The Nineteenth Byte

Conor O'Brien

3:00 AM

like Mathematica: (* asdf *)

Qwerp-Derp

@ConorO'Brien What does isolate do?

I think we should move all of this into a separate room

Conor O'Brien

yes

Stacked

For discussing the Stacked language (github.com/ConorOBrien-Fo...

Leaky Nun

CMC: n → 1+2+3+...+n

ATaco

@LeakyNun RProgN-2, 2 bytes. R+

Leaky Nun

@ATaco docs?

ATaco

3:10 AM

RProgN2 is a little light on docs.

Leaky Nun

you mean a little non-existent

Correct, yes.

stack-based?

Yeah.

alright

3:12 AM

Simple enough, RANGE then ADD, Range implicitly uses a 0 as it's first argument, which, given say... 3, would produce a stack with the content (0, 1, 2, 3), +, when fed a stack, sums it.

Downgoat

question: in a language with automatic semicolon insertion, should:

3
/3

be: 3/3 or 3;/3 (i.e. syntax error)

ATaco

Jelly, 2 bytes. RS

Conor O'Brien

The latter is more sane I think

ATaco

Idealy it should be 3/3.

Conor O'Brien

"read until you have a valid expression", not "read until you can't"

ATaco

3:14 AM

I think it's more / is the first part of the next line, so it probably wants to concatenate the line.

For example:

  "Hello, "
+ "World"

Is definitely expected to produce "Hello, World"

Or rarely, "Hello, \nWorld"

Mistah Figgins

@LeakyNun Wise, 21 bytes. ::^?[:?[?~-!-~]|!-~]|

Leaky Nun

@ATaco only in languages where "Hello, " isn't a complete statement on its own

ATaco

You can assume it's wrapped in a print or x = if you want, but + "World"definitely isn't a statement in most languages.

Leaky Nun

@MistahFiggins how do you add two numbers?

Mistah Figgins

[?~-!-~]|

i.e. "until the first number is 0, add 1 to the second number and subtract one to the first"

then get rid of the trailing 0

Leaky Nun

3:19 AM

how do you add 1?

oh, never mind

Mistah Figgins

I'm not sure if multiplication is shorter, because we don't really have a formula for that yet :/

Leaky Nun

then create one

try squaring a number

Mistah Figgins

wait, not, WheatWizard got dennis to fix the bug

Leaky Nun

what bug?

Mistah Figgins

but the stack clean multiplication that I came up with is 31 bytes alone

Leaky Nun

3:21 AM

never mind then

Mistah Figgins

it used to be that a loop (with [code]) wouldn't check if something was 0 before it entered

it was a do while, not a while loop

However, I think stack dirty multiplication might be faster

ATaco

ReRegex, 81 Bytes

Leaky Nun

@ATaco can't you use 1 as unary?

ATaco

I can, I just like Underscores.

Leaky Nun

they are confusing

here

Downgoat

3:26 AM

has anyone here taken the AP CS A test?

ATaco

If you prefer decimal IO

Mistah Figgins

@Downgoat I think so

pretty sure

Downgoat

@MistahFiggins do we need to know GridWorld?

Mistah Figgins

what do you mean, "know"?

ATaco

@LeakyNun Once. But it will loop endlessly.

Leaky Nun

3:27 AM

@ATaco how to not loop?

ATaco

Well, I should clarify.

Mistah Figgins

I'm not sure, but I don't think so

ATaco

It keeps looping as long as the match/replace changes something.

Leaky Nun

how not to loop?

ATaco

Ensure nothing is changed at the final iteration.

Downgoat

3:28 AM

@MistahFiggins is it on the test though?

ATaco

Due to the nature of it, the entire math library is horrifying in how it works.

For reference

Mistah Figgins

@Downgoat I doubt it. My memory is meh, but I don't recall so, no

Thomas Ward

@Mendeleev I haven't been around, sorry. What's up?

Leaky Nun

execution order? @ATaco

ATaco

Arbitrary.

Based entirely on Java's Hashing.

Leaky Nun

3:29 AM

how to run two replaces?

Downgoat

@Phoenix you took AP CS A quiz recently right? I have questions, I have MC final tomorrow >_>

ATaco

a/b/
c/d/
ac

Will turn into bd

If they've got conflicting matches, hope and pray.

Leaky Nun

loop order?

ATaco

As I said, execution of replacements is completely arbitrary.

But they will only ever happen once in a cycle.

