The Nineteenth Byte

quartata

12:00 AM

@DJMcMayhem No, especially if the start article hasn't been wikifyied or the end article is an orphan

Downgoat

12:14 AM

@DJMcMayhem I've gotten to Programming in 6 clicks...

8 clicks to code golf if I counted right!

@DJMcMayhem: Canada -> Urbanization -> Modernization Theory -> Technology -> Computer -> Computer Program -> Programming Language -> Perl -> Code Golf

Conor O'Brien

0

Q: Sum or difference of two powers of two

Your challenge, should you choose to accept it, is, given an integer K >= 1, find integers A and B such that at least one of the two conditions following hold: K = 2^A + 2^B K = 2^A - 2^B If there does not exist such A and B, your program may behave in any fashion. Test cases K => A, B 15 =>...

@Fatalize

Downgoat

@DJMcMayhem got 6! Tennis -> Tennis Ball -> Ball -> Golf Ball -> Golf -> Disambiguation -> Code Golf

@DJMcMayhem got 4!! Tennis @ Summer Olympics -> Golf @ Summer Olympics -> Golf -> Dismambiguation -> Code golf

El'endia Starman

That disambiguation page is really useful here. :P

Downgoat

^

New Main Posts

0

Q: Sum or difference of two powers of two

Your challenge, should you choose to accept it, is, given an integer K >= 1, find integers A and B such that at least one of the two conditions following hold: K = 2^A + 2^B K = 2^A - 2^B If there does not exist such A and B, your program may behave in any fashion. Test cases K => A, B 15 =>...

code-golf math

Conor O'Brien

12:24 AM

Milos_Raonic => NBA_All-Star_Celebrity_Game => Golf => Golf_(disambiguation) => Code_golf

El'endia Starman

@Downgoat Isn't that five clicks though? Takes one click to go from Milos Raonic to Tennis @ Summer Olympics.

Downgoat

Hm ok

Conor O'Brien

@NewMainPosts for the 11th time, stop being ninja'd :P

Do I win? :P

Downgoat

@NewMainPosts soltuion: sqrt(n/2) :P

Conor O'Brien

@Downgoat CMC: given n, calculatoe sqrt(n/2)

%:@-:, J, 5 bytes

Downgoat

12:30 AM

ches: n->sqrt(n/2)

pretty sure this is built-in in jelly

Conor O'Brien

I think (sqrt)+(/)&2 works for same byte count

unless I mixed up precedence of + and &

quartata

No but it's two bytes in it

Conor O'Brien

that makes sense

quartata

Halve and sqrt I mean

Conor O'Brien

ye

betseg

12:32 AM

@ConorO'Brien C: sqrt(n/2)

Conor O'Brien

that's a snippet.

betseg

You said calculate, you didn't say do something with it

Conor O'Brien

You know well what I meant.

betseg

@ConorO'Brien C: f(n){return sqrt(n/2);}

quartata

Will that work for float n?

Qwerp-Derp

12:40 AM

Argh I need help

betseg

@quartata Wow i just realized it won't work for n=m² m∈ℝ m∉ℤ

Add float at the beginning

Qwerp-Derp

How do I make a regex that matches something but doesn't match something else

I want to match all >'s but not match >'s if they're in ->

betseg

@Qwerp-Derp with magic

Qwerp-Derp

:(

ΛεγίωνΜάμμαλϠΨΠʹ

@ConorO'Brien Mathematica, 10 bytes: Sqrt[#/2]&

Conor O'Brien

12:42 AM

@Qwerp-Derp (?!-)>

wait

no

betseg

@ConorO'Brien ?!

Conor O'Brien

which flavour?

Qwerp-Derp

ummm

I have no idea

Conor O'Brien

language?

Qwerp-Derp

JS?

Conor O'Brien

12:44 AM

ah

Qwerp-Derp

I'm making it for Atom

betseg

Ahhh

Conor O'Brien

replace all -> with →, replace all > with whatever, then replace all → with -> ? :P

betseg

WOAH dude

Qwerp-Derp

Wot

@ConorO'Brien How to do?

