| I think most of the points he made in his original post are very valid ones, it's just the fatalist "devs are going to create bugs, such is life" approach to errors that I think I glimpsed in the original post that prompted my response in the previous thread. I'm tempted to call it the "C Stockholm syndrome". The reason I say that is that I could actually have written something like that a few years ago myself. I think it's a pretty common mindset amongst the "low level" C crowd. And there's a good reason for that, we've been burned before. People have been advertising "safe" alternatives to C since forever. Use Perl! Use Java! Use Lisp! Use Haskel! Use Python! Use Go! But none of these languages can replace C in all of its use cases. You have big runtimes, garbage collection, interpreters, you name it. You can't just bootstrap a stack pointer in 3 lines of ASM and call into them like you can with C. And even if we don't say it out loud there's a bit of snobbism to writing low level "unsafe" code. "Safe languages are for noobs and webdevs, real coders debug with oscilloscopes". And so I had this mindset that you couldn't have your cake and eat it too: you could be safe or you could be low level. Segfaults and NULL pointers were just a normal part of dealing with low level "unmanaged" environments and we'd just have to deal with it. People who complain about it are just not worthy. In my opinion the biggest success of Rust so far is in proving that you could be safe without needing a GC, a big runtime or complex abstractions. You could design a better C without having to compromise. Even if Rust fails in the long run, it would still be a very important "proof of concept". It showed that C was unnecessarily unsafe, that you could achieve the same feature set without throwing safety completely out of the window. That being said if there really are folks bugging devs by asking them to consider switching to language X or Y then I think it's fair to say that these people are idiots. Might as well ask JK Rowling to rewrite Harry Potter in Esperanto because "it's totally the future!". I mean, let's face it, at this point Curl probably has a much bigger userbase than Rust itself by at least an order of magnitude. Which one is more likely to be irrelevant in 10 years? Rust or Curl (and by extension, C)? If you answered the latter you should read less hacker news. And even if Rust ends up being successful it's not like C will go away any time soon, or even in our lifetimes. Curl (as a C library) will probably still be useful and relevant for decades to come no matter what. |