If you are consuming an API that provides an object with a destructor, you are correct, you can determine when destructors will be called.
The issue is when you produce an API that contains objects with destructors. Since you are handing these entities off to unknown code, you cannot ensure that they will be dropped. This was a problem in scoped threads in Rust.
I was a little unclear but that is of course what I meant: talking about the underlying shared data because the pointers themselves don't have particularly interesting destruction behaviour. (Although the sibling is also correct that not all Rc/Arc/shared_ptr handles to the shared data with have their Drop called.)
I think that falls into the category I mentioned in the third paragraph of my comment: a serious pre-existing bug with other consequences will potentially cause the guarantee to be violated. A similar effect would happen if you had a double free that sometimes caused a crash, which is a similar level of programming mistake to creating a cyclic reference. To me it sits outside of a reasonable definition of "guaranteed".
No, typically, a reference cycle is fine. It results in valid memory that never gets read again, which is unfortunate but not dangerous, whereas double-frees can result in memory corruption. http://huonw.github.io/blog/2016/04/memory-leaks-are-memory-...
The issue is when you produce an API that contains objects with destructors. Since you are handing these entities off to unknown code, you cannot ensure that they will be dropped. This was a problem in scoped threads in Rust.