Comment by thesuperbigfrog | Hacker News Reader

Comment by thesuperbigfrog | Hacker News Reader

thesuperbigfrog

>> C is not memory safe. It suffers from undefined behavior. These are valid complaints. However, C code can be secure!

C and C++ code CAN be secure, but most of it is not. It is too easy to write or update C / C++ code so that it is not secure or has unexpected results.

The lack of memory safety and undefined behavior that is difficult to reason about is why C and C++ are being displaced by Rust.

https://blog.regehr.org/archives/213 gives great insights into how undefined behavior in C and C++ can be difficult to reason about and cause problems.

I agree with Drew that Rust is not perfect and it still needs to target more architecture and platforms, needs a formal specification, and needs more implementations, but it is the most promising alternative to C and C++ so far and it has a bright future with the recent Rust Foundation: https://foundation.rust-lang.org/

With support from AWS, Google, Huawei, Microsoft, and Mozilla, Rust will continue to get better and will undoubtedly begin to replace C and C++ more and more.

To me, the most promising alternative to C are actually alternatives to GCC and Clang, which don't use "undefined behavior" as a wanton license to rewrite your program to do whatever they feel like at the time. See for example the cproc compiler that I linked to in the article. And, as someone who still writes heaps of C code, I feel like undefined behavior is an over-stated problem. I run into weird C issues pretty infrequently. Memory safety does cause issues more often than undefined behavior, though. Unintialized values, use-after-free, buffer overflows, I run into these more often than I'd like. Usually they're pretty loud, though, so it's apparent that there's a bug somewhere, and tools like Valgrind make dealing with them pretty straightforward.

It helps to remember that you're not clever enough to do clever things, in any language. The other most promising alternative to C is just doing it more conservatively.

Also, the Rust Foundation just looks to me like a corporate mouthpiece, not unlike the Linux Foundation, which will do little to nothing to make Rust better for anyone. 10 directors and 1 developer, what a recipe for success!

Aside: I'm working on a new programming language which I hope will be interesting with respect to these ideas. It's being kept private while it's under development, but anyone who is interested in this problem space should shoot me an email if they want to hear more.

> Also, the Rust Foundation just looks to me like a corporate mouthpiece, not unlike the Linux Foundation, which will do little to nothing to make Rust better for anyone. 10 directors and 1 developer, what a recipe for success!

The Rust foundation isn't supposed to have any control over the project, rather it is meant to serve as a liason between companies that want to fund Rust development and operational costs and the developers that work on it full time. It also serves as a holding ground for various trademarks that used to be owned by Mozilla. You are also wrong on your numbers, only 5 of the directors are from companies, the rest are members of already existing Rust leadership.

>> To me, the most promising alternative to C are actually alternatives to GCC and Clang, which don't use "undefined behavior" as a wanton license to rewrite your program to do whatever they feel like at the time.

That's one of the biggest issues with C and C++: Undefined behavior is in the standards, so compilers can do whatever they want which can be problematic.

>> The other most promising alternative to C is just doing it more conservatively.

There are other programming languages besides C, C++, and Rust which target low-level and embedded targets. Ada or SPARK might be options depending on your needs and targets.

>> Aside: I'm working on a new programming language which I hope will be interesting with respect to these ideas. It's being kept private while it's under development, but anyone who is interested in this problem space should shoot me an email if they want to hear more.

That sounds very interesting. I look forward to seeing your ideas in the future.

Reply on news.ycombinator.com