The Nineteenth Byte

Unrelated String

00:07

it's not like it;s an insignificant one :P

caird coinheringaahin g

Yeah, Vladivostok has historically been incredibly important to Russia :P

New Posts

03:33

0

Q: Imtiaz Germain Primes

Description "Imtiaz Germain primes" is not a technical name in Mathematics, but my weird creation, in the memoir of the famous mathematician Sophie Germain. These primes can be generated by the following process: 1. Take a Safe prime 2. Apply 2p + 1 on it to obtain a composite number 3. Apply 2p ...

caird coinheringaahin g

03:54

@lyxal 59 seconds ;P

lyxal

If I hadn't bothered to write an explanation first, I'd have beaten you to posting :p

caird coinheringaahin g

Honestly, I leave my explanations to after my first post, because I often find golfs in my explanations, and prefer just having an answer there, then working on it

I also just noticed:

> if bytes are equated then the earliest posted program wins.

Aitzaz Imtiaz

time is a mystery :P perhaps someone can win by nano seconds

lyxal

04:23

@cairdcoinheringaahing doesn't matter now :p

The power of collect until fixed point and lazy evaluation saves the day lol

Unrelated String

hype

Aitzaz Imtiaz

04:51

@lyxal she still does not prefer you get the bounty on that question. despite of the fact you deserved that XD

lyxal

05:37

@cairdcoinheringaahing I write explanations before posting because I hate doing captchas :p

AviFS

05:48

This is the most taxing world-building SE post I've ever read. I thought I was pretty good at this abstract mathy stuff, but clearly physics is a whole 'nother thing.

45

Q: What does it actually mean to have two time dimensions?

The trivial answer is that it just means "you have two dimensions with inverted sign in your spacetime metric". But the perceptual result of that kind of choice doesn't actually look qualitatively different from our universe. Consider the 4D case; our 4 dimensional spacetime has a a +,+,+,- metr...

time-travel universe spacetime-dimensions

I challenge you to read, and wrap your head around, each answer in one sitting :p

By the way, I was actually curious about this, because I was thinking about what it'd look like to have an esolang with two different time dimensions. If you figure out how to conceptualize >1 temporal dimensions, or how to esolang them, let me know!

(Specifically, the write-a-parser book I'm reading defines a program as an array of statements. So I immediately wondered about an esolang that defined a program as a 2d array of statements.)

type Program struct {
	Statements []Statement
}

type Program struct {
	Statements [][]Statement
}

??!!

lyxal

wouldn't that just be a 2d esolang?

AviFS

@lyxal i, uh...

interesting counterpoint you bring up there, my fair watson

but what if i say no?

lyxal

well I'll just say it again then

and I'll keep asking it until the answer changes :p

AviFS

@lyxal Amazing, haha :P TNB humor never gets old

lyxal

I try my best :p

AviFS

06:01

Or maybe... I never get old, lol

lyxal

everyone gets old eventually

AviFS

I mean, I guess it can be if you think of v and ^ as being able to move through the temporal dimensions at will

lyxal

@lyxal and if not naturally, then by force

AviFS

But, the way it manipulates through the time dimensions, doesn't seem thematic to most interpretations of >1 T dim

It seems pretty clearly designed with a spatial intuition, not a temporal one

(right?)

I mean, you have :, or whatever, that skips a command

You can sometimes, mirror, and then like skip over a "time dimension" or skip over the next event in the current time dimension, with the same primitive

I'm not really seeing how to think of it as >1 time axis the more I think about it, actually

I mean, you can rotate the whole thing 90 degrees, and it should still pretty much work the same

So what even are the time dimensions then? Every primitive?

tl;dr i think an esolang designed with a 2-time-dimension intuition/thematic would look very different from any "2d lang"/fungeoid i know

@lyxal

Two quick ideas: You might want there to be one arrow of time, but the two time dimensions can be manipulated independently. In this conception, the arrow of time is a function of the two temporal axes (a vector sum or something).

I forget the other quick idea, LOL

---
Yeah, I guess I'm thinking of two main sets of ideas.

One set is that you can't control time, just like in our universe, so the arrow of time goes through the evaluation of the program according to its own laws. However, you can manipulate the axes of time that together direct the arrow, so that the observable behavior changes.

The other set of ideas would be that you can, in a more typical sci-fi way, go back and forth through the two dimension of time, each one being its own arrow through spacetime. This may, or may not, involve things like time travel.

And if anyone would like to inform me of what the current physical theories are for how >1 temporal dimension actually work, I'd be very interested! Those are the two models I can semi-regurgitate right now, but I don't even know if any/either of these are currently posited to be plausible.

/end

Unrelated String

06:55

now i'm just thinking about an esolang with one time dimension, but also time travel within one timeline
where outright contradictions are compiler errors, but origin paradoxes are a necessary part of the language :P

PyGamer0

15:12

does anyone else see 2 stars here

because i do

but only a single person is visible

Wheat Wizard

I see the same.

