The Nineteenth Byte - 2017-11-25

« first day (2492 days earlier) ← previous day next day → last day (2647 days later) »

MD XF

01:14

Who got deleted one hour ago?

2

A: Build this pyramid

Jelly, 15 bytes 24p;€$€;/ ḅ70ị¢ Try it online! This is fairly simple: we construct the list of coordinates of cubes within the pyramid as an actual list. Then all we need to do is biject the input coordinates within the square into an index within the list, which is trivial to do via base con...

Luis Mendo

01:51

Would you guys please take a look at this question in Maths.SE, in case you can give some idea?

Eric Tressler

02:48

@LuisMendo were the existing comments helpful enough?

Wheat Wizard

03:37

@MDXF Looks like they only had two answers in Jelly if that is at all helpful

I'm always interested in these mysteries

2 hours later…

Wheat Wizard

05:24

@Downgoat You should probably have options for none on each of the questions as it is impossible to answer them if the answer is none.

Downgoat

05:36

@WheatWizard I added an other option but if a language doesn’t have data types then it’s not a lang

Wheat Wizard

I mostly meant the first one

I'm not sure that languages require data, types though.

MD XF

06:21

@WheatWizard Same here

Pavel

06:48

._. how does Community make edits

2 hours later…

Fatalize

08:36

I don't understand why ais523 left the site but actually still posts answers

Erik the Outgolfer

@Fatalize ais523 is now user62131, not user76249

Fatalize

There's no way this isn't ais523

"Brachylog v2" + long explanation + knows Brachylog + criticizes Brachylog at the end

Erik the Outgolfer

well, ¯\_(ツ)_/¯

oh wait that really looks like it's posted by ais523

especially the explanation doesn't follow the one-code-block usual format and the expressions used

Fatalize

yeah

New Main Posts

09:01

0

Q: Build an Electrical Grid

The Challenge There are N cities aligned in a straight line. The i-th city is located at A[i] kilometers to the right of the origin. No two cities will be in the same place. You are going to build an electrical grid with some power plants. Power plants must be built inside the city. However, yo...

ais523

09:47

@Fatalize I still enjoy golfing, and most of my criticisms of the site go away when you don't have any reputation (things like the moderation tools being ridiculous, etc., aren't really a concern for newbies), so I've posted a few answers anonymously (typically I delete the account a few seconds after posting the answer, to ensure that it doesn't end up in a state where it can't be deleted)

so this way I get the most of all worlds, really

(also, Stack Exchange generally sucks at deleting accounts, as evidenced by the fact that I can still post here)

wait, how on earth do I have >20k rep here, I definitely don't have that on non-PPCG sites and my PPCG account was deleted…

it kept the badges too

I give up, SE clearly don't know how to run a website

(I also considered keeping an account and just bountying away all the reputation, but I really don't want to flood the site with junk bounties, and I imagine that SE would get fed up of manually deleting an account repeatedly)

Mr. Xcoder

@ais523 Well, you should join PPCG v2 when it's finished :-)

ais523

I've been giving suggestions to it precisely because of that

although if it relies on Google accounts I won't be able to (I don't have a Google account and am unwilling to create one)

Mr. Xcoder

Awww come on you can just use a dummy one

ais523

10:10

just to clarify about why I'm trying so hard to avoid gaining reputation: the pressure to gain reputation does things like making you continuously check the site to see if people are voting on your answers, try to FGITW questions before they're fully thought through, post near-trivial questions because they're nearly always massively upvoted, and the like

if I post anonymously – unable to gain reputation – much of that pressure goes away; in a way, it doesn't matter how many votes a post gets if it's from a deleted user, so instead it should hopefully be evaluated more on its own merits

Fatalize

10:53

If you feel that pressure, I think you care too much

Mr. Xcoder

^

ais523

@Fatalize I did, and it was kind-of ruining my life, so this is the way I'm coping with it

(note that SE is explicitly designed to produce that sort of pressure…)

