Mark Van de Vyver

Also, with `#[trace("a", recurse=all)]` the `"a"` becomes a prefix added to child trace names, which default to the function name, so names in the above are: - `a_outer` - `b_defined_later_too`...

Whenever this is implemented... we will need to explicitly document that when `recurse=all|private|public` then nested `#[trace(...)]` settings are not cumulative. Specifically, for example: ```rust #[trace(recurse=all, option_a="some")] fn f() { #[trace(option_b="thing")]...

> What's the scenario is recurse for? Pure convenience when debugging the scenario is you have a bug of unknown provenance in a non-trivial `async` context. Without such a switch...

Possible workaround for the macrotest issue: Cargo virtual manifests do not support the `[features]` stanza (without also having the `[package stanza]`). See [Cargo issue 4942](https://github.com/rust-lang/cargo/issues/4942).

I think to begin with it'll be baby steps: 1. Restructure the code *layout* and introduce tests. 2. Error reporting / UI 3. Once some tests are in place would...

Agreed. The main risk here is rejecting the whole pipeline idea. I don't propose the RFC be more detailed than what it is now, e.g. the Model is undefined/unspecified -...

Long story short, when you want to spawn threads, or record within a function/span, the ergonomics are not as great as they might be. Ideally, one would like to land...

Thanks for the feedback. Initially I'd also thought about `Span::enter_with_parent::::("spawn", &root);`. However, after a while the use of switchable levels in minitrace seemed unnecessary, and possibly out of minitrace scope?...

OK, I think I understand things a little better. Parking workarounds, and levels, because neither is strictly the subject of this feature request - I'll open a separate issue for...

> My only question is: what will the user do with the span returned from `mintrace::Span::current()`? The use case is: The span is created by `#[trace(...)]` and some recording occurs...