The Nineteenth Byte

Unrelated String

00:10

@RydwolfPrograms i opened the room and immediately clicked the x on instinct

user

@Seggan Hardly, it's getting quite popular. Not as trendy as Rust, but still not that underrated

Seggan

00:57

rofl

OldSandboxPosts

Sandbox posts last active a week ago: Find the sum of the sub-list with the largest sum

Ginger

@RydwolfPrograms wdym? hashes are already taken care of by the library index and INTEGRITY files

Seggan

Ginger

What do you want to do?

Seggan

Ginger

01:07

task failed successfully

Bbrk24

One time I was playing around in the JavaScript IDE I had to use for high school, and I messed it up so badly it eval’d the error message

I just said false = true;

Caused some error starting with the word “Invalid”, but then for some reason it eval’d that and complained that Invalid is not defined

Seggan

lul

js is cursed

user

01:25

And on top of that, whoever was making the IDE must've written some really cursed code

Bbrk24

If you just throw an error it didn't do that, you had to cause a different kind of error, like the one above

New Posts

01:40

0

Q: Classify a surface from its fundamental polygon

This question is an extension of Who's that Polygon? to arbitrary numbers of sides. A fundamental polygon for a surface is an polygon with a prescribed pairing for all its \$2n\$ sides, each marked with an arrow, such that identifying the sides according to this pairing and the arrow directions p...

zoomlogo

03:03

@Seggan does kotlin have operator overloading?

@RydwolfPrograms does anyone keep it on "all rooms"? :P

Rydwolf Programs

03:27

@Ginger Where are those stored?

Also you've never told me anything about an "INTEGRITY file"

Bubbler

04:38

@ParclyTaxel Piet 27b factorial

Also the code golf submission includes a generated image by default

Cloudflare page functions is so OP

Parcly Taxel

05:00

hue!!!!

(or rather, huay – exclamation of cheering)

Bbrk24

05:54

I have been procrastinating the final part of my sorting program but I should be able to finish it tomorrow. It’s by far the largest Trilangle program I’ve written (n=21)

The threading operator I added recently has been surprisingly helpful

It’s based on selection sort. But you can’t remove the minimum element from the middle of the list. So instead it spins off two threads; if the list length is n and the minimum is at index i, one thread discards the bottom i elements and the other thread discards the top n - i - 1 elements, and then they join back together to create the full list sans the minimum

Only problem is that I forgot to store what i is, and control flow in that loop is likely too tight to refactor it. So those threads have to search for the minimum value again

Aiden Chow

06:29

@Bubbler wtf is that....

damn how u even think of these stuff

actually just too good at piet

wait why is the inN in the middle of the loop????

soo confused

Bubbler

06:58

and there goes 33b num2bin

optimizing two loops for size is so fun

Aiden Chow

@Bubbler fucking hell is that monstrosity...

i was trying to get under 40 bytes for the longest time.... now i see that my structure just sucks

at least my number of codels is lower :P, not that it matters anyways...

i really need to study this DP+ in the middle thing

u gotta make a tip or smth on this

Bubbler

07:13

DP+ is not randomly in the middle, it's specifically at A2

so you can break out of the loop through the 2nd column, which minimizes the dead space on row 3

Aiden Chow

@Bubbler huh interesting, taking advantage of the ascii piet encoding

in my answer i just put two loops right next to each other lol

Aiden Chow

07:34

so basically if im understanding the structure correctly, the first instruction A1 -> A2 is DP+, and then u have the loop #1, then u place a white on B2 so that both the way going down to the loop #2 and the path through the loop #1 can be pass through without trouble, then the path going down from A2 merge with the loop #2

ok well ig in ur number to binary answer u dont have a white on B2 cuz somehow that works out but usually u would put a white on B2 to ensure that the paths work right

wait did u specifically choose light red on B8 and B9 just so that it would work out like that???

if so that is actually big brain

like damn

yo also i was wondering if u were still working on your piet interpreter or if u r just done with it

was really hoping for an undo button lol

that wouldve been extremely helpful in case i accidentally delete a row/column and i forget what colors were there or smth like that... it certainly happen to me before :p

