James Liu

Another alternative is to follow through with the `repr(transparent)` use detailed in #111, and instead allow `LeakedExecutor` to be const-constructible (would require `ConcurrentQueue::unbounded` to be const). This way it could...

> Another alternative is to follow through with the `repr(transparent)` use detailed in #111, and instead allow `LeakedExecutor` to be const-constructible (would require `ConcurrentQueue::unbounded` to be const). This way it...

@fogti we could desugar the user-facing versions of it and force a `#[inline(always)]` annotation on the State implementations for the functions. It definitely looks worse in terms of user-facing documentation,...

I also noted the variance with `spawn_batch`, but that particular benchmark shouldn't be affected that strongly by this PR, beyond moving the try_tick/tick/run implementations to `State`. It's interesting to see...

Hacked together a quick implementation (just omits all of the active task tracking from the implementation), and there are notable improvements in both single threaded and multithreaded performance, with some...

`x_unchecked` makes my mind jump immediately to `unsafe`, which we shouldn't be encouraging unless it's paramount that users squeeze every last ounce of performance out.

Removing the controversial label, as it seems like everyone's in agreement that this is something we should be doing.

> Looks good, the only reason I didn't do this in the first pass was because I wasn't sure about the potential for overhead, if this is the same for...

On further inspection, having `FnMut` hooks may be unsound due to `UnsafeWorldCell::world_mut`'s safety invariants as DeferredWorld would only be sound to pass into the hook if the hook doesn't use...

> Why not leave it as a function pointer for simplicity and minimal overhead? If you really need to store state you can always use a resource, though I feel...