The Nineteenth Byte

Ginger

01:14

@mathcat you've got competition

lyxal

@Ginger so do you

Ginger

>:(

lyxal

if ranked ginger is too much, you can always play causal ginger

lyxal

01:37

@Ginger this reminds me of this message

in the day TNB became hell, Apr 16, 2021 at 2:42, by Redwolf Programs

I kinda want to upgrade from free trial Lyxal to full Lyxal

Ginger

me when page not found

lyxal

@Ginger well find it then

Aiden Chow

01:57

@NewPosts guys the challenge here says all coordinates are on a rotated hexagonal grid... but what exactly does that mean and how do i convert the given coordinates to regular cartesian coords?

Bubbler

02:13

@AidenChow if the given coord is (u,v), you stand at the origin facing east (slightly rotated in the challenge), walk u units forward, turn left 60 degrees, and walk v units forward

Ginger

02:26

YOOOOOOOOOOOOOOO

RABBIT PARSED A PROGRAM CORRECTLY

>

FunctionDeclaration(FUNCTION Identifier(name=hello, pkg=null)([]) -> community.rto.ginger.rabbit.tree.common.NothingType@6be46e8f, UntypedStatements(UntypedFunctionCall(UntypedVariableAccess(), ""hello world!"")))

I'm so proud :D

ATaco

Nawww

It's like a baby taking its first steps

Ginger

@Seggan thanks for all your (mostly inadvertent) help :b

it's time for me to go, but tomorrow will be... very interesting >:)

Seggan

02:46

yw

Aiden Chow

@Bubbler so in the context of the challenge i start by facing 15 degrees from the horizontal?

Seggan

@Ginger you should probably override toString for the Type class

Bbrk24

Hello! This is my first time being in any stackexchange chat, and I'm still not quite sure how to easily get here (I just googled "the nineteenth byte" and this happened to be the first result)

Aiden Chow

@NewPosts WTF is this

@Bbrk24 welcome!

Rydwolf Programs

> Saint Alan Turing only envisioned in his wildest wet dreams

@Bbrk24 Hi and welcome to TNB!

Aiden Chow

03:04

@RydwolfPrograms lmaoooo

Wheat Wizard

@Bbrk24 Hey welcome! You're on the conlangs discord right?

Bbrk24

@WheatWizard Yeah I know you from there! You left though right? Look at me, I have an esolang now! (And the github repo is starred by the creator of Befunge?!)

Aiden Chow

yo also rydwolf did u do the new lang bounty for nekomata that i mentioned in the comments of ur endless bounty?

Rydwolf Programs

Oh oops no

They're a pain so I put them off lol

I'll do it ig

Jo King

→ 1 message moved to The Nineteenth Bakery

@Bbrk24 catseye? neat, what's the language?

Wheat Wizard

03:07

@Bbrk24 Yeah, I should join back again.

Aiden Chow

@JoKing man that was some weird shit lol

Bbrk24

@JoKing Trilangle, A 2-D language that's vaguely inspired by Hexagony. I copied the mirrors and branches but it's pretty distinct otherwise -- the wrapping rules are different, letters and numbers behave differently, and the memory is stack-based. Also I made the basic datatype i24 for some reason

Jo King

ah right, i've seen your answers on that, very cool. need to give it a go myself

thank god you went with a stack rather than hexagony's convoluted memory structure

Rydwolf Programs

New bounty for an interesting new language just went up. If you're interested, check out Nekomata!

Aiden Chow

@JoKing what kind of memory structure does hexagony even have???

Bbrk24

03:11

Right now I am trying to investigate this strange issue with the online interpreter. If you write a program with a long-running loop, run it and then hit the run button a second time before the first one finishes, any further attempts to run or decompile the program after that result in “RuntimeError: Out of bounds memory access (evaluating 'original.apply(null,arguments)')”

Jo King

@AidenChow a hexagonal one

Rydwolf Programs

Ooh that's...a spicy looking error

Aiden Chow

@JoKing sounds cursed already

Bbrk24

It's the last bug that I know about so after I fix that I should in theory be ready for v1.0 (but I'm very hesitant to say that)

@JoKing It only has one stack and no registers, which isn't by itself sufficient for Turing-completeness, but I've added an indexed peek operation which makes it Turing complete modulo the limitations of 24-bit integers.

