The Nineteenth Byte

New Posts

00:01

0

Q: CGAC2022 Day 2: Self-trapping Elf

Part of Code Golf Advent Calendar 2022 event. See the linked meta post for details. An Elf is hopping on a chessboard which extends infinitely in all four directions. He moves like a generalized chess Knight: on each step, he moves X units horizontally and Y units vertically, or Y units horizont...

Bubbler

@DLosc I'd say that's why challenge-solving experience matters when writing a challenge

Ginger

01:37

Idea: a language where the syntax is formal logic

so hello world would be the formal representation of "there exists a string x such that the content of x is hello world and x is sent to stdout" or something similar

I have absolutely no clue how you'd implement this language

lyxal

Isn't that just prolog?

Ginger

never used it so not sure

Jo King

yes, that's pretty close to how prolog operates

Ginger

yea it is :p

ah well

seemed like a cool idea

Jo King

feel free to learn prolog :p

Ginger

01:47

 perfect(N) :-
     between(1, inf, N), U is N // 2,
     findall(D, (between(1,U,D), N mod D =:= 0), Ds),
     sumlist(Ds, N).

uhhhhhhhhhhh

ಠ_ಠ

@JoKing yea I'll get right on that

Jo King

that snippet isn't really that hard to understand though

Jo King

02:23

what's cool about it is that it can be used to generate perfect numbers, as well as verify them

@cairdcoinheringaahing you should check if your system comes with it pre-installed before you start buying mod lootboxes

caird coinheringaahing

@JoKing I know,but it's just so fun to gamble. Did you see the holiday prize? We could potentially get the power to force a CM to just outright fulfil one feature-request :P

Jo King

the drop rate is going to be like 0.1% though

i already have enough user skins

caird coinheringaahing

Fair enough. I'm still gunning for the Browncat skin, apparently they increased the rarity recently

A certain song just happened to appear in my random playlist, so: CMP: Have y'all heard All I Want For Christmas Is You yet this holiday season, and if so, how soon after Halloween was it? :P

Seggan

02:44

nope

lyxal

03:35

@cairdcoinheringaahing by choice or by force?

caird coinheringaahing

Yes

lyxal

5th of november

Nov 5 at 4:15, by lyxal

I see your youtube link and raise you another

mousetail

10:15

LDQ What should arithmetic operators on functions do? Like +, -. *, / etc.

Or when one side is a function but the other a number or array

Adám

@mousetail Compose them: (f+g)(x) = f(x)+g(x)

mousetail

What about f + 3? This is for a golfing lang so I'd like every possible combination to do something useful

Adám

Let 3 act as a constant function. (f+3)(x) = f(x)+3

mousetail

That makes a lot of sense thanks

Adám

Btw, this is what happens when doing tacit programming in APL and its derivatives.

mousetail

10:36

When composing with a array, should f + [a,b] return [f+a, f+b] or lambda:f(x)+[a,b])

Adám

10:52

@mousetail The latter, because maybe f(x) returns a 2-element array too.

Neil

11:37

@RadvylfPrograms are you aware of the connection to the Riemann Zeta function?

Ginger

12:15

of course there's a connection to the Riemann Zeta function, why wouldn't there be a connection to the Riemann Zeta function

mousetail

LDQ When doing logic on arrays with different lengths, should the output I only apply to the shortest one, or just NOP the extra elements in the longest array

Ginger

13:20

wait no

lyxal

13:32

winterbash2022.stackexchange.com

Ginger

YOOOOOOOOOOO

lyxal

12 days until hats

mousetail

Wasn't winter bash all of december before?

Ginger

fun fact: the site code updates the counter twice per second despite it only having second resolution

mousetail

Makes sense. Browser times frequently take just slightly longer than specified to run and the difference can add up

Ginger

13:35

also there exists code to reload the site once the timer hits zero

shame I can't find any secrets...

Radvylf Programs

14:18

@mousetail I'd go with the second

Easier to get to the first from the second than vice versa

Maybe instead of NOPing, fill the other list with some sensible default

For most things it would be a NOP, but for others you'd be able to make it a useful "overload"

mousetail

Thanks

Radvylf Programs

15:01

@Ginger I usually do something like 10-40 times per second to make it as close to accurate as possible

→ 1 message moved to Trash

Well someone's not getting the christmas spirit

(also did the bakery freeze?)

