Improving programming language performance | Hacker News Reader

Improving programming language performance

xnacly.me

58 pointsxnacly16 comments

The syntax of your language is quite nice. I’d maybe change (for …) to (each …), (fun …) to (fn …), and (let …) to (def …) or (set …) depending on implementation details of variable assignment, but those are just aesthetic preferences :)

I love '{thing}' for string interpolation.

If you haven’t already, check out clojure, janet-lang, io-lang, and a library like lodash/fp for more syntax and naming inspiration.

Keep building!

Thanks, I think this language will probably keep my hacking for a while :). you're right about the keyword naming, i tried to be unique, maybe too much for functions, simply because i love them being named "fun". Let in the current implementation definitely is confusing due to it being used for defining and updating a variable and its contents - maybe ill change it to let and set or def and set, who knows. Thanks for the feedback :)

> reducing its execution time by 7.03 times or 703%.

Shouldn't the latter be by 86 % ?

that is true, miscalculated on my side, maybe i was to ecstatic by the improvement

Yes, but ...

> The benchmark script previously took 22.2 seconds. I reduced it to 3.3s!

this yields 85 %.

Somewhat OT, but I cannot infer the semantics of let from the examples given. Among other things it seems to have indefinite scope.

The scope is clearly defined by the parentheses, though if you're not used to Lisp syntax that can be confusing at first.

    (let n 12)                 =>  let n = 12;
    (let res (square a))       =>  let res = a*a;
    (let mod3 (eq 0 (% i 3)))  =>  let mod3 = (0 == i % 3);

In Lisp, "let" is only used for binding the variable, and it comes with a scope.

  (let ((a 1)
        (b 2))
    (+ a b))

It's closer to the ML version of let, which works like this:

  let
      val a = 1
      val b = 2
  in
      a + b
  end

You can sequence expressions for side effects, as well.

  (let ((a 1)
        (b 2))
    (incf a)
    (do-it a b)
    (+ a b))

> The scope is clearly defined by the parentheses, though if you're not used to Lisp syntax that can be confusing at first.

The scope is clearly not defined by the parentheses, and I've been programming in Common Lisp for 25 years, so I'm pretty used to Lisp syntax.

A sibling comment suggests it comes from the JavaScript let (which did not exist when I learned JavaScript; it wasn't introduced until 1.7), which is scoped to the containing lexical block, rather than a block introduced by the let form itself.

Those lets in TFA seem more like Lisp’s setq rather than let. A let expression would bind values to names only within the scope of (let …) itself. The lets in TFA and in your examples have side effects. You reference the bound values outside of (let …).

In other lisps a let form has a body that the binding(s) range over and they are not available outside. As in `(let n 12 (+ n n))` here n wouldn't be accessible after. You have `(def name 5)` for making definitions that are readable outsude.

I’m not sure but I assume it works similarly to define in scheme or to let bindings in languages like rust or JavaScript where the binding lasts until the end of some kind of implicit scope.

I'm reasonably sure that it is modeled after JS `let`, so there should be a notion of implicit scopes defined by, for example, `fun`.

You must be right. I actually didn't know JS had a "let" (which is different from "var" in that it is block scoped rather than function scoped). It's also quite different from the let in most Lisps, in that it does not establish a new lexical block.

Playing with it, the binding is scoped to the entire block, but the assignment happens in statement order so that e.g. the following will be an error, rather than logging either "1" or "2" both of which there could be arguments for:

  {
    let x = 1;
    {
      console.log(x);
      let x = 2;
    }
  }

Other way round. They're implementing a language of the LISP family, which tend to use let this way. LISP languages originated around 1960. The use of let then caught on in other languages, including OCaml and Haskell, and eventually JavaScript.

(It's possible some language other than LISP was the first to use let, but my suspicion is it was LISP.)

https://en.wikipedia.org/wiki/Lisp_(programming_language)

I'm specifically saying about `let` syntax in this particular language. LISP and related languages also have `let` with substantially different semantics, but it's evident that it wouldn't make much sense given current examples.

Reply on news.ycombinator.com