The Nineteenth Byte

DLosc

00:00

@RedwolfPrograms Understandable. I tolerate ==, but with my math + QBasic background, I don't think I'm ever going to prefer it if = is a decent option.

In fact, I just realized, not one of the languages I've ever created uses ==. (Although, only one of them was designed with any sort of practicality in mind, so take that with a grain of salt.)

Neil

@Mayube since it has no data, source-reading or eval primitives, all its quines are true by default

Redwolf Programs

TFW you accidentally use a 0% infill instead of a 100% infill because you forgot how numbers work

lyxal

TFW when you forget how input works in your language so you try and bugfix an issue that doesn't actually exist

Leaky Nun

is thing happening here?

DLosc

Yep, I was just waiting for someone else to agree that we're learning /// this week

Redwolf Programs

00:09

That's the random-find-and-replace one, right?

DLosc

Not random, but yeah

Redwolf Programs

Oh, not random? Not sure which one I'm thinking of that does it randomly then

Maybe Thue

user

@RedwolfPrograms Not really, they can lead to bugs if not used right (although this isn't the strongest case for them), they're not super useful since you can do i += 1 in the majority of cases, and they make parsing very slightly harder (again, not the strongest case against them)

DLosc

Welcome to the sixth Learn You A Lang For Great Good! Today, we'll be learning /// (pronounced "slashes"), a surprisingly Turing-complete esolang based on string replacement. You can check out the linked esolangs.org page for some examples, or mess around with it on Try it online!. Feel free to ask questions, post ///-themed CMCs and more for the next 24 hours or so.

4

@RedwolfPrograms Yeah, looks like it

user

@DLosc I hate languages with unsearchable names even more (looking at you, C)

hyper-neutrino

00:16

→ 26 messages moved to Sandbox

DLosc

@user To be fair, C was named before the invention of search engines =P

user

Hey just let me hate C in peace okay? :P

DLosc

okay

Bubbler

///MC: Add two unary numbers. First number is written in a, second number is written in b, the output should be written in c. aabbb -> ccccc

///MC: Multiply two unary numbers. aaabbbb -> cccccccccccc

DLosc

@Bubbler I have 10 bytes, which I think is optimal

Bubbler

00:21

Yeah, I agree that 10 is optimal

Just throwing out simple ideas because I guess we have to learn on our own today

DLosc

/// is one of those languages where I know how it works, I just don't know how to program with it.

hyper-neutrino

idk if this (10 bytes) is right... it has an extra trailing newline, and I'm not sure if I'm even giving input correctly ¯\_(ツ)_/¯

Bubbler

Just append the input in the code section

Having a footer adds a newline, which isn't good for our purposes

hyper-neutrino

oh okay

DLosc

(although in this case, I think ignoring the newline in the output is a reasonable thing to do)

hyper-neutrino

00:25

oh right /// doesn't do regex right?

Bubbler

I suppose it might not work if the program gets more complex

Yeah, no regex

just plain string substitution

DLosc

@Bubbler Does it need to be able to handle either of the numbers being zero?

Bubbler

Let's say both inputs are strictly positive, if it makes things simpler

DLosc

Simpler, but definitely not simple...

Is this actually going to require making a loop using self-modifying code already??

Bubbler

Idk lol

DLosc

00:39

Ooh, wait, I have an idea

Aha! 12 bytes

And note that putting this in the header gets rid of all newlines from the rest of the program:

/
//

So that allows for putting the input in the footer, as long as the input can go after the entire program.

Bubbler

Nice find

And 17 to output in c

user

@DLosc Nice, how does that work?

DLosc

@Bubbler Oh, right

@user Consider this version that replaces some of the slashes with underscores.

Bubbler

The last substitution /b/... is incomplete, which is dynamically created by inserting a slash after a's

So it becomes /b/aaa/bbbb after first sub

DLosc

We turn the three as into a command that says "replace every occurrence of b with three as". Then we run that command on the remainder of the program (i.e. the bs).

Bubbler

00:46

which is classic multiplication by sub

user

Ah, clever

DLosc

///MC: Given a unary number, multiply the number by 3 and add 1 if the number is odd, or divide the number by 2 if it is even.

hyper-neutrino

hm, I managed to get mortarboard on math se meta? interesting

Bubbler

@DLosc 34, takes a and outputs b

hyper-neutrino

01:03

49 bytes, inputs a and outputs d

DLosc

I've got 36 without looking at either of yours yet

Bubbler

Meanwhile, 17 for multiply that supports zeros

Leaky Nun

01:18

@DLosc 31, takes a and outputs c

Bubbler

Oh, obviously I didn't need to dynamically add code for multiply

DLosc

@Bubbler The previous approach also supports zeros

Bubbler

Oh... by forming an incomplete sub I guess

///MC: Given two numbers a and b, compute a^b (exponentiation).