@cairdcoinheringaahing Can you unfreeze the bakery pls

@lyxal Looks like they're doing some sort of half-winter-half-summer thing this year which is cool

I guess they forgot Australia doesn't really exist

@cairdcoinheringaahing Yeah, Spotify randomly played the MCR cover of it a few days ago. A little earlier than I expected to hear it lol

mousetail

15:19

LDQ Would it be a major issue is input is available only as a lazy array?

mousetail

15:30

That got deleted super fast

hyper-neutrino

→ 1 message moved to The Nineteenth Bakery

yeah I was in the middle of destroying the account so it automatically got deleted

Radvylf Programs

I've never seen this in a review task before

mousetail

I see it all the time on SO, though 99% of the time it indicates a audit and the post is fine

Radvylf Programs

Interesting, it let me review a deleted post:

mousetail

"Looks ok"?

Ginger

15:34

can someone find me that website which let you use block-based coding to represent logical stements?

Radvylf Programs

(if you're wondering about "looks ok", since it's deleted no other action was allowed)

Ginger

it got mentioned a while ago but I've lost the link

@RadvylfPrograms ಠ_ಠ

what was that person's problem

Radvylf Programs

32 mins ago, by Radvylf Programs

Well someone's not getting the christmas spirit

Ginger

ah well

mousetail

Good thing we have plenty of challenges coming about identifying naughty vs nice kids

Ginger

15:36

lmao

2 mins ago, by Ginger

can someone find me that website which let you use block-based coding to represent logical stements?

mousetail

Pure data?

Ginger

not it

it used blockly or smth

brb

Ginger

15:48

back

Seggan

15:59

@Ginger scratch?

Ginger

nope

Seggan

snap! ?

Ginger

not it either

it was very specialised

and iirc it included challenges to represent some set of prepositions with blocks

Seggan

16:20

Blockly

Blockly is a client-side library for the programming language JavaScript for creating block-based visual programming languages (VPLs) and editors. A project of Google, it is free and open-source software released under the Apache License 2.0. It typically runs in a web browser, and visually resembles the language Scratch. It is also being implemented for the mobile operating systems Android and iOS, though not all of its browser-based features will be available on those platforms. Blockly uses visual blocks that link together to make writing code easier, and can generate code in JavaScript, Lua...

applications section

Radvylf Programs

None of those have to do with logic

Seggan

yeah just giving ginger resources

> a long list of Java/Python/JavaScript/C++ wannabes: Ceylon, Clojure, Crystal, Elixir, Elm, Haxe, Julia, Kotlin, Nim, Pharo, Racket, Rust, etc.

imo rust and kotlin are above that now

qr.ae/pr73dp

that was 5 years ago

DLosc

16:42

@cairdcoinheringaahing November 28 :P

(though it's possible it was playing at a store a day or two before that and I just tuned it out)

@RadvylfPrograms They did that last year too--at least, I remember there were two versions of the artwork on the (Wint|Summ)er Bash homepage

caird coinheringaahing

@mousetail no, winter bash is usually half of December, plus the first week in Jan

Ginger

17:56

am I hallucinating this website???

I swear it exists

HAHAAAAA I FOUND IT

mathsblocks.com

Ginger

18:17

Presenting, (maybe) Hello World in this hypothetical logic language: ∃x∈S((x="Hello World")∧(x∈O)) where S is the set of all strings and O is the set of all lines of text written to standard output

very readable, isn't it

if I'm interpreting it correctly it basically says "there exists some element x in the set of all strings where x equals "Hello World" and x is in the set of all lines of text written to standard output"

and then the interpreter has to make that statement true

the idea is to make a program such that that the side-effects of the interpreter enforcing the rules of the program are the desired results

Unrelated String

18:46

mmm

Seggan

@Ginger id star that but i dont wanna clutter the board

mathcat

@Ginger wow

Ginger

@UnrelatedString is that a good or bad noise

Unrelated String

interested

Ginger

cool

Seggan

18:51

are you gonna build it

Ginger

the interpreter?

no clue

Seggan

mind if i start on it

Ginger

it would be, how shall I put this, less than trivial

@Seggan sure, good luck

I'll make some test programs for you

Seggan

given everything else im workin on rn, i probably will do it verrrry slowly (i.e. when i feel like it)

if anything, just ask for collab on GH

Neil

@Ginger why can't you just write ("Hello World"∈O)?

