The Nineteenth Byte

Jo King

12:16 AM

standing on the pull requests of giants

6

Unrelated String

12:47 AM

@JoKing and you'd lose the privilege of naming a type FatRat :P

Seggan

1:12 AM

ok so ive got good news for yall

the first release of Rol has been released

7

partly due to the fact that i finally decided to redo the UX and so i added an interpreter mode

so you dont have to mess around with the stdlib and linking lua files and whatnot

lyxal

@Seggan the beginning of a long journey :p

Hello, World!

0

A: "Hello, World!"

Rol, 24 bytes println("Hello, World!") Rol is a new language I'm creating for the purposes of replacing Lua (because who likes Lua). I was gonna post this when the language was much more mature but I felt like reworking the CLI so I added an interpreter mode. That means you don't have to mess ar...

lyxal

1:30 AM

smh 98.66% likely to be GPT generated

Seggan

... my answer?

lyxal

can't believe you've done this

Seggan

lol

whats that userscript

lyxal

GPT detector

let me see if I can find the link real quick

one second

github.com/Glorfindel83/SE-Userscripts/tree/master/…

Seggan

ty

> It is likely that this text was generated by a human and not by a language model like GPT-3 or ChatGPT. The text contains specific information about a new programming language called Rol, its purpose for replacing Lua, and how to run it with a jar file. This level of detail and technical information is not typical of text generated by language models.

thats what chatgpt said :P

ATaco

2:19 AM

> ... (because who likes Lua)

I Like lua

Jo King

i'm ambivalent towards lua

it certainly has some design decisions

Unrelated String

it sure does

Seggan

ok, i dont exactly hate lua

just its verbosity + the entire metatable thing just puts me off

dynamic typing is not exactly something i like either

Jo King

my favourite blunder is gmatch -> global match, match -> single match, therefore gsub -> global substitution and sub -> ?

substring of course

Rydwolf Programs

2:37 AM

yikes that's a design choice alright

Seggan

coulda at least called it subs

lyxal

One of the keyword mappings ever used

Rydwolf Programs

Or replace

Seggan

@lyxal huh

that sentence doesnt make sense

Jo King

@RydwolfPrograms time to grep lace

Unrelated String

2:49 AM

it's one of the sentences ever written

Rydwolf Programs

I think Seggan's subs is probably best yeah

Then you can use gsub for "global substring" which would be a substring of the system RAM as a string

Jo King

an iterator of every substring of length n?

Rydwolf Programs

3:26 AM

Fun fact: The interpreter for my new esolang is current sitting at over 120 KB

Most of which is a bunch of data (compressed using a custom algorithm that I designed as part of a different in-progress language)

Get excited, the most challenging part of the online interpreter is now finished :D

Bubbler

How many in-progress languages do you have lol

Rydwolf Programs

Way too many

I'd guess in the low dozens but most of those will never see the light of day

I'm actually really impressed that all that data decompresses with my very unoptimized code instantly

I guess 100 KB isn't actually as much as it looks tho for a computer

Bubbler

Well, it's not that surprising

unless the decompression algorithm itself is borked and is actually quadratic, or the data is huge like some hundreds of megabytes

Rydwolf Programs

Well, part of the decompression involves working bit-by-bit and storing a few at a time as a string, which feels really unoptimal, but yeah, still a pretty small amount of data for a computer

It just looks big and scary when I scroll through it in my editor I guess :p

Bubbler

lol

ATaco

3:37 AM

Can I invite anyone to mess around with my In-progress web-app for BrainChild?

lyxal

Score: 16

so this is all compiled to brainfuck?

Rydwolf Programs

Feels a little disorienting at first

Also I don't like the scroll behavior being overridden in the editor, if you have a way to disable that

ATaco

@lyxal designed with the intent to compile to brainfuck, although interpreted at the second highest level to make it execute in reasonable time

@RydwolfPrograms I’m not sure I have much control, the editor is Ace.JS

Rydwolf Programs

I think my thumb is permanently deformed from pushing in RAM sticks

it's had a dent in it for like a month now

Those things suck

Unrelated String

wow

ATaco

3:48 AM

The ASM is also trivially transpilable to x86_64, so it can at least be used as a Prac lang, theoretically.

lyxal

it's a little laggy kind of with a lot of code

also, why doesn't the following update the snake position?

