The Nineteenth Byte

user202729

3:00 PM

@Adám timeline

Adám

@user202729 ?

Wheat Wizard

Timeline can't be used to check voting times

just comments and edit times

user202729

Doesn't work.

Erik the Outgolfer

voting times is something secret even to mods (well, maybe not to SE employees if they check some internal logs by hand)

you can only know your own vote time for your last vote if your vote gets locked in

Adám

@EriktheOutgolfer So how do they detect serial voting?

Erik the Outgolfer

3:03 PM

@Adám voting patterns are visible, individual votes aren't (that is, to mods)

dzaima

@user202729 I feel like "error" is too subjective for it to really have concrete boundaries

Anonymous

@Adám I don't know the exact details, but from how it was described to me in the past: mods don't have "user X voted on post Y at time Z", but they do have something along the lines of "user X has voted on N posts by user W in a short period of time"

HyperNeutrino

I guess you could say that the program has to exit with non-zero status?

Adám

@HyperNeutrino Not every programming language may be able to exit with a non-zero status.

HyperNeutrino

that's true

Erik the Outgolfer

3:08 PM

@HyperNeutrino what if it's a function though? functions don't have an "exit code"

HyperNeutrino

true

Anonymous

@Adám With the wide variety of nonsensical things that programming languages can choose to do, at some point we have to say "close enough"

Wheat Wizard

^

Erik the Outgolfer

@Mego nah, worst case should be "ask the author"

user202729

... which is because users (mostly on this site) design weird languages ...

Dennis

3:11 PM

@Adám No, the timeline doesn't tell us much more than it tells you. On main, we'd have the reputation tab, but on meta...

Martin Ender

Does anyone happen to know if it's possible to have a string whose length changes under a (.NET) ToLowerInvariant() or ToUpperInvariant()? The only letter I'm aware of that is represented by multiple characters under a case change is ß -> SS, but the Invariant methods leave that untouched. Are there any others I should be aware of?

Wheat Wizard

I think we have to live with the fact that some languages can't participate in every challenge

12

user202729

... I guess the stars here means "accept" ...

Erik the Outgolfer

that's something one can conclude to using multiple ways

HyperNeutrino

well, HQ9+ can't participate in most challenges and nobody's complaining about that :P some languages just don't work for those challenges

Erik the Outgolfer

3:13 PM

yeah, especially since we have abolished the need of a solution using a programming language as defined by consensus

Anonymous

@MartinEnder Since Invariant is based on English/ASCII, and only those letters are affected, I'd think not. There's no scenario in English where one letter maps to more than one via swapping case (or vice-versa). Even byte count should be unaffected, since you can't get an ASCII char to a non-ASCII char (or vice-versa) via case swapping.

Martin Ender

oh, is it really only ASCII? I assumed it's all Unicode characters that aren't ambiguous between various cultures.

user202729

The easiest way to try is to run it on all characters. There are only about a million of them.

HyperNeutrino

only

i mean that's actually not too much I guess

Anonymous

> The invariant culture represents a culture that is culture-insensitive. It is associated with the English language but not with a specific country or region.

user202729

3:16 PM

"there are only about four billion floats"

J. Sallé

@MartinEnder Actually (as of mid-2017), the es-zett ß has an uppercase counterpart ẞ

Martin Ender

@Mego yeah, I read that, but that doesn't mean it's just ASCII

Anonymous

Oh, I misread. Their example at the bottom has a bug that makes it not useful

Martin Ender

@J.Sallé I'm aware, but I'm sure there are still some (many?) upper case implementations out there that give you SS.

Adám

@MartinEnder ﬀ → FF etc.

J. Sallé

3:18 PM

@MartinEnder yeah probably, I was just commenting on that because my German teacher told me so the other day hahahahah

Erik the Outgolfer

that's the problem with substantial changes to a very widely-used applied concept, such as a very well-known language such as German

Anonymous

According to a modification I made on MSDN, applying ToLowerInvariant doesn't change the length of any of the example words

HyperNeutrino

Does anyone here know how to get the output of a subprocess.call?

