Asuna

> Not the `print to console` behavior, I mean that we can move the formatting part into `OmaFormatter`, and the printing part is obviously moved to `StdStreamSink`. However we may...

Can you provide the error message from CMake and commands you executed? Do you have the `fmt` package installed on your system?

@NotEvenANeko It's exciting to see that you will start this work. Please use the [`sink-decouple`](https://github.com/SpriteOvO/spdlog-rs/tree/sink-decouple) branch of _spdlog-rs_ as dependency in your draft PR, where there are some breaking changes...

Please have a person who owns AirPods 4 device run the program with `--trace` parameter with the lid of the device case opend, and then attach the log file `AirPodsDesktop.log`...

@uenoku I do agree that the maintainability of such options structs is poor. @sequencer If we change the options structs to something like a CLI arguments string, would that be...

> Here, the explicit use of `drop(thing)` effectively drops `thing` before the end of its scope, hence not triggering the lint. In fact, the explicit use of `drop` is not...

I have changed to use `mem::transmute(AtomicLevelFilterInner) -> LevelFilter`, since they should have the same layout and `struct Inner` contains no uninitialized bytes, this should be safe. This way we can...

In this PR we adds a `repr(u16)` to `LevelFilter`, refer to [The Cargo Book - SemVer Compatibility](https://doc.rust-lang.org/cargo/reference/semver.html#type-layout), this change may not be considered as a breaking change, so we could...

> # ⚠️ **Performance Alert** ⚠️ > > Possible performance regression was detected for benchmark **'spdlog-rs on Linux'**. Benchmark result of this commit is worse than the previous benchmark result...

Renamed `AtomicLevelFilterInner` to `LevelFilterLayout`.