It could be a bug, or maybe there

's an anonymising feature hiding a user.

Command Master

How do you sandbox cops-and-robbers? Do you write both questions into the same sandbox post?

PyGamer0

15:55

cmp: should i change my username to zoomlogo everywhere? (its my original username lol)

@CommandMaster i think many people just put both the cops and robbers into the same post because its easier to see that they are related

user

17:37

I do not want to be the programmer who writes the datetime library in that universe! — Tim Jul 17, 2019 at 16:28

Adám

17:57

@PyGamer0 Yes. PyGamer sounds like you use Python and play games.

Neil

18:09

@AviFS there is one part of relativity in that if you observe something else travelling at constant velocity then the rate it appears to age compared to you and its velocity relative to light satisfy w²+x²+y²+z²=1, and that at least would still appear to work for multiple time dimensions (note that I have handwaved the definition of constant because that itself depends on time)

@lyxal maybe the answer will only change in the second time dimension

@Adám I zoomed his logo but I wasn't enlightened...

Aitzaz Imtiaz

sorry, if I am disturbing an on-going chat, I am just announcing a 100 rep bounty to Neil, though it might not be a best gift to you, I and Aira really want to thank you for your extremely exceptional answer here

you really made a contribution for which we both had a thirst to recieve an algorithm for

bounty awarded, see now here :P

Newly Featured Posts

5

Q: Bumping Series Implementation

I have a follow-up question here from my previous question on Math SE. I am lazy enough to explain the content again, so I have used a paraphraser to explain it below: I was considering arbitrary series, springing up as a top priority, when I considered one potential series in my mind. It is as p...

sequence subsequence fastest-algorithm

Wheat Wizard

18:48

@AitzazImtiaz No need to announce bounties, there's a bot that does that already.

Sandbox Posts

0

A: Sandbox for Proposed Challenges

All languages have their strengths, right? (Cops' thread) cops-and-robbers code-golf This is the cops' thread of a cops-and-robbers challenge. The robbers' thread can be found [here] (sandbox note: will add the link when posting) The main idea of this challenge is whether there are languages whic...

Aitzaz Imtiaz

19:11

unaware of that earlier, expression of gratitude @WheatWizard :D

mathcat

19:35

@Adám I do think that was idea of the username

But I'm pretty sure Pygamer0 hates Pygame, so agreed.

Ginger

19:56

LDQ: Should I have length limits on package/version names?

Wheat Wizard

20:13

Why would you?

Adám

@Ginger yes; do not allow names that cannot fit in RAM of the current machine.

Ginger

@Adám luckily, I don't have to enforce that restriction :p

user

@Ginger Yes, make sure they have positive lengths

Ginger

5D package naming with multiverse time travel

mathcat

irrational lengths might be problematic too

Adám

20:30

@user Strictly.

I'd strongly recommend real lengths. Complex length names are slow to look up.

Wheat Wizard

Strictly positive has a different meaning than you intend ...

Bbrk24

20:53

iirc the semver spec says there’s no hard limit but a 255-character-long version number is likely more than enough

Unrelated String

21:17

@mathcat okay yeah definitely not a bad idea to change it lmao

all i can think of is pygame

ATaco

21:31

@Ginger Absurd package/version lengths should be undefined behaviour over a 255 characters, though not strictly forbidden.

Ginger

21:59

update: due to implementation reasons I have to specify a maximum package name/version limit; I'm thinking 256 bytes

ATaco

255 would be reasonable, allowing you to encode the length as one byte.

Ginger

I don't need to, I'm using nul-terminated strings

Adám

@Ginger What if I want nulls in my package name?

Ginger

too bad

~~cry about it~~

ATaco

I for one support 7bit encoded int as a length specifier for strings

pxeger

22:15

Zero one infinity rule

The Zero one infinity (ZOI) rule is a rule of thumb in software design proposed by early computing pioneer Willem van der Poel. It argues that arbitrary limits on the number of instances of a particular type of data or structure should not be allowed. Instead, an entity should either be forbidden entirely, only one should be allowed, or any number of them should be allowed. Although various factors outside that particular software could limit this number in practice, it should not be the software itself that puts a hard limit on the number of instances of the entity. Examples of this rule may be...

No

Ginger

@pxeger well, uh, I hate to say this, but I cannot do that

I have to know the size of the array containing the title/version at creation-time

ATaco

Rules of Thumbs and Principals are meant to be broken

Ginger

there's probably a way to do arbitrary lengths, but it's almost certainly going to be jank as hell

pxeger

if it's going to be jank as hell, you might want to figure out a way for it not to be jank

ATaco

What's forcing the limitation?