J. Sallé

And as Adám pointed out, there are some digraphs in the latin alphabet that do not exist in uppercase and are written as the doubled letter

Anonymous

@HyperNeutrino Use check_output instead of call

HyperNeutrino

3:20 PM

Oh okay thanks

Anonymous

Or use run (if >= 3.5)

HyperNeutrino

right okay

How do I provide custom input to a program? I need to be able to feed the input in when the program is already running BTW.

Pavel

@Mego My dad told me about a bug in a program he was working on, where the layout of the app would break horribly for European users. As it turns out, the Float.Parse function uses the system culture, which in many European countries uses a comma and not a dot as the decimal point.

It also fetches the system culture every time it's run, instead of caching it, so it's really slow, too.

Anonymous

@HyperNeutrino Then use subprocess.Popen

Pavel

@Mego Uppercase python built-in function :O

HyperNeutrino

3:31 PM

what the heck

oh

J. Sallé

@Pavel had a similar problem on a program I'm developing for work. We've got clients who use commas and others who use dots for the decimal point, while my program's expression parser expects commas for decimal points and dots for separating the numbers (we deal with monetary values, so a lot of input is supposed to be something like 10.000.000,00)

HyperNeutrino

TIO server's python command is Python 2

LOL i got confused slightly

Dennis

@Pavel That's a class.

Thomas Ward

@Dennis a callable class, but yes.

Martin Ender

@Mego hm, yeah, now I'm just wondering whether I shouldn't use the Invariant methods after all so that I get some useful conversions for other characters as well

Thomas Ward

3:32 PM

I have some holdovers from the sp.Popen days :p

Martin Ender

man, I hate when the real world leaks into my programming...

HyperNeutrino

@ThomasWard what is a "callable class"?

user202729

implements __call__, probably.

HyperNeutrino

but Popen doesn't...

Popen(...)() errors

Anonymous

@ThomasWard If by "callable class" you mean "has a constructor", sure

HyperNeutrino

3:35 PM

oh that's what it means? ok

I thought it meant "implements __call__"

Anonymous

a = Popen(...) is no different from a = SomeOtherClass(...)

Anonymous

(str is a bit different because it's a builtin function that's a wrapper around the str class's constructor)

HyperNeutrino

@Mego No I'm pretty sure str is the same case it's just that the constructor takes different things and then .__str__()s them

I guess having the new keyword only makes sense in compiled languages where class names and function names can be the same without ambiguity

Martin Ender

@Mego btw, the Invariant methods do change the case of non-ASCII characters. but even with an appopriate culture I can't get ß to change, so I guess I'll just go for the Invariant methods and assume the length stays the same.

Anonymous

@HyperNeutrino Hmm... It's listed in the docs as a built-in function, but it's actually the class constructor. Come on, Python :(

HyperNeutrino

3:40 PM

o wtf D:

Adám

@MartinEnder Wonder if there are any tools that will convert ß to ẞ…

Pavel

@MartinEnder Using the default culture is both slow and inconsistent, leading to odd bugs. Even if you don't use InvariantCulture, you should always chose a culture to use when calling culture-dependent methods.

Martin Ender

good to know, thanks. I'll be sticking to Invariant for now though.

it seems to do pretty much what I want

Adám

CMC: Given a German string and a case (lower/upper/title), fold the string to that case. E.g. Weiß, upper → WEIẞ; Die Räder sagten "ßß", title → Die Räder Sagten "Ssß"

Erik the Outgolfer

@Adám why Ssß and not ẞß?

Adám

3:53 PM

@EriktheOutgolfer Because the proper title case of ß is Ss

Maybe it is hard enough to go on main?

Erik the Outgolfer

I don't think so?

Pavel

@Adám Are there any rules to consider that weren't covered in test cases? I don't know German.

Erik the Outgolfer

@Adám can we assume there's so ss, sS, Ss or SS in the string?

Poke

tfw you restart your browser and computer several times because you're being prompted to log in to a site with kerberos auth only to realize you were sending an authorization header with an extension

blah

Adám

@Pavel You must handle ä←→Ä, ö←→Ö, ü←→Ü and of course the difficult Ss←ß←→ẞ→Ss.

