[steveklabnik]

No, people are not claiming Rust cannot have any problems.

UB is not possible in safe Rust, by design. The root cause of UB is always in unsafe code. Miri is useless if your code is 100% safe Rust.

The only exception to this is bugs in the compiler, of which there are a few. They’ll be fixed.

[chiggsy]

I have no faith in this statement. Let's see how it plays out.

[steveklabnik]

If you ever find UB in purely safe Rust, it is a very serious bug. Please report it.

[PreInternet01]

> UB is not possible in safe Rust, by design

You're available as an expert witness to that fact?

Because, eh, well, in at least one of the Rust-related situations that I'm involved in right now, someone might soon very well require the services of a person both as wise and reluctant-to-offer-any-kind-of-compromise as yourself...

[jcranmer]

The situation you've alluded to in another thread seems to involve an unsafe block (since it's using a type which is only usable in an unsafe block).

Let me be even more explicit than steveklabnik here. If your code, including any libraries you link to, is 100% Rust and free of any unsafe blocks, then (barring compiler bugs) it is impossible to execute undefined behavior. If your code has an unsafe block, then it is possible execute undefined behavior. Note that it is possible for safe code to execute undefined behavior, IF there was an unsafe block that did an operation that requires the programmer to promise something was true that was not true.

For example, there is an unsafe method that will let you convert a pointer to a reference with an arbitrary lifetime. If you wrap that in a safe function, you can return a reference to an object whose lifetime has ended, and cause undefined behavior in attempting to use that lifetime--the attempt can even be outside the safe block. But were that unsafe block that upgraded the lifetime not present, then you couldn't cause the later undefined behavior to happen.

In short, an unsafe block is where the compiler can no longer guarantee that the conditions that prevent the ability to observe undefined behavior are present, and it is up to the programmer to ensure that these conditions are met, and even and especially ensure that they continue to be met after the unsafe block completes. I do worry that too many programmers are blasé about the last bit, and it sounds like your coworker may fall into that category. But Rust has always maintained this principle.

[steveklabnik]

Yes, it is a core design tenet of the language. It's as benign a statement as "C# has garbage collection." That's not "reluctant to offer compromise."

[PreInternet01]

OK, you truly seem not to understand how much damage you're dealing to the general population using absolutist statements like this, do you? Nor do you seem to understand "compromise", like at all, because you seem to equate it with "tit for that", which is unsurprising, but still... disappointing.

In any case, I'm truly done here, in all senses of the word, but I still I wish you and your acolytes the absolute best.

[superb_dev]

Man all you had to do was bring proof, like maybe a code snippet with UB?

[keybored]

Calling Steve Klabnik (of all Core Rust background people, literally all of them) an “absolutist” proves how unreasonable you’re being.

[chiggsy]

Why do you feel it is unreasonable for this person to have human failings? What label would you find suitable?

[keybored]

You’re either reframing the statement to be about human failings overall—the lack thereof—or you’re assuming the conclusion.

[Slyfox33]

What are you talking about? Yes it's impossible to have UB in safe rust unless theres some obscure compiler bug or something. This isn't a controversial statement.