Ginger

22:19

the manner in which I'm processing the filetype containing the data

pxeger

because eventually you'll probably find something that actually needs to allow arbitrary lengths

Ginger

lemme see if I can fix it...

ATaco

Linux MAX_PATH is 4096, if that matters

Ginger

hm yeah CharBuffer wants an allocated length up-front, as does CharArray

lyxal

No string types?

Ginger

22:21

idk if Kotlin has mutable String types, lemme look...

lyxal

StringBuilder, no?

Ginger

wat that

pxeger

google it?

Ginger

huh, yeah

that could work, one second...

lyxal

4

Q: Does Kotlin support mutable Strings?

The following code doesn't work in Kotlin: var s: String = "hello" s[1]='a' Does Kotlin have a mutable string type or do I have to use an Array<Char> if I want to be able to edit?

kotlin

The accepted answer demonstrates a string builder

Ginger

22:30

well, I made it, and it is exactly as awful as I thought it'd be:

val name =
        StringBuilder()
                .also {
                    loop@ while (true) {
                        with(getChar()) {
                            if (this != Char.MIN_VALUE) it.append(this.toChar())
                            else break@loop
                        }
                    }
                }
                .toString()

lyxal

The hell kinda file format are you using?

Ginger

wait, that one's bugged, here's a fixed one:

val name =
        StringBuilder()
                .also {
                    loop@ while (true) {
                        with(getChar() ?: break@loop) {
                            if (this != Char.MIN_VALUE) it.append(this.toChar())
                            else break@loop
                        }
                    }
                }
                .toString()

lyxal

1 min ago, by lyxal

The hell kinda file format are you using?

Ginger

A custom-made one, which should tell you all you need to know :b

lyxal

Well not really. I'm trying to say maybe there's a better way to read it

Ginger

22:35

that awful mess reads an arbitrary-length nul-terminated string from a ByteBuffer

works great!

Wheat Wizard

I love the aesthetic result of writing normal code with a large number of stupid golfing builtins that hgl's source code ends up with.

x_W errorString ordering =
  rFp (Prelude.errorWithoutStackTrace errorString) <
    lF' (zd' (maW ordering) ^. p) []

Ginger

lmao

lyxal

@Ginger why does it have to be null terminated? And why a byte buffer? Surely using even something like a Scanner or BufferedReader is a better idea

Ginger

Scanner? I 'ardly know 'er!

/srs I don't know what a Scanner is and BufferedReader doesn't work for my purposes

lyxal

...

A Scanner is like only the most common class taught in intro to Java classes

And because kotlin is jvm, it has it too

Ginger

22:40

never done one

lyxal

Well that explains a lot

Ginger

lemme look it up

lyxal

@Ginger why doesn't a buffered reader work?

Ginger

It can only read CharArrays

I'm using ByteBuffer because I can read whatever I want as whatever I want

and Scanner doesn't look like it'll work either

lyxal

@Ginger it's supposed to read files though

Ginger

22:42

well now I'm in an argument with my language server

lyxal

How are you opening the file to read?

Ginger

it says the Elvis is useless because getChar() doesn't return a nullable type, but if I remove it it complains about unreachable code

ATaco

When I needed to read a file in Java recently I just let Github Copilot right the code for me.

Ginger

@lyxal I'm not, the code calling this library gives me a File object

ATaco

Straight up just // Read file.cfg and it did the hard part.

Ginger

22:44

but this is a custom format

so I have to write my own reader

lyxal

@Ginger so then you just pass that to a file reader and then a buffered reader, get a list of strings and parse those

Ginger

it's binary data

I have to manipulate the raw bytes myself

emanresu A

Why though

Ginger

well I fixed the warning by casting the result of getChar() to Char? :/

@emanresuA what do you mean "why"?

emanresu A

Why not use something nice like json

lyxal

22:46

@Ginger what library?

Ginger

@emanresuA because I CAN

@lyxal the one I'm writing right now

lyxal

3 mins ago, by Ginger

@lyxal I'm not, the code calling this library gives me a File object

So change it then to not return binary data

Ginger

...what

what do you mean

I need the binary data

lyxal

But you're parsing a package information file right?

Is that what you're doing?

Ginger

I'm parsing a library information file, but yes close enough

lyxal

22:49

Is this library information file saved locally in a directory?

Ginger

yes

lyxal

Open the file then and parse it yourself

Use buffered reader to get the file contents

Ginger

@lyxal I am!

I am writing a library which will be called by my code

lyxal

Why is it being converted to binary data at all?

Ginger

my library, my code, ~~my fountain, my world~~

@lyxal because that's the format it's in

lyxal

22:50

@Ginger Too many fountains causes the roaring

@Ginger change the format then

Ginger

don't wanna

lyxal

Because you're stubborn and don't want to save yourself a whole lot of time?

Ginger

yes

lyxal

...