@EriktheOutgolfer So?

Erik the Outgolfer

4:01 PM

@Adám I've been thinking that if so I could use an algorithm to replace ss -> ß and SS -> ẞ

while leaving Ss untouched

(sS can't occur after case conversion)

Adám

@EriktheOutgolfer You mean "no"?

Erik the Outgolfer

@Adám oh yeah

Adám

@EriktheOutgolfer No, strings may contain ss (in any capitalisation).

Leaky Nun

@Adám can a word actually begin with ss?

Adám

@LeakyNun No.

Erik the Outgolfer

4:04 PM

@LeakyNun but he didn't say there must only be words :P

Anonymous

@Adám Actually, 14 bytes: FOi5@%D"ÿûù"Eƒ

Adám

@EriktheOutgolfer In fact, a test case had a non-word.

Anonymous

Oh, Python thinks ß -> SS :(

Thomas Ward

i'm tired sue me :P

Erik the Outgolfer

@Adám technically, test cases ∉ challenge specification

Adám

4:05 PM

@Mego You don't need to take string to indicate case. Just any consistent 3 markers is ok.

@EriktheOutgolfer I disagree when it comes to CMCs.

Anonymous

Oh in that case

Adám

@Mego Notice that the cases were not in <code> tags.

Erik the Outgolfer

@Adám :P of course I'm just nitpicking there

Anonymous

1 byte: ƒ (uses "ÿ" for title, "û" for upper, and "ù" for lower). Still suffers from the incorrect translation thanks to Python.

Anonymous

I know about as much about German as the average stump

Anonymous

4:08 PM

In fact, the stump may have me beat

Adám

@Mego Oh, so Python handles the ß→Ss title casing? Nice.

@Mego But still invalid :-(

Anonymous

Unfortunately

Erik the Outgolfer

@Mego well, that's the challenge ;)

Adám

@Mego ^

Anonymous

Ok I don't need 4 pings in 10 seconds thanks

Leaky Nun

4:09 PM

CMC: given n positive integer coprime with 10, output the last 100 digits of 1/n in 10-adic integers

e.g. 1/3 = ...6666666666667, so you output 99 "6"s followed by "7"

1/7 = ...142857142857142857143

Anonymous

Well that challenge is quite different from the pre-edit challenge

Erik the Outgolfer

that's a problem with CMCs...

Leaky Nun

sorry :P

btw 1/21 = ...23809523809523809523809523809523809523809523809523809523809523809523809523809‌52380952380952380952381

Downgoat

> Meteor allows you to develop in one language, JavaScript, in all environments: application server, web browser, and mobile device.

Anonymous

@Downgoat But why would anyone want that?

Downgoat

4:16 PM

they say this like it’s a good thing...

Anonymous

> This knife allows you to do one thing, stab yourself, in all places: foot, hand, chest, leg, etc.

3

Adám

@Adám APL (Dyalog Unicode), 65 chars/97 bytes: {'\bß' '\b.'⎕R'Ss' '\u&'⍠'UCP'1⍣(0>⍺)⊢'ß'⎕R'ẞ'⍣(1=⍺)⊢⍵(819⌶)⍨0⌈⍺}

Ironically, in DejaVu Sans Mono, ß looks uppercasey while ẞ looks lowercasey.

FrownyFrog

+/@]`($:],#)@.(0<1|2!inv+/@,)

It gives me rank error

For codegolf.stackexchange.com/questions/129004/…

Adám

@FrownyFrog Is that a question/request for support?

FrownyFrog

I’m not sure yet

Yes.

HyperNeutrino

4:30 PM

@Mego it also lets you stab other people too :D

FrownyFrog

What would be the problem, could someone give this a look?

It’s supposed to be called with a zero like N +/@]`($:],#)@.(0<1|2!inv+/@,) 0

Adám

For anyone who doesn't recognise it, @FrownyFrog's code is in J.

