The Nineteenth Byte - 2023-03-21 (page 2 of 3)

« first day (4434 days earlier) ← previous day next day → last day (425 days later) »

2:02 PM

for the most part

I'm not changing the parser extensively, because it works fine, but pretty much everything after it is being modified if not remade

0

Q: Do absolutely nothing

Write a program that has at least one byte that, when run, does absolutely nothing. I thought this would be done by now, but I can't find any How to tell if it does nothing: No variables are declared Nothing is returned Input does nothing You could run a program after it and it would run normall...

VTC as no OWC

this challenge is impossible in pretty much any language

the main issue is what's stopping me from submitting thousands of programs that are just an empty file

A literal space would be valid in pretty much any language

2:04 PM

^

It said input must be at least one byte

so you couldn't submit a 0 byte solution

I could still submit hundreds of Python answers with varying numbers of spaces

"It's like you never ran it in the first place"

what's the point then

also like how do I define "no variables are declared"

I think they mean it shouldn't effect the program state, whether via the stack or variables

2:07 PM

also what's stopping me from submitting a Text answer

oh, I guess technically all programs in Text are quines

@Ginger Text would be invalid, it would output something

output isn't explicitly ruled out tho

unless that falls under "Nothing is returned"

Rydwolf Programs

@Ginger code-golf?

I’m guessing they mean no output with Nothing is returned

@RydwolfPrograms whoops

opinion-based then

2:09 PM

If you can output stuff in a challenge called "Do absolutely nothing" that would be comical

~~or just "boring"~~

Needs details or clarity

^

Eh I guess it can stay open

Rydwolf Programs

No

2:09 PM

You shouldn't VTC posts for being terrible

Just downvote

Rydwolf Programs

It's mega unobservable

It's observable

Rydwolf Programs

What is "declaring a variable"?

Besides not overvable is not a rule

What is "run normally" in "You could run a program after it and it would run normally"?

Rydwolf Programs

2:10 PM

Doesn't need to be, it's still unclear

Do I need to check what the JVM is doing when I run my Java answer?

He states it means you could insert this program inside any other valid program and it wouldn't effect execution

I want to comment "Pretty much every C-family language, 1 byte: just a space"

Rydwolf Programs

@mousetail That's not how I read it

What else do you interpret "You could run a program after it and it would run normally" as?

Rydwolf Programs

2:11 PM

E.g., the program () might do nothing in a lot of langs, but I can't insert () anywhere in an existing program and expect it to run fine

what if my language has no variables?

Rydwolf Programs

@mousetail After it like, ./nothing ./otherprogram

Actually it says you should be able to insert it in the start of any program

Not in the middle I misread

So a linebreak would be a valid answer in any language, but a space not always

Rydwolf Programs

I don't think OP's talking about inserting it, textually, into other code

the best one is "It's like you never ran it in the first place", because there is absolutely no way for that to be observable

2:12 PM

That's what the question says

Rydwolf Programs

But...this goes to show how unclear it is

It literally says

like even if I run a program in my command prompt that still moves my cursor around

Rydwolf Programs

@Ginger Actually that's precisely observable

what? how

2:12 PM

Also you shouldn't close posts for being unobservable, just downvote

Rydwolf Programs

It just doesn't say what you need to observe. "Not changing an external thing" is observable, that is an observation

@mousetail If the unobservability gets in the way of knowing what's valid, you sure should

That's not how it works

Rydwolf Programs

@mousetail That's how you read it

I am reading it, at this very second, a different way

What makes your reading of it automatically right?

._.

can we get another close vote so we can be done with this debacle?