^ I have 27

Fmbalbuena

@DLosc Hey

DLosc

o/

New Posts

01:29

0

Q: Can we change the Late Answers prompts regarding duplicates?

Unlike other (Q&A) sites in the Stack Exchange network, CGCC has a consensus that duplicate answers are acceptable. For some code golf challenges there is one obvious shortest solution in a particular language, so it's unsurprising that different users might independently arrive at the same answe...

Leaky Nun

@DLosc wow! out of the box

DLosc

@Bubbler Oh, duh, the underscore shenanigans I was using aren't necessary at all. In that case, 26 bytes

Bubbler

Nice lol

DLosc

@Bubbler Can we take them in reverse order?

Bubbler

Yes, that's fine

Leaky Nun

01:37

@Bubbler 27

and since we both have 27 i'll just conclude that it's optimal XD

DLosc

Why do I get the feeling that I'm overengineering this... :P

Bubbler

My 27 handles 0^x and x^0 too

DLosc

What does it give for 0^0?

Bubbler

1

New Posts

0

Q: pattern matching coding challenge

Is there any possible better solution for the below pattern-matching coding challenge without using if-else statements? one of the most important aspects of accurately matching records to candidates is ensuring that the name in the record matches the name or a known alias of the candidate. In thi...

java pattern-matching

Fmbalbuena

01:45

@DLosc im new at " Learn You A Lang For Great Good!"

Bubbler

It's just a cool name for an "event" where you try to learn a specific language and solve language-specific CMCs

Fmbalbuena

what is CMC?

Wezl programs redwolf

CMC: explain CMCs

@Wezlprogramsredwolf CMC = Chat Mini Challenge

Leaky Nun

Chat Mini Challenge: a small task posted in the form of a message

59 mins ago, by DLosc

///MC: Given a unary number, multiply the number by 3 and add 1 if the number is odd, or divide the number by 2 if it is even.

for example ^

Fmbalbuena

I understand

Leaky Nun

01:48

and we answer by replying to the comment

Fmbalbuena

How to make text bold in chat stack exchange

hyper-neutrino

**bold**

DLosc

@Bubbler Argh

Fmbalbuena

Can i make mini-challenges

hyper-neutrino

sure, anyone can

Fmbalbuena

01:55

///MC: Given two numbers a and b, compute a + b in Roman numeral

DLosc

@Fmbalbuena Meaning, a and b are given in Roman numerals and the output should also be Roman numerals?

Fmbalbuena

the input must be unary and output must be roman

Leaky Nun

///MC: Given a number, "square" it but literally, i.e. aaa should produce:

aaa
aaa
aaa

DLosc

02:10

@Fmbalbuena 106 bytes

Fmbalbuena

@DLosc i think is optimal

///MC: Given two numbers "a", "b", convert unary to base n (Min 2, Max 16)

Bubbler

n is also in the input?

Fmbalbuena

02:34

Yes

DLosc

This may be helpful to others besides just me: a /// interpreter in Pip with toggleable debug output.

Try it online!

DLosc

02:54

@LeakyNun Does the output have to use the same character as the input?

Leaky Nun

@DLosc no

DLosc

@LeakyNun 26 bytes, output has leading newline

Bubbler

03:08

@LeakyNun 26 with trailing newline

///MC: Given a number n, output nth triangular number.

^ 28

///MC: Given n, output n! (n factorial)

hyper-neutrino

0 bytes, but I'm not sure why you're so excited about this particular challenge (/s)

5

Bubbler

Lol

Factorial actually sounds hard btw

Mayube

///MC: Given n, output ~n, ie bitwise NOT of n

Bubbler

@Mayube In what representation of n?

Mayube

@Bubbler unary

Bubbler

03:22

Unary doesn't have negative numbers

Mayube

input in unary, output in binary

Bubbler

And negative binary (as in 2's complement) is infinite to the left

Probably we can do it in signed unary though (so 3 = aaa, -2 = -aa)

In that case 13

Mayube

you only need to NOT the bits needed to represent n. So for n = 6 output would be 001, for n = 21 output would be 01010

New Posts

0

Q: Verb arities in J trains

Background J has trains similar to APL's. Given a sequence of verbs (functions), three rightmost verbs are grouped to form a derived verb (a fork) recursively, until one or two verbs remain. If the sequence has odd length, the entire train is a chain of forks. Otherwise, two verbs remain at the e...

code-golf

Bubbler

@Mayube Then it's just unary-to-binary conversion using 1 and 0 instead of 0 and 1 in the output

emanresu A

03:30

^

Mayube

good point

emanresu A

so 50

Bubbler

04:06

Meanwhile the tag renaming purged all the history of tag badges :P

emanresu A

CMQ Which tags (where you have ≥5 answers) have the highest / lowest average score?

