Nemo157

Hmmm, it would be nice if that gave an error while loading. In any case it seems like an incorrect syntax file shouldn't allow infinite processing to occur, and this...

IIRC when trying to determine that too I realised that `bat` was using `JavaScript (Babel)` for `.js` files, not `JavaScript`, using `-l js` or `-l 'JavaScript (Babel)'` in the above...

The point of having `FutureObj` is to make `Spawn` `no_std` compatible (by allowing you to implement `UnsafeFutureObj` for your own `Box`like types). Maybe this is not a useful goal and...

I thought that’s what `ThreadPool::run` did, but it appears it’s only `LocalPool::run` that behaves this way. Would be good to have consistent naming and behaviour between pools as well.

> Tilde is not expanded for native programs (only some cmdlets, e.g. `Get-ChildItem ~`) AFAIK tilde is not expanded for _anything_, `Get-ChildItem` receives a path with a tilde in it...

No, I hadn't considered that. I'm personally unsure of the usecase for `wrap_comments` itself, I've only ever thought the doc-comment variant was useful, so I'm not sure if that implication...

Updated to use the last structure you mention, that does seem much nicer to me.

Switched to manual impls of the traits, I've pushed a commit built on top of this with the `#[bool]` attribute in the proc-macro to https://github.com/Nemo157/rustfmt/commit/d09761690a09a272cbcf87561b99db86146b09ec

The specific issue I ran into was ``` thread 'main' panicked at utils/src/ipc.rs:384:50: called `Result::unwrap()` on an `Err` value: CheckBytesError(InvalidTuple { variant_name: "Info", inner: TupleStructCheckError { field_index: 0, inner: ContextError(ArchiveError(Unaligned...

Hmmm, turns out this is maybe supposed to work automatically without config like this (https://github.com/rust-lang/rust/issues/105907) and the config doesn't actually appear to work for these paths anyway. So :shrug: this...