Taiki Endo comments

Results 938 comments of


                                            Taiki Endo

Add I/O-safe traits

Published in 1.7.0.

Bounded channel fully deadlocked when sender is forgotten

Thanks for the report and investigation! I would accept a PR to fix this.

Add cooperative task yielding to this runtime

async-io previously had something similar, but it has been removed: https://github.com/smol-rs/async-io/commit/4e7cdb6a48d94864d3106adb773634f747899f3f

Add cooperative task yielding to this runtime

I don't know the exact reason, but the tokio's cooperative scheduling had compatibility issues with various primitives in futures and tokio that do not participate in the cooperative scheduling. ([maybe...

`Timer` is paused on Android in "doze" state

> Another is to lift this issue to the standard library so it uses Android-specific `ioctl` on `/dev/alarm` as described here: See https://github.com/rust-lang/rust/issues/71860

allow waiting on process exit without requiring a mutable borrow and without dropping stdin

Thanks for the PR and sorry for the late response. I'm not sure what a good API is to support this (I feel `*_no_drop` does not sufficiently explain what it...

allow waiting on process exit without requiring a mutable borrow and without dropping stdin

Also, the signature change of `try_status` is a breaking change for users who use `.as_mut().map(Child::try_status)` on `Option` or cast `Child::try_status` as `fn(&mut Child) -> ...`.

Recover from thread spawn failure

I would accept PR to fix this. The minimal way is probably to make grow_pool failable and delay unwrap until after the lock is released. (There may be a better...

Add `dedupe()` to `StreamExt`

It is not surprising that methods that do not exist in `Iterator` are not implemented in `StreamExt`. https://doc.rust-lang.org/nightly/std/index.html?search=dedupe Also, please describe the specific behavior you need. `dedupe` has at least...

[Discussion] A more ergonomic way to deal with 'blocking' futures

I don't have a strong opinion on this. However, bounded thread pools are not necessarily suitable for this case (see https://github.com/tokio-rs/tokio/discussions/3717#discussioncomment-639205 for more).