Bubbler

07:51

@AidenChow No, that choice is irrelevant. A2B2 is fixed by B2B1 + B1A1 + A1A2

It's just that A2B2 turned out to be useful by luck

Aiden Chow

@Bubbler oh i saw the white codel on A8 and thought u added that there just to reset the color to smth more favorable lmao

Bubbler

I certainly didn't do that :P just some coincidence

while moving the codels cyclically, I realized putting a white at A8 saves one byte

New Posts

0

Q: Count the Genotypes

Description Assume you have learned the Genotypes, the concept of hereditary itself. The one concept, that is important here is: This is a Test Cross, every parent here has two genotypes related to each other in any organism. Here a parent "TT" and parent "tt" can produce four offspring in the c...

code-golf

Aiden Chow

@Bubbler u were moving the codels cyclically by hand??

wait how would that help anyways

now i want to see ur previous programs to see how that could save a byte lol

ur number to binary post could seriously be a nomination tho

lemme see which category i could put it in...

all i can see are wrong tool for the job and wild card

none of the others fit

man they need to bring the best non-text answer category back

wheres that one

Bubbler

08:25

@AidenChow With a white in the middle, so it's like moving commands one by one but with recoloring

I guess I need to add a backward mode in palette

Aiden Chow

@Bubbler if u r suggesting what im thinking then yea that would be super useful

cuz i sometimes find myself starting with an end color that i want to end at, and then trying to build the answer backwards lol

(is that what u r suggesting??)

Bubbler

Also golfed factorial to 24

Aiden Chow

@Bubbler ok bruh thats just insane

Bubbler

@AidenChow yes

Aiden Chow

nice, can u add an undo button while u r at it :P (really tho, it would be useful)

even if i can only undo like 10 times at most i would be happy with that :)

Bubbler

08:29

Yeah I guess I can do that

Aiden Chow

yay!!

also working on the nomination for ur num to binary answer rn lmao

Aiden Chow

09:02

guys i think i mightve wrote a way too long nomination explanation for bubblers piet answer....

but like its such an impressive answer lol, it really deserves it!

not sure if its in the right category but i put it in wrong tool for the job for now

@Bubbler i finish writing the nomination for ur post lol, u can read it here if u really want to: codegolf.meta.stackexchange.com/a/25609/96039

i kinda feel like its just me rambling on... anyone else think that too?

mousetail

09:15

Lol yea was just going to say, we're going to reach the post length limit really fast if everyone types that long

Good way to avoid any competing nominations if they won't fit, you'll win by default

Aiden Chow

@mousetail man i wasnt even thinking about that lol

i just wanted to fully flesh out the explanation cuz idk how impressive it is to ppl who dont code in piet (which, tbh, is pretty much everyone lol)

like the stuff bubbler comes up with in his piet answers are just like, whaaaaat?

Bubbler

lol thanks

Aiden Chow

@Bubbler can u skim it over to see if my basic explanation is correct.... idk if im fully correct on some things

also now that i wrote all of this im too lazy to reduce it down lmfao

maybe someone else whos dedicated enough could try but like its hella long lol

in fact lemme put it in a word count real quick

mousetail

I think explaining why your answer is worse is not really relavent

Aiden Chow

@mousetail it kinda highlights what bubbler does different tho...

like if i just say bubbler has a different and smarter structure, that doesnt really mean anything

mousetail

09:24

True... it's kinda hard to follow though, and the comparison is 3 long paragraphs

Aiden Chow

im just comparing it to my own answer to give some perspective, u know

im not too sure how to shorten it...

maybe i can take out mention of parcly taxel's answer but i feel like that wont be fair to him lol

mousetail

Mind if I try to shorten it a little? You can roll back if you disagree

Aiden Chow

sure thing!

i couldnt care less as long as it makes sense lol

@mousetail maybe i should add a diagram to make it easier to understand??

i kinda feel like i just put some long ramble instead of an actual coherent explanation at this point

@mousetail can u ping me or smth when u done editing, i have to go for a while...

mousetail

@AidenChow codegolf.meta.stackexchange.com/posts/25609/revisions

@AidenChow That would probably help

Bubbler

10:07

aaand it is now 29b

Ginger

12:22

@user like what?

