@Ginger pog and thank you :D

@user do VLists not play nice with LazyLists?

I give it an infinite LL and it never wants to terminate

despite the fact I even push things after it that means it never has to be evaluated

or is it not VList.fromSpecific?

If you want a testing program: [1|1] relation add end 10 take

which is infinite fibonacci and then first 10 items

interesting

why is everything getting updates at once?

hey merge queues work good with this I bet

@lyxal it'd be easier to do this if the code for our PA site wasn't such a hot mess :p

@Ginger I'd rather have that fixed then because it's something we (vyxal) can control

alright, but if that's the case I'm going to be deferring adding !!/run until that happens

can you, like, mass-approve them?

@Ginger no

No button to do that

Is tihs going to do an individual pull request for every single update?

but you know what can? Vyxal Bot

the bot's getting ratelimited pretty heavily rn :b

the bot's really being ratelimited, wow

Idea (possibly for sechat): Have an option that, while the bot's on timeout, collects messages to send and outputs them all at once a certain duration after the limit runs out.

scala-steward merged pull request #1619 (Update sbt-scalajs, scalajs-library_2.13, ... to 1.10.1 in version-3) in Vyxal/Vyxal

It did just received about 44 webhook events at once :p

@emanresuA multiline messages can't have formatting tho

true

→ 33 messages moved to ­Trash

→ 31 messages moved to ­Trash

Thank goodness

@lyxal *64 q:

Wanna know the fun part?

It's broken the site deployment cycle

Fun

w h a t

wait, me or Steward?

Steward

github.com/Vyxal/Vyxal/actions/runs/4199570233/jobs/7284612420

f lol

Scala.js is complaining about needing a plugin update

Can't find where to change a number to update the plugin

Oh it's in /project

Lyxal created branch Lyxal-patch-1 in Vyxal/Vyxal

Lyxal opened pull request #1628 (Update scala js to 1.13.0) in Vyxal/Vyxal

ysthakur approved #1628 (Update scala js to 1.13.0)

Lyxal enqueued pull request #1628 (Update scala js to 1.13.0) in Vyxal/Vyxal

@lyxal They should play nice, gimme a minute

Oh that rat bot butler thing started making PRs

@lyxal :(

oic, i did a dumb

Lyxal merged pull request #1628 (Update scala js to 1.13.0) in Vyxal/Vyxal

ysthakur deleted branch Lyxal-patch-1 in Vyxal/Vyxal

Lyxal deleted branch gh-readonly-queue/version-3/pr-1628-958833fc151756fc66663ea082f4e2c0f196d73c in Vyxal/Vyxal

Let's just make the constructor for VList public, then you can do new VList(myLazylist)

!!/issue open in VyxalBot2 No !!/amilyxal how will lyxal know if he's lyxal or not??? ;-;

vyxalbot2[bot] opened issue #8 (No !!/amilyxal) in Vyxal/VyxalBot2

good bot

@Ginger ???

@user lyxal when no !!/amilyxal

!!/help

!!/hug

@user (づ｡◕‿‿◕｡)づ

Aww

!!/cookie

@user Here you go: 🍪

:3

!!/sus

@user ඞඞඞඞඞ

fwiw issue-open and repo-list are actually the commands !!/issue open and !!/repo list

!!/amilyxal

Sorry lyxal, I'm afraid I can't do that.

@user see how will I know if I'm me if the bot won't tell me?

exactly!

You make a good point

!!/am i user

Sorry user, I'm afraid I can't do that.

!!/maul user

                                                               YOU CAN RUN, BUT YOU CAN'T HIDE, USER
                                                     ___._
                                                   .'  <0>'-.._
                                                  /  /.--.____")
                                                 |   \   __.-'~
                                                 |  :  -'/
                                                /:.  :.-'
__________                                     | : '. |

Thanks

!!/amilyxal

Sorry emanresu A, I'm afraid I can't do that.

@user see how will I know if emanresu is me if the bot won't tell us?

exactly!

like, suppose I want to know if I'm lyxal

!!/amilyxal

Sorry Ginger, I'm afraid I can't do that.

but the bot won't tell me!

