Prefer Display over Debug

[BenjaminBrienen]

Objective

Fixes #16104

Solution

I removed all instances of :? and put them back one by one where it caused an error.

I removed some bevy_utils helper functions that were only used in 2 places and don't add value. See: #11478

Testing

CI should catch the mistakes

Migration Guide

bevy::utils::{dbg,info,warn,error} were removed. Use bevy::utils::tracing::{debug,info,warn,error} instead.

[BenjaminBrienen]

@SpecificProtagonist feel free to review or point out good places to use disqualified::ShortName here.

[mockersf]

Why is there a migration guide? There's no information in the description about this change.

I would prefer this to be blocked until after the 0.15 is released, as it has high chances of conflicts during cherry picks.

[mockersf]

Controversial by extension of #16113 as it was used to do this PR

[SpecificProtagonist]

I'm not the biggest fan of mixing widespread code formatting changes with changes in behavior. Did you do Debug→Display switches for anything besides Entity (and str)?

feel free to review or point out good places to use disqualified::ShortName here.

I'll do that, or maybe I'll change them myself – but again, the PR probably shouldn't include both the code format changes too. But before that: Is there any criteria according to which we decide whether to use the short name in messages? E.g. I think the full name is useful when referring to systems.

[bushrat011899]

I agree with the changes made here, but I don't think it's worth it unless we have some lints to enforce the new standard. If we had format_debug_instead_of_display and format_parameter_instead_of_inline (made-up names) then we'd be in a better position to make this change long-term. Right now, this change will be made and then almost immediately regressed, since we have neither tooling nor policy to enforce it.

[BenjaminBrienen]

I agree with the changes made here, but I don't think it's worth it unless we have some lints to enforce the new standard. If we had format_debug_instead_of_display and format_parameter_instead_of_inline (made-up names) then we'd be in a better position to make this change long-term. Right now, this change will be made and then almost immediately regressed, since we have neither tooling nor policy to enforce it.

There is https://rust-lang.github.io/rust-clippy/master/index.html?levels=allow%2Cwarn%2Cnone&groups=nursery%2Cpedantic%2Crestriction%2Cstyle#uninlined_format_args

[bushrat011899]

Oh perfect! I would spin off just the addition of that lint into its own PR first, since that's enforceable. If there's an equivalent lint for the Debug/Display then likewise.

[BenjaminBrienen]

I refined this PR to not have the formatting changes

[bushrat011899]

I think this is good, but as stated I would open an issue for a lint which keeps this change maintained long-term.

[SpecificProtagonist]

Nice, this is much easier to review now!

I haven't seen any discussion of whether to use Debug or Display for entities – the former includes the raw bits, because that matches the serialized format, while the later is more concise.

[mnmaita]

Nice one, leaving two minor suggestions.

[alice-i-cecile]

@BenjaminBrienen CI failures are real; I think you need to revert a couple of things here.