@RydwolfPrograms the INTEGRITY file is generated by whatever tool packages the module; it contains a list of MD5 hashes of every file in the module except for itself

the library index then contains a hash of it

user

Nothing in particular, just any optional feature offered by the package you’re depending on

Ginger

the index file checks that the INTEGRITY file is correct, and the INTEGRITY file checks that everything else is correct

user

Like some parsing library could optionally include a json parser but only if you want

Ginger

ohhh

hmm

user

Or you could select a logging backend

Actually the parsing library one is a bad example because the hson parser should be a aeparate package but it looks like you know what i mean

Seggan

13:22

@zoomlogo read the docs™

but yes

zoomlogo

nice

the docs

@Seggan I do not consent to this

4

Seggan

why not

the docs

because I don't want people reading me

zoomlogo

lol

Seggan

13:25

lol when did you rename urself

the docs

just now lol

Seggan

im not Your Up To Date Kotlin Answering Machine™

lyxal

well neither am I

I'm lyxal

:p

Seggan

wait how did u rename urself so fast

zoomlogo

link to another site

lyxal

13:26

magic

or what zoomlogo said

the docs, Australia

101 3

zoomlogo

rename yourself on only one site and then link your chat account to that

Seggan

@zoomlogo ah

lyxal

The neat part is that the starboard retains the username at the time of posting

Seggan

noice

lyxal

as does the transcript

Seggan

13:27

i knew that

lyxal

which makes reading through that one time rydvylf, user and I swapped usernames and profile pictures even more confusing lol

Seggan

i suggest we unpin best of cgcc

lyxal

no point

it'll unpin itself in 3 days

Seggan

wait really

noice

wait how do u know

lyxal

pins only last 14 days

a little calendar looking shows it hasn't been 14 days since the 24th of Feburary yet

Seggan

13:35

mm

lyxal

Seggan

yes i can see that

lyxal

yes: a) it is the 8th of march and I shouldn't be here right now and b) I do use a purple accent on windows

Seggan

i didnt know they expire in 14 days

@lyxal why not

lyxal

that's why

for it to be the 8th of March, it'd need to be some time in the early morning :p

so I'll be off for the night

o/

Seggan

13:42

@lyxal ah yes

gn

Ginger

13:59

@Seggan any way to make this code prettier? if (it.version != null) it.name == name && it.version == version else it.name == name

mousetail

@Ginger it.name == name && (it.version??version) == version

Ginger

thamks

mousetail

What language is this?

Bbrk24

Kotlin right?

Ginger

yes

mousetail

14:03

Hopefully kotlin has a ?? operator

Ginger

so it'd probably be it.name == name && (it.version ?: version) == version

elvis go brrr

mousetail

Yea that would work. Unless version can be not-null but still falsy

If it's version 0 maybe

Ginger

nah, it's a SemVer object

mousetail

Then that should be fine

Bbrk24

Swift doesn't let you write return foo ?? throw bar, unlike some other languages, but you can do this:

func yeet(_ e: Error) throws -> Never { throw e }
return foo ?? try yeet(bar)

More idiomatic would be if let foo { return foo } else { throw bar }

Rydwolf Programs

14:24

@Ginger MD5? In 2023?!

Even for stuff where security's totally irrelevant...why

@Ginger Well then it serves a different purpose

Bbrk24

I've seen some stackoverflow answers bash MD5 and then provide an even less secure algorithm

Rydwolf Programs

That's just to ensure things don't get unintentionally modified

If you don't trust the source you're getting the modules from...you're not gonna trust the hashes it sends you

The purpose of the checksum along with the module name would be as a security thing, in theory

Bbrk24

Cocoapods uses a checksum of the podspec rather than a checksum of the whole library iirc

Rydwolf Programs

Does the podspec contain checksums of the rest of the code?

