« first day (4112 days earlier)      last day (107 days later) » 
00:00 - 20:0020:00 - 00:00

12:03 AM
@hyper-neutrino I don't think the english speakers know how to pronounce y
try the german one
or french
those languages should have the sound
12:40 AM
Thank you :)
1:00 AM
Sandbox posts last active a week ago: How normal is this group?
this group of people? Not at all.
@Steffan for the infinite candle sequence q, i have a solution thats 64 bytes (saves 12 bytes over ur 76 bytes answer), without just blindly copying over the previous solution :P. not gonna share it immediately to see if u can get the solution (or maybe even find a better golf!)
1:15 AM
i was just editing a gradle build script and was typing out freeCompilerArgs. just a few minutes earlier i was researching phytoplankton, and one of them produces "yessotoxin". so with those 2 words in my mind, my fingers ended up typing freeCompilerToxins...
I prefer either nosootoxin or deadly neurotoxin
2 hours later…
2:50 AM
@AidenChow well i could shorten if desmos supported recursive funuctions
by a lot
but idk
@Steffan true, but i wasnt thinking about recursion
maybe actions would make it shorter?
if i could figure out how
i dont understand the recursion answers, so i cant really help you with that :(
Well it's easy
First you start by understanding recursion
And then you know how they work
ok i can pretty much only understand the python answer and thats it
not too sure what the and is doing
2:59 AM
haven't seen the code you're talking about but it's probably for short-circuit evaluation; i.e. a funky conditional
this one: f=lambda a:a and(f(a//2)+a%2)/2
^^ yep
so like... what does it do tho
it returns a if a is falsy or recurs if it's truthy
3:17 AM
@NewPosts sorry for that mistake :/
3:37 AM
@AidenChow well actions did not make it shorter
it made it 1 byte longer
yeah well, actions tend to be a bit bulky to set up
that was porting the C answer
also i needed a newline at the beginning, but you can't have a newline in a ticker, so it wouldn't have really worked anyway
@Steffan hmmm, well ill give a hint: my answer uses the base 0.5 trick that some other answers are using
3:57 AM
This may be an odd query: Could I have some feedback on this challenge proposal, please? I have had it in the sandbox for merely a day, so this may be premature. What is wrong with it? What can I improve?
4:49 AM
in the k tree, 4 mins ago, by PyGamer0
gol in k: done!
i guess i am getting better at k :)
5:12 AM
@Pyautogui given that im already trying to think of smart ways to solve it, id say its definitely a pretty good looking challenge
i guess my main question, which may have something to do with my already planned answer, is whether the input values for T, F, and U have to be the same as the output values
I almost have a 1-byter
@Pyautogui I have a 1-byter
In many languages
And a 3-byter in some more
Simply mark F as 0, U as 1 and T as 2, and take the minimum of the two values.
1 hour later…
6:22 AM
anyone got any concurrency jokes? because that's what I'm learning rn and unlike recursion I don't have any haha funny
most of them are meta-jokes, like
6:49 AM
like what
or is that the joke? nothing?
sounds like the send process interrupted the punchline ;p
7:03 AM
sorry, i lost the thread
> Knock Knock.
Race condition!
Who's there?
@Adám Lol
> Hard right concurrency to is get.
7:24 AM
@Adám What does that do precisely?
On each word, it launches a print lambda in a new thread.
No. The ⎕DL.1 is just a delay of 100ms to let the threads finish.
@emanresuA Explanation.
1 hour later…
9:04 AM
Let's stop now
What the hell happened here?
I go away and have a shower and this is what I come back to
:61035211 What did this say?
9:19 AM
Hm, it's a ping
9:52 AM
@emanresuA it was a bad joke
so unfunny that I deleted it
1 hour later…
11:15 AM
@Adám Why does that always output in the same order?
@pxeger Not sure.
Does APL just use native Linux threads?
If so, it's probably Linux's fault
These are APL threads sharing a single OS thread, like Java "green threads". (OS threads are also available.)
Happens on ATO too. Oh, and you still need to rename "Dyalog APL" to "APL (Dyalog)".
Huh, you've got ngn/apl, but I can't see it on the site.
11:23 AM
It's disabled (see the comments /* */?), because it doesn't work
Oh, missed that.
It'd be lovely to have GNU APL there, if you can figure out how to.
Shouldn't be too difficult, I think
@Pyautogui it’s not an odd query at all, thanks for asking for feedback first
It looks like the rules section is unnecessary because those are the site defaults
You might want to just have the truth table and remove the description before it, which is mildly confusing
Maybe specify that T, F, and U are three distinct values
1 hour later…
12:41 PM
No way hidden github reference to pxeger?
oh i see it.
"github desktop removes branches after 14 days of inactivity" they said
"we clean things up for you" they said
yet here we are
4 branches in my branch list
all past 14 days and all merged
1:05 PM
@lyxal yes, I singlehandedly invented regular expressions
regexps were named after me, definitely not the other way around
of course, of course
that's like one of the fundamental things every programmer knows
1:31 PM
pxeger, the one who pxegs, invented the regexp, gexps that er
It seems to make unnecessary use of mathjax. It makes the question look way more complicated than it is.
If your challenge can be described in a sentence of normal english starting the question with "let [symbol] mean x is a substring of y" is pointless
Much better
2:00 PM
Q: Convert JSON object of directories to list of paths

