Isolated synchronous deinit

[inamiy (Yasuhiro Inami)]

+1 for opt-in deinit async as I previously commented in:

Deinit and MainActor

I like this @epam-gor-gyolchanyan's idea, and having actor-isolated deinit async is probably a better solution than trying to capture all actor properties into a new task manually for clean up as @Douglas_Gregor mentioned, which is obviously getting difficult when actor has tons of internal properties: I also think deinit async is slightly better than having synchronous actor-isolated deinit (without async) since non-isolated deinit is preferred whenever possible, and it looks symmetric to ho…

And in my opinion:

@Nickolas_Pohilets

Note that proposal does not generate isolated deinit if there is no explicit deinit. Which in my experience is the most common case. This kinda serves as an opt-in mechanism.

"Non-isolated deinit when code is not written" will mislead developers when they just want to write a debug-print to observe when actor instance actually gets deallocated.

// Assume this is implicit isolated deinit.
deinit {
    // This printing may be called slower than non-isolated deinit.
    // (Also, thread printing will get changed which might also be tricky)
    print("deinit in \(Thread.current)") 
}