Daniel McCarney

@ctz I've reworked this branch around the [`lock_api`](https://docs.rs/lock_api/0.4.11/lock_api/) dependency you found. I think it's workable, but the generics on the [`Mutex`](https://docs.rs/lock_api/0.4.11/lock_api/struct.Mutex.html) and [`MutexGuard`](https://docs.rs/lock_api/0.4.11/lock_api/struct.MutexGuard.html) types mean it's a more invasive diff...

> Sorry, this might've been a wild goose chase. No worries, it was an interesting exercise if nothing else. > I was assuming we could make the whole thing generic...

> That does feel like the best direction long-term. Do you think this work should be shelved in the meantime? based on conversations out-of-band we agreed the original approach used...

> This mostly looks good. It feels to me like the commits are too small -- in my mind there is a HashMap commit and a Mutex commit, and splitting...

> in my mind there is a HashMap commit and a Mutex commit, and splitting it up further doesn't help much. Force pushed an update with just three commits: the...

I think this branch is ready for another review pass now.

Since the impl has changed quite a bit since Ctz last reviewed I'm going to wait to merge this until he has a chance to do another pass (_avoiding a...

@sharksforarms Is this branch something you would be interested in rebasing on `main` and putting up for review?

> Hi @cpu, thanks for taking a look. I'm interested yes. Great! > Please advise if the design (callback approach) is still appropriate or if this should change I haven't...

In the effort of keeping our pull request tracker focused on work in progress I'm going to close this for now. Like https://github.com/rustls/rustls/pull/1534, were this to be revived I think...