Matteo C.Task The input consists of a JSON object, where every value is an object (eventually empty), representing a directory structure. The output must be a list of the corresponding root-to-leaf paths. Inspired by this question on StackOverflow. Input specifications You can assume that that the input...

2:29 PM
Thanks for the feedback @user. I will fix those problems. @emanresuA, is the fact that you have a one byte-er solution so quickly a mark of an uninteresting problem, or is it still a challenge worth posting?
Sorry for my very late response.
@Pyautogui It's an interesting problem, but it's just a shame that the shortest solution is not an interesting one
Got it. Thanks for the quick reply @pxeger. I do suppose that is a bit of a shame.
You could make it more interesting by requiring a specific representation of T/F/U rather than allowing answers to choose
I will do that! Makes sense.
3:25 PM
hmm i just found out copilot will be paid in the future... hopefully they make it free for the preview users or im toast 😅
They're like drug dealers, getting you addicted then making you pay for it :p
tbh i can live with intellij's autocomplete
cant stand coding in notepad tho, i need something to help me find my methods
Just get an openai account and use its edit mode as an IDE. That mode's free for now. /s
(although now I need to test this)
i need something while im typing
3:29 PM
It filled in code that's scarily similar to my actual main class, even though I only gave it the very beginning portion.
It wouldn't run because it's missing a few things and it has no ideas what functions my other classes have, but it's very close.
i wanna give tabnine a try, its supposedly gpt powered
so, I think I can change my code to let me input koths from anyone
who wants to volunteer?
what are you talking about?
my latest project
I would be willing if it works with the Java KoTHs. Gradle doesn't work on my computer for some reason though, so IDK if that would intefere.
I have King of the Holster running right now which might be a good test case.
3:32 PM
unfortunately, we're Python-only for right now
but in the future it will support all languages rto.run supports (allthough probably not multi-language)
80% of koths i saw are java(script), so might wanna work on those first
@Ginger what's that?
also GKOTH sounds dumb
please pipe your criticism to /dev/null
3:34 PM
My next KoTH will probably be Python (since it involves coordinates and vector math, which I don't want to deal with in Java) so maybe that one? Should submit something to the sandbox in a week or so (want there to be a bit of a break between King of the Holster and this one).
@Ginger tbh mathcats right
you can tell from looking at the name the two most important aspects of the software:
1. I made it
2. It runs KOTHs
i actually had no idea what the G was for
@lyxal onions
it should be catchy
if you want to get rich with GKOTH
we considered AAAAAAAAAARGH but that's confusing and hard to understand
3:36 PM
so what is the G for?
what does gkoth do?
automatically run KOTHs
BOOMKOTH (builder for overcomplicated online machines)
3:38 PM
*not amused*
AutoKOTH could even work
3:42 PM
Idea: In order to prevent homoglyphs from being a problem in usernames, display a truncated hash of the username next to it
A bit like Discord?
but how often does that actually become a problem?
Normally, not often. If someone's intentionally being malicious, probably often.
@AidenChow got it
A: Infinite Candle Sequence

SteffanDesmos, 65 bytes f(n)=total(.5mod(floor(n/2^l),2).5^l) l=[floor(log_2(n+0^n))...0] Try it on Desmos!

