The Nineteenth Byte

Unrelated String

5:00 AM

@AidenChow it can be golfier; it's always fun ;)

Nitrodon

If it's a simple enough point-free form, sure.

Bubbler

4 hours ago, by Bubbler

More sophisticated transforms are possible but learning how to do it requires some understanding of lambda calculus and combinatory logic, so as a beginner you can just remember the site pointfree.io that does the transformation for you. (Note that it doesn't guarantee an optimal transformation.)

Jo King

@AidenChow pointfree.io is helpful, there's a golfing tip for manual conversion too

Aiden Chow

so if i were to try to do a challenge in haskell, i have to make two answers, one in non point free form and one thats in point free form and see which one is shorter?

Unrelated String

you can usually tell and it can often be useful to go in between

DLosc

5:02 AM

@UnrelatedString Somehow, in a language like Haskell, it seems like everything should be point-free. I know point-full (?) can be golfier, but it just feels wrong somehow. X^D I have the same problem with BQN.

Unrelated String

lol yeah

Bubbler

"Just post it as-is and wait until others do the transformation" is always valid :P

Jo King

if you're only using variables once or twice, it's usually golfy, but more than that or if it's in a list comprehension, it gets complicated

Unrelated String

my least favorite thing in writing jelly is when i realize explicit ⁸/⁹/³/® is actually shortest

Nitrodon

Are there any functions that can't be made point-free?

Bubbler

5:05 AM

Ranges and list comprehensions are hard

Aiden Chow

is there a way to check if a certain integer is inside a list of integers?

like in keyword in python?

Jo King

elem

you could also do a hoogle search for a -> [a] -> Bool

DLosc

@JoKing Is there any general pattern that determines why it's a -> [a] -> Bool and not [a] -> a -> Bool, or do you just have to guess the order of the arguments?

Bubbler

Hoogle will search both forms for you

DLosc

Ah, nice

Bubbler

5:10 AM

It also helps to limit search to package:base

DLosc

Well, this has been equal parts confusing, enlightening, and fun, but I really need to go to bed now. o/

Bubbler

o/

Nitrodon

What is package:base, and how is it related to Prelude?

Bubbler

Learning haskell is a highly intellectual recreational activity :P

Aiden Chow

what No instance for (Num Bool) arising from a use of ‘+’ mean?

cant make sense of it

Bubbler

5:13 AM

The base package is what you can import without installing 3rd party packages. Prelude is a part of base package which is loaded automatically for any source code.

Nitrodon

It means you're trying to add a boolean.

Aiden Chow

@Nitrodon oh shoot you cant do that?

Bubbler

Haskell is strongly typed, so you can't treat bools like numbers

Aiden Chow

ah ok, what is the best way to convert true to 1 and false to 0

Jo King

Bool -> Int results in fromEnum

Bubbler

5:15 AM

Yeah, I think that's the shortest

Limiting yourself to numeric operations in the first place is a good idea too

Jo King

By the way, the duckduckgo shortcut for hoogle is !h

Bubbler

5:30 AM

...and you can do sum[1|x] where x is the boolean

Aiden Chow

@Bubbler this does the same thing as fromEnum?

Bubbler

Same as fromEnum x to be exact

Unrelated String

it usually boils down to either "which is more useful to partially apply" or "which makes more sense if it's infix"
in this case it's the latter

Aiden Chow

yay my first haskell answer: codegolf.stackexchange.com/questions/242303/…

tell me if there are any golfs :P

Unrelated String

you don't need the parens around []

or the spaces around h-1

Aiden Chow

5:36 AM

@UnrelatedString ok thanks

Unrelated String

or the parens around t

Aiden Chow

huh, didnt realize you can do that.

wait i tried ft and it didnt work, so i have to do f t?

Unrelated String

yeah

ft would be a single identifier

Aiden Chow

huh, then why does f[] work without the space

Jo King

cos that's not a single identifier

Unrelated String

5:41 AM

because [] aren't valid characters for identifiers

at least for defining them

Aiden Chow

@UnrelatedString oh yeah, doh!

