StatisMike

@Bromeon For now it seems to be a few functions affected - I think making every `Gd` an `Option` in virtuals can be a little too much? Maybe a *special...

@Bromeon I think we can begin with the ones mentioned in this thread. I will try to spot some differences in godot source code with EditorPlugin and EditorInspectorPlugin implementations -maybe...

Thanks for this, I think that this is great time for this as IMO we could really benefit from good, consolidated structure of `gdext` errors, as there are still room...

Wouldn't it be possible to handle Rust traits for GodotClass structs in other way? I mean something like: ```rust // trait declaration #[godot_api_trait] pub trait MyGodotTrait { #[func] fn my_trait_function_to_implement(&self)...

I believe @Bromeon meant it as a `XML` (by itself) is not too standardized (there are many variants, every one introducing some caveats). As unfortunately only `XML`-based methods are provided...

IMO the `itest` seems like a better place for that. Less type-changes: the better. Doctest could check only for compile errors with `no_run`. To improve maintability comment with the information...

It feel to me like this is connected heavily with #3255 . Should these issues be gathered in one? It seems that singular well-thought system would resolves both of them....