Fatalize

@ais523 But anyway glad to see that you're still using Brachylog! I got a few ideas to make that "all deltas equal" shorter (a (is a delta of the list predicate and a "all choice points lead to the same result" metapred)

@ais523 I see

ais523

but that said, is it really a bad thing that I care about the site? I really do like the community here, and the questions, and all that sort of thing

I just hate the site software itself so much, and the SE administration to some extent too

Fatalize

There is caring and caring. I care about the cool stuff I see from time to time on the site, and some of the things I contributed

ais523

10:56

also, nowadays I don't post answers unless I particularly like them or else feel they're missing (e.g. a Jelly answer to a question that badly needs one and doesn't have one)

Fatalize

But if the site were to disappear tomorrow I wouldn't cry over it

ais523

rather than aiming to get an early answer on every question for the rep

Fatalize

@ais523 Same, I don't post many answers nowadays

Anyway, you can keep submitting answers anonymously if that's better for you. I'm more interested in your answers/explanations than you having more rep than me :)

ais523

fwiw, Brachylog is one of those languages that changed my perspective on programming

4

rather more than Prolog did, actually

Fatalize

Really? blushes

ais523

10:58

although sometimes the underlying Prolog shows through and it's a pain when it does

Fatalize

Why though? It's not that much different from Prolog from a paradigm standpoint

ais523

it's because it's one of the first languages I've seen that is focused mostly on the problem rather than on the solution

Mr. Xcoder

WW's hunt for the weirdest profile picture continues

3

ais523

someone interviewed me about esolangs recently (the interview hasn't been published yet), and I went and gave three Brachylog programs for the same simple problem to explain what sort of expressivity you can get in a golfing language

the equivalent Prolog programs would have been much uglier (and in fact wouldn't work without the help of libraries)

Fatalize

Yeah the big problem with Prolog is that there are very few libraries/built-in constructs to be more expressive

(which incidentally makes developping Brachylog harder because you have to implement everything...)

@ais523 I agree, but succeeding in hiding Prolog in all cases is pretty much impossible :(

I've lost quite a bit of willpower on developping Brachylog in the last months

ais523

11:03

Brachylog's libraries are part of what defines the nature of the language, I think

operations like "prime decomposition" are not the sort of thing that comes in standard libraries normally

(the only non-golfing-language example I'm aware of is factor in coreutils; Mathematica doesn't count)

Fatalize

Started doing more sport this summer and started my 3rd and final year of PhD in october so I'm just less enthousiastic about it

ais523

I have been working on some extent to my own golfing language, but I'm pulled in way too many directions at once

so haven't got all that far

Fatalize

I was about to ask about that :p

ais523

that said, I've worked out an evaluation algorithm for it that should produce Prolog-like results in cases where Prolog works, but act better in cases where it doesn't

(it's vaguely similar to Prolog + an unbounded loop detector, and uses this sort of trick to handle unbounded loops by running them in parallel)

I'm not sure if I want to implement reference variables / the trail array, though; they're certainly useful in Prolog but I'm not sure if they're as useful here

don't feel like you need to work on Brachylog, anyway

no matter how much you work on it I'm likely never to be satisfied

I'm too much of a perfectionist when it comes to languages

Fatalize

The only thing I'm not satisfied with is that it's hard to pass variables from a predicate to another

I had plans to improve this but it's been more obnoxious than I though to implement so I kind of gave up

ais523

11:10

it's clear what the syntax would be, and equally clear that it'd be a pain to implement

Fatalize

And since it's not a groundbreaking feature, it's probably not gonna change who uses/doesn't use Brachylog

ais523

(if you have two predicates with one nested in the other, and a variable name appears in both, it should be shared)

come to think of it, a good "hotfix" for that would be to have a few global variables

which are shared between all predicates

Fatalize

yeah that would be the quick solution

ais523

those would be useful even when you had just one predicate, too

e.g. if I use a global variable inside a map, that means it needs the same value for all elements

Fatalize

But if I started to work on Brachylog again I would rather want to make it a bit less golfy and more expressive