Leaky Nun

for example?

11112222?

ATaco

3:32 AM

Perhaps I'm explaining this language wrong.

The entire language is just blocks of match/replace/match/replace.../starting text

Then, it takes each pair of match/replace, and executes it on starting text in no particular order.

Downgoat

@HelkaHomba wait what where did you get this— err I mean idk what you're talking about

ATaco

And it keeps doing this until it stops changing.

Mistah Figgins

yeah, gridworld wasn't on there at all @Downgoat

Leaky Nun

@ATaco that's quite pointless, to be honest

ATaco

It's turing complete.

Leaky Nun

3:34 AM

well I can't do two replaces one after the other

Downgoat

:/ I mean I Know Java but I keep getting really bad like 80% on MC >_<

ATaco

You can, if you get a bit smart about it.

Downgoat

though I can get a 70% and still get a 5

ATaco

Here's a factorial function

Mistah Figgins

@Downgoat There's these pages if you don't know about them yet...

ATaco

3:35 AM

And almost brainfuck

Mistah Figgins

> almost

?

ATaco

< , > are replaced with l i r, for convenient interface with the libraries.

Leaky Nun

@ATaco this is the best I got

ATaco

It's not a very good language.

But it does entirely serve it's purpose.

The issue with the way you've tried to solve it, is you're IO was going to be identical, so it would have to loop forever. Changing /#input to something like /<#input> protects against that.

Leaky Nun

I wonder why (?<=^b(1+)) gives a "Look-behind group does not have an obvious maximum length near index 9" error, but not (?<=^(1+))

Mendeleev

3:47 AM

@ThomasWard It seems like you have a lot of experience with nginx, I was wondering if you could help me with something

Downgoat

@Mendeleev wait you are in US right, have you taken AP CS A test

@Downgoat No

:(

you'll do fine

Should I?

3:48 AM

@Mendeleev live in the US, or take the test?

Mendeleev

@MistahFiggins I already do the first one,

Mistah Figgins

maybe I could have looked at your profile, or remebered that you are also in WA

Mendeleev

@MistahFiggins you are in WA as well?

ATaco

All my languages are in their roots experiments, so they usually aren't incredibly useful, but they are interesting.

Mendeleev

@ATaco Can you make the color apply behind the username, not the message?

ATaco

3:57 AM

Of course, hold on.

Downgoat

Question: can a static method access a private field in Java?

Leaky Nun

@Downgoat try it and you'll know

Downgoat

:( ok

ATaco

@Downgoat It can access a static private field, but not an instanced private field.

TMK

@Mendeleev Done.

Downgoat

Huh, apparently it can't

Mendeleev

4:01 AM

@ATaco Cool

Downgoat

class Main {
  private double a = 5;

  static void doThing(A a) {
    System.out.println(a.a);
  }

  public static void main(String[] args) {
    doThing(new A());
  }

}

I'm doing this right, right?

Mendeleev

@ATaco Still shows behind messages.

ATaco

@Mendeleev Have ya updated the userscript and refreshed..?

Mendeleev

Yes

ATaco

What version number..?

feersum

4:02 AM

@Downgoat The Main class can't access the fields of the A class, what a surprise.

Downgoat

@feersum wait what, there is no A class

... Java is throwing wrong error

@ATaco it can apparently :P

Mendeleev

@ATaco 1.0

Ruby

hello!

ATaco

@Mendeleev Should be 1.1

Ensure it's updated

Mendeleev

ok better now

ATaco

4:04 AM

Red Team has all the cool players, we've got Downgoat, Dennis, Me.

Mendeleev

But can you make it only go behind the username? Not behind the whole message?

@Ruby Do you like to take trains?

Ruby

what ?????

Mendeleev

because then it would be Ruby on Rails

ATaco

Text should be fine, because of the dimensions of the signature box, It looks terrible if I just put it under the username.

Ruby

what do u mean ?

@Mendeleev

Mendeleev

4:05 AM

it's a pun rubyonrails.org

Sherlock9

IRB (Interactive Ruby) question: How do you call the last line. In Python, it's something like _(1)

Ruby

hahahahahah

Downgoat

@ATaco team red?

oh hey I'm red :D

ATaco

Team Blue contains all the people who use Stacked.

(That I know of)

And Leaky Nun, too.

Sherlock9

Anyone?

Leaky Nun

4:08 AM

CMC: "abc" → "abcba"

@ATaco good luck using reregex to do that

ATaco

On it boss.