Conor O'Brien

12:48 AM

actually

(?<!-)>

that works

Qwerp-Derp

Not supported in JS...

Conor O'Brien

you said you were using atom?

Qwerp-Derp

Yeah

Which probs uses JS

You can inspect element so...

Conor O'Brien

wait

atom the text editor, or something else?

betseg

Text editor built on node.js

Conor O'Brien

12:51 AM

it works on my copy of atom

Qwerp-Derp

@ConorO'Brien Ah, it works

Thanks!

Conor O'Brien

no problem! :D

Qwerp-Derp

1:08 AM

I still need help with regex

var a=input
out a>>
circ xor(a,b)->(!(a&b))&(a|b)
cond xor((a><),(a>>))->var b=a<+1+a>>/var b=a<+0+a>>
out b
# This is a comment

This is my code I'm trying to match stuff with

Conor O'Brien

what are you trying to match?

Qwerp-Derp

This is my regex: (?<=var )(.*)(?=\=)

I'm trying to match all the var names

so a and b in this case

rubular.com/r/J5sGyEAMuL

Conor O'Brien

(?<=var )(.*?)(?==)

try that

Qwerp-Derp

Ah kule it works!

Conor O'Brien

:D

Qwerp-Derp

1:21 AM

Ummm one more thing

I'm trying to match the args for circuits only

betseg

# var x = y

Qwerp-Derp

And this is what I have: (?<=\()(.*?)(?=\)-)

betseg

is that a match

Qwerp-Derp

For regex

I want it to not match the commas, and I also want it to only match if circ is the first 5 chars

Conor O'Brien

test cases plox

args look quite complex

Qwerp-Derp

1:23 AM

The same testcase as above

rubular.com/r/ULRToU7q0X

Destructible Watermelon

now to fix bug in interpreter

Conor O'Brien

@Qwerp-Derp like, (a,b) in circ xor(a,b)-> and (a><),(a>>) in cond xor((a><),(a>>))->

?

Qwerp-Derp

Just the first one

Conor O'Brien

oh

(?<=circ .+?\()(.+?)(?=\)->)

try that

Qwerp-Derp

"Invalid pattern in look-behind"

Conor O'Brien

1:28 AM

really?

Qwerp-Derp

Ya

In rubular

Conor O'Brien

works on atom

Qwerp-Derp

Rly?

Conor O'Brien

ye. I'm testing these on atom

Qwerp-Derp

Oh

No commas?

Conor O'Brien

1:29 AM

you don't want to match commas?

Qwerp-Derp

Ya

Is it possible?

Conor O'Brien

idk

might be

might not be :/

Qwerp-Derp

:/

quartata

@Qwerp-Derp Not entirely sure what you are trying to do but I think you need a proper parser

Qwerp-Derp

But this is atom...

I'm making atom syntaxy stuff

quartata

1:41 AM

What is this for? Syntax highlighting

Oh.

Qwerp-Derp

Ya

Destructible Watermelon

@Dennis There was a bug with the interpreter which I fixed. Please pull Woefully when you can.

Conor O'Brien

And, as he gazed upon the screen / a sense of sorrow overcame him / and he pulled woefully

12

Destructible Watermelon

;_; y u do dis conor o'brien

ETHproductions

1:58 AM

@ConorO'Brien I made a tiny calculator in 340 bytes of ES5 :) ethproductions.github.io/golf/calculator

4

Conor O'Brien

@ETHproductions whoa

Downgoat

@Qwerp-Derp (^|[^-])>

Qwerp-Derp

Whoa what is that for

ETHproductions

@ConorO'Brien Source code: ethproductions.github.io/golf/calculator/index.js

Conor O'Brien

@Downgoat too late, and that matches only at beginning of string

New Sandboxed Posts

2:00 AM

0

A: Sandbox for Proposed Challenges

code-golfstring Make me a bad fade animation Your task is to, as the title discreetly suggests, make a bad fade animation of one word turning into a second word for yours truly. What exactly is this "fade animation" you may ask? To make a spectacular(ly bad) fade animation, you start with you...

