Interesting. It's hard to take that claim seriously, given that Rust's entire design is based around the idea that any code you write outside of 'unsafe' blocks

Interesting. It's hard to take that claim seriously, given that Rust's entire design is based around the idea that any code you write outside of 'unsafe' blocks should be free from UB. And that the only way you can introduce UB into your code is by writing an 'unsafe' block. Contrast that with C++ where UB can be introduced anywhere in your code. The only thing I can think of is that you're using a different definition of 'safety' than everyone else. Which is fine, you can define words however you wish. But that's why I specifically included "with respect to whether undefined behavior occurs," in my comment, to avoid that kind of definitional misunderstanding. So perhaps you just didn't read my comment carefully enough. But yeah, if you think there is literally no difference in safety between Assembly and Java, then I'd probably call that "totally and completely nutso." It's nuts enough that I don't think any kind of asynchronous discussion could fix it. Next time, lead with that statement, so that everyone else reading along can calibrate their expectations for just how seriously they should take you. :-) The other interesting bit here is that I was effectively asking for a steel man. But instead, you just decided to give your own answer to the question. So either you misunderstood that too, or you just literally don't have any idea what other people are actually claiming.