Aiden Chow

why 24 bit lol

@RydwolfPrograms ayy nice!

Jo King

03:15

in hexagony each memory cell is contained on an edge of a hexagon in a hexagonal grid. your memory pointer has the choice of moving forward, either to the left or right branch, or backwards either way, or flipping direction

Bbrk24

@AidenChow Well all Unicode characters fit in 21 bits, but if you want to distinguish EOF vs everything else on sign alone (which plays nice with the branching constructs) you need 22 bits, which is too close to a multiple of eight not to round up

Jo King

and presumably you want negative numbers as well?

lyxal

@Bbrk24 ayo gg on the star from Pressey!

Aiden Chow

why not just round up all the way to 64 bit :P

Bbrk24

right, they're 24-bit signed integers, so the range is [-2^23, 2^23)

Rydwolf Programs

03:17

Why not round up to BigInt :p

Bbrk24

It's far easier to implement fixed-width integers in C++, but I chose C++ because it's easier to implement arbitrary-size integers in it (just struct intN { int value : N; };) so that's a bit circular of a reason. The true reason is what I said above about Unicode being 21 bits

user

Surely Boost has something for arbitrary size integers

Bbrk24

Believe it or not I don't think I have boost installed

Here's something even more cursed about that struct intN: you can make it a template. template<unsigned n> struct n_bit_integer { int value : n; };. You'd probably do some std::conditional_t for the base type (here just int) so that way it doesn't take up 32 bits for a 3-bit integer... but if you need that template at all you should probably step back and rethink the choices that got you there

user

How about this GNU library?

@Bbrk24 What's cursed about it?

Bbrk24

@user Given that my installation of bash somehow doesn't even have man, I can say "no" without even clicking on that link.

user

03:23

oof

Bbrk24

^X ^E gives me the error "bash: emacs: command not found" which is multiple levels of wat

user

Bash has a builtin to do Emacs stuff?

Or do you have $TERMINAL set to Emacs?

Bbrk24

Bash has a builtin to open the current command line in $VISUAL or $EDITOR, but it falls back to emacs if neither is defined

user

*$EDITOR, not $TERMINAL

@Bbrk24 That's interesting, I would've expected vi or nano

Bbrk24

But I... don't have emacs. I do have vim, and vi is a symlink to vim apparently

I don't have ed either??

my installation is cursed

user

03:27

Are you a Gentoo user?

How do you have nothing on your computer? (other than Vim, apparently)

Bbrk24

I managed to install bash directly onto Windows

user

Holy heck

Bbrk24

It comes with git so that's nice

user

Oh, so is this just Git Bash?

Bbrk24

yeah

It has... I want to say most of coreutils, but it's missing some pretty common things

user

03:29

You've probably heard of it, but in case you haven't, WSL is pretty good. You can install a more minimal distro if you don't want the default Ubuntu one

Rydwolf Programs

^

Bbrk24

Also it tries to convert Unix-style paths to Windows-style paths to play nice with Win32 executables, but then I get a confusing error when Node sees my regex is prefixed with C:/Program Files/Git/

Yeah I've heard of WSL but my current setup is peak "if it aint broke dont fix it" lol. It has what I need for my workflow and that's what matters to me. (I ended up adding alias man='npx --yes tldr' to my bash_profile)

user

o.O

Bubbler

duh

Bbrk24

i ended up having to send my $PATH to sort | uniq because otherwise it contains so many duplicates that it's 2kB. Which isn't normally a huge deal, but my version of which prints out the entire $PATH on failure, and a 2kB path fills the default terminal size

not to turn this into "bbrk complains about a solvable problem"

Newly Featured Posts

03:36

18

Q: Recognize a counting tree

Let a counting tree be a rooted tree in which every node is labeled with the number descendants it has. We can represent such trees as ragged lists with each node being represented by a list containing its label followed by its children. For example the following is a counting tree: [5,[2,[0],[0]...