With the goal of writing a Brachylog interpreter in Brachylog

ais523

11:12

(whereas a local variable would be unique to the element)

Fatalize

Probably never going to happen though :p

ais523

right, I think Brachylog's being pulled in two directions

one of my early attempts at an "improved Brachylog", which I vaguely planned but never implemented, had a 6-bit codepage and the meaning of each of the 64 characters was heavily context-sensitive

for example, you defined predicates by masking builtins; builtins came in pairs, and if one was masked by a predicate, the other member of the same pair would become a variable

that way you have a large set of identifiers if you need them but they don't take up any character set space that could be used for builtins

but something like that is very much against the spirit of Brachylog's syntax

Fatalize

That would be really golfy and also really unreadable from program to program :p

ais523

my experience with golfing languages is that few people try to read them anyway

they just assume they can't read them, even with the relatively readable ones like Jelly and Brachylog

Fatalize

The problems I've had reading Jelly is that the rules for what's monadic/dyadic are crazy

ais523

11:20

right, that's a huge pain because sometimes it's calculated outside-in, and sometimes inside-out

Fatalize

When I read a chain of built-ins I have no idea what inputs they each take

ais523

actually I normally program Jelly by trial and error, when it comes to arity

Mr. Xcoder

Me too ^

ais523

normally I guess right, nowadays, except when writing my own dyads

which is normally a complete disaster

Mr. Xcoder

I sometimes hate Jelly control flow

ais523

11:24

I thought it was mostly straightforward, apart from ¦ which acts pretty differently from what you'd expect

the parsing, on the other hand, is much less straighforward

Mr. Xcoder

I find myself in trouble when I write things like <blahblah><ɓ/µ/whatever><blah><blah>

Or handling helper links

ais523

that's more a parsing issue than a control flow issue, I'd say

Mr. Xcoder

I wouldn't call it an issue, I'd just say it's a bit annoying and uncomfortable to use, although it is rather golfy

ais523

actually, I'd say one of Jelly's main weaknesses is $$ and µ…µ, that's two characters for something that's fairly common

Mr. Xcoder

Yeah, that too. We should make a feature request for $$. It should be replaced by a single char

ais523

11:30

especially as ABC$$ and AB$C$ are equivalent in most cases (for arbitrary monads A,B,C); it upsets me when I see two equivalent bits of code in golfing languages as it's a clear loss in compressibility

Mr. Xcoder

yeah, we most definitely need something like "last 3 links as a monad"

@ais523 I am afraid there is nothing to do about µ...µ

ais523

what I'd actually suggest would be a syntactic element "combine everything from here to the next quick (other than $ and friends, and Ŀ and friends) into a single link"

Mr. Xcoder

That would be harder to implement though... @Dennis What do you think about ^, ^^^?

ais523

Jelly's other largest problem, incidentally, is an explosion of € in brute-forcers; this is a problem that Brachylog solves by having lists and generators as separate concepts

Mr. Xcoder

Why is ^ a problem?

ais523

11:35

it wastes bytes

Mr. Xcoder

Example?

ais523

I'm not sure I can remember any offhand

i tried searching for €€ but got no results, I guess it's the sort of thing that SE doesn't index

Mr. Xcoder

well ok then

ais523

hmm, what about this one?:

3

A: Calculate the Number, Divisors Edition

Jelly, 17 16 bytes ×€ȷ5R¤ÆDL€€Z=Ḅi3 This is a brute force solution that tries all possible values up to 100,000. Try it online! Non-competing version The latest version of Jelly has a bug fix that allows to golf down the above code to 15 bytes. ȷ5R×€³ÆDL€€=Ḅi3 Try it online! How it works...

lots of euro signs there

Mr. Xcoder

Ok what would you think that should become?

