR depth maps, semantic segmentation masks, etc. were treated as CIImages, with conversion to CVPixelBuffer or MTLTexture only at the edges when required. Problem I’ve run into cases where Core Image transformations do not preserve numeric fidelity for non-visual data. Example: Rendering a CIImage-backed segmentation mask into a larger CVPixelBuffer can cause label values to change in predictable but incorrect ways. This occurs even when: using nearest-neighbor sampling disabling color management (workingColorSpace / outputColorSpace = NSNull) applying identity or simple affine transforms I’ve confirmed via controlled tests that: Metal → CVPixelBuffer paths preserve values correctly CIImage → CVPixelBuffer paths can introduce value changes when resampling or expanding the render target This makes CIImage unsafe as a source of numeric truth for segmentation masks and depth-based logic, even though it works well for visualization, and I should have realized this much sooner. Direction I’m considering I’m now considering refactoring toward more intent-based abstractions instead of a single image type, for example: Visual images: CIImage (camera frames, overlays, debugging, UI) Scalar fields: depth / confidence maps backed by CVPixelBuffer + Metal Label maps: segmentation masks backed by integer-preserving buffers (no interpolation, no transforms) In this model, CIImage would still be used extensively — but primarily for visualization and perceptual processing, not as the container for numerically sensitive data. Thread safety concern One of the original advantages of CIImage was that it is thread-safe by design, and that was my biggest incentive. For CVPixelBuffer / MTLTexture–backed data, I’m considering enforcing thread safety explicitly via: Swift Concurrency (actor-owned data, explicit ownership) Questions For those may have experience with CV / AR / imaging-heavy iOS apps, I was hoping to know the following: Is this separation of image intent (visual vs numeric vs categorical) a reasonable architectural direction? Do you generally keep CIImage at the heart of your pipeline, or push it to the edges (visualization only)? How do you manage thread safety and ownership when working heavily with CVPixelBuffer and Metal? Using actor-based abstractions, GCD, or adhoc? Are there any best practices or gotchas around using Core Image with depth maps or segmentation masks that I should be aware of? I’d really appreciate any guidance or experience-based advice. I suspect I’ve hit a boundary of Core Image’s design, and I’m trying to refactor in a way that doesn't involve too much immediate tech debt, remains robust and maintainable long-term. Thank you in advance!

its been a week i uploaded the documents for enrolment and there has been no reply. i have raised multiple concerns but am not receiving any resposes on them. is it normal?

Hi there, I recently updated to the latest version of macOS Tahoe 26.3. Since then, Xcode is not able to resolve my Swift Package dependencies anymore. We use SSH for all our Github hosted packages. When package resolution is running, we randomly have Failed to verify SSH fingerprint on some remote package. Nothing has changed and it worked perfectly with Tahoe 26.2. The SSH have been reconfigured and known hosts have been verified. Is something changed between those 2 versions or is it a bug ?

We are developing a video processing app that applies CIFilter chains to video frames. To not force the user to keep the app foregrounded, we were happy to see the introduction of BGContinuedProcessingTask to continue processing when backgrounded. With iOS 26, I was excited to see the com.apple.developer.background-tasks.continued-processing.gpu entitlement, which should allow GPU access in the background. Even the article in the documentation provides "exporting video in a film-editing app" or "applying visual filters (HDR, etc) or compressing images for social media posts" as use cases. However, when I check BGTaskScheduler.shared.supportedResources.contains(.gpu) at runtime, it returns false on every iPhone I've tested (including iPhone 15 Pro and iPhone 16 Pro). From forum responses I've seen, it sounds like background GPU access is currently limited to iPad only. If that's the case, I have a few questions: Is this an intentional, permanent limitation — or is iPhone support planned for a future iOS release? What is the recommended approach for GPU-dependent background work on iPhone? My custom CIKernels are written in Metal (as Apple recommends since CIKL is deprecated), but Metal CIKernels cannot fall back to CPU rendering. This creates a situation where Apple's own deprecation guidance (migrate to Metal) conflicts with background processing realities (no GPU on iPhone). Should developers maintain deprecated CIKL kernel versions alongside Metal kernels purely as a CPU fallback for background execution? That feels like it defeats the purpose of the migration. It seems like a gap in the platform: the API exists, the entitlement exists, but the hardware support isn't there for the most common device category. Any clarity on Apple's direction here would be very helpful.