Ginger

19:00

you could

mathcat

Don't ever question esolangists.

Ginger

but I didn't think of doing that :p

@mathcat no, they're right

Seggan

yeah i was gonna ask why you said hello world was in the set of all strings if its a string

Unrelated String

maybe get rid of string literals; have regex literals as predicates :P

Seggan

oooh lol

the intersection of all strings and all strings matching this regex :P

ok wait thats just all strings matching this regex

Ginger

19:03

this program should count up forever if N is the set of all natural numbers: ∀x∈N(x∈O)

Seggan

my set theory is bad, whats the upside down a?

Ginger

"for all"

hyper-neutrino

for all

Seggan

ah

Ginger

get ninja'd lol

Seggan

19:04

for all x in N print?

Ginger

@Seggan yes, pretty much

<something>∈O literally means "<something> is in the set O" and since I'm using O to represent stdout it would just print <something> to stdout

(btw I'm learning these symbols as I go so feel free to correct any mistakes)

Seggan

@Ginger yeak ik the e

~~dont tell anyone but all the set theory i know i picked up myself, i never took any course in that~~

and from algebra 2

Ginger

lol

this one should print all even numbers in the set of natural numbers: ∀x∈{y∈N:y%2=0}(x∈O)

mathcat

do rationals

Ginger

hmm

that'd just be a change of N to R in theory

Seggan

19:08

@Ginger for all x in the set of (all y numbers such that y%2 = 0), x is a member of stdout?

Ginger

@Seggan exactly!

Seggan

who needs to take set theory when its this easy? :P

Unrelated String

are we thinking prolog-esque semantics where things do execute in some known order with reliable side effects

Ginger

that's an implementation detail; I'm just doing syntax right now

Seggan

so this is basically set theory, executed?

Ginger

19:11

yes

and some formal logic as well

some things are different to make complex programs possible; an example is in progress

mathcat

What about loops?

Seggan

i think the upside down a does that

Ginger

^

mathcat

right

Ginger

conditionals are the tricky bit and what I'm working on right now

Seggan

19:13

∀x∈{y∈N:y<10}(x∈O)

print 0..9

Ginger

yup

Seggan

or {y∈N:y<10}⊆O should do the same thing

(dont ask me i just looked up a symbol table)

Ginger

the syntax really becomes iffy with programs like FizzBuzz, which should look something like: ∀x∈N((((x+5=0)⟹("Fizz"∈O))∧((x+3=0)⟹("Buzz"∈O)))∧("\n"∈O))

there are a lot of weird bits to this program; lemme explain

first: I'm using ∧ to join statements because this language will (probably) be eagerly evaluated

Seggan

yep (to the eagerly evaled part) at least thats how i plan to do it

Ginger

second: I'm using ⟹ as an if statement, which introduces another weird syntax thing: the distinction between "statements" and "questions"

in short, questions are evaluated to check their return value, but statements are enforced to cause effects

(replace all the +s with %s btw)

Seggan

19:21

is there not a way to do it with set unions/intersections?

Ginger

x%5=0 is being evaluated to check if x mod 5 equals zero, but "Fizz"∈O is enforced to make "Fizz" a member of set O

if "Fizz"∈O were on the other side, it would be evaluated to check if "Fizz" is a member of set O

and that is (the main reason) why I don't want to implement this

Seggan

wait if O is a set, then you cant print multiple of the same thing to stdout?

Ginger

fair point

O isn't technically a set

Unrelated String

you could make it so strings aren't unique mathematical objects

Ginger

it's a special object representing stdout, which can only be written to and accepts duplicate elements

Unrelated String

19:24

or maybe O's members are supposed to be tuples

Ginger

strings are unique to allow things like "the set of all strings" to exist

Seggan

"a" = "a" tho

Ginger

@Seggan that's my point

those both represent a single string, a point in the space of all strings

so of course they're equal

Seggan

@Ginger then wdym by this

nvm

i get it

Ginger

things really become interesting when you consider the ramifications of things like set complements

Seggan

19:27

hmm to actually implement this sets will need to be ordered

Ginger

@Seggan "sets" in this language will probably be similar to "lists" in python, but maybe w/o duplication

Seggan

yeah

Ginger

actually they'll be closer to a cross between generators and lists

Radvylf Programs

So lazy lists

/iterators

Seggan

@Ginger stringset^c = all strings not in stringset

Radvylf Programs

19:29

Y'all aren't going to be able to implement this like a normal imperative language, even with laziness

Ginger

who's this "y'all", kemosabe?

