Torvalds weighs in on 'nasty' Rust vs. C for Linux debate
theregister.com
theregister.com
Pre-reform Linus 1.0 had his own flaws, but at least had a clear opinion on things, be it good or bad.
I think so too. From some comments in this discussion, it seems some Rust enthusiasts are mainly interested in sidestepping any discussion on the real world output, technical merits and tradeoffs of using Rust in kernel work, and were instead eager to have an authority figure dictate adoption to further push the bandwagon. Marketing and self-promotion are important, but without a solid technical argument what's the point really?
"So I see the whole Rust thing as positive, even if the arguments are not necessarily always [so]."
IDK, that seems pretty clear to me. He thinks that it's a good thing to try regardless of the outcome because it's generating good technical discussion and unsettling overly-settled assumptions. He does also seem to think that it hasn't failed and won't fail. Mostly he seems surprised by the sheer religiosity of the arguments that crop up.
So, he's not taking sides, but he doesn't regret the decision to attempt to integrate Rust into the kernel.
Rust-in-kernel project is now years old, it is about time to take a stance. IMO he would actually weigh in with a clear opinion if he would say something like "Rust code goes in, if it improves the kernel! Everybody get out, it's already merged" or "So far Rust brought only more drama in the kernel, show me the code finally or get out!".
But this "meh, I couldn't care less what happens" answer is on one hand disappointing, and on the other hand isn't unexpected, unfortunately.
I disagree with that interpretation. To me it reads as “the arguments are silly and unproductive. So far Rust looks positive to me but it’s early to say for sure”. That sounds perfectly reasonable to me.
And honestly, I hope Linus ignores criticism like yours¹. I hope that after the “reform” he feels happier and more adjusted and that he has a renewed love in doing good over being efficient. Software ultimately exists to serve people, so we’d all do well to put humanity first.
¹ Though I don’t mean to single you out.
I think this is the sort of unhelpful emotional comment that justifies Linus' position. Arguments need to stand on solid technical ground, but the emotional and almost religious fervor some less-technically inclined of some proponents is unhelpful and detrimental to make the case.
> Rust-in-kernel project is now years old, it is about time to take a stance (...)
You see, this is the kind of emotional drama that should stay way outside the whole debate.
You're not arguing technical merits of tradeoffs (which there are, and important ones). You're arguing for an authority figure to force-feed your personal choice in spite of its technical merits, and you frame the only scenario where your option isn't pushed as an irrational outburst that shields your personal choice from any criticism.
The Linux kernel project has been a huge success in a big part because the discussions on its progress stayed mainly on a technical ground, and calls were made in spite of emotional outbursts or attachments to technical choices. In line with that principle, Rust is still an exploratory technology used in proofs of concept,and the conclusions so far were clearly not decisive. That's perfectly fine. Don't you agree?
That's just reality
vi was a joke. vim/emacs is worth debating. I mention this because the quote is talking about the old days, as if the same debate is continues.
emacs/emacs is also worth debating, because for all its power, emacs usage does not live up to emacs potential.
my problem with the Rust/C debate is that the Rust crowd hates C. I think Rust would be a better language if the community loved C.