The Nineteenth Byte

cp a b is <a>b

@dzaima I can relate to that :)

@Pavel doesn't work for me, b is empty

$ echo foo > a
$ ls
a
$ <a>b
$ cat b
foo

@ngn Oh, it doesn't work in Bash apparantly. Just zsh.

@Pavel ah, ok

Q: Finding Factorials with Gamma

4:05 PM

1

Introduction We know that the factorial notation is valid for all natural numbers. However, Euler had extended it for all positive real numbers, as well as for complex numbers by defining a function, which is known as the Gamma Function. It is represented by Γ. Challenge You will be given a no...

@Zgarb that could be implemented independently of the language

@ManishKundu FYI you don't need to link to newly posted challenges, we have a bot to do that for us :P

It's too slow lmao

Whoops, I didn't see the updated version and Bubbler's and Dennis' comment.

As the function value is very large or very small near negative integers.

Probably "It's guaranteed that the result has absolute value <= 7!"?

@ManishKundu ^

@cairdcoinheringaahing False... (1) (2)

4:17 PM

Huh. It might be the primorial base builtin

primorial*

There's a difference.

Can anyone improve my Jelly answer?

@Mr.Xcoder ¯\_(ツ)_/¯

@user202729 Erik most certainly can

See, I have done something wrong.

Q: Finding Factorials with Gamma

4:19 PM

en.wikipedia.org/wiki/Primorial [no onebox]

> Not to be confused with primordial (disambiguation). 😂

New Main Posts

2

Introduction We know that the factorial notation is valid for all natural numbers. However, Euler had extended it for all positive real numbers, as well as for complex numbers by defining a function, which is known as the Gamma Function. It is represented by Γ. Challenge You will be given a no...

code-golf number factorial floating-point

... Set difference. I don't think that's valid.

For example large string = "abcdce" and small string = "ce". Can anyone check?

I'll check.

It yields ce

(it should be "ecabcdec")

Wait

4:22 PM

Have you messed up the order of those 2 strings? ...

Nevermind

@user202729 It gives (ಠ_ಠ) ecabdcecabcdceecabdcecabcdceceabcdceecabdcecabcdceecabdcec

@EriktheOutgolfer ^

@Mr.Xcoder: If you still have Safari 10, could you try navigating to staxlang.xyz/#p=94ef8fc713a48c&i=&a=1 Thanks

@recursive I still have Safari 10. What should I see?

Does it say Hello, World!?

4:36 PM

Nope, everything is blank.

I see this [anti-onebox]

ok, i guess my webpack config didn't solve it.

Thanks.

This is what I see immediately after loading the page. Np

I get the exact same error: SyntaxError: Cannot declare a let variable twice: 'e'.

Yeah. There's a bug in safari 10 that interacts badly with the default UglifyJS settings.

There's a config switch to mangle identifiers Safari-10-compatibly, but I guess I didn't set it right.

Ah, I see :(

Hi.

@user202729 no. n <=7 is guaranteed

4:43 PM

@ManishKundu The reason why the condition n<=7 is put is that to limit the value of n!.

And that, unfortunately, doesn't imply |answer| <= 7!.

Why not? Well it does..

Result is going to be within 7! Range

@ManishKundu No, it's not.

It's already mentioned that n will be non negative

I'm wanting to setup my computer at home so I can remote into it. (the files as well as the screen). Any suggestions on software? (Both are on windows)

@Mego ಠ_ಠ

4:55 PM

@NathanMerrill TeamViewer.

Yeah, teamviewer ftw

hmmm..I've really only considered that as a "let me debug your computer".

how does that work with long-term access?

like, does the ID/password persist over long periods of time (like months)?

@NathanMerrill Fine. I set up a computer for remote access before leaving my old job in Canada. I still remote into it now two years later.

@Adám ...should you be saying this in the open?

:P

@NathanMerrill … so that I can keep helping them.

5:03 PM

so, that works for remote screen...but I typically hate typing over remote desktop: Is there another solution for remote file share?

In C, how can I read a line of input without saving it? I tried scanf("%s",NULL); but it didn't do anything at all

are you trying to do "press enter to continue?"

Yep

Q: Press Enter to Continue in C

9

How can you do a "Press Enter to Continue" in C?

c input

Q: Are backward array and inputting array via inputting the last element's pointer allowed?

