Login
You're viewing the functional.cafe public feed.
  • Nov 10, 2026, 8:02 PM

    @liberty

    The problem with simpler languages is that you have to sacrifice a lot in order to make a language simple: run-time performance, safety, and so on.

    Rust is by most accounts a very non-simple language, but I wouldn't part with it. Why? Because it makes my job more secure? No; my job isn't in any danger. It's because Rust very often stops me from making mistakes, and it does so without making my program slow and waste memory.

    @aparrish

    💬 5🔄 0⭐ 11

Replies

  • Nov 10, 2026, 8:12 PM

    @argv_minus_one @aparrish @liberty Any time someone tells me that one language—any language—is the best, I get suspicious.

    Admittedly, I don't know a lot about Rust, though I have dabbled in it. It's my experience that the best programmers know multiple languages and can pick the one that best suits the task at hand.

    That said, I also strongly believe that the "best" language is the one that works for you, so in your case... that'd probably be Rust. 🙃

    💬 1🔄 0⭐ 0
  • Nov 10, 2026, 9:46 PM

    @me

    Yeah, I'm always anxious about bugs, and Rust does a ton of correctness checking, so I tend to use it when feasible.

    Although I am worried about its long-term viability, seeing as its LLM policy is not a complete ban. It'll lose its reliability real quick if they start vibe-coding the compiler…

    @aparrish @liberty

    💬 0🔄 0⭐ 7
  • Nov 10, 2026, 9:49 PM
    @argv_minus_one Yeah, I use Haskell for similar reasons (the type system saves me from a lot of potential bugs) though I've looked at rust for scenarios where system resources are scarce. I actually wish I understood it better.
    💬 2🔄 0⭐ 0
  • Nov 11, 2026, 3:39 AM

    @me I read somewhere that there's also LLM output in GHC these days. (one of the reasons I want to eventually make my Spiteful Lisp C Compiler share a backend with a Spiteful Lisp Haskell Compiler and a Spiteful Lisp Rust Compiler - the other reason being that I tried and failed to port GHC to Haiku)

    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 10, 2026, 8:12 PM

    @argv_minus_one @aparrish We'll always have higher and lower level languages. We can get rid of some of the cruft of the more complex languages, but people don't have to be expected to use them or deal with complexities like memory management.

    Some concepts are fundamentally harder and will require more expertise and domain-specific tools. But that's not all programming.

    I agree that Rust is pretty nice for the more complex situations (but more complicated than it needs to be still IMO).

    I am fine sacrificing some run-time performance in direct control over the computer for accessibility. I am a fan of declarative programming and functional programming as avenues for this.

    💬 1🔄 0⭐ 7
  • Nov 11, 2026, 11:35 AM

    @liberty
    @argv_minus_one @aparrish
    AFAIK the most popular programming language is Microsoft Excel. A lot of non-programmers use it for stuff that I would do with a shell one-liner or a one-off python script.
    I think there's room for improvement here.

    💬 0🔄 1⭐ 0
  • Nov 10, 2026, 10:09 PM

    @argv_minus_one @liberty @aparrish
    "Simpler" languages almost always just shift the complexity onto the programmer. So they achieve a superficial sense of simplicity, with the complexity distributed all over codebases.

    Rust is a great example. There's a fair bit of complexity in the types, but the result is that everyone who can follow its discipline benefits from overall clarity and simplicity. Similar programs in other languages distribute thinking about aliasing and ownership.

    💬 0🔄 0⭐ 6
  • Nov 10, 2026, 11:32 PM

    @argv_minus_one @liberty @aparrish

    Writing safe accessible languages is hard, but we should keep trying.

    I once taught a kid at my local library how to code using Scratch.

    Within about 20 minutes, he re-invented fork-bombing. (Future Cybersecurity engineer, I imagine.)

    The Scratch engine handled his fork bomb perfectly - filling his screen with the expected randomized bat animations, while quietly cleaning up extras when there were too many for thre processor to handle.

    Building good accessible languages is certainly a challenge, but I believe it is achievable. And it's so important to help us teach fork bombing to our next generation of Cybersecurity engineers.

    💬 1🔄 0⭐ 2
  • Nov 11, 2026, 4:42 AM

    @EdTheDev

    Writing a safe accessible language seems pretty impossible, though, because to do so would be to reconcile two mutually-opposing forces.

    Safety requires a sophisticated type system, with generics and bounds and all that. Rust has that, as do some other languages, but such languages are famously non-simple.

    A dynamically-typed language is simpler, but then you lose type safety, and have to keep track of types in your head, which is easy to screw up.

    @liberty @aparrish

    💬 1🔄 0⭐ 2
  • 💬 1🔄 0⭐ 2
  • 💬 0🔄 0⭐ 0
  • Nov 11, 2026, 5:24 AM

    @argv_minus_one @liberty @aparrish

    As I mentioned, the solution is animated bats.

    (Sorry, I'm not treally trying to argue, just trying to get a laugh.)

    I find that new programmers that I teach need accessible languages, rather than safe ones, or even fast or powerful ones.

    They need languages with great tutorials, interesting uses, and high quality error messages.

    💬 0🔄 0⭐ 3
  • Nov 10, 2026, 11:42 PM

    @argv_minus_one @liberty @aparrish

    Scheme is a very simple language, safe, and has native compilers that make it fast enough for most of what people actually need (If you need scientific number crunching Julia is a secret LISP that does that full speed)

    💬 1🔄 0⭐ 2
  • Nov 11, 2026, 2:27 AM

    @dlakelan

    I'm not familiar, but Wikipedia says Scheme is dynamically typed, which is not great for type safety.

    Full-featured static type systems are complicated, what with generics and bounds and all, but once you learn them, you can rely on them and so can everyone else who works with your code. Dynamic typing increases cognitive load and bug probability *a lot*.

    @liberty @aparrish

    💬 1🔄 0⭐ 1
  • Nov 11, 2026, 4:53 AM

    @argv_minus_one
    I have not found that to be true at all. Ive worked in statically typed languages, Pascal, C, C++, a tiny bit of ML and in dynamically typed languages, Scheme, R, Python, Prolog, CommonLisp, Erlang, Julia

    In general bottom up programming in REPL based languages has been far less bug prone for me. The kinds of bugs type systems prevent are rarely the ones I am concerned about. its more like doing wrong behavior that worries me. Seems like a preference mostly.
    @liberty @aparrish

    💬 0🔄 0⭐ 2