Good to see this happening, although aside from the memory-safety, thread-safety bit, how does it compare to SPARK[1], which I am currently using due to its legacy of being used for real-world applications for a couple of decades. SPARK has a verification toolset.

I hope the collaboration between AdaCore and Ferrous Systems brings Rust closer to SPARK's current capabilities in producing high-integrity software. I didn't find a formal document from NIST listing what criteria have to be met to be put on this list other than the mention that "ISO/IEC/JTC 1/SC 22/WG 23 is working on"technical report (TR) 24772 Guidance to avoiding vulnerabilities in programming languages."

[1] https://www.adacore.com/about-spark

I'm old enough to remember a time when a lot of folks considered NIST to be compromised and wouldn't trusted their elliptic curves. Kind of makes one wonder about rustc.

Wait till they find that it downloads untrusted code from crates.io.

You're using the word "safe" in the wrong context here.

No, rust has stolen the word. Personally, haveing not drunk the coolaid, I'm not ok with that. They could have been explicit, and just stuck with "memory-safe", which is a subset of "safe". But now it's basically impossible to have a conversation about safety and rust when safety is something other than how rust chose to define.

No one is confused about how the word safe is used except you, and your deliberate misunderstanding.

What? This entire thread is about exactly that confusion.