Thanks

New Meta Posts

5:27 PM

0

https://codegolf.stackexchange.com/a/157447/76323 without STD is backward array, with STD/CLD is last element's pointer

discussion

@NathanMerrill TeamViewer.

@Pavel Done

Thanks!

Also, thanks for making it single digits because of vim lol

@DJMcMayhem I ended up seeing my first ever piece of SNOBOL code because of it, so totally worth

5:37 PM

@Adám sorry to keep pestering you: Have you used file transfer mode? Does it work like a remote drive (When the file is saved, it will save it across the network)?

hmmm...nevermind, I found a screenshot

oh this is hilarious

If your system can't find files, a glitch in the system may be responsible. To fix the glitch, uninstall and reinstall the program, or powerwash your computer. After doing so, the system should be in working order.

source (at the bottom)

A: Sandbox for Proposed Challenges

lololol

New Sandboxed Posts

6:07 PM

0

Same code, different work Write a code that do these few things: Input a positive integer n, check whether it's a prime. (2 -> true, 5 -> true, 8 -> false) Input an integer n>1, output an array with fewest positive prime as element, where the sum of the array is n. (2 -> [2], 5 -> [5], 6 -> [3...

6:17 PM

Is there a way to get a pointer to a constant value (0) in C other than int zero = 0; &zero? Everything I can find is about const int* or about null pointers rather than a pointer to null.

@Pavel I don't think there is

It sounds like you need new in C++

god dammit

Or malloc I guess?

create an array, initialize it to zero

that would work too

Huh... I wonder if malloc(0) is safe if I don't free it.

6:25 PM

valgrind:

==11685== LEAK SUMMARY:
==11685==    definitely lost: 0 bytes in 1 blocks
==11685==    indirectly lost: 0 bytes in 0 blocks
==11685==      possibly lost: 0 bytes in 0 blocks
==11685==    still reachable: 0 bytes in 0 blocks
==11685==         suppressed: 0 bytes in 0 blocks

Poke

you definitely lost 0 bytes 1 time

uh oh!

So I can use malloc(0) to get a free pointer to 0 without decalring a variable. Awesome.

Why do you need a non-declared pointer to 0?

@Pavel with malloc() there's meta information about the allocated block just before the returned pointer

@ngn Valgrind says nothing leeaked tho

6:27 PM

@Pavel it's lying :)

@Pavel or counting only the amount of memory that the user is supposed to use

@DJMcMayhem getline takes a pointer to an int as it's second arg. If the first argument is a pointer to NULL, then the second should also be a pointer to 0.

@Pavel (void*)0 is golfier than malloc(0)

@betseg malloc vs (void*)0.

malloc(0) still points to somewhere in the heap

Woooh segfaults! That means it likes you.

6:30 PM

(void*)0 is NULL

@Pavel I think by malloc(0) you really mean: int*p=malloc(sizeof(int));*p=0;

Anonymous

@Mr.Xcoder That's weird

@ngn The point is to not declare a variable

@Pavel well, if you do malloc(0) do you have any idea what the pointer points to?

0, right? malloc always points to 0.

6:34 PM

@Pavel it points to something beyond the end of your (0-sized) chunk of memory

Mm

wait, seriously, how would a memory block of a size 0 work?

@betseg why wouldn't it? :)

a memory block is something like [metadata] + [malloc argument bytes] right?

malloc(0) actually does make sense though, right? If you have an empty array you want to fill and realloc later.

6:38 PM

@betseg right

@Pavel right

so if i malloc(0);malloc(0);, in theory, it can result in something like [metadata1][metadata2]

Is there a simple way to get the size of a malloc'd chunk by reading its metadata?

won't it segfault if you try to read outside its designated space tho

free gets away with it without segfaults

because it's a system call

6:42 PM

@betseg likely not, as access permissions are given to pages (usually of size 4k), not to malloc-allocated blocks

@betseg free()? a system call?

huh it isnt

@Pavel That's a good question. I don't know the answer.

@Pavel apparently gcc has one

msize()

Supported by gcc, clang, and msvc, but not part of the C standard

or malloc_usable_size(), apparently not standard API

Anonymous

6:47 PM

@Mr.Xcoder Oh nevermind, that's not weird. If the command fails, the stack is left alone, as if the command was a NOP.