3:48 PM
that does not work
who said?
seems to work fine?
@Ginger I used to be a snake_case hater like you until I took an arrow in the knee
I'm sorry used to be?
3:56 PM
Now I tolerate it when writing Python, although I still prefer camelCase.
finally, another camelCase lover
Or kebab-case, though that doesn't work in languages that aren't Lisp
snake case is unequivocally the best case
I don't like typing underscores.
snake_case is exactly the same as normal english, where spaces are replaced with underscores
3:57 PM
Finally, someone gets it
counterpoint: I don't like typing capital letters
countercounterpoint: snake case requires shift + underscore but camel case only requires shift
@DLosc yep
3:58 PM
but camel case is less readable than snake case because there's no gap between the words
countercountercounterpoint: there are more capital letters than underscores
If you're that fussy, you could swap = and _ on your keyboard or something
countercountercountercounterpoint: camel case is better for code golf
@pxeger Fun fact: in QBasic, variable names cannot contain underscores. They can, however, contain periods.
3:59 PM
countercountercountercountercounterpoint: single letters are better
@Seggan when do you ever use multi-word variables in code golf?
@pxeger This is also something I like about Haskell: you can have apostrophes in names as well
when the challenge is reducing clutter in my code
@Seggan clutter ≠ byte count
capital letters are clutterier than underscores
but shorter
i hate having lines longer than 60 chars and split in weird spots
and this is the code golf chat, isnt it? ;)
Personally I think both are fine as long as you don't mix them
4:02 PM
countercountercountercountercountercounterpoint: Just allow for spaces in variable names already!
but spaces are useful for so much other syntax
Use nbsps in variable names
@pxeger I take it you don't like Fortran then
Y'all are overcomplicating this
@Romanp causes confusion
4:03 PM
var x y = 1; var x = n -> n + 1; var y = 2; x x y == y
Yes, but as long as you don't have both implicit multiplication AND spaces in variable names things should be fine.
and force a strict order of operations. Doesn't have to be APL-level, but still make it read in one direction primarily
countercountercountercountercountercountercounterpoint: what is the var name in a = 1
@Romanp Or have a Haskell-like syntax but with a glyph for high-precedence function application instead of juxtaposition.
Can we stop with the /(counter)+
4:06 PM
@BgilMidol Spaces would only be allowed internal to names, not on the ends
or at least golf our syntax. counter^8point: your syntax is either invalid or the expression name is "a "
Ruby lets you use spaces in variable names: ato.pxeger.com/…
so does kotlin iirc, if wrapped in backticks
@pxeger I thought it would be something like this ^_^
Ugh I hate those backtick variable things kotlin and scala do
4:07 PM
radvylf why is that whenever you say you hate something you're wrong
JS sort of has that...window["this variable name has spaces in it :o"] = 1
@RadvylfPrograms me too
@RadvylfPrograms Python equivalent: vars()["name with spaces"] = 1
@pxeger I'm clearly right here. `a b c + d` + name_of_thing is just stupid
it's useful if you want to name a variable as a keyword
4:09 PM
Allowing spaces in variable names, unless it's the only option, is totally wack, especially in C-style syntax
@pxeger Then you have `class` or whatever which is ugly. Just add an underscore to the end or pick a different name
Like input_class or function_on_finish or var_to_inc
useful for `in` for naming InputStreams
Counterpoint: in is a bad variable name anyway
val `in` = FileInputStream(...)
const in_stream = FileInputStream(...)
Much nicer
4:11 PM
@RadvylfPrograms no
Oops, s/const/val
Wtf kind of name is val for a keyword anyway...
@RadvylfPrograms that's too long for a function which works with lots of streams
@RadvylfPrograms Kotlin also has const
but only for compile-time constants
It looks almost identical to var at a glance, and a variable is a value too
Kotlin’s const is a modifier on val
4:13 PM
idk lol
@pxeger Huh
It’s kinda dumb which is uncharacteristic of Kotlin
@pxeger I disagree, but input or in_s would also work. Both just as short as `in` and without ugly punctuation.
@pxeger val is immutable
var is mutable
but const is also immutable
you can't have a const var
4:15 PM
@RadvylfPrograms adding an underscore at the end is uglier imo
another idea: make all builtins 1-character expressions which LOOK like the whole word, but underlined, so you can have variable in and the 1-byte symbol "in" which can be entered with something like `i.
Someone kick romanp
@Romanp i̲n̲?
@user Yeah, probably worth it.
But yes, something like jelly-style codepages, but with actually readable symbols
> another idea: make all builtins 1-character expressions which LOOK like the whole word, but underlined, so you can have variable in and the 1-byte symbol "in" which can be entered with something like `i.
4:17 PM
@pxeger const means a compile time constant in kotlin
@RadvylfPrograms btw, backticks are necessary to match on existing variables in scala . I hope that makes your day worse :)
BQN has entered the chat
Pretty much that, except with c-style syntax.
@Romanp I have considered making all keywords uppercase in my language, which is a somewhat similar idea