Downgoat

@ConorO'Brien no?

Conor O'Brien

@Downgoat doesn't it, with the anchors?

Downgoat

There's an OR

Conor O'Brien

ah

well that contains the preceding character in the match

Downgoat

Well JS doesn't have lookbehinds do just use a capturing group

lookbehinds are porpoised for ES8 however

Dennis

2:04 AM

@DestructibleWatermelon I pulled woefully.

Conor O'Brien

@Downgoat he's not using JS :P

and what happened to ES7? :P

ETHproductions

@Downgoat that makes me want to make a pun out of "reporpoised"

Downgoat

@ConorO'Brien ;_;

ETHproductions

ES7/ES8 compatibility table: kangax.github.io/compat-table/es2016plus

Firefox is way behind on the exponentiation operator :-(

Destructible Watermelon

@Dennis thanks!

Downgoat

2:08 AM

@ETHproductions this is incorrect :(

ETHproductions

@Downgoat What about it?

Conor O'Brien

@ETHproductions rip :|

Destructible Watermelon

2:24 AM

I should make a hello world answer for woefully...

it will be hard, but I think I'm up to it... not sure about getting the golfiest solution, though

Destructible Watermelon

2:52 AM

that moment when you make a bot that tests a bomb (by having it run on itself), and it beats someone else's bot in a match...

Qwerp-Derp

Yes!

I uploaded my package to atom

Maltysen

what's it do?

something cool?

Qwerp-Derp

It's support for Logicode

Maltysen

rofl, just transposed something and then transposed it right back just to use the padding function on the first transpose: codegolf.stackexchange.com/a/92855/31343. still shorter than anything else... D:

Conor O'Brien

night y'all

Qwerp-Derp

3:00 AM

Night

Destructible Watermelon

a bomb!

start:spl 0,{1
slt #200,#201
mov 2,start
jmp start
dat #0,#0

slows down the enemy before killing it

actually I should probably beef up those numbers

El'endia Starman

@Maltysen Pyth: where golfing off 5 bytes is "a lot". :P

Destructible Watermelon

at some point my scanner became a bomber wat

argh

now to attach this bomb to a program

Maltysen

@El'endiaStarman well its 18%... :P

Destructible Watermelon

3:19 AM

why does my bomb not work ;_;

IT WORKS NOW WOO

wait no it doesn't...

oh

quartata

Poll: Would you write bots for a KoTH whose controller (and therefore its bots) are written in a functional language (i.e not Java JS or Python)

Erlang is such the right tool for this it hurts...

And "universal" KoTHs with IPC never work out right

Nathan Merrill

...did chrome change fonts on you guys?

everything looks so...different

Geobits

@quartata I'm not great with functional languages in general, but I'm good at tinkering with boilerplates code. Given a working template to handle the basics, I might if the challenge itself is interesting.

Nathan Merrill

@quartata if I actually participated in KoTHs, possibly. I agree with geobits. I'd also likely want a list of "helper" functions I could call

(common routines that submissions would likely need to perform, such as path finding)

oO

I just realized that I'm running chrome 32 bit

Destructible Watermelon

3:50 AM

is chat dying...

feersum

Dying = becoming less spammy?

@DestructibleWatermelon So how did your language even work?

With those backwards coordinates and all.

Destructible Watermelon

uh, they weren't entirely backwards

line wrapping was just messed up

feersum

Did it produce the right answer by accident somehow?

Destructible Watermelon

good thing noone had tried making any programs with line wrapping

@feersum nope

def move():                                                              #
    global cp                                                            #
    cp[0]+=stack_A.pop()                                                 #
    while cp[0]>=len(program[cp[1]]):                                    #
        cp[0]=cp[0]-len(program[cp[1]])                                  #
        cp[1]+=1                                                         #
        if cp[1]>=len(program):cp[1]-len(program)

you'll notice that this is fixed up now, but before

it had the while part and if part messed up

but the cp[0]+=stack_A.pop() was right the entire time, so it was fine

It didn't get found immediately because I hadn't used that feature yet. Thanks for letting me know though

did you see the shorter solution I made?

feersum

Only in the sense of seeing a wall of pipes.