Anonymous

And I don't have an overload for Gamma

Then feature-request: overloads for gamma

Anonymous

Soon(TM)

@Pavel are you implementing a golfing language in c?

It's not a golfing language, it's just an esolang

6:54 PM

@Pavel do you have a link with more info about it?

@ngn Nope

@Pavel I have an impl of a (non-golfing) language in c and I decided to do my own memory management instead of using malloc()

Seeing as how I'm still figuring out how to read input, I think I'm going to do not that.

@Pavel oh, so that's what the getline() is for...

@Pavel why not just char buf[256]; read(0,buf,sizeof(buf)) ?

@ngn Fails for lines longer than 256 characters

6:57 PM

@Pavel I mean, it's easiest to just read from stdin and write to stdout and let rlwrap handle user-friendly line editing

Or just call readline directly. getline is for the case that !isatty(fileno(stdin))

@Pavel "just"? that sounds like more pointless work for you

@ngn It really isn't. readline returns a dynamically allocated char*, so I don't have to worry about buffer overflows and it saves a malloc call.

@Pavel ok, as you wish

Anonymous

Have you heard the good word of our lord and savior, C++'s std::string?

7:02 PM

Nooooooo

Anonymous

How to read in a line of input from STDIN in C++: std::string foo; std::getline(std::cin, foo);

Begone, demon of unnecessary memory overhead

Anonymous

@EriktheOutgolfer You saw nothing

uh, what? (also imho 11 and generally "you saw nothing" has been done to death)

Anonymous

@Pavel Unless you're programming for an ancient computer, memory overhead is not a concern

7:05 PM

Pure C runs faster than C++

Anonymous

Bad C code is slower than good C++ code, though

Anonymous

And the speed differences are too minute to matter in most cases - I/O will almost always dominate the runtime

You don't have to write bad C code. All you have to do is not write the best C code, and it'll be slower

Anonymous

The simple act of reading in bytes from STDIN is much more of a factor in runtime than whether you use C or C++ to do the reading, past a certain threshold of bytes.

because there are tons of ridiculously smart people that make C++ as blazing fast as possible

Anonymous

7:07 PM

That threshold is probably somewhere in the 4-8 region

I may not write good C, but I don't write good C++ either

yeah, but its far easier to write idiomatic C++ than C

Anonymous

@NathanMerrill As evidenced by the previous several minutes of discussing the best way to read a line from STDIN in C :P

Dennis

@Mego Which is why compiling golfed code is faster.

lol

7:09 PM

all I need is c and its read() syscall

I get it know.

Anonymous

@Dennis I'm not sure if you're trolling or not :P

I once spent a good hour trying to add an element to an std::vector. I think I'm just allergic to C++.

@Dennis what compiles faster, 1+1*1/1 or 1 /* blah blah blah blah blah blah blah blah */ + 1? ;P

Dennis

Partially. If you have a 100 MB C project, I think golfing would actually improve compilation time

7:10 PM

anyone here familiar with select()? i've heard it's faster than read() when reading from several places

@Pavel I can agree that C++ likes to be really pedantic about their names, which makes finding the function you want difficult

@Dennis This gives me an argument for tabs vs spaces: Code with tabs takes less time to compile :P

Anonymous

@Pavel It actually takes longer, because the compiler curses you out every time it encounters a tab

Anonymous

@Dennis Depends on the golfs. Heavy macro abuse could make it take longer.

@Mego Only in Haskell

7:13 PM

somebody should do a performance test in Python in spaces vs tabs

well, of course tabs would be faster, but the difference would be virtually nothing

tabs > all

F# compiles extremely fast with tabs: It'll finish almost instantly. The fact that it finishes with exit code 1 and doesn't produce an executable is beside the point.

8

@EriktheOutgolfer I want numbers. Like, are we talking a 1% performance increase, or a .001%?

I'm guessing even less than that, but I swill want it tested.

7:16 PM

I think that the inaccuracies will be way more than the actual difference, so sorry but I can't provide numbers :P

@Dennis Ooh, is Jelly 2 being actively developed?