It says you could "run" a program after it, not insert it a the beginning of the code

2:14 PM

That could benefit form more clear specification

But the intent is clear

IMO OP wanted to say "write a program that does not affect the state of your computer in any way", as emphasized by his last point

Rydwolf Programs

@mousetail How's it not how it works? Unobservable requirements mean there's fuzziness/a lack of clarity over what's a valid answer, which is the precise purpose of the NDC close reason

@mousetail It most certainly isn't if two reasonable people interpret it two entirely different ways

^^^

Nobody else interpreted that sentence at all

I was the only one who had a reasonable explanation

but there's no way to run a program that doesn't affect the state of the computer, which is why this is unobservable

Rydwolf Programs

4 mins ago, by Rydwolf Programs

@mousetail After it like, ./nothing ./otherprogram

2:16 PM

running any program changes a whole load of things, from the OS's PID counter to the position of the cursor on-screen (if you ran it in a shell)

That's not really reasonable, because then declarting variables wouldn't effect the next program

Rydwolf Programs

"You could run a program after it"

My interpretation of "running a program" is running a full program

:|

Rydwolf Programs

Not inserting a second program into the same file and running it all at once

But you could run loads of programs that do stuff before another one, and the second would work fine

so it doesn’t really make sense

Rydwolf Programs

2:17 PM

It's sheer, unbeliebvable assholery to declare that you have the only reasonable interpretation of some vague sentence in a context where both options are questionable

@Fatalize Same's true of pasting it into a second program tho

It's the only thing that makes sense in context, unless you are purposefully trying to interpret in a way that makes no sense and makes the question bad

Rydwolf Programs

Most programs that do nothing will just be statements that do nothing

assume good faith

Rydwolf Programs

@mousetail Except I don't think yours makes sense with the question either

Yes the question should be clarified but that doesn't mean the intent isn't clear

Rydwolf Programs

2:18 PM

Nowhere does it talk about the actual physical source code

would yall mind having this argument somewhere else?

Rydwolf Programs

They're all about effects of the code, not the code itself

This site it about source code

I’m not sure it’s quite appropriate for a room owner to outright call someone an asshole

^

that's a chatiquette violation

Rydwolf Programs

2:19 PM

Not calling the person an AH, just an action of theirs AH-y

you’re as nitpicky as mousetail

> Be nice [...] means avoiding name-calling

:P

Rydwolf Programs

I like mousetail but asserting your interpretation is the only reasonable one is incredibly rude

When I read that rule, I read it the way I read it

please discuss this in TST or a dedicated room; I'd like to talk with @Seggan about Rabbit

2:20 PM

You admit yourself your interpretation is not reasonable

Rydwolf Programs

When I consider both options, neither makes sense to me

Mine makes more sense to me tho

The fact we can even have this argument renders itself moot. It shows it's unclear.

I am part of this question's intended audience. You are part of this question's intended audience. We do not have a shared understanding of what it means. End of discussion, it's unclear.

It would take a two-word edit to fix this, and that's what needs to happen. Until then, the question shouldn't be answered.

Thus, VtCs.

are we done?

Rydwolf Programs

@Ginger This is more on-topic than Rabbit

In the hierarchy of on-topic-ness, discussion of moderation of the site should always come before side projects that aren't even intended for code golf

5

Rabbit is on topic, normally

@RydwolfPrograms okay, fair enough

Rydwolf Programs

2:23 PM

Also, no rule against having two conversations at once and playing a game of musical pings :p

I think we should insult each other on the topic of whether Rabbit is on topic

oh no

(but first: what is Rabbit?)

@Fatalize ~~hey, VSauce! Michael here.~~

Language Ginger likes to talk to themselves about

2:24 PM

wait, myself?

has Seggan been a hallucination this whole time?

Rydwolf Programs

Anyone who thinks Rabbit is on-topic is LUDACRIS IDIOT

Here, Ginger talks to themselves while me and Lyxal have a unrelated dicussion: chat.stackexchange.com/transcript/message/63218268#63218268

I enjoy reading it but I don't really understand anything you say

okay yeah I guess I have been doing that

@RydwolfPrograms I forgot I said that lmao

how tf did you remember it

Rydwolf Programs

It's okay you had a headache

*takes Tylenol sexily*

Rydwolf Programs

2:28 PM

To the rest of TNB I promise we're referencing an inside joke and not just very weird

Can't it be both?

technically it's a meta-inside joke

I still don’t know if Rabbit is a new esolang or what

It's a praclang I'm working on

Praclang? Ew

2:29 PM

user image

I made it primarily because I was infuriated with the dumpster fire that is Python's packaging ecosystem

@Catija hi

Pythons eat rabbits though

Rydwolf Programs

@Fatalize Uh oh

no, this is a giant, skyscraper-size rabbit that squishes pythons whole

Rydwolf Programs

2:30 PM

Ginger you didn't think this through

yes I did

Rydwolf Programs

Time to write a language called Eagle which forms a symbiotic relationship with Rabbit

I've been waiting for at least two months for someone to point that out so I could use that response

Would you rather fight 100 small Python packages, or one skyscraper-sized badly implemented Rabbit?

