Oli Scherer

> Is this a thing? rustc already reports whether something was * defined by command line * defined by default settings * defined by code so we just add another...

The argument against `Cargo.toml` is that you can have a single `lints.toml` for multiple projects. Or even hierarchical `lints.toml`s

That would mean that companies need one enormous workspace for all the crates in their company if they want to have a consistent lint setup. This is the very use...

> down to the statement-level that part is no problem, you can configure lints in source code just fine. And if you have one enormous workspace, then the `Cargo.toml` solution...

Maybe we could come up with a scheme where you can import the lint configuration from another crate? So making it a language feature where you can define a crate...

I'm fine making it `Copy`, if we switch to a more stringly typed representation we are likely at a stage where a new major bump is a good idea anyway.

r? @oli-obk

I think this is again completely in T-types territory, but maybe we should make it both teams this time to ask T-lang if they also think that ^^

``` 309 | static RT: Lazy = Lazy::new(Default::default); | --------- ^^^^^^^^^^^^^^^^ the trait `~const Default` is not implemented for `Mutex` | required by a bound introduced by this call ```...

> This error occurs in a static, so it is a const context. It just has to be fixed I think. yes, but inference would figure out that we require...