If not that's pretty useless isn't it

Bbrk24

No, and it's typically handwritten

Rydwolf Programs

14:29

Sounds like that's intended for something other than security

Bbrk24

Example podspec and lockfile that references it

Rydwolf Programs

Looks like that's for integrity purposes not to protect against intentional modification of the code to do malicious stuff

Bbrk24

Yeah, in fact I have intentionally modified the local copy of a library before to test a bugfix

Rydwolf Programs

Well that would be after the checksum check

Bbrk24

true

Rydwolf Programs

14:31

The point is just to check it when it's installed, what happens after that is a you probelm™

Bbrk24

14:45

mfw I accidentally delete an hour’s worth of work because the file wasn’t checked into git

user

@Ginger Similar to mousetail's it.name == name && (it.version?.let { it == version } ?: true). tbh what you have right now isn't terrible, although you really should name your parameters. it is meant more for tiny little lambdas, I think

Seggan

@Ginger yeah this is best

user

@Bbrk24 Huh, I'm mildly surprised podspecs are in Ruby

@Seggan I'm not a huge fan of the unnecessary version == version check in case it.version is null

Seggan

15:01

it doesnt affect performance (at most a few nanoseconds)

and the code is clear

user

I know, but it still makes me uneasy

Seggan

@RydwolfPrograms its fast

user

(my alternative with ?.let also makes me uneasy because it's needlessly complicated, I honestly would go with Ginger's original version personally)

Seggan

yeah the let version isnt very clear

@Ginger you really should just override equals

user

^

Seggan

15:08

@user thats a bad usecase

unless the optional feature is the default backend

@Ginger bad

user

@Seggan Why's that?

Seggan

the default logger should activate by itself

like slf4j

user

Yes, but you should also be able to choose a different one if you don't want the default

Seggan

yes, in that case the default shouldnt activate

if you specify smth other than the default

like slf4j

user

What makes it a bad use case then?

Seggan

15:21

because you shouldnt have to specify whether you want the default or not

it should activate/deactivate itself based on whether you added your own backend

user

I didn't say you should explicitly tell it to use the default or not

Seggan

thats what i read from your message

why would a logging lib have multiple backends in it?

user

I thought logging libraries generally supported multiple backends

Seggan

they do

but the backends are generally provided as separate libs

user

It would be convenient to specify which backend to use while listing the logging library as a dependency

Seggan

15:26

yeah

but why cant you just add a backend as a dependency urself?

user

Guess you could

Bbrk24

@user Cocoapods itself is in Ruby and the Podfile format is based on the Gemfile format

Ginger

@RydwolfPrograms no, it's to check accidental modifications

I'm trying to outsource as much of the "make sure the build environment is sane" stuff as possible to carrot, so the interpreter doesn't have to check that anything is broken

@Seggan ~~cry about it~~

@RydwolfPrograms because I like it

Seggan

i once used md5 hashes for location markers in a minecraft plugin just because i liked the hex

i eventually switched to using hashcodes when other devs berated me for the performance problems thatd cause

Bbrk24

Other people: Don't use MD5!
Me: https://github.com/bbrk24/Trilangle/blob/master/program_walker.hh#L402

In all seriousness I tested a few different ways to combine the three into a single hash and this formula caused the fewest collisions in real-world use cases

Seggan

15:35

@Bbrk24 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa c++

Bbrk24

The entire reason I chose C++ is so that I could do this: github.com/bbrk24/Trilangle/blob/master/int24.hh#L9

Seggan

~~you can do that in kotlin or rust~~

seriously, c++ is one of the worst languages

Sandbox Posts

0

A: Sandbox for Proposed Challenges

Calculate the p-adic square root of -1 Given a prime number \$p\$, a \$p\$-adic number is a number whose representation in base \$p\$ may have infinitely digits to the left of the radix point, but only finitely many digits to the right of it. A \$p\$-adic integer has no digits after the radix poi...

Bbrk24

It's also one I'm rather familiar with

In no particular order, I'd say I'm most familiar with Swift, C++, C#, and TS/JS

Seggan

Kotlin, Rust, Java

CMP: three favorite languages in order?