FrownyFrog

If the sum is not triangular, append the length of the right to the right and recurse

Adám

@FrownyFrog Don't you need parens around the entire function when calling it with N and 0?

Anonymous

Gah, what is up with all of the chat flags today?

Anonymous

4:33 PM

I've seen 5 different messages flagged in the last hour and a half or so

Thomas Ward

@Mego it is the beginning of the end.

that's what's up.

FrownyFrog

hmm

Thomas Ward

Chat is exploding. Doom is coming to it. Enough chat flags will happen and then the global Diamond Storm begins.

5

yes i'm a little crazy :)

FrownyFrog

it’s not a train, so I don’t think so

Anonymous

@ThomasWard Maybe then SE will remember that chat exists

Adám

4:35 PM

@FrownyFrog See? I don't know enough J to help you. You even have a gerund in there, something I've never used. I do have a colleague who knows J well, though, but it isn't in his job description to do code golf. I can ask him after we go home, in about an hour or so.

FrownyFrog

Thank you.

Ok, this part 0<1|2!inv+/@, returns rank 1, but @. wants an atom or something like that, so 0{ fixes it

@Adam I figured it out

Adám

@FrownyFrog I don't understand why that returns rank 1. Doesn't +/ return a scalar?

FrownyFrog

the inverse of ! is literally 3 : '(-(!-y"_)%1e_3&* !"0 D:1 ])^:_ <.&170^:(-:+)^.y'

I have no idea what that would return

probably the culprit

or maybe it’s <

< is the last one

Riker

@Mego s/chat/ppcg/

HyperNeutrino

lol

Adám

4:45 PM

@FrownyFrog < has rank 0 0 so that isn't it.

@FrownyFrog ! also has rank 0 0, so its inverse must have rank 0 0 too.

Neil

don't worry, Diamond Storm only has 5 PP (8 with max PP Up)

2

eaglgenes101

I think I have an idea for the moby dick challenge

On first run, it decompresses itself, then a file embedded in itself, then "exits" while leaving a daemon running in the background

Subsequent invocations ask the daemon to give it another byte of the decompressed file, which it gives

Erik the Outgolfer

I don't think we can consider it "exits" then? that's definitely cheating and will most probably be downvoted

FrownyFrog

@Adám Well $2 !inv 2 gives 1

So it’s rank 1

eaglgenes101

Alright, let's then drop the daemon part

Adám

4:56 PM

@FrownyFrog Looks like a bug to me. It gives a scalar 2.56 in APL.

FrownyFrog

that’s what it does

an integer means the number is triangular

Adám

@FrownyFrog No, if $ gives 1, then it is a vector, not a scalar.

FrownyFrog

look at the inverse, it’s like a feature-bug

eaglgenes101

Is there a fuzzer that can optimize for a goal other than crashing the executable it's handed?

FrownyFrog

Oh I see, it works in APL

eaglgenes101

4:57 PM

Might be useful for the moby dick challenge

FrownyFrog

I don’t know what I’m talking about either, but yeah, I see how it’s a bug

Mr. Xcoder

5:25 PM

I get the feeling this is biased towards 05AB1E... — caird coinheringaahing 20 secs ago

@cairdcoinheringaahing No, I don't think so. It's (perhaps) more biased towards Pyth.

caird coinheringaahing

It's definitely not a nice challenge for Jelly

New Main Posts

1

Q: Minimise the count of prime factors through insertion

Given two positive integers A and B, return the position p that minimises the number of prime factors (counting multiplicities) of the resulting integer, when B is inserted in A at p. For example, given A = 1234 and B = 32, these are the possible insertions (with p being 0-indexed) and the corre...

Mr. Xcoder

I don't think it's so hard for Jelly either.

caird coinheringaahing

The insert part is long in Jelly

Unless there's a trick I'm missing

Mr. Xcoder

Hehe I have 12 bytes in Pyth (first try)

I think the insertions might be longer in Jelly, indeed.

