A regression :) Been dealing with Apple regressions for about 15 years, often never fixed. But I'm not bitter... ;)

We have our own licensing system for web store sales.

Huge thanks for replying. I have no idea how they would do this, or how I would test it. If the fundamental comms with StoreKit/App Store don't work, how can I write code to verify anything? Looking at the code I originally wrote to handle this case of AppTransaction failing, I put some headscratching comments in, basically saying WTF do I do now? I went for asking users to check their connection, restart, and retry.

Thanks for the speedy reply. Are you saying my example above should work? I can't see how to include any bridging header(s) in my C++ header file. Factory method - you mean initialise the Swift variable in the MyCppClass class constructor? I can't get Xcode to recognise my Swift type in the context of a C++ header.

Same. Please let us know.

Please see post following this

Same problem in July 24. After several weeks of trying to bully the Storekit/Xcode Testing/Sandbox debacle to do something meaningful for a Mac app, I've come to the conclusion that Apple really don't care about Mac apps at all. The stuff just DOESN'T work in any predictable or reliable way on the Mac. The documentation is pretty useless, and virtually everything in the WWDC videos is targeted at iOS.

The upgrade notifications seem to vary. Sometimes I get the correct ones, sometimes not. Sometimes one, sometimes two. The two are sometimes the same, sometimes not. Very inconsistent. Pretty tricky to use them to make any reliable decisions in code.

Thanks for replying. I'll give that a go. I omitted to mention I'm developing a Mac app, but your suggestion sounds good. I had looked at transaction.offer.paymentMode == freeTrial and assumed that it was indicating the subscription contained an offer as part of it, rather than indicating the trial is active. It would help if the documentation actually said these things :) Thanks again.

Thanks for this - the example is perfect. Is there a reliable way to test this? I'm trying both Xcode and Sandbox with StoreKit. Using Xcode StoreKit testing I get the current version of my app back (the one I'm building and testing), which doesn't help. Is there a way to inject something into a test environment to simulate a pre-existing user?