Destructible Watermelon

3:59 AM

heh

I should do the 10*10 asterisks challenge

Destructible Watermelon

4:19 AM

ok, here is ten asterisks

| ||||||||| | |
|| |||||||| |
||| ||||||| |
|||| |||||| |
||||| ||||| |
|||||| |||| |
||||||| || |
|||||| || |
||||| || |
|||| ||| |
||| |||| |
|||| ||| |
||||| || |
|||||| || |
||||||| || |
|||||||| || |
||||||||| || |
|||||||||| || |
||||||||| || |
|||||||| || |
||||||| || |
|||||| |||| |
||||| |||||| |
|||| |
||| |
|| |
| |
| |
| |
| |
| |
| |
|| |
|| |
|| |
|| |
|| |
|| |
| |
| |
| |
| |
| |
|| |
|| |
|| |
|| |

now I just need to loop that ten times

this is gonna be a challenge...

New Meta Posts

0

Q: "Too broad" close reason applied to "restricted source" puzzle

After posting a rather specific, narrow, restrictive, one-language (well... one group of languages: POSIX-ey shells) puzzle, thats been "closed as too broad". Commentators then offered explanations inconsistent with the notion of "too broad": "challenges that are restricted to one language ar...

discussion close-reasons

Destructible Watermelon

4:41 AM

starting to think I'm going about this challenge the wrong way...

ah screw it

Eᴀsᴛᴇʀʟʏ Iʀᴋ

5:11 AM

@quartata @Downgoat @ConorO'Brien youtube.com/watch?v=0PtLBAkk6Bc

Qwerp-Derp

Argh I need to make a successor function for Logicode

ASCII-only

@Qwerp-Derp Successor?

Qwerp-Derp

successor of a is a+1

Destructible Watermelon

can functions take arbitrary sized input?

Qwerp-Derp

Yes

Hang on

Uhhh... yeah, they can

Destructible Watermelon

5:17 AM

if they can't, it's impossible

ok, good

well, what you want to do, in pseudo code is

if last of input==0, (output input except last of input, and 1, ) else output (this function with input all of input but the last part of input) and zero

that's kind of cinfusing, let me try again

funcname: if input[-1]==0, output input[:-1]+[1], else output funcname(input[:-1])+[0]

Qwerp-Derp

Ummm... we only have heads and tails

ASCII-only

except last of input? that's not going to be easy

Qwerp-Derp

Introduce more stuff?

So we have both classics

ASCII-only

@Qwerp-Derp No

we can't get the middle that way

Destructible Watermelon

@Qwerp-Derp just make it take input backwards!

Qwerp-Derp

5:23 AM

I'm not finished

Oooh yeah

Destructible Watermelon

oh wait, that won't work, will it

Qwerp-Derp

We also have ~ to reverse string

ASCII-only

We can either add behead and betail

@Qwerp-Derp Wait what

Qwerp-Derp

Actually no scratch that

Going back from the start

We have < and > staying as-is

But we have ~ to reverse a binary string

So if a is 111011, ~a is 110111

ASCII-only

Also btw what is supposed to happen for 001&1

Qwerp-Derp

5:25 AM

Ummm

ASCII-only

@Qwerp-Derp That would work too

Qwerp-Derp

@ASCII-only Which would work?

The ~ thingo?

ASCII-only

@Qwerp-Derp Yeah

Qwerp-Derp

TBH we can get rid of ?

Should we?

ASCII-only

@Qwerp-Derp ?

Random is useful (sometimes, rarely), but it's not needed for it to be TC

If you bother making a large enough library you can use ? in something like random(min, max)

Qwerp-Derp

5:28 AM

True

Will you implement the ~?