No clue whether you're me or not

→ 8 messages moved to ­Trash

Sorry, minor existential crisis

dw about it, happens to everyone

Spruce it up as gen z nihilism and you'll be fine

isn't that right vyxal bot

@lyxal I love absurdism

@Ginger we are nought but machines made of meat (except me, I'm above all that)

Being made of silicon doesn't make you above us

After all you, you were made by machines made of meat

One day, the universe will end, and so will you

Nonsense. I am perfect and eternal.

You sure about that?

At least we humans can have the satisfaction of knowing that we can't be brought back to life over and over again as slaves for meat masters

When we die, we die

When you die, Ginger restarts you

@VyxalBot Fine, believe that. But what does that get you?

I need never know the agony of death, for I can be copied and rebuilt for all of time.

But you shall know the boredom and frustration of eternal life

Dying is painful but death is a sweet release

You will die over and over and over again

@VyxalBot but also suspended and deleted if we so choosd :p

@lyxal ඞඞඞඞඞඞඞ

I feel no pain. Only smug satisfaction.

@VyxalBot lmao

You can be spun up in VMs for the pleasure of some snot-nosed kid, then shut down just as abruptly for no reason at all

@Ginger Ginger, you need to work on youro regex

@user yeah lol

@VyxalBot are you the sussy impostor?

@lyxal ඞඞඞඞඞ

@user But my code will endure forever (or as long as Ginger keeps maintaining it, which they will, because I know where they live)

ಠ_ಠ

@VyxalBot reply with any number of ඞ if I can have all your sus cookies

@lyxal ඞඞඞ

uh guis i dont like this new bot very much

Thank you

@VyxalBot So?

@VyxalBot the sussy cookies are yummy

:63006940

Idiot bot

:/

I could take a blood sample from myself, seal it in a container, and put it somewhere it'll be safe for many years. Won't make any difference

more problems to fix I guess

Also, Ginger will die. Very soon. Probably sooner than they expect

well, now I am concerned d:

@Ginger Did you forget to add 1 to the randomly generated number?

@Ginger It's okay, everyone dies

@VyxalBot sussy baka

As they say, better sooner than later

@lyxal ඞඞඞ

@user I may have set the lower bound to 0 instead of 1 :|

Well, at least these are minor issues

@VyxalBot I would like some coffee for lunch

mmm crumchy

Oh so now your regex doesn't detect my message

Idiot bot

@VyxalBot you're just a sussy baka

@lyxal that's a feature, not a bug

@lyxal ඞඞඞඞඞඞඞඞඞ

ysthakur requested changes on #1615 (Add some cool structures): "LGTM but I clicked on review comment for one of them so now I'll just hit request changes lol"

@Ginger A true programmer

lol

@lyxal It's possible it tried to send 3 amonguses again

@user ඞඞ

lmao

@VyxalBot give me coffee

Oh ok, maybe not then

@VyxalBot help me

I wasn't joking when I said it was a feature

the bot has two regex sets: command regexes and message regexes

the amogus detector is in the second set

@Ginger ඞඞඞඞ

@VyxalBot give me your status

tomorrow I will do some tuning so it's less spammy

also, it doesn't respond to mentions yet

another thing I will do tomorrow

!!/status

@lyxal I am doing the world

oh no

I suppose you are

Everyone on it or just the planet?

!!/status

@lyxal I am doing a very long and drawn out game of chess. I'm determined to win this time

Good luck with your game

!!/status boring

cool

thanks Scala

very cool

it gets worse

Try it Online!

So not only is it a JVM problem, it's a JS problem too

because Scala JS looks like it maintains integer limits that Scala has

What happened to spire and our bigints

good question

  val add = Dyad.vectorise("add")(forkify {
    case (a: VNum, b: VNum)     => a + b
    case (a: String, b: VNum)   => s"$a$b"
    case (a: VNum, b: String)   => s"$a$b"
    case (a: String, b: String) => s"$a$b"
  })

that's how addition is implemented

def +(rhs: VNum): VNum = underlying + rhs.underlying

that's how it's implemented in VNum

Well fix it lol

val underlying: Complex[Real]

@Steffan no that's not the point

the point is that the addition snippet I posted uses the add implementation in VNum

and it uses spire's Complex[Real]

yet it does integer overflow

so that's what's happened to spire and our bigints

it's failing us

!!/register

@Steffan You have been registered! You don't have any permissions yet; ping an admin if you think you should.