@cairdcoinheringaahing I think a rotation-based approach would work pretty well

caird coinheringaahing

5:31 PM

Insertion code

Only works for 3 inputs :/

Mr. Xcoder

Ugh... longer than the whole Pyth code

I guess it's extremely biased then ¯\_(ツ)_/¯

wizzwizz4

@eaglgenes101 Yeah, there was one that made valid JPEG2000.

*jpegs

I don't believe that iOS 9 autocorrect does that. :-/

Anonymous

I've bumped up the bounty offering on the Actually segfault bounty to 200 rep. I've also made another branch that calls you out for cheating if you try to use eval.

caird coinheringaahing

5:52 PM

@Mr.Xcoder ಠ_ಠ I have 41 bytes

Mr. Xcoder

@cairdcoinheringaahing Bloody hell

@cairdcoinheringaahing a wonderful definition of Outgolfed!

caird coinheringaahing

Answer deleted :/

Erik the Outgolfer

why? as long as you made a honest attempt at golfing it's all good

caird coinheringaahing

Usually I agree. But 41 bytes to 22 is just not worth keeping

New Meta Posts

1

Q: Why is this non-serious-contender answer still around?

As explained in our policy on serious contenders, this answer is not a serious contender because it is deliberately crafted to get a low score (the only way it could win is if there were no other entries). Several others seem to agree. As per our policy on invalid/unacceptable answers, I flagged...

discussion policy site-health specific-answer

Mr. Xcoder

6:10 PM

Yay someone doesn't underestimate Pyth and got 13/12 bytes \o/

Christopher 2EZ 4RTZ

@Mego done with your updated bounty

Anonymous

@Christopher2EZ4RTZ You do realize you have to post the code, right? :P

caird coinheringaahing

> The bounty will be awarded on this challenge to the first solution posted there

ಠ_ಠ Ninja'd

Christopher 2EZ 4RTZ

@Mego wait i was wrong

didn't notice that I can't just stack overflow

(That is a segfault btw)

well

Anonymous

It's not

Christopher 2EZ 4RTZ

6:15 PM

it causes segfaults

but dumb python won't let my ruin my computer :(

caird coinheringaahing

@Christopher2EZ4RTZ Do you just hang around the indefinite bounties page as a way to get rep?

Anonymous

While some environments may segfault as a result of a stack overflow, Python 3 (at least, on TIO, running Actually) does not

Christopher 2EZ 4RTZ

@cairdcoinheringaahing no

I post old sandbox for rep

(joke) kinda

I mean with the werewolf KoTH i am giving the winner most if not all the rep i got from posting it

Mr. Xcoder

So I was right thinking that posting Mafia was a rep-grab

caird coinheringaahing

@Christopher2EZ4RTZ Hmm, I'm still kinda salty about that, but I'll not get into a fight

Christopher 2EZ 4RTZ

6:17 PM

does anyone read the 2nd part of my message

Adám

Announcement: APL learning session in 12 minutes in the APL chat room.

caird coinheringaahing

@Christopher2EZ4RTZ Yes, but it doesn't stop it from still looking like a rep grap

Christopher 2EZ 4RTZ

@cairdcoinheringaahing I did put it in the challenge

And who cares what it looks like, I am giving 150 rep that I got for posting the challenge away

Anonymous

@cairdcoinheringaahing Accusing someone of trying to rep grab isn't really Being Nice. It's not too far from calling someone a rep whore (which is definitely not Being Nice).

Christopher 2EZ 4RTZ

@Mego can I use python flags?

Anonymous

6:23 PM

@Christopher2EZ4RTZ No. Not just that it's disallowed - it's impossible on TIO, because there's no way to pass flags to the Python interpreter.

caird coinheringaahing

What about if it can be done locally?

Christopher 2EZ 4RTZ

@Mego it is possible iirc, through bash

Anonymous

@cairdcoinheringaahing Doesn't matter. The bounty specifically states that it needs to work on TIO.

Anonymous

@Christopher2EZ4RTZ But then you wouldn't be using Actually on TIO - you'd be using bash + Actually

Christopher 2EZ 4RTZ

@Mego :(

caird coinheringaahing

6:25 PM

@Mego Oh, its for the Actually segfault. I thought he meant just in general

Christopher 2EZ 4RTZ

@Mego uhh in python 3 i don't think you can segfault

Pavel

@Christopher2EZ4RTZ Sure you can.