function moveTo(int x, int y){
    this.Pos = pVec2.Make(x, y)
}

ATaco

It looks like it should?

lyxal

I placed it in the Body class, called it in the frame update thing and it doesn't move it to 0 0 I think

if(snake.Pos.X > 63){snake.moveTo(0, 0)}
if(snake.Pos.Y > 31){snake.moveTo(0, 0)}
if(snake.Pos.Y < 1){snake.moveTo(0, 0)}

ATaco

That'll move the head, but not the tail.

lyxal

if(snake.Pos.X > 63){snake.moveTo(4, 4)} works

it keeps going in a loop

just like I wanted :p

ATaco

3:52 AM

@lyxal If it's really laggy, try pausing the interpreter. That uses a LOT of machine time

lyxal

oh

yeah that makes it better

ATaco

4:31 AM

Shortest quine I've managed is

include term.bc;macro('t'(expression)){{$1;Term.Write(#1)}}t{Term.Write("include term.bc;macro('t'(expression)){{$1;Term.Write(#1)}}]t")}

ATaco

5:07 AM

Slightly shorter,

include term.bc;macro(':'(expression)){{$1;Term.Write(#1)}}:Term.Write("include term.bc;macro(':'(expression)){{$1;Term.Write(#1)}}]:")

l4m2

6:23 AM

CMC: Infinite loop. Never reach same state. Take log(elapsed time) memory.

emanresu A

6:38 AM

So pretty much for(i=0;;i++) with arbitrary-size ints?

l4m2

Yeah, so RAM model used

Bubbler

x=0
while 1:x+=1

(python)

pxeger

8:15 AM

this is upsetting

(does anyone know what determines the order of the Linked Questions list?)

mousetail

Seems to be mostly sorted by votes with the exception of 2021 which is further up for no reason

Unrelated String

8:36 AM

might be some starboard order type thing

a bizarre compromise

Fatalize

9:22 AM

@UnrelatedString Yeah I think you're right

JNat

12:18 PM

Hey, @RydwolfPrograms, the Discord invite at area51.stackexchange.com/proposals/127456 has expired — mind linking me to a new one, please?

Ginger

er, I didn't know we had a Discord invite there :p

lyxal

same

JNat

when proposing a new community, the system asks for an existing community — a link to this chat room and a Discord invite were posted there

Ginger

huh

JNat

this is visible only to mods (which, for Area 51, means SE staff)

Ginger

12:20 PM

@lyxal any idea what said Discord could be?

lyxal

oh yeah that's right

no

Ginger

hmmmmm

lyxal

I only remember there was a community link requirement thing

Ginger

we forgor 💀

lyxal

I don't think we ever knew

JNat

12:21 PM

Maybe Redwolf will remember, since they provided it

mousetail

Probably the nineteenth byte discord?

Ginger

we have a Discord?

mousetail

discord.gg/utPdRQwZ

Could also be the programming languages discord: discord.gg/GPHB5V2W

lyxal

@JNat is the link you're trying by any chance discord.gg/WevqpTe6

JNat

@lyxal it is

lyxal

12:24 PM

one second

I reroll you a new one

discord.gg/bNPMxGcQFj

JNat

thanks~~

would language design questions be off-topic on SO?

mousetail

Would probably be closed as opinon based or need more focus

JNat

do any of y'all know if they've ever been discussed on MSO and officially off-topic'd?

@mousetail looking through a few MSO posts, it looks like discussion often points in that direction, yeah

mathcat

relavent?

Ninja'd lol

JNat

yup — just read that

mousetail

12:30 PM

There is this: meta.stackoverflow.com/questions/399215/… but there is no conclusive consultion

JNat

yeah, that's what I meant — all point to them being disallowed due to vagueness etc., but no discussion outright concludes "yes, these specific questions should be off-topic"

mousetail

There is stackoverflow.com/questions/tagged/language-design with 1,300 questions

parz

hi

JNat

In any case, if I'm reading the proposal correctly, it wouldn't so much be for "why" questions for existing languages, but moreso for questions folks might be designing themselves, right?

mousetail

Both, but yes focus on the latter

JNat

12:33 PM

@mousetail ^^

and Software Engineering wouldn't work here, because it is specifically about software, and a programming language is not software...?