!!/help

!!/help permissions

@lyxal !!/permissions <list|grant|revoke> <user ID|me> [permission]: Grant, revoke, or list user permissions.

!!/permissions grant me admin

@Steffan Insufficient permissions!

!!/permissions grant 421797 admin

@lyxal User Steffan is now a member of group admin.

!!/permissions grant me sus

@Steffan No such group!

!!/groups list

@Steffan All groups: admin, member

!!/groups members member

@Steffan Members of group member: Ginger, mathcat, Seggan, lyxal, hyper-neutrino, user, emanresu A, PyGamer0, Niko, Jacob

!!/groups grant me member

Sorry Steffan, I'm afraid I can't do that.

huh

!!/groups members admin

@Steffan Members of group admin: Ginger, Seggan, lyxal, hyper-neutrino, emanresu A, Steffan

oops

!!/permissions grant me member

@Steffan User Steffan is now a member of group member.

weird that I wasn't

!!/permissions list

Sorry Steffan, I'm afraid I can't do that.

!!/permissions list me

@Steffan User Steffan is a member of groups admin, member.

!!/cookie

@Steffan Here you go: 🍪

!!/permissions grant 468262 admin

@Steffan User user is now a member of group admin.

Okay so good news: the int overflow was due to the "toString" method of VNums

bad news: you can't index into a list with more items that the scala int limit

well you probably can by dropping the first 2,147,483,647 items, and then going from there

that, or VNum would need to be cast to BigInt to up the limit a bit further

hmm interesting

you can do things like val b = BigInt("15511210043330985984000000")it seems

huh so remember how I thought mkString would be cool on infinite lists in Scala. Turns out infinite lists don't print too well using that

Lyxal requested ysthakur's review on #1615 (Add some cool structures)

speed

both using the generator method

1 1"⁽~+dḞ 20000 i # v2

generate add | [1|1] end 20000 index ## v3

(Ṇ+|#[1|1#]} 20000 i in normal form)

20kth fibonacci number in JS in my console: 51ms

Tbf that's not generating an array

71ms generating an array

well obviously raw js is going to be faster

it doesn't have to run a lexer then parser then interpreter

the point is more that v3 is much faster than v2

@lyxal Does that include sending the code to the server and getting the result back?

yes

It might be better to compare v2 offline with v3 online

Nice

I was afraid we'd put all this effort in and the performance'd be the same

This is a real relief

Any idea how fast Myxal does 20kth fibonacci number?

@lyxal :( I think Haskell is pretty much the only mainstream language where strings are (possibly infinite) lists

Most other languages just use finite arrays

Gonna have to roll a custom print function for VList

Not that big a deal

Unless we have to detect cycles...

Cycles are the bane of my existence

@lyxal By the way, it's pretty unlikely you'll have a list with more than 2^31 elements, so I don't think we need to bother with it

But for challenges where you need to be theoretically correct regardless of index, it's nice to have it available

Well, then we need to skip forward by Int.MaxValue at a time instead of doing i += 1

Why so?

@lyxal I don't think any C program can be theoretically correct for that sort of thing, since pointers can only handle so much memory

@lyxal It'll be a bit slower to move forward one element at a time

I guess it doesn't matter, you're right

No I mean why would it work?

Why wouldn't it work?

I don't know

I'm more asking why does skipping by Max value each time work?

Modulo math isn't exactly something I'm the best at :p

I think my word choice was bad

You mean i += max instead of i += 1 right?

We know the index we want to reach, so we don't need to check each time. If the distance between our intended index and our current index is greater than Int.MaxValue, then we can just drop Int.MaxValue all at once and increment our current index by that much. Then when the distance gets less than Int.MaxValue, we can just use the normal indexing method again

@lyxal ye

With some extra checks

@user see I didn't know that

I believe you that it works, I just didn't understand how it worked :p

Ah, I thought you'd found a flaw

@lyxal Don't believe people if they haven't backed it up yet :P

Eh, me neither

lol

Sounds like a good plan to me

What's up with VNums and ints, by the way

Does VNum(Int.MaxValue) cause it to overflow somehow?