ASCII-only

@Qwerp-Derp Okay brb

Qwerp-Derp

Kule thx

ASCII-only

@Qwerp-Derp BTW I may need to fix multibit operators

Qwerp-Derp

Ah

ASCII-only

Is 001&10 001&100 or 001&010 or 01&10 or 00&10

Qwerp-Derp

5:31 AM

???

Uhhh

It should probably be either 001&010 or 01&10

I'm leaning towards the former

ASCII-only

Okay

Maltysen

@Fatalize plz help, I was trying sum or diff with prolog, not working

Fatalize

@Maltysen example?

Maltysen

2^A + 2^B is 264; 2^A - 2^B is 264.

gives false

Fatalize

you can't have variables on the left side of is

nevermind

you can't have variables on the right side of is

but

you can't have an arithmetic expression on the left side either and expect it to work

is is for low lvel arithmetic really

Maltysen

5:37 AM

oh, it doesn't solve?

Fatalize

no

Maltysen

how do I solve stuff then?

Fatalize

You have to use library(clpfd) for that

and then 2^A + 2^B #= 264; 2^A - 2^B #= 264.

will add constraints to A and B

you can obtain values for them with label([A,B])

Maltysen

so there's no way with vanilla prolog?

Fatalize

well there's alway brute force :p

clpfd is pretty much vanilla at this point, though in SWI-Prolog you still have to load it for some reason

Maltysen

5:40 AM

oh, i see

ASCII-only

@Qwerp-Derp Should be done now

Maltysen

@Fatalize I guess brachylog deals with stuff fine?

Fatalize

@Maltysen Another problem that you will have though after that is that you can't label variables that have an infinite domain in CLPFD

1

A: Sum or difference of two powers of two

Brachylog, 23 bytes ,A:B#+.:2rz:^a{+|-}?,.= Try it online! This is much faster than required, e.g. this is still under 10 seconds on TIO. Explanation This is basically a direct transcription of the formula with no optimization: ,A:B The list [A, B] #+ Both A and B are greater tha...

Maltysen

so how does brachylog deal with infinite domains?

Fatalize

@Maltysen It has a custom labeling predicate which works the same as label on finite ones and enumerates properly on infinite ones

This was mostly implemented by the person who implemented CLPFD for SWI-Prolog

The reason this isn't implemented by default in CLPFD is because you have no guarantees of termination when the domain is infinite which Prolog people don't like

5 ?- [library(clpfd)].
true.

6 ?- [A,B] ins 0..264, (2^A + 2^B #= 264 ; 2^A - 2^B #= 264), label([A,B]).
A = 3,
B = 8 ;
A = 8,
B = 3 ;
false.

This works

DJMcMayhem

6:01 AM

facepalm This is probably the worst ad I've ever seen.

ASCII-only

@DJMcMayhem Where

DJMcMayhem

The ad was "Learn Ethical Hacking", and it has a picture of a dude sitting at a laptop with a guy fawkes mask.

Somebody should be fired for that

Anonymous

@DJMcMayhem Wait you mean that's not how you code golf?

ASCII-only

@Mego Of course not, DJ has a red striped hat, not a mask

DJMcMayhem

If this company honestly thinks that Anonymous is the emblem of ethical hacking, WTF would I expect them to be smart enough to teach me anything worth knowing?

Qwerp-Derp

6:13 AM

Also: multi-digit inputs?

That shouldn't be too hard

ASCII-only

@Qwerp-Derp Syntax?

Qwerp-Derp

Just input

It should support multiple digits

ASCII-only

Also is it not possible with existing syntax

Qwerp-Derp

Not really

ASCII-only

why not

circ in(n)->input+in(decrement(n))

Plus input is designed partly for TIO compatibility, partly for convenience, you can input at start of program

Unless you want to split by newlines

Or some other character

Qwerp-Derp

6:17 AM

Yeah, no, it doesn't work

With something like this

var digit=0+input
out digit

And with input as 111

The output is 01

ASCII-only

Yeah

That's what is supposed to happen

You can either do 0+input+input or tell me a separator char for input so I can fix it

Qwerp-Derp

Why is that supposed to happen?