I'm just the SYN_TAX guy

Radvylf Programs

You and Seggan

Even as the syntax guy you've still got to think about implementation at least a little

Since implementation will change how that syntax works and what syntax is possible

Seggan

@RadvylfPrograms and how do you suggest we do it?

i was thinking a depth first approach where the interpreter tries to make it true

Radvylf Programs

E.g., you'd probably be able to explain the collatz conjecture to the interpreter in terms of set theory

Unrelated String

@Seggan sounding a lot like prolog again :P

Radvylf Programs

19:32

At some point turning programs in this lang into nice imperative code is going to run into some halting problem stuff

Ginger

it is

which is why this interpreter will need some stuff to make sure that no paradoxes arise

Seggan

@UnrelatedString eh, its a fun take on prolog :P

Unrelated String

:P

Seggan

what should we call it

Proset

Unrelated String

high on math

Radvylf Programs

19:34

"Unimplementable" so someone else gets nerd sniped into writing it

Unrelated String

ahah

yes

Seggan

@RadvylfPrograms whats why you interpret it, not (trans)(comp)ile

maybe i should learn prolog for this

Ginger

"Complement"

~~"Aristotle"~~

Seggan

"I Got Nerd Sniped"

TIL the "highlights" button

Ginger

I don't think there's any easy way for this language to take input

Radvylf Programs

19:46

"x is in input"

Then you can just use x like anything else

Ginger

makes sense

Seggan

@Ginger i was thinking of that lol

Radvylf Programs

IeO (where e is the subset symbol) is even a somewhat un-rigorous cat program

I=O would actually work too wouldn't it

Ginger

hmmmmmmmm

the second one should, the first one...

I have no idea what that'd do

you'd be saying that input is contained within output

Radvylf Programs

Yep

Seggan

19:50

also print the input then

Radvylf Programs

So it's UB

Ginger

so I suspect it'd just print a representation of the input object and exit?

either that or error

Seggan

@RadvylfPrograms the second one has UB if you add something to stdout

Ginger

because input and output are both infinite "sets"

Seggan

i.e. what is I=O(1eO) where e is element of

Radvylf Programs

19:51

Well they're not infinite are they? Just have the potential to be?

Ginger

anyway, here's a program which would cause the interpreter to error with something along the lines of "unenforcable program": ~1=1

Seggan

@RadvylfPrograms imo the first one should be the answer

Ginger

remind me what "UB" means?

Radvylf Programs

Which first one

@Ginger Undefined behavior

Ginger

ah

Seggan

19:52

3 mins ago, by Radvylf Programs

IeO (where e is the subset symbol) is even a somewhat un-rigorous cat program

Radvylf Programs

Yeah but that could also mean "print the input plus some other stuff"

I=O is the one that means "the output is the input"

Ginger

technically it means "make the output equal the input"

Seggan

3 mins ago, by Seggan

i.e. what is I=O(1eO) where e is element of

Ginger

and the interpreter has to enforce that

Radvylf Programs

@Ginger No, it means I is equal to O, right? It's only since the compiler is actively trying to make that true that it turns into a command like that

Seggan

19:55

^

therefore the program is O=I?

Ginger

@RadvylfPrograms yes, I guess

Seggan

wait this aint assignment

it means "I is the same as O"

Ginger

here's a question: would the program min N (where N is all natural numbers) exit?

I think it should

Radvylf Programs

@Seggan yeah

Seggan

so since it cant change I, it will change O, no matter what order

Radvylf Programs

19:57

Yep

= is commutative

Seggan

okie i get it

Ginger

ideally, if the interpreter were given the program ∃x≠0(x=0) it would error instead of running forever

Radvylf Programs

but okie isn't here right now

Ginger

@RadvylfPrograms :|

Radvylf Programs

(also, that's who I was confusing you with, not Niko)

Seggan

20:00

@RadvylfPrograms aww

@Ginger english please?

Ginger

@Seggan "there exists a value x, which cannot be equal to 0, that equals 0"

Radvylf Programs

Well you'd need some sort of constraint-based thing anyway

Seggan

yeah thats an obvious contradiction

Radvylf Programs

So infinite looping would be indicative of a very poor implementation

Ginger

(btw I'm using mathsblocks.com/html/builder.html to make these programs, it's very useful)

interestingly, mathsblocks lets me make this block:

what exactly would that do?

