LispE: Lisp Interpreter with Pattern Programming and Lazy Evaluation

> LispE provides an alternative to parentheses with the composition operator: "."

That is a... Choice.

Breaking the pair operator in favour of something new.

arethuza · a month ago

After programming in Common Lisp for a few years (a long time ago) and then later on having a brief period where I was fond of Python, I did also become fascinated with the concept of lisps where indentation replaces parenthesis such as Wisp:

https://www.draketo.de/software/wisp

Mind you - I usually end up concluding that Lisp syntax is actually pretty good as it is...

vindarel · a month ago

There's a new one, pretty good, resembling Python/Julia syntax, check it out! https://moonli-lang.github.io/

    defun multiply-thrice(x):
      print(x * x * x)
    end

    multiply-thrice(23)

petre · a month ago

It all ends up like YAML. Thanx but I'd rather take the parens any day.

shakna · a month ago

I've always been tempted with wisp. Ever since I saw SRFI-110. Love the concept.

I just never quite manage to grasp the new syntax.

bunderbunder · a month ago

They’ve got a page on that. They did away with linked lists and chose to represent them as vectors. With some of the usual stuff you see going on under the hood in this style of list on imperative languages, like pre-allocating a little room for growth.

I can’t opine on whether that’s a good choice. But I will observe two things: first, singly linked lists aren’t as great on modern computing architectures as they were 50 years ago. Locality of reference matters a lot more now. And second, both Hy and Clojure abandoned the traditional focus on dotted pairs, and in both cases I found it was fine. (Disclaimer, I didn’t spend a whole lot of time with Hy.)

shakna · a month ago

Uh... Most Schemes don't use linked lists under the hood. That doesn't impact the syntax, just the implementation.

Guile uses vectors, for example. [0]

Chicken... Is Cheney, so your list disappears entirely, half the time.

Gambit uses vectors. [1]

[0] https://www.gnu.org/software/guile/manual/html_node/Cheaper-...

[1] https://github.com/gambit/gambit/blob/master/gsc/_back.c#L11...

sph · a month ago

Yeah, that's pretty unclean on two aspects: breaks pairs, and breaks the orthogonality of s-expressions

A simple macro would've sufficed, say:

  (compose
    sum
    (numbers 1 2 3))

shiandow · a month ago

I don't think it's too bad orthogonality wise, though it is a bit weird to introduce infix notation. It would almost make more sense to write

((. sum numbers) (1 2 3))

jnpnj · a month ago

and beside multiple-args, there's the usual threading macros

    (-> [1 2 3] f g)

mchaver · a month ago

It's not too bad. I like it! Haskell uses "$" to do the same thing.

shiandow · a month ago

Technically $ means something slightly different, it is more somilar to putting parentheses around the right half of the expression. For function composition it uses the same '.' .

shakna · a month ago

Well, you could use $ in Lisp, too. Thats a standard valid symbol, that doesn't have a builtin meaning.

tkrn · a month ago

I agree that there is maybe too much potential for confusion with that, but is the dot operator (or read syntax?) actually used that much these days?

Personally I have mostly sometimes used it with Emacs Lisp, but in general relying too much on plain cons cells and cadring down the cars of their cddars feels like a code smell to me and if I need a pair I can always just use cons? As the (only, I think?) infix operator in traditional lisps it has always felt extra-ordinarily useless to me (outside of Schemes use of it lambda lists), but maybe I'm just missing something.

shakna · a month ago

Pairs are used by about 3-quarters of the standard library of Scheme, so I really would not consider its use to be a code smell.

You should be using the pairs when using make-hash, for example.

Cons also doesn't always return a pair. Its main purpose is for prepending to a list. Only when "the second argument is not empty and not itself produced by cons" does it produce a pair.

Which means '(a . b) is clearer code in intent, than (cons a b).

wiml · a month ago

I think you need something like it in order to have a print representation of a cons cell whose cdr is not a cons-or-nil. And it's nice if your print representations are readable.

mghackerlady · a month ago

I honestly would've prefered someone try and turn xml into a lisp, at least that has a cool hack value