"Jelly 2" is something far, far in the future

@Pavel 0.00003% with Clean

Removing both tabs and spaces and using brace style layout is slower because the compiler yells at you for it.

Tested by swapping between spaces and tabs for the entire shipped module package.

I feel like most of this is because IIRC the compiler has the case for tabs before the case for spaces.

@Οurous wait what? how did you manage that so fast

7:31 PM

@Οurous I'm pretty sure reading from a file is slower than a conditional jump.

@EriktheOutgolfer Well you can use expand to conver tabs to spaces, so you can just run it on an entire directory.

talking about Clean? hm, I definitely need to learn that language

@Pavel oh, I was doing a 1-for-1. 4-for-1 is about 0.00007% slower.

I think your random variance is higher than your observed difference.

The case work is slow compared to what minimal assembler would be because the compiler is built with stack, index checking, memory and time debug harness, etc.

Probably.

@Pavel I don't have a larger sample to use though.

And the startup and winddown time for the timer probably counts here too.

I mean basically everything would. I'd have to have a huge amount of source to actually measure this well.

Maybe we can try with the entire C++ standard library

7:39 PM

That's probably be large enough.

IDK how to build it though

Maybe put it on a ramdisk too?

7:57 PM

TFW the footer is 850 bytes longer than the actual code :P

Ok, decoding works..

@moonheart08 decoding what?

Working on a golfing language. It has a 9 bit codepage :P

How? Doesn't that mean you can't claim 1 byte per character?

How? By writing my own encoder and decoder :P

And yea, it means i cant

8:12 PM

So doing it for the fun of it? :P

Pretty much. But you can do quite a bit with the doubled amount of characters you can use

AdmBorkBork

@Mr.Xcoder I awarded my bounty for the digital clock in GoL.

Challenge for anyone who cares: I can golf 10 bytes off this answer. Cookie emojis for anyone who can golf any bytes from it, extra kudos if you can beat or match 10 bytes golfed :P

Dennis

@Pavel Let's say it's being actively thought about. :P

@cairdcoinheringaahing I'll look into it.

I want a cookie, but i've never add++ :P

Q: What answer-chaining scoring criteria promote collaboration?

8:18 PM

I'd recommend looking at the add++.py for commands, the wiki really needs to be updated

New Meta Posts

0

I have an answer-chaining question in the sandbox, but I still haven't come up with a winning criterion. What scoring criteria are there for answer-chaining questions that encourage users to collaborate for a longer chain?

discussion tips answer-chaining

I just realised this existed. codegolf.meta.stackexchange.com/questions/14337/…

That means that one idea i had on the backburner to try out, which was flags that change which 'module' libraries are loaded into the golf lang for quick usage, no longer works as well

@cairdcoinheringaahing Oh nice, I will try to learn Add++ for this :)

@AdmBorkBork Great! I'll update the Gist later

8:35 PM

someone seems to be having fun with vote buttons...

Is someone upvoting/downvoting a lot of posts? (Sock-esque behaviour?)

wishes I could be a mod

for me it was just -10 for a user being removed :)

@cairdcoinheringaahing no, I just saw two answers being downvoted and then immediately undownvoted.

Much less interesting :P

-10 user was removed :(

8:40 PM

Same thing.

Wow I saw -440 in my rep list on Feb 20 and I was like Whaaaat?, but then I realised I placed a bounty :|

Also ಠ_ಠ I have lived to see Wumpus beat 05AB1E

Where? The infinite loop, no output question?

No... Kolakoski repost

Which is even more surprising given the difficulty of the challenge

Anyone here has experience with .bnf files? (brb)

@Mr.Xcoder I see a 12 byte 05AB1E answer, and 13 Wumpus (although beating the 15 byte answer is decent) :P

Q: Tips for Golfing in Brain-Flak

8:46 PM

Wait, there's a 12-byter

ಠ_ಠ Emigna posted 2 answers so I didn't bother to check if there's anything that beats their submission

Newly Featured Questions

23

Brain-flak is a stack-based turing-tarpit language, written collaboratively between me, DJMcMayhem, and 1000000000. Some users are very experienced in the mysterious ways of Brain-Flak. So I thought it a good idea to set up this question as a way for us, and hopefully others too, to share our k...

code-golf tips brain-flak

is this question really named "tips for golfing in brain-flak"?

