Thomas de Zeeuw

> @Thomasdezeeuw Do we want to wait on v1 for figuring out io_uring support, if this is something we are interested in? I took too long with v1, so I...

Closing this in favour of the milestone: https://github.com/tokio-rs/mio/milestone/4.

In #1748 I decide not to list all versions, but instead saying we support the same versions that rustc supports.

After https://github.com/tokio-rs/mio/pull/1688 all that is left are some AFD related types and constants (see updated issue description), so looking pretty good.

Closing this as done, we can switch to using AFD constants from windows-sys if/when those become available, no point in keeping this open.

> Ah, I see the others. Will be interesting to get these into metadata, given they're undocumented/"private" structures. I suspected as much. I'm happy to keep the definitions in Mio...

I think we should close this in favour for an issue that follows the `read-buf` Rust feature (https://github.com/rust-lang/rust/issues/78485). We decide a while ago to follow API defined in std lib,...

I don't think there is anything actionable here, so closing.

I have a runtime that is based on io_uring: [Heph](https://github.com/Thomasdezeeuw/heph), which is based on [A10](https://docs.rs/a10/latest/a10). I'm trying to port Hyper's client to it, but its I/O traits will not work...

To add another use case is storing the waker atomically. You can split the data and vtable into two `AtomicU64` and store those atomically separately (with some careful programming) or...