I don't use Rust and I've bounced off trying to learn it 3 times so far, so take this with a mountain of salt, but... I think that Rust having unsafe is fine a

I don't use Rust and I've bounced off trying to learn it 3 times so far, so take this with a mountain of salt, but... I think that Rust having unsafe is fine as there will be edge cases where the compiler can't quite work out whether some code will work fine or not and where the programmer can vouch for it. That's no problem. The issue I have with it is that the unsafe block then gets kinda swallowed by the supposedly safe wrappers around it. And there's no clear auditable trail back to it. I find that a bit surprising as I'd expect it obvious to have some kind of 'uses unsafe' declaration (annotation? I don't know what Rust calls those statements with a number sign and square brackets in front of a subroutine) which would then propagate upwards. A bit like how in Java (the most elegant and refined of all programming languages) you can use an annotation to state that a function 'throws XYZException' which then needs to be propagated up to a point where it can get handled. Not having such a mechanism feels a bit icky to me. It's like if there's spiders crawling out of your ears it's useful to know that they're coming out of your ears so you don't have to wonder 'are those spiders creeping out of my ears? Or out of my nose? Or out of my eyelids?', which would be a bit inconvenient.