> it’s hard to imagine why more than a handful of researchers would use them If you could automatically prove that your concurrency protocol is safe, or that y

> it’s hard to imagine why more than a handful of researchers would use them If you could automatically prove that your concurrency protocol is safe, or that your C program has no memory management mistakes, or that your algorithm always produces the same results as a simpler, more obviously correct but less optimized algorithm, I think that would be a huge benefit for many programmers.