slog VS log

I only recently learned about rustwide myself when implementing sandboxed rustdoc builds for Shipyard.rs. After spending a good amount of time with the codebase, I have found it to be generally high quality, but the way the code is organized makes it fairly difficult to adapt for different purposes than it was intended for (not modular). I have a fork that I have changed to do what I need but not sure whether the two codebases can be reconciled because I ended up needing to put stuff very specific to my purposes in there. I also ran into a weird issue where the logging from rustwide was conflicting with the slog-based logging from my code, which feels to me like there is some UB lurking somewhere.

Slog Core Crate : core package to the gateway of logging modules.

log: Logging interface with various levels.

Depends on your logging crate. log does not depend on any async runtime.

See the corresponing PR: https://github.com/rust-lang/log/pull/543

log

It's a for-fun project that provides a configurable implementation for the log crate. This language and community is awesome, and I hope some of you get a kick out of this!

Firstly I would start by taking a look at how crates using log actually resolve this. Particularly the set_boxed_logger which simply calls Box::leak.

You might want to look at log and simple_log.

4) Another route that I tried is to develop a simple terminal app using SwiftUI with a Xcode project to build that app + link against a Rust library compiled for iOS with the actual logic. I used swift-bridge for this and it works really well, to the point where I have a custom logger that you can simply use the print stuff to SwiftUI from Rust using the log crate. Once I have a bit more time, I will probably try figuring out how to clean this up a bit more.