Can we assume only letters will be used?

Mendeleev

@LeakyNun Ruby: ->s{s+s.reverse[1..-1]}

Sherlock9

Python 3, 20 bytes: lambda x:x+x[-2::-1]

Leaky Nun

anyone java the CMC?

Downgoat

@LeakyNun s->s+s.tail(-1).rev

Sherlock9

4:11 AM

Oh that reminds me. Have you seen this challenge @LeakyNun? codegolf.stackexchange.com/questions/95100/…

Leaky Nun

@Downgoat cheddar :D

Downgoat

:D

Leaky Nun

@Sherlock9 I gave that up

Sherlock9

Ah dang

Leaky Nun

@Downgoat what about s->s+s.rev.tail(1)?

wait, yours doesn't work

feersum

4:13 AM

Um... how about s->"abcba"

Leaky Nun

@feersum not funny

ATaco

@LeakyNun Done.

Downgoat

@LeakyNun ah, yes that work :D

@LeakyNun what version are you using?

Leaky Nun

@ATaco how on earth did you use a and b and c in the code?

user165474

@LeakyNun 1+¡0

Leaky Nun

4:15 AM

@HyperNeutrino ?

ATaco

a b and c were always followed by a <

Leaky Nun

@Downgoat v1.0.0-beta.45

user165474

@LeakyNun Sorry, missing context, that's the Fibonacci code from earlier today

user165474

I think I finally figured it out

Leaky Nun

@HyperNeutrino well have you tested it?

Downgoat

4:16 AM

@LeakyNun oh ok that's old, try npm i -g cheddar-lang@1.0.0 IIRC

user165474

I've tested it; it appears to return [1, 1, 2, 3, 5, 8, 13, ...][argument] so it differs from the built-in by an index of 1, but it's valid in certain implementations of the sequence itself.

Downgoat

I think newer versions I replaced .rev with a longer version

Leaky Nun

@HyperNeutrino can you fix it?

user165474

I'll try.

user165474

Shouldn't change the bytecount though, right?

Leaky Nun

4:17 AM

@Downgoat still not working

@HyperNeutrino right

user165474

Alright. I'll try to fix that.

Downgoat

@LeakyNun is there an error?

Leaky Nun

@Downgoat no, it gives me "abccba"

Downgoat

ohhh

ok, I guess .slice(1) at the end

wow, I am surprised I haven't fixed

Leaky Nun

you don't need the tail then

@Downgoat I'm surprised you haven't tested the code before giving me

@Downgoat what is wrong about tail?

Mendeleev

4:21 AM

yayy openscad

Leaky Nun

@Downgoat PR

ATaco

@LeakyNun Brainfuck, 12 bytes. ,[.>,]<<[.<]

Leaky Nun

@ATaco you picked the right language

ATaco

Not the first time I've done something like this in Brainfuck.

One of my favourite answers is very similar to this

Albeit with a fair bit more header.

user165474

@LeakyNun I can't seem to figure out how to fix it to match the built-in. I could try to reduce the number of iterations, but that adds extra bytes, so I'm pretty confident that that's not the way to go about it

Leaky Nun

4:28 AM

@HyperNeutrino it's easier than you think

user165474

1+¡0 My original thought was to change the numbers but that doesn't seem to work.

Leaky Nun

@HyperNeutrino well it works for me

user165474

Hm. Okay.

Leaky Nun

@Downgoat are you here?

user165474

I think the 0 is correct; not sure.

user165474

4:37 AM

Anyway, gonna go to bed now; 1237. GN!

Thomas Ward

4:49 AM

@Mendeleev Probably, if I weren't dead tired.

Mendeleev

@ThomasWard Oh, sorry abuot that

I'll ask you tomorrow

Leaky Nun

5:05 AM

CMC: nCr

Destructible Lemon

yo

ATaco

@LeakyNun Math.JS, 20 bytes C(n,r)=n!/(r!(n-r)!)

Leaky Nun

@ATaco you can use n=>r=> I guess

ATaco

Not in Math.JS

Math.JS is, surprisingly, not a library.

Which is why n! is valid syntax.

Leaky Nun

@ATaco C(n,r)=n!/r!/(n-r)! is shorter

ATaco

5:17 AM

Oh, I wouldn't have thought to do that.

Found the builtin!

combinations

Link

Leaky Nun

that's nice

ATaco

Not sure why it's not just nCr, but I'm not the developer.