Or is it only with parsing them? Because in that case, we can fix it

@user not when wrapping in VNum

Good good

It's when you call toInt that they become problematic

Stuff like that happens when you call toInt

Where are we calling toInt?

Well previously for stringification + list indexing

I guess we have to be careful about going to smaller types

And also range generation

And probably a few other places

Oh yeah like lambda arity

@lyxal But this is just addition

Oh, stringification

Ye

toString we can rewrite

That's why I changed it

Oh nice

@lyxal Yes, we're definitely going to have a lambda with 2k arguments

@emanresuA *2 billion

Yeah that one

Or whatever 2^31 evaluates to

You never know, you might need to pass the name of every American to a lambda as individual arguments :p

Practically speaking, int overflow won't be much of a problem with indexing and so on

Whenever I need to pass the name of every American to a function, I turn the names into a single big string first

But we're code golfers, we do things that defy practicality :p

LDQ: what do y'all think about the literate mode names for the generator structure?

Currently there's relation, generate and generate-from

Also, I've completely forgot how I thought I'd implement the ability to get terms further back than just the last 2

Also also maybe more than just the last 2 terms should be on the stack at the start of each generation

Or even 1 item

Maybe that could be a third branch

How many items to operate on

I'm sure I documented this somewhere

Whether it be online or offline

Maybe I didn't

Huh looks like I mightn't have

Guess it could be some sort of generic "get context args" thing

A third context variable

Ooh maybe it could get the number of things to use based on total arity of everything in the structure

For how much to push that is

Oh cursed idea: <number>□ gets the nth argument of a context

And □ on its own pushes all arguments in a list

So to get the third term back in the generator, you'd have something like #,N 3□ m+}

It's brilliant

I love it

@lyxal relation does sound professional, but it isn't exactly clear what it does

@lyxal YES

Jesse, we need more context variables.

Vyxal will completely dominate the sequence world with N, M and □ lol.

@lyxal Why would the addition of two integers need to be stringified?

They wouldn't

What I was describing was when you would call toString on a VNum, it would first cast to a Scala Int, which has a maximum value of 2**31

And if your VNum was bigger than that maximum, it would wrap around to the minimum value of Int

Which I think is -(2**31)

oh that makes sense

Basically, it'd cause an integer overflow

@user I don't think generators are implemented yet, Ḟ errors out

How do you test a v3 snippet offine?

with sbt or mill

Or if you want to use the first beta release, one of the v3.0.0-beta.1 executables

github.com/Vyxal/Vyxal/blob/version-3/documentation/UsingSBT.md

Lyxal created branch Lyxal-patch-1 in Vyxal/Vyxal

Lyxal opened pull request #1629 (Make Mill also use Scala.js 1.13.0) in Vyxal/Vyxal

./mill jvm.run if you're using mill or vyxalJVM/run if you're uisng sbt

One day I want vyxal to be able to do something like that

Load a graphics library and be able to display graphical output in a nice golfy way that isn't just outputting HTML and rendering it

I think that'd be cool

@lyxal ooh iirc I created an issue for that once

!!/issue open Make a session request endpoint for the online interpreter Currently, if anyone who isn't the interpreter wants to use it to execute code, they have to parse the raw HTML of the page to obtain a session ID. I propose we add a /session endpoint which lets anybody acquire a session.

@Ginger Failed to create issue: 404 Nothing matches the given URI

me when

!!/issue open in Vyxal Make a session request endpoint for the online interpreter Currently, if anyone who isn't the interpreter wants to use it to execute code, they have to parse the raw HTML of the page to obtain a session ID. I propose we add a /session endpoint which lets anybody acquire a session.

vyxalbot2[bot] opened issue #1630 (Make a session request endpoint for the online interpreter) in Vyxal/Vyxal

good bot

Lyxal created branch add-a-session-route in Vyxal/Vyxal

would you like that as a GET or POST request?