How could it be improved

ais523

11:38

hmm, it's not a perfect example because it depends on matching up each kx with the corresponding kz

but if you change the first R to a hypothetical range generator rather than producing a list

then you would only need L€ not L€€, and you wouldn't need the zip, and =Ḅi3 would almost certainly be shorter than four bytes if it used generator primitives (although you'd want to change the algorithm)

Mr. Xcoder

I see

@Adám Pyth, 4 bytes: .fn3

Jelly, 5 bytes 1n3$#

Also 5 bytes: R>2+R

dzaima

@Mr.Xcoder isn't the 9 supposed to be a part of the program?

Mr. Xcoder

Oh, first 9 numbers... I read it as first N.

ais523

I like the R>2+R solution

Mr. Xcoder

@dzaima Then 4 bytes: -ST3

Thanks!

ais523

11:50

you can adapt the same general idea to print 3 twice, too: `9n€3+\`

Mr. Xcoder

Indeed

Reading the challenge again properly, Jelly, 4 bytes ⁵Rḟ3

ais523

interestingly, the integer 12456789 has a 6-byte representation in Jelly

so just hardcoding the output isn't far off the other solutions

anyway, I should go to bed

bye everyone

Mr. Xcoder

@ais523 “İL&’ works too

dzaima

@ais523 huh, 4 bytes in SOGL then

Mr. Xcoder

@ais523 Bye!

So “İL&’D also works

BTW your chat info just got updated

dzaima

12:00

CMC: bracket triangle. Example for 3:

[]
[][]
[][][]

Mr. Xcoder

12:20

@dzaima Pyth, 8 bytes jm*"[]"h

Jelly, 6 bytes R⁾[]ẋY

Pyth, 6 bytes jm*`Yh

Python 3, 45 bytes lambda a:"\n".join(-~k*"[]"for k in range(a))

Proton, 32 bytes a=>"\n".join(k*"[]"for k:1..a+1)

caird coinheringaahing

12:57

@Pavel If an anonymous user suggests an edit, then Community takes credit for it

Mr. Xcoder

Or it edits links, from http://stackexchange.blah to https://stackexchange.blah

user202729

13:47

The next OEIS sequence oeis.org/A000636 is related to chemistry. X is some atoms (!= H, of course) with valence 1.

New Main Posts

14:43

0

Q: Calculate pi, but

The Challenge: Calculate pi, using a spigot style algorithm, but do so, in a domain specific programming language that isn't intended for computation. Acceptable languages include (but aren't limited to): Makefile, Vimscript, Grub config, the C Preprocessor etc.

0

Q: A bank for less Trustworthy Friends

Introduction You have gotten a job as the minister of finance in your made-up country in your back yard. You have decided to make your own bank in your country for you and your less trustworthy friends. Since you don't trust your friends, you have decided to write a program to validate all trans...

Simply Beautiful Art

15:23

3

Q: BigNum Bakeoff Reboot

Some of you may be familiar with the BigNum Bakeoff, which ended up quite interestingly. The goal can more or less be summarized as writing a C program who's output would be the largest, under some constraints and theoretical conditions e.g. a computer that could run the program. In the same spi...

math code-challenge busy-beaver

Think I should write out a few explanations on the fast growing hierarchy, at least for small values?

Neil

15:46

@dzaima Charcoal, 7 bytes: ＥＮ×[]⊕ι

ETHproductions

16:43

@dzaima Japt, can't believe I managed to fit this into 5 bytes: õç"[]

Note: returns an array of lines, -R flag added to join them on newlines

dzaima

@ETHproductions Nice. That language that I'm making gets 5 bytes (currently) too, and SOGL's staying behind with 6 bytes

ETHproductions

I should get back to working on Crayon so it's actually reasonably finished by the end of the year

Man, there's a lot of things I need to finish up...

dzaima

huh, looking at Crayon - it and my language both are for ascii-art, made in JS and are stack based :p

ETHproductions

haha, nice

it's not online anywhere is it?

dzaima

@ETHproductions nope, it barely can push strings and the only string manipulation is repeat :p

ETHproductions

16:52

#relateable

The version of Crayon available online was hardly more advanced than that for several months

I had major changes just sitting in my local repo for four months before I committed them

dzaima

I don't think it will have a main canvas though, but that might be optional. And I'll leave cursor/positioning stuff for Charcoal