Bbrk24

> RuntimeError: Out of bounds memory access (evaluating 'original.apply(null,arguments)')

I discovered that something somewhere somehow calls `unreachable()` and that's causing problems... but also that that's not even the first error logged in the dev console

Ah, a destructor isn't getting called. nice.

Wrapping the wasm call in setTimeout(..., 5) fixed it and that's the slightest bit concerning to me

...did I kill chat

lyxal

04:00

nah, it just gets quiet sometimes

some of us probably have chat open in the background while we do work and stuff ready to jump in if there's an interesting conversation (interesting here is subjective and means different things to each people)

Bbrk24

That's fair. Like I said, this is my first time in chat

Rydwolf Programs

@Bbrk24 Maybe being queued as a new task/microtask (and thus having less stuff on the stack already) is the cause?

Bbrk24

So the thing is if I set the delay to 0 it doesn't work, but if I set the delay to 5 it does

Rydwolf Programs

Oh weird. Might be your browser optimizing something out? Or else something really strange is going on

Bbrk24

and that's what worries me. Maybe 5 is a number that just happens to work on my browser, and not a guarantee.

requestIdleCallback might be better if it works

Which it does sometimes, but not as reliably. Weirdly enough, hardcoding a 5ms delay is more reliable

Rydwolf Programs

04:08

Maybe something related to GC?

(does wasm have GC?)

Bbrk24

Well what's happening is that every time I "print" something, the C++ code waits a moment so that the page has a chance to render the new output. The click of the run button happens during that wait time, so if I start the new run immediately it reuses the same chunk of memory the old one was using. Once I hit the stop button, the destructors race and usually one ends up trying to access memory that's already freed.
If I add a delay in the run button, I'm giving the previous run a chance to run its destructors.

(I say "waits a moment", but the delay ends up being barely noticeable. It's probably on the order of single-digit ms.)

Rydwolf Programs

Have you considered learning Rust?

It's pretty C++-like in syntax (as an outsider to C++ they look similar to me, at least), and eliminates so many weird race conditions and memory issues and stuff

Bbrk24

I think I have rust installed? But that doesn't fix the problem at hand and I don't particularly feel like rewriting a couple thousand lines of code for one minor wasm/js interaction bug

Rydwolf Programs

Fair enough lol

Bbrk24

I'll probably just slap a 5ms delay in there and call it good enough. If it really is a problem you can just, reload the page

ATaco

04:18

sleep(5000); // Decrement to get raise for improving code performance

Bbrk24

```
$ rustc --version
rustc 1.60.0 (7737e0b5c 2022-04-04)
```
ah, what do you know, I do have rust installed

Bubbler

04:49

@Bbrk24 Maybe try instantiating a new Wasm instance every time you click "run"?

Bbrk24

See that would be the smart thing to do. Right now the run button just activates a Module.ccall basically

B. Wilson

06:17

Was pointed here from another room. Meta question: Is there a reliable IRC bridge for SE chat? Or perhaps even just a CLI client?

This is one of the few applications that forces me to context switch out of the CLI, so I'm looking for remedies.

user

People have made libraries for doing chat stuff. I know there's Discord and Matrix bridges out there

Matrix probably has CLI clients, right?

88

Q: IRC access for the chat?

Will there be any way of accessing the chat over IRC? As beautiful ajax interfaces are ... IRC is still #1 for chats.

discussion chat irc

Kinda tempted to write a TUI for SE chat now. Authentication would be a bit tough. Would you have to store your password in a plaintext file or something?