Leaky Nun

@DJMcMayhem hi

ASCII-only

@ATaco because it's not a golfing library

ATaco

It's not a library.

Well, it is, but the eval is the interesting part of it.

Riker

5:25 AM

What's Pavel called now?

ASCII-only

Phoenix

ATaco

eval uses it's own Compiler/Parser/Interpreter

It's intention is just to process math.

Riker

Oh really?

Didn't realize

@Phoenix reddit.com/r/WritingPrompts/comments/67hbjg/…

ASCII-only

@ATaco Well yeah, Math.JS doesn't interpret JS, it inteprets a slightly modified language, and the language is not meant to be a golflang

ATaco

It's not intended to be TC but here we are.

ASCII-only

5:29 AM

@ATaco Wait what I thought it supported all JS syntax constructs

ATaco

Nope, very few of them. It just follows a similar guideline.

' is the transpose function, not a string deliminator, ` does god knows what...

x=>x doesn't work, etc. etc.

However, {} is objects, and [] is matrices, which serves the same purpose as arrays.

New Meta Posts

0

Q: Currying with Programs

We allow functions to return functions: f = lambda x: lambda y: x + y f is called like: f(3)(4) This is called currying. However, we also allow them to be called in other ways: lambda: lambda x, y: x + y # f()(3,4) lambda x, y: lambda: x + y # f(3,4)() lambda x: lambda: lam...

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Find the Calling Pattern number function cops-and-robbers I've written a program to add two numbers a and b: f=lambda:lambda a:lambda b:lambda:a+b This has a very specific calling pattern. It must be called as: f()(a)(b)() In this case, it was fairly clear, but sometimes the calling patte...

Leaky Nun

5:53 AM

@Phoenix have you seen my answer to your challenge?

Phoenix

I have

It's pretty good, (I think)

I don't read Pyth

Leaky Nun

thanks

ATaco

6:04 AM

Ey, Pavel's on the Red team!

Phoenix

?

What red team?

ATaco

I made a Userscript that arbitrarily assigns chat users to teams.

github.com/TehFlaminTaco/TacosUserscripts/blob/master/…

Phoenix

Is it deterministic? Will We see everyone on the same team?

ATaco

Yeah, it's based off the UserID.

Literally a userID%2 ? RED : BLUE

Phoenix

I see.

@ATaco feature-request for the caret-reply script: Just carets with noothing else on the line star the message pointed to instead of sending anything.

ATaco

6:10 AM

That... Is rather smart.

Anonymous

I don't know if that's such a good idea

ATaco

Would you prefer messages of that vein to have ^ messages, or to be starred?

Phoenix

Says someone on team blue :P

Anonymous

@ATaco Preferably neither, but if I had to choose, I'd prefer carets

Phoenix

It's just a convenience tool.

I highly doubt it will cause a problem.

Anonymous

6:13 AM

There are times when one would want to caret something without starring it

Phoenix

<space>^^

ATaco

Rarely is just carreting something beneficial to the chat.

Challenger5

^ I disagree. This use case might be fine.

Anonymous

@ATaco I agree, but people use carets to mean "I agree with this". That doesn't necessarily translate into "I think this is important/interesting/funny enough to be starred".

Phoenix

You can put a space in front

No one is making you use the script.

ATaco

6:15 AM

Alternatively, I could make ^^* or similar star the target.

Phoenix

I like that.

ATaco

Of course, still supporting things like ^?cool message?*

Challenger5

On this answer, Destructible Lemon commented to the effect that a 1-byte solution would necessarily use a built-in. Is that true?

Anonymous

IMO the syntax should be <carets>*, and the script should suppress actually sending any message to the chat

ATaco

That was the idea

Anonymous

6:21 AM

@Challenger5 Any single command that accepts a list and outputs a sorted list would necessarily be a sorting builtin.

ATaco

However, ^*10 and ^?cool message? are valid selectors for the Caret script

Dennis

@Challenger5 Practically, yes. In order to solve the problem without using built-ins, you must carry out at least two operations. Unless both of them fit into a single byte (a certain behavior before the code is executed or sub-byte tokens), a single byte will be a built-in.

Phoenix

Unless the language does magic like take input as an unordered list, and then happens to output it sorted.

Anonymous

@Phoenix In that case the 0-byte program would still be using a builtin. It just wouldn't have an explicit call.