Christopher 2EZ 4RTZ

@Pavel everything i find points otherwise

Anonymous

@Christopher2EZ4RTZ There are plenty of ways to segfault in Python 3. Several of them involve C extensions. I don't know if it's possible to segfault using Actually (and I put that disclaimer in the bounty notice).

caird coinheringaahing

@Christopher2EZ4RTZ This may help you

Christopher 2EZ 4RTZ

6:27 PM

@Mego hmm

is there a brainfuck interpreter in actually?

Anonymous

@cairdcoinheringaahing The only answer there that works on Python 3 requires using the os module to send SIGSEGV to the process, which won't work in Actually (no access to os).

Anonymous

@Christopher2EZ4RTZ Built-in? No. Implemented? Yes.

Christopher 2EZ 4RTZ

@Mego where is the implementation?

Pavel

Do you know of a clever brainfuck segfault we're not aware of?

Christopher 2EZ 4RTZ

I plead the 5th

Dennis

6:29 PM

12

A: Shortest code that returns SIGSEGV

Python, 33 characters >>> import ctypes;ctypes.string_at(0) Segmentation fault Source: http://bugs.python.org/issue1215#msg143236 Python, 60 characters >>> import sys;sys.setrecursionlimit(1<<30);f=lambda f:f(f);f(f) Segmentation fault Source: http://svn.python.org/view/python/trunk/Lib/te...

caird coinheringaahing

@Mego I have a Whispers segfault, which is written in Python

Christopher 2EZ 4RTZ

Hello dennis

caird coinheringaahing

Dennis' linked answer works in Python 3 as well (answer just mentions Py2)

Anonymous

@Christopher2EZ4RTZ I posted it somewhere on the site. I'm having trouble finding it.

Christopher 2EZ 4RTZ

@Mego same here

Anonymous

6:32 PM

@Dennis The commentary on that answer implied that it only worked on Py2.

Adám

A bit late notification:

3

caird coinheringaahing

1 min ago, by caird coinheringaahing

Dennis' linked answer works in Python 3 as well (answer just mentions Py2)

Riker

@cairdcoinheringaahing how does that work?

caird coinheringaahing

@Riker Explanation

Infinite recursion basically

Anonymous

@cairdcoinheringaahing Yes, I read that. I was explaining why I (incorrectly) stated that it only worked in Py2.

Christopher 2EZ 4RTZ

6:41 PM

Try it online!

closest i have to braking it

Anonymous

A stack overflow in Actually is trivial

Christopher 2EZ 4RTZ

yeah

actually it really is

caird coinheringaahing

Most likely why it's not a valid answer to the bounty :P

Anonymous

When I first posted the bounty in chat, I also posted a 3-byte SO: chat.stackexchange.com/transcript/240?m=42020156#42020156

AdmBorkBork

Yeesh ... post my PowerShell answer which feels long, and there is a Python answer that is just about half the code length. :-/

Christopher 2EZ 4RTZ

6:53 PM

@Mego is it cheating to submit a PR and get it in without you noticing I opened a backdoor for segfaults?

Anonymous

@Christopher2EZ4RTZ Yes. Also I wouldn't accept it. Also I would disqualify you from the bounty.

figures, meh , :(

Also that's mean.

I broke the pipe

you dare to mess with the penguin?

Anonymous

6:55 PM

That wouldn't be you writing a segfaulting program in Actually - it would be you writing a segfault into Actually

Christopher 2EZ 4RTZ

wait what

AdmBorkBork

@Christopher2EZ4RTZ Don't break the pipe. I need that for coding practically anything.

Christopher 2EZ 4RTZ

the problem i see with segfaulting in actually is you have to work with python so there is not a way to disable the protections

@AdmBorkBork The pipe is down

eaglgenes101

I'm writing my own xz fuzzer in julia, but it's a slow process of typing what I think is right, then running it, then checking the docs to see what went wrong

Transcript for