@dzaima Sorry for the ping, but you had Discord and Matrix bridges going for the APL Orchard, right?

lyxal

06:50

@user we had one once

Hyper might know something

mousetail

@user You could probably look at how all the chat bots work

lyxal

@user see here and here

user

@mousetail Well, Ginger seemed to put the password and email for their chatbot in plaintext. I guess you could encrypt it but not hash it

@lyxal imma steal that Python one

@user Forgot to ask an actual question here. Do you happen to have the code for that in a public repo?

mousetail

@user Even important bots like smoke detector?

lyxal

07:08

@user they're both js though

How you gonna steal that python one when there ain't no python one

@user lol what

user

@lyxal Oh, I somehow read the first one as Python

@B.Wilson There's probably an "API" or two for SE chat out there in case you feel like making a client/IRC bridge yourself

emanresu A

From what I've heard chatbots are a huge pain because SE sus

Aiden Chow

07:31

@emanresuA SE sus how?

emanresu A

07:50

Midnight hiccups, random disconnects / timeouts, websocket weirdness, and some things people aren't even sure about

Ginger

12:03

@emanresuA correct

@emanresuA also correct

@lyxal yeah, Vyxal Bot has a file called config.json that contains all of the important bot settings as well as at least 4 secret values

lyxal

I thought user meant in a file in the repo

a file in an upper directory is what hyper did anyway :p

Ginger

no ._.

why tf would I put my email and password in the repo, how 🤓 do you think I am

lyxal

you can never be too sure

given all the times you've leaked things in chat

:p

Ginger

those were accidents

it requires effort to put a secrets file in a GH repo

lyxal

58 secs ago, by lyxal

you can never be too sure

B. Wilson

12:46

@emanresuA What does "sus" mean here?

lyxal

Sus: "Giving the impression that something is questionable or dishonest; suspicious."

in the way of:

5 hours ago, by emanresu A

Midnight hiccups, random disconnects / timeouts, websocket weirdness, and some things people aren't even sure about

B. Wilson

Ah, sus(picious). Gotcha.

lyxal

2020 was a wild time :p

B. Wilson

I'm so out of touch with English zeitgeists.

lyxal

we just call it gen z meme culture :p

B. Wilson

12:52

Anyway, so it seems like CLI-ing this chat boils down to DIY-ing it, IIUC.

lyxal

it does

because SE has a fantastic lack of chat API support

meaning you have to write any chat interaction stuff yourself

B. Wilson

A really dumb workaround might be brow.sh

lyxal

and hope like crazy SE doesn't change something under the hood that breaks everything

dzaima

13:06

@user I do have an SE APL Orchard←→Matrix bridge that I run locally for myself, but it's not published anywhere and is decently incomplete

B. Wilson

Using brow.sh is surprisingly less janky than one might suspect!

Okay. I take that back.

Ginger

@B.Wilson If you want to make a CLI-based chat client, may I recommend sechat? It's a Python library I made to interface with SEChat and is in my completely unbiased opinion the best one available as of right now :b

The version on PyPI is actually out-of-date; I'm going to be releasing version 2.0.0 soon, which is a complete rewrite using asynchronous programming

@Seggan get this: NothingType and NeverType are special singleton objects that inherit from the Type class :p

and their definitions are just as cursed: object NothingType : Type(FullIdentifier(Identifier("NOTHING", "//rabbit-internals")), listOf<Type>())

Yes, I set the package name to //rabbit-internals. Yes, // is the comment symbol.

lyxal

13:22

@Ginger I also strongly recommend sechat in a totally 100% unbiased way :p

Ginger

IF YOU DON'T BELIEVE ME, JUST LISTEN TO [Our Satisfied Customers]!

lyxal

But actually I would recommend it even without bias

Mostly because it's the only working bot library I know of

Ginger

I mean, you've never used it, so it could be a pile of garbage for all you know q:

also whatever Smokey is doing seems to be working fine, and I have no clue what's happening over in SOBotics

lyxal

I have used it

Ginger

what?

lyxal

13:25

Yeah I used it to send like 3 messages once

Ginger

uhh when was this

because I bet you it was before the rewrite d:

lyxal

Last year in the sandbox chatroom

Ginger

~~last year?~~ oh right it's 2023

lyxal

Around the time of Schroedinger's uno I think

Ginger

hmmm

B. Wilson

13:26

@Ginger And in your completely unbiased opinion, where can this amazing library be found?

