(foldl (q(A B) (* A B)) (1to 10)) What am I doing wrong?

I know I could just give * to foldl but I'm trying to figure out how to supply an arbitrary lambda.

@chunes Looks like you're passing multiple arguments to q. Try (q ((A B) (* A B))) instead.

Oh, something really dumb. I'm pretty new to lisp, you can probably tell. Hard to keep track of all the parenthesis

Yeah, it can be a bit confusing because the two ways to define a lambda are (q ((A B) (* A B))) and (lambda (A B) (* A B)). lambda is a macro that takes two arguments, the function args and the function body. q is just the quoting builtin, so it always takes one argument; to make a function, that argument should be a two-element list.

Makes sense

Also, your commented/ungolfed answers have been golden for getting up to speed. Nice work on those

codegolf.stackexchange.com/questions/242257/… the lack of negative numbers makes this unendingly painful

@DLosc that moment when string lib with no split function

(d split (q ((A C S) (if A (if (e (h A) C) (c (reverse S) (split (t A) C ())) (split (t A) C (c (h A) S))) ()))

done

@chunes Hey, thanks! Glad to hear that it's helpful. I'm a big proponent of explanations.

@Razetime Yeah, string support is... "basic" is too generous. It was very much an afterthought--tinylisp was always supposed to be just integers and lists, but somebody asked for strings at one point and I thought, "Well, I guess it wouldn't hurt to treat names as strings."

names as strings is so funny

i love that idea

but the result is you always need a quoted codepoint array to represent string

which isn't the most terrible thing

@Razetime It can lead to some funny situations if you play around with it. For example, the empty string is a perfectly valid name that you can define; you just can't refer to it directly, you have to construct code and eval it. TIO

@Razetime Not sure what you mean--any string without spaces or parens in it is just (q your_string_here)

@DLosc yeah but obviously we need both

and newlines lmao

DLosc might have to rename it to Sorta minimal lisp afterward

@Razetime Yeah. That's why the little that is in string.tl does include space, newline, and converting a codepoint to a single character. (I'm a bit disappointed looking at it now that I didn't think to include an empty-string definition, but (string ()) isn't so bad I guess.)

well tinylisp ain't made for golfing

@chunes This is part of why I've historically resisted adding more features to the core language. It's supposed to be tiny. :)

but that bracket autocompletion is such a godsend sometimes

True dat.

HBL is me trying to take that idea as far as it'll go and make a Lisp that's actually golfy. I think it could potentially end up being the golfiest Lisp dialect (right now it's suffering from being severely underdeveloped), but I don't think any Lisp will ever be as golfy as the tacit and stack-based languages. Parentheses just cost bytes, there's no way around it.

Unless... we use capitalization to encode nesting levels

please forget I said that

It wouldn't even work anyway if all the names are 1 byte

But a lisp program that looks like forth would be a trip

Also everything would have to have specific arities, I guess. So total failure of an idea

That's the one thing that is really different about lisp that will take some getting used to. The idea that you get varargs 'for free'

@chunes Failure is just part of the process of generating good ideas.

Oh, I guess tinylisp doesn't work that way.

(a 1 2 3) throws an error where (+ 1 2 3) works in most lisps I think

Yeah, tinylisp is unusual among Lisps in that way. I did it that way to make it simpler to implement.

But there are variadic versions in the library.

I see

mul2 and div2 are dyadic, while * and / are variadic (and product takes a single argument, a list).

(d C(q((N R)(i N (i R (a (C (s N 1) (s R 1)) (C (s N 1) R)) 0) 1)))

is this a correct combinations function

sometimes i wish tinylisp pointed to the paren that was wrong

@Razetime ♪ CS tinylisp will make each day a quest to find a missing close-paren ♪

@Razetime Looks like the parentheses all match up correctly (though you're missing one at the end if you don't want to rely on autocomplete parens). The recurrence relation looks right, but I think the base cases are wrong. This should give (1 4 6 4 1).

For comparison, I implemented it using the definition.

That probably won't be the golfiest way, but it gives you something to check against.