Unrelated String

not sure how [] as the name of the list type of kind * -> * works but you get the idea

Bubbler

elem (h-1) t -> h-1`elem`t

Unrelated String

most symboly stuff works like that

Bubbler

though I now realize it is the same length as elem(h-1)t

Jo King

5:42 AM

@AidenChow technically you could also move the f[]=1 down so you could do f _=1, though that doesn't save any bytes

Unrelated String

@Bubbler same length as just removing the spaces but yeah probably more idiomatic

Aiden Chow

@Bubbler oh wow ok, that was pretty obvious

what do backticks do????????

Unrelated String

infix application

Bubbler

takes any function and lets you use it like an infix operator

Aiden Chow

@Bubbler wait how does this work?

Bubbler

5:47 AM

If x is True, it evaluates to single-element list [1]; otherwise it becomes []

Then you sum it, exploiting the fact that sum [] is 0

Unrelated String

you can sort of think of list comprehensions as starting with a 1 element list the element of which can't be accessed and doesn't exist, so if all you have is a condition you get a singleton list if it's true and empty if it's false

Aiden Chow

@JoKing what does underscore mean?

Unrelated String

fairly intuitive if you think of it in terms of what multiple lists do

or even just how multiple lists' lengths relate to the result's length

Jo King

@AidenChow just a variable name, usually indicating an unused variable

Unrelated String

_ is a special ignored variable, isn't it

you can match f _ _ no problem and neither _ is accessible in the body

Jo King

5:51 AM

oh, that's actually a special syntax? i assumed it was just a stylistic thing

Unrelated String

i forget if it originated with other languages using _ prefix to silence unused variable warnings or the other way around

Jo King

aww, sum[1|r<-t,r==h-1] is one byte longer

Unrelated String

but yeah since with no conditions a list comprehension binding from one list of length a has a elements, with a list of length a and a list of length b it has a*b elements, etc., it's clear that a list comprehension that doesn't bind from any lists should have 1 element unless there's a condition to bump that down to 0

New Posts

6:04 AM

0

Q: Recursively rotate a ragged list

Given a ragged list of positive integers return a full cycle of recursive rotations starting with the unchanged input and ending with the state immediately before revisiting the initial state. Examples: [[2,3],4,5,5] -> [[2,3],4,5,5] , [4,5,5,[3,2]] , [5,5,[2,3],4] , [5,[3,2],4,5] [1,10,[2,2,4],...

code-golf

pxeger

8:12 AM

Proof Assistants is now into public beta

3

Unrelated String

🥳

pxeger

@JoKing try using an expression with _ in it, rather than a pattern with _

you'll get a type error: _ is called a "hole"

you can use them to find out what type the Haskell type checker wants an expression to have

like if you have [] >>= (what goes here?), then write [] >>= _ and it will tell you what type is wanted

Unrelated String

i never knew you could do that
that's cool

SegFaultPlus4

10:05 AM

Chessle 32 (Normal) 4/6

🟩⬛⬛⬛⬛⬛
🟩🟩🟩⬛⬛⬛
🟩🟩🟩🟩🟩⬛
🟩🟩🟩🟩🟩🟩

https://jackli.gg/chessle

PyGamer0

@NewPosts wow there is a lot of ragged list questions now...

SegFaultPlus4

what is your bullet rating on lichess

PyGamer0

i read that as itichiness lmao

SegFaultPlus4

oof

Unrelated String

i just always parse lichess as lich-ess instead of li-chess

SegFaultPlus4

10:09 AM

well what is it

PyGamer0

idk sounds like lychees

SegFaultPlus4

here it is

lyxal

10:27 AM

I pronounce it as li as in likesal and chess as in chess

SegFaultPlus4

yes but what is your rating

ophact

@SegFaultPlus4 For some reason I was expecting a rickroll

5

SegFaultPlus4

i've learned to wait

PyGamer0

11:04 AM

i have been thinking to disable autocompletion and lsp on my editor..

Ginger Industries

~~oh no~~ yay haskell

Razetime

oh yum haskell

i wonder where i left off in the tutorial

PyGamer0

do you has kell?

New Posts

0

Q: Randomly Rounding

Input a decimal number, round it to integer. It randomly round up or round down based on its fractional part. If input \$x\$ is an integer, the program should output it as is. If input \$x\$ is non-integer, the program has \$x-\left\lfloor x \right\rfloor\$ probability output \$\left\lceil x \rig...

code-golf math integer random

Razetime

not very well, pygamer.

