eskimo

[quote='870431022, vishalkalola1, /thread/786553?answerId=870431022#870431022, /profile/vishalkalola1'] when i set usesClassicLoadingMode = false. its stopped preparing form request encoding [/quote] I’m sorry but I don’t understand this. Please elaborate. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

[quote='869492022, yuvalishay, /thread/809084?answerId=869492022#869492022, /profile/yuvalishay'] is there some mechanism to filter packets in advance? [/quote] Just to be sure we’re on the same page here: You want to implement an NE packet filter. But only be called for specific packets. For example, you might want to apply a filter that means that you only see TCP packets with a specific remote port. Is that right? If so, then, no, the current packet filter provider has nothing like that. [quote='869492022, yuvalishay, /thread/809084?answerId=869492022#869492022, /profile/yuvalishay'] do you think there is a point in opening a feature request? [/quote] Yes. The industry has a history of providing general filter support like this (for example, BPF). I don’t know if such an approach will actually improve performance given the constraints of the NE architecture, but it’s a perfectly reasonably request. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo

[quote='869978022, MikasaAlen, /thread/805248?answerId=869978022#869978022, /profile/MikasaAlen'] iOS 26 only provides an API to jump to the [Settings app] path? [/quote] Not quite. It has APIs to get to various locations within Settings. I’ve includes links to all the ones I know about in Supported URL Schemes. [quote='869978022, MikasaAlen, /thread/805248?answerId=869978022#869978022, /profile/MikasaAlen'] Will attempts to jump to other submenus in the system settings be redirected here by default? [/quote] If you attempt to navigate to specific Settings pages by opening undocumented URLs, you’re exercising implementation details. The exact behaviour you get isn’t documented and may change from release to release. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

Tracking this stuff down can be tricky. I have specific advice about this in Determining Why a Symbol is Referenced. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

There’s a lot to unpack here but I’m still grinding through the holiday backlog so I’m going to focus on the critical bits. First, don’t test stuff like this on your day-to-day development machine, because cached state can confuse you. Rather, run tests like this on a fresh Mac, one that’s never seen your app before. I generally do this in a VM, where I can restore to a fresh snapshot between each test. Second, there is a supported way for a sandboxed app to achieve this goal on modern systems [1], namely, via the CGEventTap mechanism. See CGPreflightListenEventAccess, CGRequestListenEventAccess, and CGEventTapCreate in . This relies on the Input Monitoring privilege, rather than the Accessibility privilege, and that’s available to sandboxed apps. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com [1] macOS 10.15 and later, IIRC.

It’s hard to offer insight here without more information. Are you able to capture a sysdiagnose log while your app is hung? Well, get your user to capture that log, I guess. See Bug Reporting > Profiles and Logs for advice on how to capture a sysdiagnose log. Also, if your user is reluctant to send you the log, they can use Feedback Assistant to file the bug directly with Apple and then pass along the bug number for you to share here. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

How are you cloning these VMs? Specifically, are you calling Virtualization framework from your own code? Or is your infrastructure relying on VM infrastructure provided by another third party? Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

[quote='811013021, pjfloyd, /thread/811013, /profile/pjfloyd'] On macOS I see an extra mutex. [/quote] That doesn’t surprise me, but that’s probably the result of my not understanding your setup properly. When I run your cond_timedwait_test.c code natively and set a breakpoint on pthread_mutex_lock, I hit that breakpoint a lot inside the dynamic linker and various loaded-by-default libraries. But I suspect that you’re doing stuff to avoid that, like statically linking the executable [1]. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com [1] Not something we support in general but tools like this often have to stray from the supported path.

I’ve not seen reports like this previously, but that’s not in any way conclusive, especially as this seems to happen relatively infrequently. Some questions and comments… Is this macOS or iOS? [quote='811445021, emilsp, /thread/811445, /profile/emilsp'] Stopping and restarting the packet tunnel does not help. [/quote] What about restarting the device? [quote='811445021, emilsp, /thread/811445, /profile/emilsp'] sysdiagnose was not captured on a device with the network extension diagnostic profile [/quote] Honestly, that’s not a showstopper. While it’s best to get a sysdiagnose log with additional NE debugging enabled, the NE team can often make meaningful progress with just a vanilla log. So, if you do see this again, please make sure to grab a sysdiagnose log even if it’s on a non-developer machine. If you have a reasonably large beta tester pool, you can try working with them to capture a sysdiagnose log. I posted some suggestions for how to approach this in your other thread. [quote='811445021, emilsp, /th

The general model for apply(_:completionHandler:), as explained in the docs, involves three steps: Apply the Wi-Fi configuration. Join the network. Bring up the TCP/IP stack on that interface. Problems with step 1 are delivered to your app via the method’s completion handler. Problems with step 2 are communicated to the user via an alert. Problems with step 3 can result in a variety of different behaviours. If you’re seeing a case where the devices fails to join the network and that’s not reported to the app or the user, that’s something we should investigate via a bug report. See this post for my advice on filing Wi-Fi bugs. If you do file any bugs about this, please post your bug number, just for the record. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

What platform is this? Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

Are the problematic networks considered to be local, per the definition in TN3179 Understanding local network privacy. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

[quote='811866021, KrishnaveniNalka, /thread/811866, /profile/KrishnaveniNalka'] Apple Feedback ticket: FB21464147 [/quote] I think that’s the right path forward here, especially given this: [quote='811866021, KrishnaveniNalka, /thread/811866, /profile/KrishnaveniNalka'] this issue is seen only in Mac OS Tahoe. [/quote] So: Your product was widely deployed on macOS 15 and earlier. You’re now seeing reports of weird problems that look like the system breaking its API contract. But only on macOS 26 and later. If so, that’s eminently bugworthy. Oh, and looking at your bug I don’t see any sysdiagnose logs attached. It’s gonna be hard for your bug to get traction without that. Are you able to reproduce this in house? If so, I recommend that you enable extra NE logging [1], reproduce the problem, grab a sysdiagnose log, and attach it to your bug. OTOH, if you’re trying to debug this solely based on reports coming in from the field, lemme know and we can discuss your options. Share and Enjoy — Quinn “The Eskimo

[quote='811186021, tramo, /thread/811186, /profile/tramo'] Is there a way to prevent batch operations … for Secure Enclave keys? [/quote] As far I’m aware there’s no such facility. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com

I want to start by understanding your goals. The com.apple.developer.ubiquity-kvstore-identifier value starts with your Team ID, and the app’s Team ID changed as part of the app transfer. That means that your new version of the app won’t be able to access settings created by the old version. Is that OK? If so, we can dig into why Xcode is being snarky. But if it’s not OK — if, for example, you’ve stored critical user data in the iCloud key-value store [1] — then things get more complex. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = eskimo + 1 + @ + apple.com [1] This is something I strongly recommend against. The UserDefaults subsystem, and thus the key-value store, is intended for preferences, not critical user data.