Luis Mendo

@EricTressler Yes, they were (although a sharper answer would always be welcome). Question 2 is answered negatively by the counterexamples in the Wikipedia article. As for question 1, I gather there is no formula or algorithm

I will soon post a challenge about this :-)

ETHproductions

Barely related: I spent a few hours learning the concepts of V a while back and it really changed my perspective on ASCII-art golfing langs

The only memory it has is the "canvas" (called a buffer), but it's really freakin' good at manipulating that

even though it can only insert/overwrite text horizontally like a normal text editor

Riker

@ETHproductions ?

ETHproductions

or am I wrong about that?

New Main Posts

16:58

0

Q: Append and erase

Given one line that consists of only letters, process as following: You maintain a string that's empty at the beginning. If the next character exists in the string, erase it. If the next character doesn't exist in the string, append it to the last Output the final state of the string. Pseudo...

code-golf string

Riker

I'm fairly certain you can do it vertically in vim

it's just tricky

and i think vim is compatible with V (or the other way around)

ETHproductions

You can't just point the cursor in any desired direction though was my main point

V is backwards-compatible with vim IIRC

Riker

yeah thanks

I knew what i meant I just can't word things today

@ETHproductions ah, yeah

@DJMcMayhem idea: 4D V

Eric Tressler

@LuisMendo You did get all of the help already that I could have offered without doing research. I don't know how you came up with the question, but you stumbled onto a pretty well-studied difficult problem

@LuisMendo I don't remember as much about additive combinatorics as I should, but I think there is a lot more you can say about |A + A| than |A - B| (where that is {a + a' : a,a' in A}, {a - b : a in A, b in B})

ETHproductions

17:17

Is it bad that I always read apostrophes as string literals...?

DJMcMayhem

@Riker aaaaaagh

@ETHproductions Correct, you can't change the direction of the cursor

You can sorta do it in reverse, but that's not super convenient

@ETHproductions V is like 95% backwards compatible, with a few very minor differences (implicit endings, M, H, and L are different, and 0 doesn't always work)

caird coinheringaahing

\o/ Verbosity can print Hello, World! (and takes about 150 bytes to do so :P)

DJMcMayhem

I'm really glad to hear you for found V helpful and interesting. :)

Erik the Outgolfer

@ETHproductions yes, definitely

dzaima

@cairdcoinheringaahing ಠ_ಠ not enough

Erik the Outgolfer

17:22

for example 'a in Japt isn't an unterminated string literal, but a perfectly terminated 1-char string literal

caird coinheringaahing

@dzaima It's designed to be readable, as well as long (like a better version of Java)

So something like 1 repeated umpteen times wouldn't be a good language to make

Luis Mendo

@EricTressler Funny thing, it's a real application. For some radio propagation measurements, I want to transmit as few frequencies as possible and get as many differences between them as possible, all differences being multiples of a certain frequency step

New Main Posts

0

Q: Swap the Endianness

As most of you probably know, (byte-addressable) hardware memories can be divided into two categories - little-endian and big-endian. In little-endian memories the bytes are numbered starting with 0 at the little (least significant) end and in big-endian ones the other way round. Fun fact: Thes...

code-golf number integer hexadecimal

Luis Mendo

But it will also make a good challenge

Erik the Outgolfer

@LuisMendo hmm

Eric Tressler

18:00

@LuisMendo that's really interesting. If you need a large number of differences (say, dozens or hundreds or more), where the exact optimal sets are impossible to calculate, there might be good constructions that drop the optimality condition

Simply Beautiful Art

18:19

22

Q: Calculate 500 digits of pi

Write a program to calculate the first 500 digits of pi, meeting the rules below: It must be less than 500 characters in length. It cannot include "pi", "math.pi" or similar pi constants, nor may it call a library function to calculate pi. It may not use the digits "3", "1" and "4" consecutivel...

code-golf restricted-source pi

Someone should figure out how to answer in base 16 or base 2.

