If it does not currently, it would be nice to use this library on watchOS 2 (and update the Podspec file to declare this support). I’m happy to help do this, just putting this here to see if anyone else is already working on this.
No one is currently working on Watch support (though I might get to it in the next couple months if no one else does). We'd welcome a PR, assuming that you have signed our CLA.
Here's what needs to be done to get watch support:
• Create a shared Watch scheme in Xcode
• Update the Podspec to include Watch
• Create a sample app with a shared Xcode scheme that tests each Valet type on watch – ValetWatchTest. Use ValetTouchIDTest as a guide, though because there is no unit test target on watch ValetWatchTest will need significantly more features (for example, the ability to test migration).
• Test to see if VALSecureEnclaveValet will work on the watch (my guess is it won't). If it won't, we'll need to compile it out on watch the same way we do on other unsupported OS versions.
• Test to see if VALSyncronizableValet will work on the watch. Is there a way to opt into iCloud keychain on the watch? If not, does VALSyncronizableValet sync anywhere? It'd be awesome if it syncs to the phone (and other iCloud keychain devices).
• Revisit VAL_IOS_8_OR_LATER and VAL_IOS_9_OR_LATER to make sure they target iOS and not the Watch. We can use #if TARGET_OS_WATCH to compile things specifically into the Watch target. Unfortunately TARGET_OS_IPHONE is set to 1 on Watch. The only way to select iOS and not Watch is to select TARGET_OS_IOS, which of course only exists in the Xcode 7 SDK, which means we'll need to do some work to make any macros targeting iOS and not Watch backwards compatible.
I know that's a lot, but it takes a lot to vet Valet for a new target – you wouldn't believe the fun bugs we found when getting Valet working on the Mac (there are wonderfully subtle differences between how Keychain works on the Mac and iOS), and at least we had unit tests to help us discover the issues. We're certainly around to help with this – happy to give comments or answer questions while you're developing, or even hop on your fork and help out a little. We just won't have time to tackle this project on our own in the short-term.
That sounds like something I would like to help with. I’m a little busy with a watchOS book right now, but that should ship in mid-October, at which point I’ll have plenty of time to start tackling the list. I’m looking into the CLA. 👍
Sounds great @SlaunchaMan. Thanks for offering to help! We'll be sure to message you if we start working on it first.
According to @digoreis, VALSecureEnclaveValet works on the watch!
Since Valet is supported on watchos, maybe the .podspec should be modified to include watchos support?
@nemesis Valet isn't supported on watchOS. Some things may work on watchOS... but we don't have any tests (regression or otherwise) proving it yet. I'd prefer not to release Valet for a platform that we haven't thoroughly vetted/tested.
@dfed I get that, but to build something on watchOS that's linked with Valet (not necessarily using it on watchOS, just linking against it) will be impossible. In my case, there are source files that I want to import on watchOS but I can't since they are linked with Valet in the main iOS target. Is there a resolution to that? I know it's not directly related to Valet, so I would appreciate any help on this one.
@nemesis you shouldn't be linking Valet on watchOS, period. If you want to share a file between your iOS and watchOS targets, and you want the file on iOS to talk to Valet and on watchOS to do something else, you'll need to use compiler flags to compile out your reference to Valet on watchOS. That said, I really don't recommend that approach. Better to make your classes smaller and more single-purposed (use delegates to break up your secrets implementation from your secrets API), so that you don't need to share a class like that across library/target boundaries.
That said, this really isn't the right forum for this kind of help. StackOverflow is a better place for this kind of question. Let's keep this Issue tracking implementing a tested Valet on WatchOS. All other issues should be new Issues.