Ginger

also, I think the decoder works now

ATaco

22:53

@lyxal Okay Bigshot

Ginger

wow, only 72 lines right now

now I just have to do the encoder

easy(?)

here, have some more cursed code:

fun serialize(): ByteArray {
    return ByteBuffer.allocate(3 + 16 + name.length + 1 + version.length + 1)
            .put(this.recordType.identifier.toByte())
            .putShort(this.installSource.toShort())
            .put(this.hash)
            .put(this.name.toByteArray())
            .put(0x00)
            .put(this.version.toByteArray())
            .put(0x00)
            .array()
}

*laughs in magic constants*

Unrelated String

23:08

lol

Ginger

okay, it's done

95 lines of jank

Unrelated String

thought of this last time i tried minecraft modding; you should really never have to write serialize and deserialize separately

Ginger

including such gems as stream.write("🐇📕".toByteArray()), with(this.getChar() as Char? ?: break@loop) {, and of course ByteBuffer.allocate(3 + 16 + name.length + 1 + version.length + 1)

Unrelated String

or at least there should be some kind of shared format object/type

Ginger

@UnrelatedString wdym?

oh, I forgot to add an internal modifier to that function :\

yeah that's for internal use only

@RydwolfPrograms come see what I hath wrought

Unrelated String

23:11

like, you shouldn't have to write the format out twice, both as a matter of convenience and not being able to fuck it up

Ginger

look upon my works, ye mighty, and despair

let's see if it buil- oop it's done

my favorite line is with(this.getChar() as Char? ?: break@loop) {, because (it looks like) I'm casting a type to a nullable version of itself only to immediately turn around and check if it's null

Unrelated String

lol

Rydwolf Programs

@UnrelatedString How would that work?

Unrelated String

i tried to make some kind of format class and it failed miserably so i'm not sure how in java :P

Ginger

the real story here is that this.getChar() can return null, because it's a Java function, but my typechecker apparently doesn't know that

Unrelated String

23:15

iirc it just like
named the fields and enforced the order

Ginger

weirdly, it does know that it could be null if I use getChar elsewhere, but that instance is inside an extension function, which I guess the typechecker has a hard time with?

now, let's see if it works

Bbrk24

23:28

I don't remember how nullability works in Kotlin but Swift has three levels of nullability: T (not nullable), T? (nullable), and T! (an Objective-C function didn't annotate which one it is). That last one can be converted to plain T without having to explicitly check for nil, but then the compiler unwraps it for you (equivalent to !! in Kotlin, but implicit).

To unwrap nil like that is slightly more dangerous in Swift though, since it causes SIGILL rather than just throwing an exception

Ginger

mfw gradle manages to build the library... without including the classes somehow

oh

I'm stupid nvm

> 'break' or 'continue' jumps across a function or a class boundary

oh no

user

@Bbrk24 Kotlin has a similar system: T for not nullable, T? for nullable, and T! for a type from a different JVM language

@Ginger I would recommend against putting stuff like break or return inside expressions like that. That's just asking for bugs

Bbrk24

One fun part of the Swift type system is explaining the difference between as T? and as? T

user

Does the first error if the expr is null?

Rydwolf Programs

@Ginger Wait what does that mean out of curiousity

Like how does that work

user

23:39

@Ginger This does not be that complicated

@RydwolfPrograms They used a break inside a lambda inside a loop

Bbrk24

as T? can be used to do things like Int? -> Any? or Int -> Int?. as? T can be used to do things like Any -> Int?, where it returns nil if the value isn't actually an int

user

val sb = StringBuilder()
var c = getChar()
while (c != Char.MIN_VALUE) {
  sb.append(c)
  c = getChar()
}
val name = sb.toString()

It's not the prettiest code ever but it sure is easier to understand than that stuff you have right now (@ Ginger)

What exactly is getChar()?

Bbrk24

C brain is saying that it returns int and not char

user

Ah

Bbrk24

But I'm probably wrong

user

23:48

No, you might be right. According to the docs, MIN_VALUE is a Char, but Ginger did do toChar in their code

Bbrk24

char is barely a type in C. Most functions that could take or return a single char use int instead, and the default type for character literals is even int. (I've seen someone abuse this: sizeof 'c' == 1 ? "C++" : "C")

at least it's not bool

user

lol I thought you could use macros to determine that

ifndef CXX_VERSION or something

Bbrk24

__cplusplus, but yes

1==sizeof 'c' is shorter for code golf

user

Ah

In that case, you could use > or <

Bbrk24

yeah they probably did, idr

1<sizeof 'c'?"C":"C++"

ATaco

23:57

Would 1<sizeof'c'?"C":"C++" not work?

Bbrk24

No, that'd be parsed as a single literal in C++

L'c' is valid, MY_MACRO"C" is a compile error, so I can only imagine sizeof'c' would error too

Transcript for