> the end of a period characterized, and not incorrectly, by the opinion that a safe programming language guarantees... I don't think that there was ever such

> the end of a period characterized, and not incorrectly, by the opinion that a safe programming language guarantees... I don't think that there was ever such a period. Provable correctness always had the caveat of highly idealized assumptions, and we've known from the start that hardware vulnerabilities such as timing attacks, rowhammer, gamma rays, and power loss can undermine those assumptions.