Rydwolf Programs

I wonder if I can kick-mute myself

2:32 PM

It’s called disconnecting and going outside

Rydwolf Programs

That would be illegal

Truancy

@Fatalize if nobody uses Rabbit, nobody will ever look at the internals; if it gets popular, thanks to Godwin's Law people will decide to go and fix it themselves :b

the classic Xanatos Gambit

This approach has gone terribly wrong in the past, just look at PHP

Rydwolf Programs

Aw man, I can't kick-mute myself

I was about to mention PHP

2:37 PM

Or squirrel

Rydwolf Programs

> $.post("/room/kickmute/240",fkey({userId:347075}))

Doesn't seem to do anything

if a challenge is really bad should you flag it, or just downvote?

If it's spam or offensive, flag it

If it should be closed, VTC

Otherwise, downvote

@Ginger gimme a few mins

Rydwolf Programs

Users who don't have 3k rep yet can flag to close, but it's pointless if there's already close votes

Since IIRC all it does is put the question in the VTC queue

2:44 PM

Are the rep levels different between stackexchange sites? I don't see how to VTC on codegolf.se, but I can see the review page on stackoverflow even though I have lower rep there

Rydwolf Programs

Reviewing only takes 500

VtCing takes 3000

@mousetail i don't have enough rep to vtc :(

Then just downvote

I also can't see the review page on codegolf.se, fwiw

ok

2:45 PM

or at least it's not as obvious

And post nome high quality answers so you can VTC the next one

easier said than done :p

Rydwolf Programs

@Bbrk24 It should just be here:

user image

Unfortunatly true

Rydwolf Programs

On SO they might put it in a more noticeable place, but if so that's a special case

2:46 PM

Oh, I see. It's in the same place, but on SO it's red so it stands out a lot more

i get many ideas for challenges but most of them are just really bad or unoriginal

Rydwolf Programs

The red dot indicates there's stuff in the queue

On CGCC there's like five of us with review scripts so we easily keep the queues cleared

Review queues are always full on SO

And always empty here

Rydwolf Programs

(If you want the best™ review script, here's mine)

What does it do?

Rydwolf Programs

2:48 PM

(Just pin /review and it'll open tasks in new tabs when they show up, you can also configure it to play a ping sound)

i did have an idea for a challenge but i don't think it would be a good fit: write an obfuscated hello-world "program" that could is a text block to run into chatgpt and get that

but i feel like it's kind of boring and chatgpt isn't really an interpreter

Rydwolf Programs

ChatGPT isn't deterministic is it?

And even if it is it periodically changes

@RydwolfPrograms Of course that prompts me to download a random script instead of letting me look at the code

not really but certain prompts will work consistently

Rydwolf Programs

@Bbrk24 Oh sorry here: github.com/Radvylf/userscripts/blob/main/…

2:50 PM

obfuscated is also not really measurable

Thanks

Rydwolf Programs

Most of the userscript managers show you the code under the download button

I've also never set up a userscript on desktop before

(I have configured a userscript for an in-app browser for an iOS app I was working on, but I assume the API is completely different)

@Ginger ok im here

what do you need

busy rn :p

sorry lol

Rydwolf Programs

3:07 PM

user image

LIES

css colors be like

seriously, gray is darker than darkgray

@RydwolfPrograms Brownwolf Programs confirmed?!?!/1//!?1??

@Bbrk24 ~~VTC as duplicate~~

Rydwolf Programs

@Ginger Were you around when I wqs Browncat Programs?

no

Rydwolf Programs

Can't find my browncat pfp unfortunately

user image

There

3:17 PM

0

Q: A randomly colored walk

Given two inputs, a distance \$d\$ and a number \$n\$ output a list of \$n\$ random colors which each have distance \$d\$ from the previous. Background A random walk is a path which is defined by choosing a random direction and (usually) fixed distance to go at each step. We will be taking a rand...

code-golf random color

Me, trying to undo a git add: git subtract .