Seggan

20:03

tuple decomposition?

Ginger

maybe???

Radvylf Programs

@Ginger "for all x, y is a subset of ...", right?

So you could use it to say that y is the intersection of everything in x, for example

Ginger

but I can add any number of extra variables

Seggan

theyre probably all equal

Ginger

assuming you were good enough at set theory, you should theoretically be able to make a program in this language to generate (eventually) any series of numbers

Seggan

20:09

which im not

Ginger

me neither :b

Unrelated String

or just all of them are unrelated members of the set

i'm like 99% sure that's a common notational convenience

Radvylf Programs

@Ginger Nice, your language has achieved like, the bare minimum for an esolang :p

Ginger

yay

Seggan

20:24

should numbers and strings be one and the same?

or should we do like "x is in input and in the set of all numbers" to convert number to string

Radvylf Programs

A string could be an ordered multiset of char code integers

Seggan

which is a list

Radvylf Programs

yep lol

emanresu A

Osabie makes strings and numbers behave identically. IME it doeesn't work very well

Radvylf Programs

You could do numbers set theory style, and represent strings as a set of (char code, index) pairs

Seggan

20:26

@Ginger your input?

Seggan

20:46

trying to do primality, im lost

{p∈N:(p≠1)∧∀n∈N(((n>p)∨(n=1))∧(p%n=0))}⊆O i think?

Ginger

sorry, was afk

@RadvylfPrograms I don't really like this

strings and numbers should be separate

@Seggan I dunno man :p

Radvylf Programs

So you're saying this lang will have data types?

Ginger

I see no reason why not

do you? I might've missed something q:

Radvylf Programs

I guess not

Ginger

cool

Radvylf Programs

20:52

Feels kind of weird to have strings, a collection type, as a data type in a language built around sets tho

Ginger

yeah, that's fair

but how'd you represent strings then? remember, they're supposed to be a point in the space of all possible strings

Radvylf Programs

You'll need some way to turn strings into a set representation anyway

For iteration purposes

So why not just store them that way all the time

Ginger

fair enough

Radvylf Programs

@Ginger something like {'a', {'b', {'c'}}}

Assuming your sets are unordered and non-multi

Which you probably want

Seggan

@RadvylfPrograms only strings and numbers

Ginger

20:54

sets need to be ordered for infinite sets (like that of all natural numbers) to work

Radvylf Programs

@Seggan and sets...

Ginger

which reminds me: we already have an Array Programming Language, why not a SEt Programming Language?

Seggan

@RadvylfPrograms and that :P

Radvylf Programs

@Ginger How?

Ginger

@RadvylfPrograms I'd assume that if you want to iterate over all natural numbers you don't want to get 1, 214894732189473219000000000000000000000000000000000, 377, 281432, etc

Radvylf Programs

20:54

I mean internally they'd need to be ordered, but set theory uses unordered infinite sets all the time

Seggan

@RadvylfPrograms so the program wont print 34, 247689, 2345, 0, ..

ninja'd

Radvylf Programs

@Ginger The order that things are iterated over internally doesn't matter

Ginger

???

Radvylf Programs

Give me an example of where it matters, where I/O is unordered

Ginger

how the [$!$$] would unordered I/O work

Radvylf Programs

20:56

The same way unordered anything works

Raw input and output would be through sets, just like how raw BF input and output is integers

You'd just use some set-based representation of an ordered list if you needed one

Like {'a', {'b', {'c'}}} for strings

Ginger

...

it makes sense but I really don't like it

Radvylf Programs

Having ordered sets but only in one place is going to be really really weird and break a lot of stuff

Seggan

@RadvylfPrograms like what, i can write World! before Hello, ?

Radvylf Programs

And if everything uses ordered sets, it's not going to be very set theory-y of a language

Seggan

true

Radvylf Programs

20:58

@Seggan There is no "before"

Everything happens at once, the language is just a statement/set of statements about the relationship between the input and output

Ginger

I guess the entire program would kinda have to run all at once

Radvylf Programs

Yeah, exactly

Ginger

(also, this is nit-picking, but I already said that I/O objects aren't actually sets, they're special objects that have a few set-like features)

(you can't read from stdout for example)

Radvylf Programs

That doesn't really make much sense does it?

Ginger

how'd you implement iterating over stdout then?

Transcript for