> On Wed, 7 Aug 2024 11:04:03 -0700 > Kent Borg <kentb...@borg.org> wrote: > >> And...that was nicer to the cache, to the tune of being ~32% faster. > > I recognize that this is an aberration. That was the point, after all. > > Generally, though, I think my point about LLVM in general still stands. > What makes Rust interesting to me as an outside observer (I'm a > sysadmin, not a programmer) is that Rust performance is competitive > with C performance while producing much safer binaries. More Rust means > less stress for me.
I wouldn't be so confident. The last few big security issues I've had to fix in products have been attacks that have nothing to do with programming languages. The last sshd issue where an ifdef/endif was removed from code which allowed a privilege escalation. The XZ library/systemd/sshd issue, and a few others I can't recall. The problem with languages that are supposed to be safer is that they don't address the root cause of the problem: People who don't know how to code. Java was supposed to take care of a lot of this, but it didn't. I doubt very much rust will either. You may reduce the number of one class of problem, but I am quite sure we'll see another class of problem for rust. As for government recommendations on software development, I'll wait. _______________________________________________ Discuss mailing list Discuss@driftwood.blu.org https://driftwood.blu.org/mailman/listinfo/discuss