PyGamer0

11:14 AM

@Razetime did you know the flax is going to be evaluated from right to left now

Razetime

nope i did not

PyGamer0

11:26 AM

argh i hate my editor auto importing

user

@pxeger ooh, i thpught? Was just a thing agda had

Ginger Industries

12:14 PM

> thpught

In celebration of the Haskell LYAL:

The problem with Haskell is that it's a language built on lazy evaluation and nobody's actually called for it.

5

mathcat

12:33 PM

haskell!

I've always wanted to learn haskell

(what is haskell?)

PyGamer0

a functional, statically typed language

i think

its cool

lyxal

Is it really pointfree though?

I can see 4 decimal points right there :p

hmmm

@lyxal lmao

sūs

12:35 PM

@mathcat Haskell Brooks Curry was an American mathematician and logician best known for his work in combinatory logic.

@lyxal wiki.haskell.org/Pointfree#But_pointfree_has_more_points.21

lyxal

@lyxal I've waited all day to make that joke, looking for an opportune time when there wasn't on-topic haskell learning

pxeger

> A common misconception is that the 'points' of pointfree style are the (.) operator (function composition, as an ASCII symbol), which uses the same identifier as the decimal point. This is wrong. The term originated in topology, a branch of mathematics which works with spaces composed of points, and functions between those spaces. So a 'points-free' definition of a function is one which does not explicitly mention the points (values) of the space on which the function acts.

Adám

@lyxal No point in an APL solution, is there? ⊃+×/1↓⊢

PyGamer0

of course there isnt

lyxal

@Adám hence why APL is superior

mathcat

12:36 PM

@pxeger I'm his #1 fan now

PyGamer0

^^

lyxal

@pxeger wow it's almost as if that's the joke

Adám

@pxeger We have Haskell and Currying. Why no "Brook(s)" in programming?

Ginger Industries

@Adám about to say "what is it with you and APL" and then I remembered you work for them

pxeger

BrookGPU

The Brook programming language and its implementation BrookGPU were early and influential attempts to enable general-purpose computing on graphics processing units. Brook, developed at Stanford University graphics group, was a compiler and runtime implementation of a stream programming language targeting modern, highly parallel GPUs such as those found on ATI or Nvidia graphics cards. BrookGPU compiled programs written using the Brook stream programming language, which is a variant of ANSI C. It could target OpenGL v1.3+, DirectX v9+ or AMD's Close to Metal for the computational backend and ran...

Adám

12:46 PM

@GingerIndustries Not the whole story. APL has always been a part of my life, way before I got an APL job. I think I can rightfully call myself a native APL "speaker".

pxeger

Dec 19, 2018 at 22:09, by Zacharý

> Adám prides himself on having got APL with the breast-milk, as he attended his first APL conference at the age of one

Ginger Industries

._.

lyxal

1:01 PM

So turns out I'm so used to VS Code autoformatting code on saving and Copilot's autosuggestions that I get confused why MS Word isn't do either those things when I'm doing report writing

I sit there with an uncompleted sentence thinking "hey wait where's the suggested auto-fill?"

and think that something's wrong when nothing gets reformatted when I save the file

mathcat

wut

You don't write text in VS Code?

is everything ok?

lyxal

Not when it needs actual rich text formatting

and not when it's for uni

gosh what is it with everyone questioning why I'm using MS Word?

like really, is there something so terribly wrong with using the Office suite in 2022?!

Adám

No, Microsoft Word is quite good.

mathcat

Try Notepad, it might auto-fill

lyxal

@mathcat not like copilot does

pxeger

1:07 PM

MS Outlook can do autofill

lyxal

@Adám I'm glad someone understands it

Ginger Industries

1:19 PM

that is not where that goes

lyxal

well where do it go?

Ginger Industries

on the outline

go to bed

lyxal

well without any prompting from anyone tonight, I'm going for bed new

*now

o/ nerds

Ginger Industries

@lyxal oh really

mathcat

\o

PyGamer0

1:29 PM

÷/

mathcat

1:55 PM

CMC: Make a rickroll detector: Given a link, check for instances of dQw4w9WgXcQ in the get request and return True/False based on that.

