When you talk to people about Rust, you often hear "it's a great, fast language - but so hard to learn, and so much memory management, fighting the borrow checker, etc.".
So far, we've barely done any memory management at all. You've seen a few details, but haven't really had to do any yourself. You've run into:
'staticlifetimes, attached to strings stored permanently in memory (in Axum).Box, used to wrap a dynamicResulttype (which we promptly replaced withanyhow).impl (trait), specifying that you are returning a type that implements a trait (e.g.impl Future) rather than a concrete type.unsafein an example of how not to share data.
Other than that, you haven't had to add lifetime annotations, use unsafe code, allocate or de-allocate memory, or even free up resources (file handles, channels, etc.) when you are done with them.
Why is that?
"Idiomatic" Rust is very powerful and can go as low-level as C (you can even inline assembly!)---typically with safety guarantees on top. But it's also very high-level, and you can write code that is almost as high-level as Python or Ruby. "Normal" programming---building a web server, a command-line tool, a GUI app, a game, etc.---doesn't require you to do any memory management at all. The underlying systems libraries are taking care of it.