Beta testers outside the US, for our MacOS App, are being told their ID is not valid in the US Store and that they must switch to a store in their country. Yet the store switch fails Essentially beta testers outside the US cannot do testflight sandbox eCommerce for the Mac version of our app. Note that eCommerce on the Mac works for US based testers and eCommerce for the iOS/iPadOS works for testers in all territories. Many of these testers are in India, the UK and Canada. We believe that this is incorrect, that storeKit is not correctly detecting the AppStore Region for mac based testflight eCommerce. At this point we have 382 testers, most outside the US and we can only Beta Test our app with US users. Attached are images of the messages that are coming from storeKit: Here is a link to a video from a user in Canada who is demonstrating the problem (cut and paste into browser) https://youtu.be/kB818wfVld4 Here is another link to a video from a user in Canada who is demonstrating the problem (cut and paste into browser). https://youtu.be/7uAZKo8wpfU We see that there is another post with a similar problem. Similar eCommerce Problem Because eCommerce works in all territories on iOS/iPadOS but ONLY in the US for Mac we suspect that this is an error that either a DBA or a coder will need to fix. Any insights from anyone would be appreciated.

I’m using Network Framework with UDP and calling: connection.receive(minimumIncompleteLength: 1, maximumLength: 1500) { data, context, isComplete, error in ... // Some Logic } Is it possible for this completion handler to be called with data==nil if I haven't received any kind of error, i.e., error==nil and the connection is still in the .ready state?

We use Jamf Blueprint to deploy the managed app and identity to the iOS device (iOS 26.3 installed). Our managed app can access the identity via let identityProvider = ManagedAppIdentitiesProvider() let identity: SecIdentity do { identity = try await identityProvider.identity(withIdentifier: "myIdentity") } catch { } However, the app extension cannot access the same identity. Our app extension is notification extension that implemented UNNotificationServiceExtension APIs. We use above code in didReceive() function to access identity that always failed. The MDM configuration payload is: "AppConfig": { "Identities": [ { "Identifier": "myIdentity", "AssetReference": "$PAYLOAD_2" } ] }, "ExtensionConfigs": { "Identifier (com.example.myapp.extension)": { "Identities": [ { "Identifier": "myIdentity", "AssetReference": "$PAYLOAD_2" } ] } }, "ManifestURL": "https://example.net/manifest.plist", "InstallBehavior": { "Install": "Required" } } Is there any problem in our MDM configuration? Or the notification extension cannot integrate with ManagedApp FM?

Hi there, Starting with iOS 26.2 RC, all my DeviceActivityMonitor.eventDidReachThreshold get activated immediately as I pick up my iPhone for the first time, two nights in a row. Feedback: FB21267341 There's always a chance something odd is happening to my device in particular (although I can't recall making any changes here and the debug logs point to the issue), but just getting this out there ASAP in case others are seeing this (or haven't tried!), and it's critical as this is the RC. DeviceActivityMonitor.eventDidReachThreshold issues also mentioned here: https://developer.apple.com/forums/thread/793747; but I believe they are different and were potentially fixed in iOS 26.1, but it points to this part of the technology having issues and maybe someone from Apple has been tweaking it.

Can the vehicle‑side UWB module, acting as a Nearby Accessory, perform ranging with multiple iOS phones simultaneously while iOS cell uses background ranging with the vehicle‑side UWB (which involves BLE GATT services)? I mean, when the IOS cell works in the background, it will read the nearby gatt characteristic automatically. If there were multiple phones perform ranging with the accessory, the read characteristic from the accessory side, what shall its content be? Or does the accessory only perform ranging with one phone at a time?