Bbrk24

15:39

Chat mini... yeah I got nothing. What's the P stand for?

Seggan

poll

Bbrk24

ah okay

Seggan

mine are Kotlin, Rust, and Java/Python tied

and yes i know yall know i hate python but the other langs i know are just that much worse

python is quite nice for quick single file scripts, but nothing more

Bbrk24

I sometimes like to do cursed things in JS

Here's one of the package.json files of all time:

{
    "dependencies": {
        "express": "^4.18"
    },
    "scripts": {
        "start": "node -e \"(e => e().use(e.static('.')).listen(3000))(require('express'))\""
    }
}

Ginger

oh god

Seggan

15:47

i cant wait when webassembly takes over the web and we wont have to use js anymore

Bbrk24

I found this in a Microsoft C++ header:

#if _HAS_NODISCARD
    #define _NODISCARD [[nodiscard]]
#else // ^^^ CAN HAZ [[nodiscard]] / NO CAN HAZ [[nodiscard]] vvv
    #define _NODISCARD
#endif // _HAS_NODISCARD

Seggan

that reminds me: kotlin is getting a webassembly target in 1.9

Ginger

:O

Bbrk24

Hasn't Kotlin had a JS target?

user

ye

Seggan

15:49

so kotlin will target native, jvm, js, ios, android, and wasm

@Bbrk24 yep

but now also wasm

user

And a Native target, which could've been turned into WASM ig

Seggan

iirc it was

but sadly theyre removing the arm32 target

user

Very excited for Kotlin to become the next Java

Seggan

so no more kotlin native on my pi zero :(

^^

Ginger

@Seggan ;-;

user

15:50

(by next Java, I mean I want it to be used everywhere)

Seggan

it fixed so many of javas mistakes, and added so much more

@user yes

user

And will continue to fix, since it's still pretty young

Still kinda sad Scala didn't become the next Java :(

Seggan

also java is getting rust style pattern matching in 20

@user its too complex

Ginger

Scala is ~~impossible~~ nontrivial to set up, in my experience

user

Rust-style?

@Seggan Yeah, I guess

Seggan

15:51

@user ye

Bbrk24

I always want to pronounce "Scala" with a long A

user

@Ginger Your setup is cursed

@Seggan No I mean what makes it specifically Rust-like?

Seggan

@Bbrk24 i do

Bbrk24

It's officially pronounced with a short A

Seggan

@user Point(int x, int y)

@Bbrk24 wat

user

15:53

What makes it Rust-like?

Seggan

the syntax

well, actually, ig all pattern matching is rust like then

user

I mean, Rust is a pretty young language, so it's definitely not where that syntax started

Bbrk24

Scrolling through my old Discord conversations, I found two highly cursed hypothetical languages: "Objective-C#" and "visual basic but with all the features of C++"

Seggan

aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa

user

I thought by Rust-like, you meant they were going to let you not use the case keyword

Ginger

15:55

^^ what he said

user

Or that you could do FooStruct { ... }

zoomlogo

@Seggan ngn/k

python

c/c++

then rust :P

Seggan

why k

zoomlogo

because k

user

@Seggan Java 1.1, Java 1.4, Java 1.5

zoomlogo

15:59

i think python should be third for me

i have stopped liking python

(or even ... fourth)

CMP: three favorite esolangs in order?

Rydwolf Programs

Brainfuck, the water clock one, and Trianguish

user

Brainfuck, HQ9+, ///

Rydwolf Programs

HQ9+?

It's like, the least interesting esolang that exists lol

Seggan

@user why

just why

Rydwolf Programs

Intended as nothing other than a cheaty lang

user

16:03

That's why it's my favorite

Whenever I think I'm uncreative, I just look at the esolangs wiki page for HQ9+ and remind miyself someone was a bigger doofus than me

Bbrk24

I haven't written anything in it but the Taxi language is pretty interesting

Rydwolf Programs

I like Help, Wardoq! because in many conlangs q is pronounced as ng

5

user

@Seggan I like the simplicity of the old days

Bbrk24

@RydwolfPrograms I always read those languages wrong, like I consistently read "Toaq" as /to.æk/

Seggan

@user no generics.. ugh

zoomlogo

16:06

@zoomlogo flax (obviously), Vyxal, Jelly

user

@Seggan No enums either

Seggan

@Bbrk24 yep

Bbrk24

Before Trilangle, I've never intentionally built large NOP slides

Rydwolf Programs

I like how Trilangle and Trianguish have easily confusable names, Trianguish is basically a synonym of Hexagony, and Trilangle and Hexagony look similar. It's a triangle :p

Bbrk24

16:26

I have the fun task of merging two threads such that they always reach the "join" instruction in the same order... but I don't have any way of knowing how long either one has run for

Rydwolf Programs

As in like, so that one thread always finishes first?

You could maybe do like, a oneshot that one thread sends that the other thread waits for before joining

So if the first thread finishes first it sends the oneshot then finishes, and nothing changes, but if the second one finishes first, it has to wait for the first to send the oneshot

Bbrk24

Like what I described here? github.com/bbrk24/Trilangle/blob/master/… I didn't mean it was conceptually difficult, just that it's going to be annoying to implement

Aiden Chow

@Bubbler wtf how....

actually insanity

@mousetail damn thats like a whole edit right there

im gonna need to add the 29 byte answer into the nomination post now lol

well not much change other than the second loop but damn its under 30 bytes now

Rydwolf Programs

Ohh wait this is in Trilangle lol. I thought you were like, implementing it, and trying to get some sorta multithreading to work lol

Aiden Chow

thats some crazy stuff

Bbrk24

16:42

@RydwolfPrograms No. I guarantee that ticks are synchronized across all "threads", which practically means that the interpreter has to be single-threaded

Aiden Chow

16:59

wait bubbler i just realized... when did u add the svg thing to ur interpreter??? how r u able to generate a new image for any answer???

also not a huge issue, but stack exchange dont take svg image for some reason... can u have some sort of export to png option in the explain section?

i see the import/export/share section have a png option, but not the explain section

Sandbox Posts

17:12

0

A: Sandbox for Proposed Challenges

Triangularly embed a graph on a surface fastest-codegraph-theorytopologydecision-problem This challenge arises from a claim made in a MathOverflow answer and a paper linked in that answer which seems to back up the claim: Searching for triangular embeddings is much quicker than enumerating over ...

Bbrk24

I am trying to debug my WIP sorting program, but I'm running out of colors

Okay I found the bug I was looking for, now back to the regularly scheduled "attempting to merge the red-violet and yellow-green paths"

Seggan

17:34

> To do this, I solved the full Friedmann equation numerically (well, Python's astropy did it for me)

programming go brr

source

Parcly Taxel

17:49

0

A: Sandbox for Proposed Challenges

Triangularly embed a graph on a surface fastest-codegraph-theorytopologydecision-problem This challenge arises from a claim made in a MathOverflow answer and a paper linked in that answer which seems to back up the claim: Searching for triangular embeddings is much quicker than enumerating over ...

somebody review this proposed challenge?

Bbrk24

@ParclyTaxel I don't even understand what that is trying to ask. You lost me before the example triangular embedding

Bbrk24

18:14

$ trilangle sortingAlgorithm.txt -w <<<'105 110 100 115'
100
110

I seem to have implemented ThanosSort

3

Seggan

lol

Bbrk24

$ trilangle sortingAlgorithm.txt -w <<<'100 105 110 115'
100
105
110
Warning: Attempt to print from empty stack.

Okay, I think I fixed it, let's see:

$ trilangle sortingAlgorithm.txt <<<'105 110 100 115'
100
105
110
4976

okay I actually fixed it this time

Parcly Taxel

18:30

@Bbrk24 Post edited. Is this better?

Bbrk24

Yes, that explanation of a combinatorial embedding is much clearer

I don't know how to approach the problem, but it's very clear what it's asking now

Transcript for