Ginger

@lyxal ~~link or it didn't happen~~

@B.Wilson well, that depends if you want the latest 2.0.0 version (which uses async and is far better IMHO) or the 1.x version (no async)

1.x is on PyPI, 2.0.0 is on the GH repo

I'm not technically done with 2.0.0 yet, but it's stable enough for use, so if you want it just point Poetry or whatever to github.com/GingerIndustries/sechat/tree/v2 (make sure it's the v2 branch)

lyxal

in Sandbox, Oct 14, 2022 at 12:41, by ľýxäł

14

A: Reimplementing square root

Java (163) Implementing a double precision square root calculator by making use of the fast invert square root stuff from quake and a new constant for the 64bit floats. In java. Yay for verbosity. public double i(double a){double b=a/2;long c=0x5fe6ec85e7de30daL-(Double.doubleToRawLongBits(a)>>...

+ the message above and below

Ginger

oh right, back when we weren't sure why oneboxes were being weird

B. Wilson

@Ginger Code on GH, definitely.

lyxal

@Ginger yee

Ginger

13:29

(it turned out to be because the links were being sent so fast that SE hadn't had time to make onebox data for them :p)

@B.Wilson cool, that's what I'd do too

advance warning: I haven't added any docs yet, but if you want an example of a functional bot using it take a look at this project ~~that I am also 100% unaffiliated with~~: github.com/Vyxal/VyxalBot2

lyxal

@Ginger Hey, that's the funny! I use that bot on a regular basis!

Ginger

it is very the funny

lyxal

13:45

→ 28 beep boop funnies moved to the funniest chat room

Ginger

looky what I did:

FunctionDeclaration(FUNCTION Identifier(name=hello, pkg=null)([]) -> Type(FullIdentifier(identifier=Identifier(name=NOTHING, pkg=//rabbit-internals), child=null), <[]>), UntypedStatements(UntypedFunctionCall(UntypedVariableAccess(FullIdentifier(identifier=Identifier(name=print, pkg=null), child=null)), ""hello world!"")))

New Posts

0

Q: Largest Binary Area

Take the sequence of all natural numbers in binary, (1, 10, 11, ..) then write them vertically beside each-other like this: 1 1 1 1 1 1 1 1 1 11 11 11 11 1111 1111 11111111 11 Notice there are areas of connected 1s in the binary representation of the number. Your ...

code-golf number sequence binary binary-matrix

Ginger

that's right, I parsed a Rabbit program

lyxal

and summoned a new post too :p

Ginger

and that

lyxal

13:46

Well now that I've done the thing I was elected to do, imma head out for the night

o/

Ginger

bye

Rydwolf Programs

14:46

@Ginger NPSP has that problem too lol

The solution it uses is probing the SE API every hundred milliseconds or so, since that seems to update at the same time oneboxing starts to work

Ginger

I just added a fixed delay of like 1s

Rydwolf Programs

I would do that but for NPSP 2.0 I really wanted to do things right™

Ginger

lol

Rydwolf Programs

Also I wanted to make it as fast as humanly possible

(I could make it a tiny bit faster by having it post immediately, then edit it once the API updates if it doesn't onebox initially)

Ginger

too much effort for too little payoff

blugh I hate waiting 34s for Rabbit to compile

but at least it works:

FunctionDeclaration(function hello() -> Type(NOTHING), UntypedStatements(UntypedFunctionCall(UntypedVariableAccess(print), ""hello world!"")))

LDQ: Alternatives to static (the keyword)?

mousetail

14:53

In what context?

Ginger

static function x

Rydwolf Programs

Why do you want an alternative

Ginger

because I don't like static

mousetail

I'd suggest either static or just make every function where the first parameter is not self static automatically like rust

Rydwolf Programs

^

Ginger

14:55

or more accurately I feel that there's got to be a better alternative

Rydwolf Programs

"Better" how

Ginger

@mousetail except: self is a keyword in Rabbit, you don't have to add it to the function list

Rydwolf Programs

But if you did, then it would fix the issue

Ginger

@RydwolfPrograms more descriptive; when I look at static I don't think "oh this function is a member of its type, not of an instance"

mousetail

@Ginger Then I'd use static, it's the common word used everywhere, clear, consise

Rydwolf Programs

14:56

@Ginger Skill issue :p

Ginger

alright, but I'm going to resent it :b

mousetail

Though I'd recomend having self as the first paramter since you could wrap it in a monad and accept Option<Self>, Result<Self>, Promise<Self> etc. in addition to a plain Self type

Rydwolf Programs

mousetail's suggestion is the best option IMO

Ginger

@mousetail why would I want to do that again?