Dennis

^

Challenger5

6:23 AM

I'm thinking of a language called SortOf whose main operations are sorting lists. At the end of the program, the top array on the stack is sorted and then output. A one byte solution in this language would be the instruction to delete the input (which is on top of the stack).

Anonymous

@Challenger5 That's still using a builtin without an explicit call.

Leaky Nun

@Dennis seriously, you are 18 bytes now. I'm still trying to golf down my 22-byte Pyth program before posting it, just to see that the goal is farther and farther as time passes.

Challenger5

@Mego So how do you define builtin?

Anonymous

@Challenger5 A functionality present in the language/library/whatever that is not part of the program/function/whatever.

Dennis

@LeakyNun If another language outgolfs Jelly at array manipulation, I have failed miserably. ⁹f€F is rather clumsy though: really slow, and it feels to long.

Anonymous

6:27 AM

Consider this Python program. There are 5 functions used: f, print, sorted, int, and input. All but f are present in the language - ergo they are builtins. f is not a builtin because it is defined in the program (however it uses the multiplication builtin).

Leaky Nun

@Dennis it took you quite a long time to realize ẋ though

Anonymous

In other words, if you don't have to define something yourself, it's a builtin

Leaky Nun

@Dennis so I'm wondering maybe I also missed some obvious things

Challenger5

@Mego So would you call ] in this question a builtin?

Dennis

@LeakyNun Indeed. Way too long.

Anonymous

6:29 AM

@Challenger5 ] is not a builtin, but the fact that the default, implicit behavior is to count the number of matches of the first line's regex is a builtin.

Challenger5

@Mego Yes, and in SortOf, implicit behavior is to sort the list at the end.

Can you differentiate between those two cases objectively?

Anonymous

@Challenger5 Yes. You didn't write a program in SortOf to sort a list. The language did that for you. If the language/library/whatever does it for you, it's a builtin.

Challenger5

In the Retina answer, the programming language counts the number of instances of ] in the input and prints that number for you as well.

Anonymous

@Challenger5 Yes, and I said that behavior was a builtin. ] isn't a builtin because ] isn't a command - it's a parameter for a builtin.

Challenger5

ZISC vs. OISC. What's a parameter, what's its own instruction?

Anonymous

6:35 AM

Going back to Python examples: for sorted([1,2,3]), you wouldn't say [1,2,3] was the builtin - it's a parameter to the sorted builtin.

Anonymous

@Challenger5 Now you've gone completely off the rails. That tangent isn't going to further this conversation.

Challenger5

My point is just... shoot, I don't have a point.
Let's just [ban](https://codegolf.meta.stackexchange.com/a/11228/61384) non-observable requirements and be done with it.

OK I don't get the formatting in chat

Anonymous

@Challenger5 Builtins aren't non-observable requirements.

Challenger5

Looking at the program as a black box, I couldn't tell whether a language was using a builtin or solving using its own implementation.

Phoenix

@Challenger5 No formatting in multiline messages

Challenger5

6:40 AM

@Phoenix Oh, ok

Anonymous

@Challenger5 But you don't look at the program as a black box. You can see the source code. It is very clear when a program uses a builtin.

Leaky Nun

@Dennis what do you think I can golf down from here?

Challenger5

A person with common sense could tell, yes. But it's not objective what's a builtin.

Anonymous

@Challenger5 I literally just gave you an objective definition for a builtin. What more do you want?

Challenger5

@Mego What's part of the language, and what's part of the program?

Anonymous

6:44 AM

@Challenger5 Did you implement it yourself? If so, it's not a builtin.

Anonymous

I really don't know why that's so difficult

Challenger5

So is lambda x:sorted(x) not a builtin?

Leaky Nun

@Dennis for your answer, would it be longer using outer product?

Anonymous

@Challenger5 It's not a builtin, but it uses a builtin (you didn't implement sorted yourself)

Challenger5

I could argue that sorted is implicitly implemented at the beginning of every Python program.

ATaco

6:47 AM

That's my way of saying QuickStar is done.

Anonymous

@Challenger5 But then you'd be making ridiculous arguments. It's implemented in python/Python/bltinmodule.c. Did you write the sorted code as part of your program? No? Well it's a builtin.

Phoenix

@ATaco Stars and TeamSpirit don't work well together

ATaco

TeamSpirit was never designed to work well.

Phoenix

Fair enough

Challenger5

So they're builtins for CPython, but I could write an implementation in which my argument would be true

ATaco

6:52 AM

And no, trying to star your self doesn't work.

Anonymous