@Seggan I know
@Seggan pretty sure pxeger knows that. It should be const foo = bar instead of const val foo = bar
4:17 PM
but why is it const val x = 1 instead of const x = 1?
There's no need for that
yeah, that's what Mathematica does.
and every sane language
@DLosc Wow I hate a lot of stuff. 145 stuff, in fact
@RadvylfPrograms More than that:
Jan 28 at 18:18, by ta Radvylf srik su shilani
I hate everything
4:18 PM
Redwolf, say you hate the riemann hypothesis
@RadvylfPrograms A few of those are about someone else hating something, to be fair
Easiest proof of its being false :p
@pxeger Kotlin is a pretty sane language other than const (which is a rather minor thing tbh)
> With const val, const is a modifier on val rather than a keyword. Modifiers > keywords. More examples of this same design are, annotation/enum/data class, private val, inline fun, etc.
quote from Aro
Oh it’s a soft keyword
Oh, so const can be used as a normal name in most cases?
That's a good idea tbh
4:21 PM
I don’t see why you’d need const as a name tho
Make all keywords soft except a keyword keyword
Java has const as a reserved word itself
const keyword x = 1; var keyword y = 2; y += x; if keyword (y == 3) return keyword;
@user really? never saw it
4:22 PM
@DLosc including blursed_cat
Older versions of java use const instead of final
@RadvylfPrograms more seriously, you could have a prefix like \​ or something
@Seggan goto too for some reason
Maybe java 25 will have gotos
Bytecode supports it already
ye just searched it up
strictfp is also unused since java 17
4:25 PM
Even Go has goto
it's useful for breaking out of multiple loops
although idk why they don't just use labelled break/continue
@RadvylfPrograms Or just don't have keywords
@pxeger ye, labeled loops are pretty much goto
Waiiiiiiiit hang on:
1: while(1) {
2: while (1) {
3: while (1) {
continue 2; // goto 2
although IIRC Go's goto can only forwards
@RadvylfPrograms You want to remove my goto for readability purposes? Read this :p
4:29 PM
Theory: Every programming language allows you to write code that you should never, ever actually write in practice.
operator overloading and goto languages all fit in there
Proof: These programming languages are (practically) Turing-complete, which means they allow you to write any code you want
Wrong, RSNBATWPL only allows clean, functional, executable code
What's wrong with operator overloading?
and then there are just languages that you should never. ever actually write in practice
4:31 PM
@pxeger nothing. i actually quite like it
@pxeger Nothing on its own but it's easily misusable
but sometimes it can get messy
E.g., making things that should very clearly be functions, be operators
or overloading unary + to be size of a set or some garbage like that
Oh yeah, new language going well already
can you paste that whole license text for me?
Just some standard copyright stuff, you can find it online all over the place
4:38 PM
The second line, which I'm removing because it's not subtle enough, was this:
what the rgksp
4:53 PM
If you ever feel useless, just remember that Firefox has a menu bar.
5:16 PM
yay i read it on a tuesday
erm, whats the rest of the line?
Not going to spoil it
Once I post a Hello, World! program in raisin-batwitchs the source code will be made available though
5:33 PM
why is it called raisin-batwitchs
It's how you pronounce the advebritation
what does it stand for
It depends who you ask. Like, pxeger would probably guess something different than caird, for example.
> advebritation
I'm using that to sidestep the problem of abbreviation vs. initialism vs. acronym, because it's none of those. It's an advebritation.
5:35 PM
well, back to work on GKOTH
does anyone have a python KOTH that they're willing to share?
frantically begins typing out my KoTH
Technically posts here are licensed in a way you can share one from whoever you want
I know, but I need the arena source
5:54 PM
I've not made a KotH yet, sorry
Is there like a test stack exchange?
@Ginger no, unfortunately
what do
6:16 PM
SnowballFight looks like a simple one for testing with a lot of bots
6:27 PM
wt...wth...I'm working with a rather complicated recursive sequence, having to do with binary trees with some complicated properties, and there's a closed form formula for it
Not nice, 'cause I was going to use it as a code golf challenge :|
A: Sandbox for Proposed Challenges

Radvylf ProgramsCreate a program that may or may not determine if a list is in this sequence code-golf sequence code-generation random Your task is to write a program which returns another program which has a nonzero chance of correctly determining if an inputted list is in a sequence described below. This can e...

I am playing Beneath a Steel Sky
oh my god I'm 2 minutes in and there's this piece of comedy gold:
Me: *clicks on press*
Character: "It's wheezing and banging... like an asthmatic dinosaur in mating season."
comedy gold
6:59 PM
Okay I'm getting pretty tired of my school wifi. All day it's been in the like, kilobytes per second.
Every message is taking multiple retries to send
@RadvylfPrograms wdym this is how the internet is meant to be viewed
7:50 PM
@RadvylfPrograms Setup a script to run speed tests in the background
It's so slow speed tests time out 100% of the time
... interesting
00:00 - 20:0020:00 - 00:00

« first day (4112 days earlier)      last day (107 days later) »