3:51 PM

I just had an incredibly cursed idea

so yall know how Carrot has repositories (package sources) and libraries (package sinks)?

well what if they were one thing

to upload a package to Chinchilla you'd do carrot install --to https://repository.chinchilla.dev me.mypackage

(I keep forgetting that I renamed Burrow to Chinchilla)

@RydwolfPrograms what are your opinions on this idea

(although there'd be one difference: carrot wouldn't be able to directly run programs using packages from "remote" libraries, so that it doesn't re-download everything every time it's run)

actually, that's kinda an arbitrary restriction

Rydwolf Programs

I thought you removed the concept of both repos and libraries

I did not :p

Rydwolf Programs

@Ginger Just use a cache

Explain to me what they are again then please

sure, one second

a library is a directory on your computer that contains module source code (and a database file of what modules are installed in it); a repository is anything that can provide packages to install in a library

but what I realized is that the two are basically the same thing

so what I'm considering doing is combining them into one concept

if I did so, uploading a module to Chinchilla would look to carrot like installing a package in another library, except the library isn't on your computer

and installing code from Chinchilla would be the same thing in reverse

@RydwolfPrograms does that answer your question?

Rydwolf Programs

I think they serve two entirely different purposes, merging them wouldn't make much sense

4:05 PM

they don't really

Rydwolf Programs

A repository should be able to have a random assortment of modules

each one is a store of modules

Rydwolf Programs

A library has a semantic purpose

A repository has a much different semantic purpose

what do you mean "semantic purpose"?

Rydwolf Programs

As in, what you use them for

A library should be runnable; it should have all the stuff you need, and focus around a specific main/lib module

4:06 PM

just to be clear, by "library" here you mean "thing that contains modules" not "module with utility code"

right?

Rydwolf Programs

A repository could just be a random assortment of modules I've made

Actually I don't really see why you need the concept of a "library"

I guess "library" is a bit of an ambiguous name here

Rydwolf Programs

Well, capital-L library, as you described them

1 min ago, by Ginger

just to be clear, by "library" here you mean "thing that contains modules" not "module with utility code"

Rydwolf Programs

What would the distinction be?

Could I have a "library" that's just a random collection of unrelated modules?

If so, that's kinda pointless

4:09 PM

I'm using "library" here to mean what "repository" generally means; perhaps I should just refer to "repositories" instead

Rydwolf Programs

A repository should be able to contain arbitrary, unorganized modules

correct

and what I'm going to do is treat everything as a repository, from modules on your computer to modules on a remote website

Rydwolf Programs

If I want to install a module as an end-user, I download a bundle file containing it and its (static) dependencies

wrong

Rydwolf Programs

:o

4:10 PM

fricknuts, brb

4:23 PM

@lyxal Yes

5:03 PM

LDQ (not really): what setting to I need to change for bash to interpret my output as UTF-8 and not CP-437

Wait what
I changed nothing and it works now

okay my environment is just cursed i guess

Oh no it depends on where I open the terminal from. If I open bash directly it's fine, but the integrated terminal in vs code uses CP-437

@RydwolfPrograms if you want to install a module as an end-user (assuming you have carrot installed, which you should) you'd just do carrot install goatsinc.goatRecognizer and Carrot would download all the dependencies and install them into... probably your user library if you didn't use sudo

@Ginger yeah. you forgot the hive exists?

D:

LMAO

@Seggan cursed type: sealed class ResolvedType<T: RabbitObject<this>>(name: FullIdentifier, visibility: Visibility) : Type(name, visibility)

because of course this is a valid type parameter

5:19 PM

@Ginger wait how does this work

I'm currently checking to make sure I'm doing it right, which I'm probably not, but it amuses me that it's legal Kotlin (I think)

it looks like I am, somehow

There's two ways I can think of this working: The TypeScript this type, which can only be the value this and nothing else; or the Swift Self type, which is any member of the same type as the object (useful for writing interfaces)

oop no it's an error

e.g. addition is defined as func +(lhs: Self, rhs: Self) -> Self

nvm lol

fixed definition: sealed class ResolvedType<T: RabbitObject<T>>(name: FullIdentifier, visibility: Visibility) : Type(name, visibility)

much better

I've got a lot of recursive type definitions in my code

well frick

now I've really gone and done it

I have two types, each of which has a generic parameter that must refer to the other type

and I don't know how to fix this

@Seggan halp

Unrelated String

5:33 PM

user image

@Ginger NICE

me when Kotlin doesn't have self types

I have a way to fix the issue but it requires self types

that issue can maybe be fixed by a recursive type bound but that makes it even worse