Bgil Midol

@mathcat lambda x:"dQw4w9WgXcQ"in x

PyGamer0

VTC as unclear :p

Bgil Midol

^

mathcat

why did I click delete lol

PyGamer0

lmao

in the k tree, 1 min ago, by PyGamer0

idea: k plotting library using braille characters

plotting library using braille characters sounds like a nice idea to me

mathcat

1:58 PM

@mathcat Given a Link, invoke a get request, and in the HTML code of the link check for instances of "dQw4w9WgXcQ"

Bgil Midol

@mathcat tio.run/##K6gsycjPM/7/…

(should work)

100 bytes

mathcat

nice

working

[Vyxal, 10 bytes](https://‮(mathcat.surge.sh/JiIsIiIsICSrJCLiICLiIyW/moc.erehwynanohtyp.laxyv

damb

link somehow not working

Ginger Industries

@mathcat needs to start with https://

mathcat

hmm no

you'll have to copy-paste it then

SegFaultPlus4

2:29 PM

i crossed 1100 rep

mathcat

gg

I crossed 2894 rep

SegFaultPlus4

only 8900 more until ultimate power

Ginger Industries

I need a name for my tiny microcontroller board

mathcat

Mr. Chill

Ginger Industries

serious names only please

mathcat

2:38 PM

aw

Ginger Industries

currently it's called the Pyco but I don't really like that name so

anyone else?

PyGamer0

3:02 PM

@GingerIndustries Mircಠ

or CnYB

DLosc

5:18 PM

I just realized, during the Haskell LYAL, I've mostly been typing HMQ when I meant to type HMC ._.

SegFaultPlus4

@GingerIndustries miCro

since C was written in assembly

DLosc

HMC: ascii-art Given a printable ASCII character c and an integer n greater than 1, return an n by n square of characters where the outer borders are cs and the inner characters are spaces. For example, if the character is # and the integer is 5, the output should be:

#####
#   #
#   #
#   #
#####

SegFaultPlus4

@DLosc c=input();n=int(input());print(c*n+"\n"+(c+" "*n-2+c+"\n")*n-2+c*n)

ez

Nitrodon

I don't suppose there's a faster way to write take n . repeat?

SegFaultPlus4

5:36 PM

idk i have a pretty solid 67 bytes

pxeger

First attempt, 63 bytes:

Haskell, 63 bytes: f n=q"#"n:q['#':q" "(n-2)++"#"](n-2)++[q"#"n];q=flip take.cycle

SegFaultPlus4

@pxeger lambda c,n:print(c*n+"\n"+(c+" "*n-2+c+"\n")*n-2+c*n)

53 bytes

your move

pxeger

@pxeger Oops, didn't notice there was an input c. Here's 1 byte shorter:

Haskell, 62 bytes: f n c=q c n:q[c++q" "(n-2)++c](n-2)++[q c n];q=flip take.cycle

SegFaultPlus4

it appears i have outgolfed pxeger

this is my greatest moment

pxeger

in a different language

doesn't count

SegFaultPlus4

5:46 PM

ye but still

pxeger

and your answer doesn't even work

you need brackets around the n-2s

ato.pxeger.com/…

SegFaultPlus4

i have an avearge score of 2 on my posts, let me have this

@pxeger ok then, 57 bytes

DLosc

Haskell, 67 and pretty clumsy: f c n=[[[' ',c]!!sum[1|any(`elem`[1,n])[x,y]]|y<-[1..n]]|x<-[1..n]]

SegFaultPlus4

oh wow, 1st of 3

57 is the score to beat

pxeger

I mean yours can be pretty easily be shortened by outputting a list of lines (which is a standard allowed I/O method)

45 bytes: lambda c,n:[c*n,*(n-2)*[(c+" "*(n-2)+c)],c*n]

SegFaultPlus4

5:59 PM

@pxeger you can't have a * at the start of an expression

pxeger

@SegFaultPlus4 I mean you literally can, it literally works

SegFaultPlus4

i did not know that

DLosc

@SegFaultPlus4 It's a form of argument unpacking

SegFaultPlus4

i learned something new today about Python

ophact

@DLosc What's HMC??

I know what CMC is but what's HMC?

DLosc

6:10 PM

@ophact Haskell Mini Challenge

ophact

Uh... okay.