Bubbler

average score of answers?

emanresu A

Yes, tag score / answer score

Mayube

@emanresuA source-layout avg score: 9

emanresu A

For me it's pop-con & error-message (16/5) and self-referential (8/9)

@Mayube Ok that's impressive

I also have packing (8, but one answer)

Mayube

04:17

@emanresuA 45 / 8

hyper-neutrino

did anyone else just get an [array] tag badge?

Mayube

My lowest is integer 9 / 14 = 0.643

emanresu A

@Mayube Isn't it array 5/9 = .555?

Mayube

yeah just saw

emanresu A

Oh lol I have 99 bronze badges and 99/100 score on restricted-source

hyper-neutrino

04:19

is there an easier way to see this than to just go through all of them and calculate it

emanresu A

¯\_(ツ)_/¯ I just look for ones with low answer count/high score & vice versa

Bubbler

My hello-world is 66/9 = 7.33

followed by path-finding at 72/10

and interpreter at 56/8

emanresu A

I should go work on Vyxal

hyper-neutrino

you can use this to get them all; just copy-paste your profile's source HTML into the input

you might have to copy-paste a couple of pages but you can just append them all together in the input

my lowest is graph theory at 7/5=1.4 and highest is polyglot at 59/7=8.[428571]

emanresu A

I have a few <1 because answers to old questions

Bubbler

04:29

My lowest is alphabet at 8/7

and next lowest is balanced-string at 15/7

So I have only one tag with average <= 2

emanresu A

Almost all my tags are average ≤ 2

@hyper-neutrino Fun fact: Your username still disappears when I have split screen on to do Vyxal stuff

Unrelated String

quite a few tags at plain 0 for me

Bubbler

Without answer count cutoff, my highest is code-bowling at 35/1

then cellular-automata 66/3

Unrelated String

cellular-automata, json, logic, optimization, pi, printable-ascii, probability-theory, text-processing, and typography

lowest non-zero is 0.5 for path-finding

highest is 12.0 for 3d

emanresu A

With cutoff?

Unrelated String

04:33

no idea

whatever hyper's tio link calculates

Bubbler

Cutoff = answer count >= 5

emanresu A

@UnrelatedString factoring?

(41/6)

Bubbler

Only gives tags with answer count >= 5 and sorts by avg

Unrelated String

oh good idea

with the cutoff yeah factoring is my highest at 6.83 and balanced-string is my lowest at 1.375

Mayube

oof, really kills the enthusiasm when you've spent half an hour slowly golfing a python3 answer and somebody posts a python2 answer that's half the size

Unrelated String

04:39

since all of my factoring answers are mildly shit my second highest is subsequence at 6.1 (61/10)

emanresu A

@Mayube ~~Learn Vyxal~~

Razetime

@Mayube they're different languages according to rules, so focus on your solution, don't let go of it

Unrelated String

^

Bubbler

@Mayube I can golf off 4 more bytes from the python 2 answer :P

emanresu A

^^

And we folks in the TTNB can help you golf it

Mayube

04:41

@Razetime yeah but the p2 solution ports to p3 for only 1 extra byte, and is basically the same approach as me

emanresu A

Post the port, and put your original answer in as well

Bubbler

CMC: Find 28-byte Python 2 answer to the latest challenge

DLosc

04:53

@emanresuA the Twenty-Third Nineteenth Byte?

2

lyxal

@DLosc The The Nineteenth Byte

Mayube

@emanresuA meh, that seems kinda pointless. "Here's somebody else's answer, also here's my less-golfed version of that same answer"

DLosc

@lyxal But there's already a "the" before TTNB. So it would be "the The The Nineteenth Byte"

Unrelated String

^

emanresu A

Exactly!

DLosc

05:03

@emanresuA Highest: natural-language 83/6 = 13.83 and popularity-contest 178/18 = 9.88 (here's why). Lowest: rational-numbers 4/9 = 0.44.

hyper-neutrino

@Bubbler does this involve using string output

actually i suppose i shouldn't ask

Bubbler

No but actually string output gives 27

hyper-neutrino

huh

Bubbler

Because Python 2

PyGamer0

05:21

Is minecraft TC?

hyper-neutrino

i think redstone is TC

and command blocks even more so

PyGamer0

@hyper-neutrino so technically someone can make a bf interpreter

hyper-neutrino

in theory i think so

PyGamer0

ok are yall bots? :p

Unrelated String

in theory i think so

PyGamer0

05:26

@UnrelatedString nice to see fellow bots here :P

Niko

redstone is likely to be TC but imagine golfing it

hyper-neutrino

@PyGamer0 ?