Rydwolf Programs

@mousetail Ooh that's interesting. Does Rust allow that?

mousetail

14:58

@RydwolfPrograms Rust allows a few speefic types like Box and Cow, but I think it would be great if it was extended to every monad

Ginger

"Cow"???

mousetail

Cow = Copy on Write

Ginger

great name

mousetail

It's fantastic

Rust also has yeet

It's fantastic

Rydwolf Programs

@mousetail Yeah but not as a permanent solution right

I thought it was just an anti-bikeshedding temp name

mousetail

15:00

Nothing as permanent as a temporary solution

@RydwolfPrograms That's the intention yes

mathcat

15:10

@Ginger Somebody nuke this user

New Posts

16:15

0

Q: Euler's identity

Create a program that calculates \$e^{i\pi}\$ and outputs the result. It should output -1 but it can have an error margin of 5%. And it needs to calculate it, it can't just output -1. For example, in APL: *○0j1 That outputs ¯1.

math

Seggan

@Ginger whats a FullIdentifier :P

@Ginger lol

@Ginger youve got a very slow computer

rol takes ~10s

on my potato laptop

Ginger

I'm compiling on gingerdrop, which is a pretty powerful server, but that time also includes shadow jars and such

Seggan

thats why you dont shadow

just use run instead of runShadow

Ginger

bold of you to assume I'm using run

Seggan

im not

im telling you to use run

anyway, i think youve summoned me in another room

Ginger

16:24

I fixed that problem already q:

misunderstood how string substitution works in Kotlin

Seggan

"a$b c" == "a" + b + " c" afaik

Seggan

17:14

so ive got some good (for me) news

theres a mobile game i wanna clone

and the copyright database isnt showing anything for the dev's name

so even though if he did have copyright i couldve cloned it, this means i can have no fear at all of retribution

hyper-neutrino

there's a copyright database? I thought you could copyright anything you can legally copyright by just saying it's copyrighted (maybe jurisdiction dependent I guess but that's what I've heard)

Seggan

copyright.gov/public-records

@hyper-neutrino i think thats for licensing, not copyrights

hyper-neutrino

ah

Seggan

17:34

> In general, registration is voluntary. Copyright exists from the moment the work is created. You will have to register, however, if you wish to bring a lawsuit for infringement of a U.S. work.

so ure somewhat right

source

hyper-neutrino

oh, I see

Rydwolf Programs

It would be really ironic if someone made a super permissive open source license but the text of the license itself was proprietary and fiercely protected by a team of lawyers

Seggan

lol

user

18:10

@dzaima Oh ok. Thanks anyway

I think I'll look at SmokeDetector as mousetail suggested and try making a TUI based on that

Ginger

18:23

just use sechat man :p

it's been endorsed by such trustworthy figures as Lyxal

user

Oh hey Ginger

@Ginger How do you handle authentication securely with sechat?

Ginger

@user what do you mean "securely"? there are a lot of ways to define that

Rydwolf Programs

@user Like as in not having your credentials in the source code?

user

Like not having your password in plaintext

Rydwolf Programs

Using environment variables is best practice

But worth noting you'll always have to have them somewhere on the machine in what is trivially convertible into plaintext unless a user is always present when the bot starts up

user

18:26

So I'd still have to put my SE password into my .zshrc, right?