F-bounds are what are normally used (you seem to be using them already)

@Ginger nvm it doesn't fix anything

~~Come to the Scala side, we have self types~~

@user what?

5:40 PM

What you're already using (wikipedia article)

@Ginger remove the recursion

Like when you do interface Foo<F extends Foo<F>>

@Seggan can't

rework the code

don't wanna

5:40 PM

you seem to have made a lot of cursed code

^

the problem is that an object class needs to know its type class, but its type class needs to know the object class

Perhaps you should try planning out what exactly you want to do before jumping into the implementation

Often helps

because both of them need the other's type

@Ginger Does it really, though?

5:42 PM

No (technically), but if I fix it then I have to add lots of awful turgid code

One exists at compile-time, one exists at runtime, right?

Unrelated String

maybe you don’t need them to be separate to begin with

am typing please wait

Unrelated String

@Ginger also wait what

Your compiler should be doing the typing for you :P

5:44 PM

basically the type class has an instantiate function which creates a new object object from it, which obviously needs to know the type of what object will be created; I can fix this by passing the type to the object object upon creation but that 1. adds lots of extra code wherever I want to make a new object object and 2. makes object classes do things that types should handle (like visibility)

wow that's unintelligible, and I wrote it

me and the lads passing types to our object objects upon creation

I guess I'm slowly going insane

@Ginger more adv kotlin

program.opcodes
        .mapIndexedNotNull { index, (opcode, args) ->
            if (opcode == BaseOpcode.LABEL) index to args[0] else null
        }
        .toMap()
        .let {
            IntArray(it.values.maxOrNull() ?: 0) { i ->
                it[i] ?: program.opcodes.size
            }
        }

okay

I think I fixed it

our class definitions are now:

sealed class ResolvedType<T: ResolvedType<T>>(name: FullIdentifier, visibility: Visibility) : Type(name, visibility) {
    abstract fun instantiate(parentContext: Context): RabbitObject<T>
}

and

abstract class RabbitObject<T: RabbitObject<T>>(val type: ResolvedType<T>)

good

@Ginger Okay, that seems like a really bad idea

Types exist at compile-time, their instances exist at run-time

Unless your language has reflection (either compile-time or runtime), there shouldn't be any connection between the two

If you only have runtime reflection, consider having one class for types that exist at compile-time and another class for reified types that you use for runtime reflection

idk how compile-time reflection would work

@Seggan Why not do a filter followed by a mapIndexed?

program.opcodes
        .filter { (opcode, _) -> opcode == BaseOpcode.LABEL }
        .mapIndexed { index, (_, args) -> index to args[0] }
        .toMap()

The second feels clearer to me, unless I've misunderstood your intent

@user ~~performance~~

i didint think of that, actually

5:56 PM

Oh, an extra intermediate collection

Does Kotlin have something like Java's stream? I assume it does

i wasnt serious

Oh

@user yeah, sequences, but i never use em

ty for the tip

np

0

A: Sandbox for Proposed Challenges

Qat Equation Solver code-golf Qat is a tool that lets you search for words matching a given pattern, often used by crossword and other word puzzle writers and solvers. In this challenge, you will have to implement a specific part of Qat, namely a subset of the "equation solver". Equations The out...

5:59 PM

@user uhhhhhhhhhhhhhh

I understand what you're saying but I do not understand how to implement it

« first day (4434 days earlier) ← previous day next day → last day (425 days later) »

Transcript for

Mar '2321

The Nineteenth Byte

The Nineteenth Byte: General discussion for codegolf.stackexc...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile