I'm building a caching system for my textures and noticed that setPurgeableState (using MTLPurgeableStateKeepCurrent) always returns MTLPurgeableStateEmpty on the simulator. Is this a bug? I'm guessing the texture implementation is different than on the device but the debug strings are more or less the same so I'm wondering if there's a setting the can get the textures on the sim to behave like the textures on the device. It looks like the inner purgeable state is actually set correctly because it will complain about volatile/empty textures when used in a command buffer.

I want to unit test my iOS app running on my M1 machine. The tests generate snapshots and I'd like to be able to save them on the host machine running the tests but I can't seem to come up with a good way to do it. I tried saving to the sandbox and using a post action in Xcode to just copy the files from the sandbox to where I need them to be. Unfortunately I have no idea how to reliably retrieve the sandbox url from the iOS app. Instead of the bundle id it seems to use some random UUID for the sandbox identifier (e.g. /Users/[USERNAME]/Library/Containers/D354FAB7-AEA9-4351-BE46-648BF0D13357/Data/Documents/) and I have no idea where that's coming from. I also can't seem to opt out of the sandbox (which I figured but still). I can't use NSOpenPanel or any user-driven interaction for getting a system shared folder to save anything to since it's all supposed to be automated. Any ideas of what I'm missing or what else I might try?

I can't get my stencil buffer to track one state for back and front facing triangles. Assigning the same descriptor as suggested in the headers does not work: MTLDepthStencilDescriptor *depthDescriptor = [[MTLDepthStencilDescriptor alloc] init]; [... asigning values to back stencil] depthDescriptor.backFaceStencil = depthDescriptor.frontFaceStencil; Creating a stencil yourself and assigning it to both back and front also doesn't work. The backFaceStencil (and front) property is marked as copy so assigning them yourself will inevitably lead to two different objects. So if equality is evaluated purely on the pointer basis of the descriptor - I don't think it's possible to assign the same descriptor here. Is this a bug or am I missing sth?

I have been using unit tests for iOS apps running on a mac (Designed For iPad destination). The reason is that I need to be able to test Metal code on an Apple Silicon GPU and I can't do that using the Simulator since it's treated as a different GPU. Since upgrading to Xcode 14.3 that no longer seems possible: It's telling me Unable to initialize test bundle from file .... The console seems to suggest it's a Sandbox violation. Can anyone from Apple confirm that this is intended? Are there any other ways to unit test iOS code on an Apple Silicon GPU?

I have an XCTest unit test that's running in an iOS host but with the My Mac (Designed For iPad) Target. Enabling mergeable libraries causes two errors: Symbol not found: _$s10Foundation3URLV9RVAPIPackE8isGCSURLSbvg And "loading failed because of an error Error Domain=NSCocoaErrorDomain Code=3588" for the Test bundle because one of the frameworks that's supposed to be merged couldn't be loaded. Disabling mergeable libraries fixes it. It also works with mergeable libraries enabled but with a regular simulator target. So not sure if I'm missing something or if I need to disable mergeable libraries now:/

So I ran a promo for my app that got me 30k downloads or so. The app was free for a day and so it got lots of "purchases". There were two errors from StoreKit2 that I can't explain (I'll detail the other one in another post) This one: StoreKit.StoreKitError.systemError(Error Domain=NSCocoaErrorDomain Code=4097 "connection to service with pid 19497 named com.apple.storekitd" UserInfo={NSDebugDescription=connection to service with pid 19497 named com.apple.storekitd} There are various theories online that the display name of the purchase could be an issue. Another was that this indicates that this indicates that a helper process has crashed. Generally my guidance was to restart the device and that seemed to resolve it for most if not everyone. But not everyone reached out and it rained bad reviews and I was accused of a bait and switch and so on. Is there any way to mitigate this? Any way to address it? It doesn't happen for everyone but it happens very frequently still.

So I've run a promo for my which offerred a free purchase for a while. Some people - and it looks like mostly in Germany - ran into an issue that the purchase would fail with SKError.Code.unknown. One user noted that if you cover FaceID and use your password when making the purchase it would succeed. That was then my guidance and it seemed to have worked for everyone. Is there a way from my side to prevent that error?

I'm working on a catalyst video editor and I'm using my wacom graphic tablet to work. The wacom input gets translated into a pencil touch. Whenever I hold down a modifier (shift, cmd etc) the touch gets ended and also ends all gestures. The mouse (indirectPointer touch) doesn't exhibit this kind of behavior. Is this expected behavior? If so is there a way to opt out? Any way to prevent this? This basically makes the typical transform gestures impossible to do when using the graphic tablet.

I'm seeking to a specific sync frame in a video file (HEVC, recorded on iPad). When I feed the buffers from that sync frame on to VTDecompressionSession it consistently drops the 2.,3.,4. buffer with a kVTVideoDecoderReferenceMissingErr (or no error but no buffer on the simulator). If I feed all the buffers from the penultimate sync frame prior to the desired frame the buffers come out fine but that would just create a massive overhead to always do it. Tried multiple OS versions, devices etc. Seems a consistent problem. Here's a sample project with the offending video (disregard memory handling etc): https://github.com/marcuseckert/vtSample I've filed a radar FB18228296 but would appreciate any feedback on circumventing or at least detecting this behavior prior to decoding.

I can't download my app via TestFlight in macOS Tahoe. Earlier versions work. Some other apps can be downloaded via TestFlight on Tahoe. The app is a MacCatalyst app that's not yet released on the App Store. The console prints the error: Download failed with error: Error Domain=ASDErrorDomain Code=710 "Invalid hash 'XXX' expected 'XXX'" But no error is displayed. Is there a solution?