"Garbage collection" is ambiguous, actually; reference counting is traditionally considered a kind of "garbage collection". The type you're thinking of is called "tracing garbage collection," but the term "garbage collection" is often used to specifically mean "tracing garbage collection."

Note that this post is from 2014.

Notably, this article is from 2014.

Is Fortran really your favorite language?

I agree, but if you look at the specific email linked, it very clearly crosses the line into direct abuse, whereas most of Linus's rants do exist in a slightly greyer area (even if they'd be grounds for a discussion with HR at an actual company).

Delete prior iterations of the loop in the same timeline? I'm not sure there's anything in quantum mechanics to permit that...

You're not wrong, but not everything needs to scale to 200+ servers (...arguably almost nothing does), and I've actually seen middle managers assume that a product needs that kind of scale when in fact the product was fundamentally not targeting a large enough market for that.

Similarly, not everything needs certifications, but of course if you do need them there's absolutely no getting around it.

To be pedantic, I didn't ask a question, I just said I was surprised! I am still surprised.

... I am currently asking if there is a language that fits the description. And I'm clearly not the only one interested in such a language.

I personally have not found Rust onerous to learn, especially by comparison to C++. But a lot of people do find the learning curve steep.

the author in his 11y of writing Rust never once heard about the philosophy of Rust/Unsafe Rust.

Are you...aware of who the author is? He literally co-wrote The Book, aka The Rust Programming Language.

The standard differentiates between "unspecified" behavior, which is as you describe, and "undefined" behavior, which may be completely nondeterministic at runtime.

That's an extremely niche set of requirements, largely because interoperating with C++ is, well, a nightmare, and partly because "fast code" means something very different to C++ devs than to most devs who use managed runtimes.

Also, there are different definitions of "safe". Rust takes a very C++ style view: it's "safe" in the sense that UB requires either a compiler bug or an explicit opt-in to something unsafe.

For other definitions of "fast" and "safe", sure, most garbage collected languages count. Java, C#, and Go are "fast enough" for most application code, and they mostly guarantee that errors will result in crashes rather than UB. Zig is as fast as C/C++/Rust (and integrates much more easily with C than any other language except C++) and has a very different approach to safety (mostly runtime checks in debug mode that are not included in release mode).