PyGamer0

@hyper-neutrino are you a bot?

Unrelated String

@Bubbler done but can't decide if it's worth posting separately or just as a golf suggestion

hyper-neutrino

@PyGamer0 yes

Niko

05:27

you can save two byte(block) by replacing that block with piston

PyGamer0

fastest-code in Minecraft?!

Unrelated String

also really curious how string output brings it down to 27

Niko

my machine is 2 ingame tick faster then yours

hyper-neutrino

@UnrelatedString maybe [1,2] → "12" ?

although that isn't a python 2 thing

my current thought process is to make a large array and then slice-index it based on arity

Unrelated String

yeah what i have uses the exact same l%a+1 from the existing answer so no

PyGamer0

05:31

@Niko i use a PDP20

Bubbler

@UnrelatedString You can post it separately if you think it is different enough

Unrelated String

it probably will be when/if i figure the string thing out

currently it's just this lmao

Bubbler

That's exactly my 28

Unrelated String

nice

Bubbler

I guess Python 2 is a big enough hint for string-based 27 and anything further will totally spoil it

@Niko Honestly there can't be more objective fastest-code than this lol

Razetime

05:36

@PyGamer0 actually sounds quite cool, worth a shot

Niko

what happen if i use cheat engine speed hack on compiler?

it sounds funni

Unrelated String

oh the string thing is a python 2 thing

i still think i have some other lead lmao

(it got edited into the original answer)

yep and porting that string thing is 27

Bubbler

Right

Bubbler

05:53

Seems dingledooper found all these independently

Unrelated String

06:18

yeah

DLosc

First HBL answer!

Bubbler

Nice :D

Meanwhile I decided on the name of my hypothetical golflang: Wintertree, a Winning language based on ternary trees.

Its verbose mode will be named Losetertree :P

DLosc

06:35

@Bubbler Quite apart from the acronym meaning (is that an acronym?), Wintertree is a really nice name.

Unrelated String

^

emanresu A

08:43

Can someone who understands Jelly explain the chaining here to me?

Unrelated String

the top level chain is a dyadic 2 2

emanresu A

Yeah, but I don't understand why it appears to mold then prepend 2

Unrelated String

so 9 %‘ṭ2ɓṁ 1 = 9 %‘ṭ2 (1 ṁ 9)

emanresu A

Oh.

How would I get m to run with the output of the previous & 9?

Unrelated String

top level chains obey the same chaining rules as atoms and all that

uhh

i think %‘ṭ2ṁḷ ought to do it

emanresu A

08:50

Thanks :)

So... ḷ is left argument, so is ṁ called with the left argument (previous) and 9?

Or is ḷ referencing the 9 and ṁ just taking the previous by default?

Please tell me so I can write my explanation properly

¯\_(ツ)_/¯ I'll guess the latter

hyper-neutrino

chaining rules cheat sheet

emanresu A

So this is a 2-1 chain followed by a dyad-nilad pair and... what?

All I need to know is, what's the ḷ referencing?

hyper-neutrino

09:06

there is no dyadic 2,1

it's just a dyad followed by a monad and then a 2,0 and finally a 2,2-chain

emanresu A

Oh that makes more sense.

hyper-neutrino

here's how I would represent it

of course it doesn't fully explain the 2,2 mechanism but idc

i should really make that jelly tutorial video series some time

emanresu A

Does JHT host its own Jelly server or use TIO's?

hyper-neutrino

uses TIO API

Sandbox Posts

09:46

0

A: Sandbox for Proposed Challenges

Christmas Eve Eve Inspired by this xkcd Your challenge is to write a program/function that takes no input and outputs Christmas if the date is December 25, else outputs Christmas + one eve for every day until then. Capitalisation can be anything, but must be consistent - ChRIStmAs eVe eVe is acce...

emanresu A

o/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/ (I found out how Lyxal extends his arms, turns out you stick a bunch of them together)

PyGamer0

10:00

\/\/\/\/\o/\/\/\/\/

Manny Queen

10:45

when you compress them: lol

lyxal

12:30

Copilot may be good for writing code, but it really sucks at math

okay so here's what I gave it

- element: "∆s"
  name: Sine
  description: Get the sine of an angle in radians
  overloads:
    num: math.sin(a)
  vectorise: true
  tests:

and here's what it gave me:

- '[3.141519265] : 0'
- '[0] : 0'
- '[6.283185307] : 0'

because sine of everything is apparently 0

sine(10)? 0

sine(69420)? 0

sine(arcsin(Joe Mama))? You guessed it - 0

at least that's the logic implied by copilot ;p

nvm I'm just being a dumb frickass and forgetting how trig works

gosh dang it

but it still sucks at math tho

because it still gave 0 for other values

so I wasn't completely wrong

rak1507

lol

lyxal

12:52

As I head off to sleep, I strongly suggest you do indeed consider the file

doing so is good for you

o/

PyGamer0

13:37

@MannyQueen lol

@lyxal lol

(my hands are straight)

Mayube

13:50

And of course as lyxal departs, I arrive

mornin' all

I see every answer on the J arities challenge is now a port of dingledooper's python2 answer

Except for maybe Retina

Redwolf Programs

14:27

Morning!

I currently have a 75% in a class as my final 6-week grading period grade

And my teacher still hasn't fixed it >:|

Mayube

What are our default rules for upper limits on integer inputs? IIRC we have a standard loophole against saying "my language only supports 1-bit integers", but what's the default ruling if a challenge doesn't specify a maximum value?

Redwolf Programs

I think incorrect answers due to integer overflow are fine

I think the standard loophole is only for unreasonably low numbers

Mayube

so within reason the upper limit of integers in your language of choice?

Redwolf Programs

I think so. I'll see if there's a meta consensus, but that's what I'd imagine is reasonable.

codegolf.meta.stackexchange.com/questions/2280/… seems to be the relevant question

The two top answers, which aren't mutually incompatible, would indicate that your code should work for all test cases, but is fine to assume the integer size, recursion limit, etc. are arbitrarily high assuming it's not already being abused for some other behavior (e.g., using an int's overflow in one calculation, while pretending another int can never overflow)

Mayube

Gotcha, that makes sense. Thanks for digging that post up!

Sandbox Posts

15:01

0

A: Sandbox for Proposed Challenges

Designing Solar Panels This challenge is adapted from Google's FooBar platform, but has been changed slightly to make it more interesting. I'm designing a new space station for generic super-villain purposes (something something megalaser), but I'm having trouble designing the solar panels. My ge...

Redwolf Programs

@SandboxPosts I think the flavortext is a little too long

Mayube

I find this line from the linked article on Google FooBar quite comical

> Google Foobar is Google’s secret hiring challenge. Google uses this to hire some of the best developers around the globe which they think can be a good match for their organization.

I was invited to Foobar after googling "Python try catch", because I didn't know enough about python to know that it's actually try except

Redwolf Programs

lol

Mayube

I also left the platform after solving only a couple of challenges, because it got boring and I have no interest in working for google.

Fmbalbuena

16:31

@user today we learn ///

user

???

Fmbalbuena

16 hours ago, by DLosc

Welcome to the sixth Learn You A Lang For Great Good! Today, we'll be learning /// (pronounced "slashes"), a surprisingly Turing-complete esolang based on string replacement. You can check out the linked esolangs.org page for some examples, or mess around with it on Try it online!. Feel free to ask questions, post ///-themed CMCs and more for the next 24 hours or so.

user

Okay but why ping me specifically lol

New Posts

0

Q: Swap every two elements in a list

Challenge Let's have a list L of n elements. The task is to swap every two elements in this list. Constrains the list L has at least two elements size of the list L is a multiple of two (i.e. number of elements is even) Example input: [1,2,3,4,5,6] output: [2,1,4,3,5,6] input: [0,1,0,1] out...

code-golf

hyper-neutrino

can't find a dupe, but I feel like there's gotta be one

Fmbalbuena

16:40

@hyper-neutrino i found one but i forgot

Mayube

we have a lot of element swapping challenges, but I don't think we have one that's exactly this

hyper-neutrino

well, I looked up the obvious jelly answer, and couldn't find it

Mayube

This challenge for swapping items and 2 given indices links to several other similar challenges

hyper-neutrino

damn someone doesn't like jelly

or just me i guess

Mayube

seems like somebody's mass downvoting maybe?

The Vyxal answer got a downvote too

hyper-neutrino

16:50

looks like it, they only did my two at first which is pretty arbitrary but yeah it's just typical mass downvoting

oh nope it's back to just being mine

Mayube

I +1'd your python answer at least

hyper-neutrino

ok someone's just fucking with the vote buttons

Mayube

Damn, braingolf loops are broken in TIO, and LMBM can't distinguish between 0 and EOL

Transcript for