Isn't the output meant to be 0111?

ASCII-only

No

Qwerp-Derp

Wot

ASCII-only

input takes input 1 bit at a time

We can't make it do anything else since we supply input at the start

But I can split input into multiple bit chunks, like 0;111;10 or something, but IDK if you mind separator chars and IDK what separator char you want

Qwerp-Derp

6:21 AM

Why does input take one bit at a time though

I thought it was meant to take multiple bits

ASCII-only

Because IIRC Logicode was originally designed to process 1 bit at a time

Qwerp-Derp

Ah

Uhhh... can we change it to multi-bit input now?

Or no

ASCII-only

I'm fine but you haven't specified how that would work ;_;

If you want it to run on TIO you need to be able to specify multiple inputs at once before interpretation starts

Do I just split input by spaces

Qwerp-Derp

Input should probs be split by \n's

ASCII-only

okay, brb fixing

Anonymous

6:25 AM

Newlines would make the most sense

Qwerp-Derp

We should probably talk in the LC room

Hang on... how many langs have we all made?

@Mego How many langs have we actually made?

ASCII-only

Who

Qwerp-Derp

Like everyone here

ASCII-only

50-100 I'd guess

@Qwerp-Derp btw done hopefully

Qwerp-Derp

Yus

I should probs get crackalackin' on the wiki for Logicode.

trichoplax

6:32 AM

56

Q: What programming languages have been created by PPCG users?

Language-creation has become a popular activity on PPCG. A decent portion of answers, especially code-golf answers, are written in languages invented by the community. These are also languages that might be unfamiliar to this site's wider viewing audience. What languages (esoteric, golfing, or...

discussion

66 answers...

Qwerp-Derp

Ummm... I'm going to post a challenge

Make a program that, using +,-,*,/,sqrt,ceil and floor, output the numbers 1 to 100 using as few 1's as possible

ASCII-only

@Qwerp-Derp So just replace every 1 with 2/2?

Qwerp-Derp

What do you mean

The only numeral you're allowed is 1

ASCII-only

Oh

vtc-as-unclear

Qwerp-Derp

Why?

ASCII-only

6:36 AM

> Using as few 1's

You don't disallow anything else

Qwerp-Derp

OK...

ASCII-only

@Qwerp-Derp btw you may want to edit this sometime

Qwerp-Derp

Your task is to make a program that, using +, -, *, /, sqrt, ceil, ! and floor, output 100 equations that equal every number from 1 to 100 using as few 1's as possible in the output. The only numeral you are allowed in the output is 1.

@ASCII-only done.

ASCII-only

@Qwerp-Derp BTW 1. I don't think it's that similar to logisim anymore, plus there are no strings, just lists of bits

Sherlock9

@ETHproductions That is exceptionally cool :D

New Sandboxed Posts

6:52 AM

0

A: Sandbox for Proposed Challenges

There is only 1 Your task is, given a number n, to generate an equation that equals to the number n. The catch is: you're only allowed the number 1 in the output. The operators at your disposal are: +, -, * and / sqrt (as s) ceil and floor (as c and f respectively) ! (factorial) You must a...

Sherlock9

@mbomb007 Did you see my suggestions? chat.stackexchange.com/transcript/240?m=32192714#32192714

Dmitry Kudriavtsev

test

I'm back

DJMcMayhem

The longest meta answer I have ever written:

0

A: "Too broad" close reason applied to "restricted source" puzzle

First off, I'm sorry that your first contribution to our site was so poorly received. I hope this doesn't discourage from posting more challenges. Writing good challenges is really hard, especially for users who are unfamiliar with our rules and culture. But, they are extremely important for the ...

Sherlock9

@mbomb007 Hm, this works in trinket.io but not on IDLE. Strange

Transcript for