Pinning continues to be the most difficult aspect of Rust to understand

I think that ideally someone writing safe Rust drivers using already existing bindings would not need to really care about how Pin is implemented; they could play type Tetris and just have everything work.

But I also think that since Pin and its motivation are noticeably different from how C does things, and since kernel developers frequently care about low-level details of memory layout and access, it's worth having in one's mental model.