Use wildcard imports in bevy_pbr

[nicopap]

Objective

• the style of import used by bevy guarantees merge conflicts when any file change
• This is especially true when import lists are large, such as in bevy_pbr
• Merge conflicts are tricky to resolve. This bogs down rendering PRs and makes contributing to bevy's rendering system more difficult than it needs to

Solution

• Use wildcard imports to replace multiline import list in bevy_pbr

I suspect this is controversial, but I'd like to hear alternatives. Because this is one of many papercuts that makes developing render features near impossible.

[alice-i-cecile]

Hmm. I don't mind it. Ideally it would be possible to create a tool which detects and automatically resolves and formats trivial import merge conflicts like this.

Perhaps a GitHub bot like bors? It could even run fully automatically.

I agree that these are a major paper it to contributing, both in and outside of rendering so I'm enthusiastic about improving the situation by whatever means we have.

[nicopap]

I think the "ideal" would be some sort of subset feature preludes. In the case of rendering, it's especially egregious because there are so many wgpu-related types. I think it makes sense tbh to always import render_resource items as a wildcard import.

[nicopap]

In my own crates, I've been experimenting with the idea of never having multiline imports. It's purely esthetic preference, since I'm the only one working on them, it's difficult to tell if it is any help.

[nicopap]

The other "ideal" solution would be to configure rustfmt to have single item per line.

• https://rust-lang.github.io/rustfmt/?version=v1.6.0&search=import#%5C%22HorizontalVertical%5C%22%5C%3A
• https://rust-lang.github.io/rustfmt/?version=v1.6.0&search=import#Module%5C%3A

[JMS55]

I'm in favor of this, but:

• We should configure a rustfmt config for import style. Currently I've been choosing between stuff like crate:: vs super::, whether to put newlines between imports from different crates, etc according to personal preference. Ideally this is made consistent across bevy automatically, so I no longer have to think about it or manually change imports.
• We should probably hold off on merging this until Automatic batching/instancing of draw commands #9685

[mockersf]

Not controversial for me. This change has no impact except on contributor experience, and can be changed back if it doesn't work for us in a few minutes with an IDE.

It's becoming more painful to contribute on the renderer, we should merge this soon