Hmm ok

Rydwolf Programs

You can obfuscate them all you want but if it can be automatically decoded by your code it can be automatically decoded by anyone else's who has access to your machine

Ginger

there's no way to avoid having your credentials in some easily-decodable form somewhere on your machine

user

I think I'll just prompt for a password while logging in every time

Ginger

sechat, for its part, doesn't care where the credentials come from; you just give the Bot.login function an email and password

user

@Ginger I guess so yeah

Rydwolf Programs

18:32

One possibility could be a script with 100 permissions

user

Ooh

Rydwolf Programs

It would just print your password, but since it's execute-only there's a pretty low chance some automatic tool manages to leak your password

Just made a simple one with Rust

It's like 4 MB to print a string but it works :p

radvylf@penguin:~/password_guard$ cp target/release/password_guard .
radvylf@penguin:~/password_guard$ chmod 100 password_guard
radvylf@penguin:~/password_guard$ ./password_guard
RydwolfIsCool
radvylf@penguin:~/password_guard$ cat password_guard
cat: password_guard: Permission denied
radvylf@penguin:~/password_guard$ du -sh password_guard
3.8M    password_guard

Okay stripping it drops it to a more reasonable 300 KB

Ginger

ayo this is a family safe chatroom, put its clothes back on this instant

emanresu A

Whyyy is rust so chonky, I want my C back

Rydwolf Programs

Source-code-wise or executable-wise

C would probably be just as big statically linked

Bbrk24

18:45

Swift is somewhat infamously unsuitable for wasm because when you statically link it you pull in massive libraries like Foundation, and even so much as a simple print statement requires runtime reflection to determine how to stringify the argument

Literally just print("Hello, World!") produces a 6MB wasm file

New Posts

0

Q: Custom Games of Life

Custom Games of Life There's a lot of questions about Conway's Game of Life, but nothing about customization (Related: 2-Player Life). Given an input, determine the result on a 100x100 grid after some iterations. The input is given as a string. For example: Using the character #, follow the follo...

code-golf ascii-art cellular-automata game-of-life

DLosc

19:07

Random esolang idea: Brain-Flak, but mismatched brackets mean something (e.g. [...] is a different command from [...> is a different command from <...] etc)

Rydwolf Programs

Ooh interesting

You could replicate normal brainflak with just two bracket types, while using all four would give you four times more operations to work with

Maybe removing the "adjacent things are added" rule and making the monads into variadic functions would make those additional operations more useful

Bbrk24

I kinda want to make a vs code extension for Trilangle syntax highlighting if only so it stops turning my increment instructions red for being mismatched parens

Yeah in vscode 1.76 they started coloring mismatched parens/brackets red even in .txt files

Ginger

F

Bbrk24

Honestly it's not that big a deal. Since I can't think of any really helpful IDE features I mainly use notepad as my editor for trilangle programs

Seggan

19:23

@Bbrk24 what

you need reflection just to tell that a string literal is a string?

Bbrk24

@Seggan Well the thing is, the unlabeled arguments to print are boxed in Any, and to recognize it's a string you need to unbox it

Any is a 4-word box: the first 3-ish words contain the value if it's small enough or a pointer to heap-allocated memory if not, then there's a bit to indicate which it is, then the fourth word is a pointer to the type metadata

Ginger

> Exception has occurred: "java.lang.AssertionError: ?!"

oh no

Bbrk24

The error message is ?!

Ginger

I know, because I wrote it :b

> line 4:14 mismatched input 'true' expecting {STRING, NUMBER, 'not', BOOLEAN, '(', IDENTIFIER}

...what do you think a BOOLEAN is again

Bbrk24

19:40

Clearly true is a boolean. If it wants a BOOLEAN you have to say TRUE

Ginger

oh right q:

@Bbrk24 ...I know you were joking, but that actually worked, and now I'm very concerned about my parser

Bbrk24

wait what

Ginger

exactly

mathcat

...

Bbrk24

@Ginger what did it think "true" was then? If not BOOLEAN my next guess would've been IDENTIFIER but judging by that error message it's not that either

Ginger

19:53

no clue

I'm trying a fix

I've come up with names for some of Rabbit's tools: rabbundle for the bundler, rabbabel for translation, and rabbodge for the debugger

great, right?

Bbrk24

rabbabel is slightly difficult to say/type

Transcript for