I'm currently assessing the level of overlap with existing sites, in case that wasn't clear ^_^

mousetail

A compiler and or interpreter is technically software, but the language itself is more of a abstract concept

A lot of questions won't be about a specific implementation

JNat

Software Engineering doesn't appear to allow stuff only specifically about implementation, though. The tour page says software development methods and practices; requirements, architecture, and design are on-topic

mousetail

What I mean by implementation here is software that compiles or interprets a language

JNat

gotcha — you mean that talking about language design, in general, would fall outside the scope of software design, unless it specifically applies to the compiler, for instance?

mousetail

Yep

JNat

12:43 PM

'cause "programming language" per se is higher level than software?

mousetail

There are basically 3 levels:
Designing the language in terms of syntax, built-ins, etc. This wil be on topic for langdev.

Designing the software that will run it, in terms of design patterns, arcitecture, etc. This would be on topic for both langdev and software engineering

and then really nitty gritty details like using some library to manage stacks, which would be on topic for stack overflow.

JNat

that's helpful, ty~

lyxal

Also, PLDI could cover programming languages that don't run on software, like mathematical notations

mousetail

Not 100% sure about that one but maybe

lyxal

E.g. APL was originally a mathematical notation for manipulating arrays

mathcat

12:48 PM

@lyxal I feel like that is a topic for private beta

lyxal

it wasn't originally a "programming language" as such, but it was still capable of writing "programs"

mousetail

It's not really a "programming language" if it doesn't run on software

lyxal

@mathcat I know, but for the sake of overlap exploring

@mousetail would you consider a turing machine software or hardware?

mousetail

Both I guess? I'd consider languages for specifying turing machines as programing languages

lyxal

that grey area should be enough to make it so that PLDI is more than just software :p

mousetail

12:53 PM

Hmm woner if there are any stack overflow questions on accomplishing specific tasks on a pure turing machine

mathcat

Probably closed, however.

lyxal

just a quick diversion:

2

what the hell copilot

mousetail

Web developers hate this one weird trick

mathcat

Step 1: Download more ram

mousetail

You won't need any ram after this, time to upload your excess ram

0

Q: How to construct binary multiplication turing machine?

I think I have trouble figuring it out. Can someone please tell me how to construct binary multiplication Turing machine?

turing-machines

mathcat

12:58 PM

RIP zero answers

mousetail

To be fair it's a very low effort question, but it's' not closed (and shouldn't be)

mathcat

~~Imagine not having 100% questions answered~~

mousetail

Actually odd that "the workspace" has ~100% answered, guess people just love drama

lyxal

In other other news, copilot knows about vyxal flags

mathcat

@mousetail I am surprised that Puzzling just has 99% of the questions answered.

mousetail

1:09 PM

Me too, and my one question there is unanswered. Guess I just like to break the mould

lyxal

rebellious! I like it :p

mathcat

@mousetail All Star plug? :d

Oct 13, 2022 at 18:16, by Ginger

time to go into my nuclear bunker

Hey now

you're an all-star

Get the show on

1:13 PM

Get paid

okay we probably do not want this to escalate lol

mousetail

All that glitters is gold

lyxal

val ⭐ = Seq(1, 2, 3, 4)
⭐.foldLeft(true)((a, b) => a && b)

that is what y'all mean by all(star), right?

mousetail

If you write this in PHP we could get paid $ too

lyxal

Y'all know that goofy github clone website that tells you what "functionalities" a repo has using "AI"? Well turns out the list there isn't the full extent of just how goofy it is

I requested the full list and well it gets better

let me quickly gistify my results

mathcat

@lyxal real pros use forall

val ⭐ = Seq(1, 2, 3, 4)
⭐.forall(_ > 0)

lyxal

1:20 PM

gist.github.com/Lyxal/21f70922275a9c0968c91a18281117ca

mathcat

> Execute vxal

ah yes, vxal

lyxal

Vyxal and Jelly have some truly groundbreaking features, like "Splits the needle in the array", "Implementation of the REPL", "Shift a number left to right" and "Returns True if the generator has at index"

I do love how Jelly can "Join a list together together"

because joining it together once isn't enough

and I love how it "Makes the given shape of content"

mathcat

AI revolution.

mousetail

Why doesn''t worldbuilding have a twitter?

mathcat

@Staff who decides which sites get an automated Twitter account?

lyxal

1:39 PM

no one gets bots

there are no automated twitter accounts for any SE site

mathcat

… are the challenges manually posted?

lyxal

23

A: Are the various site Tweetbots affected by the upcoming Twitter API changes?

Not to put too fine a point on it but… Yes. Regardless of whether Twitter allows "good bots" to continue to post for free, it seems reasonable to assume that some amount of work would have to happen to get that "good bot" designation, and with over 100 Twitter accounts, that's a lot of work for w...

They disabled them a few days ago :p

28

A: Why is it so difficult to find the twitter account for a Stack Exchange site?

Update: Twitter bots for all Q&A sites have been shut down. See here for more details. Not all sites have Twitter accounts. We stopped creating them for all public beta sites by default because a) Twitter sees very little engagement for us; b) Twitter-the-company started blocking new accounts fo...

worldbuilding must have been created after they stopped making new automated accounts

mousetail

@lyxal In the all sites page some sites list a twitter account and some don't

lyxal

the joke was that no one gets them anymore now

because they've been stopped altogether

the second link explains why some do and don't

mousetail

Thanks

lyxal

1:44 PM

@TheThonnu your challenge is the last tweet that'll ever be posted by the Code Golf Twitter bot

twitter.com/StackCodeGolf/status/1622973473013518340?s=20

mousetail

Press F to pay respects

mathcat

F

lyxal

F

Ginger

@lyxal what does that say?

lyxal

It's just a link to codegolf.stackexchange.com/questions/257458/…

@NewPosts get dupe hammered nerd :p

Ginger

1:55 PM

also jesus elon can't you go three weeks without pissing somebody off

mousetail

@NewPosts Can I get some close votes please?

Ginger

already done

lyxal

1 min ago, by lyxal

@NewPosts get dupe hammered nerd :p

mousetail

Thanks

lyxal

I sure love it when I can't vtc as dupe on code golf questions 99% of the time because it'd be instantly closed :p

Makes that 1% of times all the more worth it

Ginger

1:57 PM

kachow

lyxal

→ 1 disproportionately sized pile of salt moved to The Nineteenth Bakery

Hope y'all like salty cakes lol

Ginger

mmm crumchy

good stuff

lyxal

Thank you

I'm glad you appreciate my cooking skills

Ginger

👍

lyxal

That totally exist

o/

mousetail

2:00 PM

Does serving a baked pile of salt count as cooking?

Ginger

it sure does!

anything is cooking if you say "COOKING!" loudly enough while doing it

the same thing applies to science

mousetail

Remind me never to come within 1000 square parsecs of Vyxal

Ginger

imagine not having a Lyxal In Direct Air Range sensor :p

mousetail

Lyxal is from Australia right? They'd be more likely to come burrowing up through the floor. A air range sensor would not be able to detect that

Ginger

good thing I also have a Lyxal Infrared Defense Automated Robot

Seggan

2:04 PM

@RydwolfPrograms im excited

Ginger

~~no, you're Seggan~~

mousetail

I'd like to buy that but that would mean coming within 1000 square parsecs of Ginger which I don't want to do either

Seggan

@RydwolfPrograms rol is 22 MB :P

Ginger

@mousetail shame I pervade the entire universe :b

mousetail

Ah but not a infinitely thin disc with a area of 1000 square parsecs

It has no volume so you can't pervade it

@cairdcoinheringaahing can you feature the voting post for best of 2022?

mathcat

2:12 PM

Fedback?

added some test cases

Also, the way mathematics finds an integer solution by leaving a rational number of apples behind is just beautiful.

(testcases #5 and #6)

Rydwolf Programs

@lyxal Or maybe things like UML?

That doesn't really "run", it's more for talking about software, but could probably be on-topic

mousetail

UML would probably fit better on SoftEng

Rydwolf Programs

Well, the design of something like it. I don't really think that'd be on-topic for SE, but yeah, it'd probably be pushing it on PLDI

I actually think that a rule that a "programming language" must be something designed to give instructions to a computer would make sense

Seggan

@mousetail lol same

Rydwolf Programs

It rules out a lot of things that don't feel like programming languages to me, but includes pretty much everything that does

mousetail

2:21 PM

There is a open question on meta on whether mark up langauges would be on topic

Rydwolf Programs

APL as a mathematical notation wouldn't be on topic, but as soon as it was a programming language it would be

@mousetail IMO the answer should be yes, fwiw

caird coinheringaahing

@mousetail status-featured

Rydwolf Programs

It's still instructing a computer in how to display something

mousetail

Maybe, I do feel like they should be included but struggle a bit to come up with a concise definition that would include those but exclude things we want to exclude

Rydwolf Programs

What's an example of something we might want to exclude?

I'm struggling to think of counterexamples but know there's probably tons

mousetail

2:24 PM

Like APL as a mathematical notation

Rydwolf Programs

But that's not to instruct computers. Programming is explicitly for computers.

mousetail

@cairdcoinheringaahing You probably want to lock it again

caird coinheringaahing

Locking it stops votes

mousetail

@RydwolfPrograms What about non-human readable data transfer languages like msgpack or protobuff?

Rydwolf Programs

I'd be fine with those, but they're def. worth discussing. "Non-human readable" can't be a criterion tho, since wasm and bytecode and stuff.

"Data transfer" could be considered more of "providing information to the things that tell a computer what to do" than "telling a computer what to do" but as soon as we get into that, everything is the former depending on how far you abstract the idea of "computer"

What about human readable data transfer/storage/representation? JSON/XML/YAML?

mousetail

2:30 PM

Good question, I don't know what I'd prefer

Rydwolf Programs

IMO they're fine. Many of the same questions about syntax, parsing, etc. would be relevant

mousetail

True, but those questions are not really specific to the category

What if your question is "What are the disadvantages of having both opening and closing tags on fields of my data transfer language", would that be on topic?

Rydwolf Programs

I think so

You can draw parallels between that and issues you'd have when designing human language

I wouldn't call it a programming language, but I think the same sort of people would be able to answer it, and it's close enough that arbitrarily excluding them seems pointless

mousetail

Fair enough

caird coinheringaahing

Did the color of links in chat change recently, or has the gradscript finally died?

Rydwolf Programs

2:34 PM

This is how it's always been for me

Gradscript issue probably

caird coinheringaahing

All the links are blue for me now, even after turning the gradscript off

mousetail

They are green for me, I'll reload

Rydwolf Programs

Maybe try disabling all userscripts/userstyles?

If you have any others

caird coinheringaahing

Okay, looks like its from the Chat Improvements script, weird

mathcat

@cairdcoinheringaahing same

wait it's green again nvm

I'm pretty sure it was blue yesterday

yeah I've got tampermonkey turned off

New Posts

3:57 PM

0

Q: Generate an emoticon

Your task is to generate an emoji of three characters, the first character being one eye, the second the mouth, and the third another eye. Each eye is a random one of these characters: $'*+-;O^`ox~, and the mouth is a random one of these: ,-._coux. All eyes and mouths must have an equal chance of...

code-golf random generation

Sandbox Posts

0

A: Sandbox for Proposed Challenges

Time to shortest transposition code-golfmath Yesterday, as part of a IQ-style test, I got this interesting question: The time on a 24-hour clock is 11:43. What is the least number of minutes I should wait before the same digits are on the screen again (in any valid 24-hour time order)? The answ...

The Thonnu

4:13 PM

@lyxal F

That is kinda cool though

mathcat

Congratulations, you just acquired a CGCC world record!

2

The Thonnu

@mathcat what other world records are there?

Seggan

6

Q: Questions with the longest shortest answers?

I'm just curious about any pure code golf questions (ie, not ones that build on other answers, just straight up, solve the problem with fewest bytes) where even the shortest golfed answers (especially in golf languages) are particularly long. Do any problems stand out as ones that were particular...

discussion statistics

mathcat

@TheThonnu ⁽ᴵ ˡᶦᵏᵉ ᵗᵒ ᶠˡᵉˣ ʷᶦᵗʰ ᵐʸ ᵐᵒˢᵗ ᵘᵖᵛᵒᵗᵉᵈ ˢᶜʳᵃᵗᶜʰ ᵃⁿˢʷᵉʳ ʰᵉʳᵉ ˡᵒˡ⁾

Sandbox Posts

5:29 PM

0

A: Sandbox for Proposed Challenges

Generate an emoticon Your task is to generate an emoji of three characters, the first character being one eye, the second the mouth, and the third another eye. Each eye is a random one of these characters: $'*+-;O^`ox~, and the mouth is a random one of these: ,-._coux. Each eye and the mouth must...

l4m2

5:42 PM

How strong is an ideal Mathematica(every expression get evaluated even if undecidable)?