Erik the Outgolfer

@SimplyBeautifulArt huh?

Simply Beautiful Art

@EriktheOutgolfer The challenge doesn't specify what base the output must be in

And there are some nice algorithms for finding pi in base 2 or base 16.

Erik the Outgolfer

not sure how would that make things golfier

Simply Beautiful Art

Not sure if it would either, but it'd be interesting.

Eric Tressler

@SimplyBeautifulArt it's not "2 or 16", it's a spigot algorithm for base 16 (which automatically gives you a spigot algorithm for base 2^k)

I guess it's not really appropriate to even say it's base 16, it's just any power of 2, so maybe it's more natural to think of it as a binary algorithm

Nobody knows whether there's a decimal equivalent, but an answer either way would probably be very interesting

caird coinheringaahing

18:40

Verbosity makes you hate it so much :P You have to define a STDOUT variable before printing any output :P

Simply Beautiful Art

lol

@cairdcoinheringaahing not always though

Erik the Outgolfer

what

what "always"

Simply Beautiful Art

:-/

Maybe idk what I'm talking about, so ignore me

caird coinheringaahing

@SimplyBeautifulArt Yes, always. You can't output unless you have Output:DeclareVariable<name; 0> beforehand

caird coinheringaahing

18:55

FWIW, this is the golfed Hello, World! in Verbosity :P

Erik the Outgolfer

@cairdcoinheringaahing excessive whitespace?

caird coinheringaahing

@EriktheOutgolfer Nope. That's required :P

New Main Posts

0

Q: Find the center through this alternating grid

Introduction I was trying to find a way to calculate the shortest amount of moves for a path through a grid where the adjacent tiles alternate between two colors, white and red. Also this is my first question, if I there is anything wrong with my question or something to improve, please let me k...

code-challenge path-finding

caird coinheringaahing

Welp, just gave Calvin's Hobbies another gold badge :P

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

I already posted this but was adviced to use the Sandbox because it is my first question here. Sorry for switching the order. Find the center through this alternating grid Introduction I was trying to find a way to calculate the shortest amount of steps for a path through a grid where the adj...

Riker

19:42

> The introduction of the page says you are allowed to ask questions "Q&A for programming puzzle enthusiasts and code golfers" so I thought it would fit here.

yet another reason why that thing needs changing :/

(source, deleted)

2 hours later…

caird coinheringaahing

21:45

CMP: What should I call a programming language which uses this program structure?

Fatalize

No idea how you're supposed to read this "structure"

dzaima

@cairdcoinheringaahing shouldn't the 5th line be divide(line 3, line 4) or am I misunderstanding the idea?

caird coinheringaahing

@dzaima Yes, it should. Sorry

@Fatalize Basically, it works by each line either returning a constant value, or returning a value dependant on the argument(s) it's given, and the operand between them. If the line begins with a single > it returns a value and the characters 0-9 are interpreted as numbers, otherwise they are seen as line references.

So a Hello World would be:

> "Hello, World!"
> Output 3

(Lines with Input and Output are exceptions to the reference rule, and always reference lines)

FrownyFrog

22:33

Allusion

Fatalize

@cairdcoinheringaahing "Unexplainable" because I didn't understand anything

FrownyFrog

It creates references to every statement's return value

one per line

So it's like a conversation with a lot of in-jokes

call it The Implication

Innuendo

caird coinheringaahing

Also, can someone familiar with Python RegEx explain why this doesn't make a match?

@Fatalize That could be an option :P I'm not very good at explaining things :P

@FrownyFrog Yeah, that's basically it in a nutshell

FrownyFrog

23:11

@cairdcoinheringaahing missing a ? after the comma

caird coinheringaahing

23:42

@FrownyFrog I figured it out, but thanks anyway!

FrownyFrog

np

« first day (2492 days earlier) ← previous day next day → last day (2647 days later) »

Transcript for

Nov '1725

The Nineteenth Byte

The Nineteenth Byte: General discussion for codegolf.stackexc...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2025 Stack Exchange Inc; legal

mobile