Ryan Levick

This produces a similar error to the one referenced in #239 but the cause is difference. While the error in that other issue stemmed from the wit incorrectly referring to...

Looks like we might need @m-ou-se to reroll the review.

Turns out the issue is that Windows SystemTime [does not support time intervals smaller than 100 nanoseconds](https://docs.microsoft.com/en-us/windows/win32/api/minwinbase/ns-minwinbase-filetime). The tests are [using intervals much smaller than that](https://github.com/rust-lang/measureme/blob/b9cccd7ad4c859a5e0e3dd6bff3daac7a190bdd7/analyzeme/src/stack_collapse.rs#L120) and when we add...

I'm going to throw my hat in the ring for using raw identifiers. It's not the most pretty solution, but I do think it strikes the right balance of reflecting...

How do we handle the following then: ```wit type param = list; no-borrow: func(param: param); record my-record { field: param } ``` This would require two types: a borrowed type...

I've opened #621 to address the generated code preamble. We really need to add the entire `component-type` section not just the producer information. I'll adjust the issue to reflect that.

As @emschwartz suggests, I think having a dedicated troubleshooting/gotchas/FAQ section would be helpful.

What are the particulars of file system access that make it worth having a dedicated section? I think it should be a part of I/O with Tokio (and the cookbook)...

@eholk @vincenzopalazzo @fmease the regression seems legit. It seems the primary benchmark affected by this `html5ever` seems to spend lots of time in `expand_crate` which could be impacted by this...

@lann what's the state of this? I thought this had already been merged.