Wink Saville

@dignifiedquire, would it be be better to discuss this on issue #78, as there maybe better solutions and/or other related issues?

> This appears to reduce the amount of completions (e.g. `Poll::Ready(Ok(()))`), which irritates me because the outer stuff should call it again if expected events didn't occur. What irritates me...

In my initial comment on this PR I say: > It is my understanding that when placing a ticker value into Direction::ticks there is a guarantee that the waker.wake call...

I agree with you, the code has been queued, but there is no guarantee it ran. So the act of sending the wake event only says at some point in...

@zseri, Ok, I'll remove the second usize. Is it your opinion I have identified a bug?

I'll try to do that tomorrow. Can you point me to a test that might be an appropriate to use as a model the new test?

Txs for the info, it was fairly difficult to narrow done the problem, but one good thing was that as I added more debug logging the problem didn't go away!

As we both suspect creating a test is likely going to be non-trivial and take significant effort on my part. It would be nice if someone with commit rights could...

> @winksaville you said you added copious amount of logging to the library stack, do you still have the patches for that laying around? I would like to try debugging...

I ran into this problem while following the [`hole_punching`](https://docs.rs/libp2p/0.44.0/libp2p/tutorials/hole_punching/index.html) tutorial for [`rust-libp2p`](https://github.com/libp2p/rust-libp2p). This has been discussed in this post [relay_v2 performance anomaly on small machines](https://github.com/libp2p/rust-libp2p/discussions/2619). One other thing that I...