Yep. We have a Java one, so BrainFlak isn't that crazy

Yes. Because it includes tips for shortening Brain-Flak submissions.

H.PWiz

@MartinEnder I do this semi-frequently on mobile, because I have a dodgy screen

8:53 PM

@labela--gotoa Brain-flak can be surprisingly golfy sometimes

ಠ_ಠ I suspect that'd beat Add++ if I took a crack at it

@cairdcoinheringaahing I'm curious... Do you have a link to Add++ docs? I might try to learn it

:o I didn't know Leaky used Brain-Flak.

Leaky uses everything :P

@DJMcMayhem They're in the Wiki of the repo, but they're reaaaaaly outdated.

8:55 PM

And anywho, the reason that particular answer is so short is because brain-flak is accidentally shockingly good at polynomial (triangular, square, etc.) numbers

@DJMcMayhem Here you go but they aren't very complete. I'm updating them at the moment

I strived to understand Dennis' answer to the Kolakoski sequence challenge but I still have absolutely no clue how it works.

then ask Martin Ender :P

@Mr.Xcoder His Jelly answer?

which one...

8:56 PM

@cairdcoinheringaahing Yes, of course.

@Mr.Xcoder Yeah, I have no clue either :P

@EriktheOutgolfer I'd really like to figure it out myself, but I think I'll eventually give up :)

remember Dennis is a math professor

And a really good one, it seems :)

It's not that mathy, compared to some of his other answers

8:58 PM

ಠ_ಠ

@Mr.Xcoder as with Emigna's 05AB1E answer, there are two Jelly answers by Dennis.

Well, I was referring to the more impressive 7-byter

maybe I should post another Jelly answer too? :)

Unless you come close to 9 bytes, I think it's not worth it :P

Q: Why do people post two answers in the same language?

8:59 PM

oh wait Jonathan Allan has covered me for that one

@Mr.Xcoder you mean 6

No, I mean 9.

@cairdcoinheringaahing Because 1) they use fundamentally different algos and 2) because the people who upvoted the first will almost certainly upvote the second one too => more rep :P

@Mr.Xcoder algorithmically, I actually find the 10-byter more interesting and impressive :D

I vividly remember Dennis's 7-byte answer because he used it in a discussion we were having

It's time to design Triangularity v2... I'll name it Cthulhu :P

@Mr.Xcoder what?

New Main Posts

9:02 PM

0

Q: Wanna be an Alchemist

Wanna be an Alchemist I've been watching Full Metal Alchemist these past days and I've been wondering, It is really possible to trans-mutate matter? In the real world, I don't know if it is possible, but in the machine world, we think we "can". Problem Your task is, Given an input of N chemi...

code-golf binary bitwise

@labela--gotoa Long, long story...

@Mr.Xcoder But, don't you already have a language named like that?

@NewMainPosts "I'll do it later"™

It was never really a thing

^^

TIO looks neat transparent

9:09 PM

> Meta.codegolf: random shit

3

+1 :P

codegolf.meta.stackexchange.com/questions/1324/… :P

Well, encoding and decoding works. Time to make a test """"program"""" for basic parsing.
{1{p}{2×p}⤙}e

i.imgur.com/7heoVzB.png `

oops

:P

@totallyhuman oops that's the wrong post

that link's supposed to be codegolf.meta.stackexchange.com/a/10923/68615

That "explains" it :)

ye my bad lol

put myself in a weird spot there XD

9:14 PM

shoot

enter key

I'm just going to note, because why not, i imagine a truth machine program, with the current sketchy concept, would look like this: 0{0p}{{1p}⥀}⤙
Breaking that down:
0 Push 0 to the stack, to give the necessary data to the upcoming switch command.
{ Begin a sub block.
0p Push 0 to the stack and print it.
} End the block. '0p' is now on the stack as a function.
{ Begin the next block. This block will be the catchall for the switch.
{1p} Shouldn't need to break it down again. This pushes '1' to the stack and prints it out.

Yes, the switch command is a little over the top :p

But this golf lang is mostly oriented towards constant operations, but a switch operation should exist.