DLosc

because of the Learn You A Lang event mentioned on the starboard

SegFaultPlus4

then i propose that LAYL #16 is Python

ophact

I don't really get what the benefit of the LYAL is, could you explain?

caird coinheringaahing

Basically, it's a mini LOTM

ophact

6:11 PM

But with no bounties right?

DLosc

@SegFaultPlus4 Propose it here

caird coinheringaahing

Yeah, there are no bounties (unless people want to offer them)

SegFaultPlus4

@DLosc alright then

DLosc

@ophact The original proposal goes over the whys and wherefores

ophact

I wonder how angry everyone would get if I were to propose JS

-15 score probably

caird coinheringaahing

6:13 PM

@ophact The general aim is to address the fact that LOTM is kinda restrictive in the languages that work. Common/popular langs are discouraged from LOTM, as are languages that are harder to use such as Malbolge or Lost. With LYAL, we just encourage people in chat to talk specifically about that language, post CMCs relating to the language, and answering with the language, as a bit of a "Give it a try", and because it only lasts a day, people can try out weirder languages "easier"

SegFaultPlus4

JS v Python, which one gets more downvotes

DLosc

@ophact Given that LYAL is explicitly open to languages that don't qualify for LotM (either because they're too hard to use or because they're too popular), JS would be fine. I'd upvote it.

ophact

@cairdcoinheringaahing Maybe I should learn Haskell then

SegFaultPlus4

this will solve the debate of which one is more popular

@DLosc is Python open to LoTM

if so, i'm gonna be making bank

ophact

Question: which language powers this website? JavaScript obviously.

DLosc

6:14 PM

@SegFaultPlus4 No, because it's commonly used already.

The idea of LotM is to promote less-used languages.

caird coinheringaahing

753

Q: Which tools and technologies are used to build the Stack Exchange Network?

What tools and technologies are used to build the Stack Exchange Network? See also: Which tools and technologies are used to build Data Explorer? Return to FAQ index

support faq stack-exchange technology

SegFaultPlus4

@DLosc oof size: large

ophact

Why does this Haskell thing error? I'm probably missing something really obvious here

SegFaultPlus4

@ophact have u heard of P y t h o n 3

ophact

We wouldn't be chatting if not for JavaScript.

caird coinheringaahing

6:19 PM

@ophact I believe Haskell needs a main function for full programs

e.g. The hello world

SegFaultPlus4

@ophact gotta remake TNB in Python, cya in 5 years

ophact

@cairdcoinheringaahing Okay... because I just pulled off the printf thing off the wiki

caird coinheringaahing

Huh weird, apparently printf isnt a thing in Haskell

Keep in mind, I know very little about Haskell :P

• Variable not in scope: printf :: Integer -> t
• Perhaps you meant ‘print’ (imported from Prelude)

DLosc

@pxeger 57 bytes using replicate

ophact

Haskell isn't really for me sorry

I'll wait another two weeks

DLosc

6:24 PM

@ophact See my question here and the response here

Or TL;DR: main = print 5 works

ophact

Yeah I sorta gave up on learning Haskell ;) as I said, I'll wait another two weeks

If I'm correct the next language will be F#

DLosc

@ophact That's a pretty quick give-up. :( But yeah, it's not the easiest to learn (which is why I was looking forward to this event--I'd never been able to learn Haskell from the online tutorials, so I wanted to learn it from people I could ask questions of).

caird coinheringaahing

@ophact There are 5 langs all currently tied on 4 votes, so it could be any of those

DLosc

@ophact Possibly. Depends what the vote totals look like in two weeks, and whether someone's available to teach F# at the time.

caird coinheringaahing

Plus, there's no requirement to pick the highest voted nomination, that's just what's normally done :P

DLosc

6:28 PM

It could even be some other language that hasn't been nominated yet, if it gets nominated and a lot of people vote for it.

ophact

If I eventually wanna learn Haskell, I'll probably have to devote a lot of time to it ;)

caird coinheringaahing

@ophact Technically, C# has played a pretty large role as well :P

DLosc

@DLosc Sadly, refactoring makes it 3 bytes longer

@pxeger Alternately, 57 bytes by using q one more time

emanresu A

7:02 PM

@lyxal Vyxal, 3 bytes: ÷*+

@mathcat That's evil

@cairdcoinheringaahing Malbolge LYAL. I don't think even Kamila understands it enough to teach.

caird coinheringaahing

@emanresuA It's nomination is currently at -3

emanresu A

Yeah lol

caird coinheringaahing

@emanresuA Kamilla wrote a goddamn Lisp interpreter in Malbolge, I think she understands it plenty :P

DLosc

@DLosc 52 bytes! Lexical scope FTW

emanresu A

@cairdcoinheringaahing She's created some ingenious toolchain that compiles assembly to Malbolge. She does understand it plenty, but I don't think even she can just type up a program that adds two numbers.

