4 ms·
There is not one group of developers C folks won’t throw under the bus (eg by calling them sloppy) to defend C. Like here you have one of the lead devs of the
by Ar-Curunir 1y ago
There is not one group of developers C folks won’t throw under the bus (eg by calling them sloppy) to defend C.
Like here you have one of the lead devs of the Linux kernel saying that Rust solves many problems that he has seen happen repeatedly in the kernel, and you’re saying “hm well they must just have been sloppy”.
- cmrdporcupine 1y agoYep. Frankly there is a whole subsection of our profession which is composed of engineers who have differentiated themselves on their elite skills with difficult low-level languages, and if you pry that cultural marker away from them, they get defensive. I see this in C/C++ forums all the time: "Rust doesn't solve a problem I have". Actually, yes it does. They might not think it, but I've been in enough codebases over the years by some very skilled C & C++ programmers to know better. The mistakes are preventable by a decent modern language that doesn't let you just null and leak all over the place in entirely preventable ways. Rust has all sorts of problems, it's definitely not perfect. It is not the final form for systems programming languages. But it's getting tiresome to see people who won't admit that the C legacy is a real problem at this point. And it's not like it (lack of safety, rigour, etc.) wasn't known along the way (or Ada would never have been a thing, etc.) -- it just wasn't considered important. Well it should be now. Hell, in the kernel in privileged execution space more than anywhere else it is most important.