So i invented that monstrosity of a command :P It'll likely get better

@moonheart08 it has less steps than argument allocation for invertible binary operators in Dirty :P

:P

I kind of want to write a challenge that involves wikipedia...

tl;dr 0{0p}{{1p}⥀}⤙ is a truth machine in some arbitrary language moony made up

codegolf.stackexchange.com/questions/34662/… Craaapp... I more specifically wanted to write this challenge that already existed.

9:26 PM

oh, and ● pushes the number 360 to the stack, because circles

is there any place where you could put all the docs on?

Me? Dunno. Probably on hellomouse.cf as a download

Create a github reop

I'm still working on it, i don't even have 256 characters for the first half of the codepage yet

lol

you can still post the ready parts of the language, Jelly isn't finished yet, but it's still one of the best golfing langauges out there

9:29 PM

@EriktheOutgolfer It isn't? I mean... are any languages "finished" by that definition :P?

True

I feel like no language is ever finished until people stop using them and updating them.

@MagicOctopusUrn it even states so in the README.md file, which is true

@EriktheOutgolfer Maybe if you can convince me to write out my concepts into the doc early

i'm still formulatig them, that's just a rough concept of a truth machine

> Please note that Jelly is currently under development; features will change or disappear without prior notice.

9:31 PM

@moonheart08 You could make a dump of the effects of each control character on an empty stack to start?

@Mr.Xcoder Oddly, I've never seen a feature change or disappear :P

@moonheart08 a table with the ready features and the ones you plan to implement can be beneficial

also Dennis is pretty conservative about breaking backwards-compatibility

@Mr.Xcoder I mean, int terms of PPCG, Turing completeness is being "finished" technically. That's only if "finished" means "able to compete", if "finished" means "no further updates will follow" I really can't list an esolang with a 256 codepage that is "finished" AFAIK O_O.

> That's only if "finished" means "able to compete"

that's actually not a requirement anymore as of a long time ago

Well I think languages like GS2 and the like are finished

9:33 PM

CJam, Golfscript and possibly Pyth

Even when 05AB1E ran out of single-byte commands they still updated the 2-bytes and list commands constantly hah.

Also Brachylog, Ohm, Pyth, Cjam and Gaia are pretty much stable

Not too many though haha, was my point.

Brachylog 1, not sure about v2

There's Brachylog and Brachylog v1...

9:34 PM

v2 is under development

@moonheart08 If you want to see how a language evolves rapidly while a person is using it just look at Dzaima's languages haha.

If I remember correctly he was competing with it before he made the documentation.

And v1 is pretty much just an ASCII only version.

@MagicOctopusUrn that's a strategy so that you can start using it sooner

LPT: cd - goes to the previous cd'd-to directory

I've gone through 6 version numbers in the last month with stax. github.com/tomtheisen/stax/releases

9:37 PM

@totallyhuman "cd -" <--- this?

LPT?

Life Pro Tip?

@recursive Linux* Pro Tip?

Oh ok thanks

both :P

9:38 PM

@totallyhuman is that every dist of linux that that works on?

@MagicOctopusUrn what are you asking me? :P

I've wanted that before

@totallyhuman I was asking if the command was "cd" plus a "hyphen" or if you were saying an empty CD did that.

hyphen yeah

@totallyhuman Niiice, I'll use that when checking websphere logs...

/opt/IBM/WebSphere/Profiles/DefaultCustom01/logs/dev/ is a mouthful...

@MagicOctopusUrn i think cd's a part of bash, so anything with bash i guess?

also works with zsh, so probably most sh flavours

9:41 PM

I'm bored.

Anyone want me to polish/post any of these challenges: codegolf.meta.stackexchange.com/… ?

CMP: en.wikipedia.org/wiki/List_of_chemical_elements would a challenge asking you to take in a number between 1 and 3 then sort the list of chemical elements according to that number as follows: 1=Element #, 2=Density, 3=MeltingPoint be decent?

Maybe even making the list include Atomic Weight or Boiling as well.

2 hours ago, by caird coinheringaahing

Challenge for anyone who cares: I can golf 10 bytes off this answer. Cookie emojis for anyone who can golf any bytes from it, extra kudos if you can beat or match 10 bytes golfed :P

@cairdcoinheringaahing My ADD prevents me from competing in ADD. Also the fact that I don't know it O_o

Only I know it (no one has bothered to learn it :/) but you can read the source code :P