Hello everyone, My app (ID: 6758039899) was submitted on Feb 17, 2026, and has been stuck in "Waiting for Review" status for over 8 days. I contacted Developer Support (case #20000108234172) and received confirmation that it's proceeding normally, but no update since. On average, Apple reviews 90 percent of apps within 24 hours. However, there might be cases that need more review time, but mine exceeds a week. Any recent experiences with long queues? Thanks!

During Apple Pay in-app provisioning (EV_ECC_v2), our iOS app successfully obtains the issuer provisioning certificates and generates cryptographic material. The flow fails when Apple posts the card blob to Apple’s broker (card creation step), returning HTTP 500 from .../broker/v4/devices/{SEID}/cards. Steps: Call issuerProvisioningCertificates?encryptionVersion=EV_ECC_v2 → 200 OK; returns ECC leaf + Apple Root CA chain; nonce=2a831be4. 2. Build {encryptedCardData, activationData, ephemeralPublicKey} 3. POST /broker/v4/devices/{SEID}/cards Expected: 200 OK on /broker/v4/devices/{SEID}/cards, or 5xx with a descriptive error if payload/cryptography is invalid. Observed: 500 Internal Server Error from Apple broker on /cards (labeled “eligibility” in PassKit logs), causing a terminal failure in Wallet UI.

Hi everyone, I’m considering integrating External Payments into my iOS app. I’ve done some research, but I’m finding it difficult to pinpoint the exact commission rates for different regions. For those of you who have already implemented external payment links: Could you share what the actual fees paid to Apple look like in your specific Geos (e.g., US, EU, or others)? Does the total cost—including Apple's commission and the payment processor's fee (like Stripe)—end up being significantly lower than the standard 15-30% IAP?

I’m building a Swift Playgrounds app (.swiftpm) in Xcode 26.2 and I consistently see this warning, even on brand-new Swift Playgrounds apps created from the default template: “Skipping duplicate build file in Compile Sources build phase: /Users/.../.swiftpm/Media.xcassets” Everything builds and runs normally and all images/assets appear correctly, so it seems like it’s just an annoying warning. I’ve checked for duplicates and don’t see the asset catalog listed twice, but the warning still appears whenever I add an asset catalog to any playground project in Xcode. Is this warning safe to ignore for submission, and it won’t affect Swift Student Challenge eligibility?

Hi everyone, I subscribed to the Apple Developer Program on Friday evening, February 20, 2026. The payment has already been charged to my bank account, but my account still shows the status “Pending” with the message “Subscribe your membership”. It’s now been several days, and I haven’t received any confirmation email or any request for additional information. I already contacted Apple Support by email, but I’d like to know if other developers have experienced the same situation and how long it took before their account was activated. Thanks in advance for your help and feedback! — Rashid

Hi, it's been almost 3 days since I submitted an app update for review. Is anyone else also experiencing such a long waiting for review time ? It's the first time in years this has happened

Updated iOS to 26.4. Gmail folders have disappeared in Apple Mail. Can‘t transfer email messages to folders as consequence. Any ideas?

Where do i find or create my key ID and issuer ID. Im lost.

Steps to reproduce: Open Xcode 26.3 → Settings → Intelligence → Claude sign-in Click the sign-in button — spinner begins, never completes An email the arrives with a magic link. The magic link opened a browser page which displayed a 6-digit verification code with instructions reading "enter this verification code where you first tried to sign in" — i.e. back in Xcode. However, Xcode was showing only an endless spinner with no code entry field anywhere in the UI. This is the core bug. I did since manage to complete authentication sign-in through a second browser verification field that eventually appeared after about 10 minutes and did get signed in, but the Claude Intelligence agent still returns "Your request could not be completed" even after successful sign-in and a full Xcode restart. Prior to this bug starting at 10 am on February 19 I had been using the intelligence agent successfully for about a week. Anthropic did have some sort of event on their system around February 18/19 so maybe this has been a result of that. I have notified Anthropic support and Apple Feedback Assistant. Does anybody have a workaround until either anthropic or Apple get back to me?