Bjarne Stroustrup’s Plan for Bringing Safety to C++ (The New Stack)

One thing I'd suggest may help with that, try being unnecessarily explicit until you're more comfortable.

When you write &str in Rust, the inference engine (or if we're writing a function signature, specifically Lifetime Elision) tries to figure out what lifetime that reference has and silently assumes that's what we meant, but while this is idiomatic if you're learning lifetimes it may be helpful to write more of them explicitly instead.

Also, much like Physicists writing Fortran think v and k are perfectly good names for important long-lived variables, the Rust lifetimes you see in real code often have absurdly terse names, but that's not mandatory, it's fine to give them longer names.

fn greet<'brief>() {
let s: &'brief str = "Hello there";
println!("{s}");
}

There's no need to name this lifetime 'brief, but by doing so maybe we get more used to seeing lifetimes around and we learn where they're needed even though the language is good at inferring them so you can get away without writing them down.

As a halfway point, you can write the underscore to say you know a lifetime goes here but that lifetime can be inferred.

fn greet() {
let s: &'_ str = "Hello there";
println!("{s}");
}

I found lifetimes very easy to get used to, but I think they're the "weirdest" syntactical change from other semicolon languages. There are a bunch of important semantic changes (notice the semicolon in Rust isn't even doing the same thing as in a language like C or Java) but the syntax is what gets in your eyes immediately.