And I doubt she'd be willing to share said toolchain.

DLosc

7:10 PM

51 bytes

caird coinheringaahing

@emanresuA I think the main problem with Malbolge as a LYAL is that it isn't really a language you can mess around with. Most of LYAL is just playing around with a lang to get it to do things, Malbolge, you don't really write programs, so you can't really play around with it

emanresu A

Yeah, that's pretty much what I was trying to say

Ginger Industries

rate my board layout

emanresu A

Sus.

Ginger Industries

???

what makes it sus

there are no impostors here

emanresu A

7:16 PM

Ginger Industries

ಠ_ಠ

mathcat

@emanresuA he he

gonna use that one more often

Neil

7:46 PM

hmm, several downvoted answers on int, float or neither

mathcat

8:08 PM

CMQ: Should I use comic sans in a comic?

emanresu A

Only use comic sans ironically.

Unrelated String

8:27 PM

and sparingly even then

pxeger

TIL Haskell main doesn't have to be IO (), it can be IO anything!

emanresu A

Cursed idea: Recursive program

pxeger

main = main?

emanresu A

That's even more cursed

pxeger

I don't see anything cursed about it

it's the most obvious and most idiomatic way to write a full program which loops forever in Haskell

emanresu A

8:37 PM

I guess...

Is there such a thing as a niladic function in Haskell?

If so, how do you call it?

pxeger

All expressions in Haskell are lazily evaluated

so there's no difference between a "niladic function" and any single value

emanresu A

oic

I meant something like Try it online! btw, although that doesn't work for some reason :(

Unrelated String

@emanresuA tio.run/##K6gsycjPM/7/… ?

emanresu A

9:01 PM

Ooh, thanks :) never seen a BlockingIOError before

Unrelated String

same

user

9:34 PM

@cairdcoinheringaahing It most certainly is

smh y'all here pointing out "flaws" in Haskell and they're not even the real flaws

caird coinheringaahing

oh right, but it takes a string, not an int

user

The example has printf "%s, %d, %.4f" "hello" 123 pi

Unrelated String

i would hope that printf would take a string lol

user

It seems to use that cursed Haskell style varargs WW mentioned once

caird coinheringaahing

> but Haskell's type system makes this hard

Unrelated String

9:36 PM

if you want to print an int just use print since ints are Show

caird coinheringaahing

Sum up Haskell in one sentence

Unrelated String

@user yeah i'm lowkey horrified by this

user

lol

This seems like it's massively abusing the type system

Which I am 100% okay with because it's at least doing it in a smart way

Unrelated String

yeah

user

Having a unified type system (is that what it's called?) like Java would help with this, but it'd ruin the amazing type inference and stuff

Actually wait no

Can you have existential types or implicit conversions in Haskell with language extensions? If so, it could be doable with a list of arguments instead of varargs

Unrelated String

9:39 PM

there's an existential types extension yeah

user

Actually, existential types aren't enough because you also need to pass in an instance of the typeclass or whatever

Unrelated String

been a while since i messed with it but it should work for this

user

The real challenge is having the format string be verified at compile time without macros. TS is the only language I know that can do it

Unrelated String

yeahhh

user

Which is sad because if you could match on strings in Scala 3 it'd be seriously OP

Unrelated String

9:41 PM

i'm pretty sure you just couldn't

you could probably implement something that works like printf that's correct at the type level but it would definitely be a different syntax than format strings

user

Can you derive an instance for a tuple H *: T if you have an instance for T?

Unrelated String

like in general? probably not

user

Aww

I guess you'd have to hack up your own tuple type then

Unrelated String

i may not quite know what you're asking about

also "hacking up" your own tuple type is sort of what product types are lol

user

Like data MyTuple = EmptyTuple | Head *: Tail?