Rydwolf Programs

6:53 PM

@TheThonnu "Shortest code" is a pretty popular one around here :p

We should start calling our OWCs "world records" instead to get people interested :p

2

Seggan

this is way overdue for how long ive been programming, but im starting to use logging :P

Rydwolf Programs

I prefer not to, to reduce my carbon footprint and save rainforests

Rydwolf Programs

7:05 PM

@mathcat You're not crazy:

Seems there's some CSS floating around that's making things blue

That's a :hover but it's usually not there

Ginger

uh wtf

ffs SE

emanresu A

Yeah same

Rydwolf Programs

@l4m2 Mathematica's TC, right?

Ginger

yes iirc

Rydwolf Programs

So if every expression got evaluated instantly you'd have a solution to the halting problem

If I rubbed a lamp a genie came out I'd ask it for a black box that solves the halting problem. Then after we sit down in front of a whiteboard and I explain what the halting problem is, we'd probably go scare little kids and buy donuts

Ginger

7:14 PM

~~chaotic neutral~~

mathcat

@RydwolfPrograms huh good to know

@RydwolfPrograms what does that stand for again? :d

hyper-neutrino

objective winning criteria

mathcat

“The shortest answer gets a code golf world record.”

DLosc

Only if we can verify the same question hasn't been asked on any other golf sites ;)

mathcat

7:32 PM

true

Rydwolf Programs

Or that we're objectively smarter

I thought of something interesting while waiting in line for lunch

"Exploding lizard solution to the halting problem"

Remember Mayube's story about a lizard that would only eat papers with halting programs written on them?

Well, imagine a lizard that, if the program doesn't halt, will violently explode after consuming it

It eats all programs indiscriminately

Ginger

@RydwolfPrograms oh yeah, the one that you accidentally moved to DBA.SE :p

Rydwolf Programs

But once one foesn't halt, it's gone

You have a single lizard. What do you do with it to maximize the information you retrieve?

You could solve any problem in 2sqrt(O) time where O is its complexity

But that would get slow quickly since the constant factor of a lizard eating paper is pretty big

You could repeatedly ask it questions we're pretty sure are true, to verify them

You could extract an arbitrary amount of information from the lizard, actually, it would just take O(2^n) time (and paper)

Where n is the number of bits of output data

You would just take a program that returns a number, then repeatedly make the lizard explode (with an infinite loop in an if loop) only if it's x where you increment x until you get the answer

That does actually leave you a teensy little chance the lizard survives, tho (if the very last option is the true one)

Does information theory have a way to deal with probability? Like, if I'm 99% sure that something is true, and I'm told it's true, is that still 1 bit of data?

If you only ask the lizard problems which either halt or don't with 50% odds, you get an average of 2 bits of data from it

But I'm curious if that's more, less, or equally optimal compared to asking it questions we're fairly sure about already, or fairly sure we're wrong about

Ginger

7:47 PM

I am confused

Rydwolf Programs

Abiout what part

Ginger

yes

I lost track at the bit where you began talking about time complexity

Rydwolf Programs

11 mins ago, by Rydwolf Programs

You could solve any problem in 2sqrt(O) time where O is its complexity

This was a bit of a jump in logic but basically how I got this is:

Imagine we're trying to find a collision with a SHA-256 hash

It would take guessing 2**128 numbers to find one

Instead of guessing all 2**128, we could, e.g., ask the lizard if there's a collision in the top half, or the bottom half (pretend it's only one of the two idk)

That would cut it down to 2**127+1 operations. Pretty good, but not optimal

Let's say you ask it if the collision (once again pretend we know there's exactly one) is in the first 2**64 options

You can do that 2**64 times, then you only need to iterate another 2**64 times to find the answer

So 2**65

And that would generalize; any decision problem that takes x steps could instead be done in 2sqrt(x) steps using the exploding lizard

Ginger

TNB should build a quantum computer and name it Exploding Lizard :b

Rydwolf Programs

Now let's say you had a second-order exploding lizard, that could eat exploding lizards

Ginger

7:54 PM

I'll provide the quantum, you provide the computer