Comment by junon
1 year ago
You're welcome to read the rustnomicon to learn about the topic you're discussing. Having written C and C++ for almost 15 years and doing extensive embedded work with it, I'm very secure in my decision to use Rust. But I'm capable of doing research to learn about it and to be somewhat involved in the development, mostly as an observer, to see both the direction it's moving and the overall process and meticulousness with which it's developed, to make an informed decision.
It doesn't seem you're making an informed statement at all anywhere in this thread, choosing instead to be hung up on semantics rather than the facts plainly laid out for you.
If that makes me an "enthusiast" then so be it.
Well, if you come from C++ then Rust might look nice to you. But I come from languages like CommonLisp and Ada, and so Rust just looks like a horrible abomination to me because that's what it is. It's also not surprising. A good programming language simply cannot be designed that fast.
Common Lisp is an amalgamation of every lisp they could find, they slammed it all in. Calling it well designed is funny because every single CL developer openly accepts that its a fucking weird language with hell lot of warts that cannot be polished away.
Ada is fine, just verbose, kinda fun, no comments about it except that its kinda sad how weak their formal verification is. I prefer Frama-C over it. You can compare Ada and rust but ada is horrible, sincerely horrible at working with ownership. Frama-C can run laps around it as you can verify EVEN arbitrary pointer arithmetic.
Calling rust a horrible abomination is weird. As someone who dabbled in CL for an year, I love the fact that it has proc macros and even tho its harder to use it, i can make my own DSLs and make buildtime compilers!!
That opens up a world of possibilities.We can actually have safer and stricter math libraries! Maybe put us back in era of non-electron applications?
The horrible part might be syntax but eh, its a stupid thing to care about.
CK standardization was before my Lisp time, but from the historical accounts that I read about, many people in the Lisp world were unhappy because Common Lisp didn't have this or that from whatever they were working on, and because CL was standard they would have to use it.
CL is fairly carefully designed with regards to compiling. This is why math functions are not generic for instance. Redefining standard functions is undefined behavior, as a self-modifying code. It omits features that don't integrate well with conventional run time and machine models like continuations. It doesn't even require implementations to optimize tail calls.
I have no idea why ANSI CL has such a large page count. In my mind it's such a small language. I think it could have benefited from an editorial pass to get it down to 600-something pages. But that would have delayed it even longer.
Once the horse escapes the barn it's risky. When you rewrite technical text you can very easily change the meaning of something, or take a particular interpretation where multiple are possible and such.
1 reply →
> Common Lisp is an amalgamation of every lisp they could find, they slammed it all in.
Not really. It's mostly a modernized version of Zetalisp. In many cases simpler as that, with some added new stuff (like type declarations).
1 reply →
> Common Lisp is an amalgamation of every lisp they could find, they slammed it all in.
Absolutely NO. Take a look to the book by Guy Steel, and you will see is not like that.