« first day (2595 days earlier)      last day (928 days later) » 

4:00 AM
we have rust-like speed so you can use that to compare
Like I'm guessing all the parts that make C++ slower than C aren't in VSL yet. Exceptions, virtual calls, std::memory_overhead
How does VSL do memory management?
@Pavel mix of RAII/RC
If we can determine object lifetime to a certain degree of certainty we will not use RC but otherwise we will use some form of RC
Our goal is for zero memory overhead on objects though, that includes type ids, vtables, etc.
C++ does virtual calls a little weird but we resolve the stuff at compile-time so you don't have to do searching and stuff executable with RTTI
@Downgoat So like smart pointers. Does it support recursive references?
@Pavel C++ does exceptions weird, our exception model tracks 'unsafe' functions so it makes the programmer consider error handling (though it's pretty trivial) and this simplifies it quite a bit, you avoid all the longjmp BS
@Pavel yeah, we'll try to identify ref loops at compile-time and emit a warning but I can't tell you 100% if this always works
we have weak refs, not sure if we want an equiv for unowned yet
4:05 AM
I mean they are bascically the same I'll probably abstract them into type system
why did I always think this wouldn't work o_O idk I thought a if b else c was strictly an expression and not a statement lol
@HyperNeutrino tio.run/##K6gsycjPM/7/… it does
yeah I know that works but like still
@HyperNeutrino ... Because print is a function in Python 3?
yeah I know why it works now but I just never thought it was valid
because a for b in c isn't valid as its own statement
4:09 AM
@HyperNeutrino and how would you detect a generator expression without parens?
ehh good point
4:36 AM
While a user is suspended, why is it necessary to display the reputation as 1?
Maybe so they don't show up in the rep leaderboards?
@user202729 Removes their privledges.
@WheatWizard Well, that has to be done by other means anyway. With 1 rep, you still have the privelage of asking questions.
@WheatWizard changed avatar again
@Pavel I was just going to say that.
@Pavel Can suspended users not ask questions?
4:40 AM
@Downgoat This is called zero-cost exceptions, which C++ does support
Pretty sure they can't
@WheatWizard They should not be able to.
but it won't do it on -Os I suspect
Hm, ok then I don't know why rep gets lowered to 1
because they're expensive size-wise (need a big table)
4:40 AM
@WheatWizard What's with this trend of motion-blurred cat photos?
@WheatWizard how many cats do you have?
I have no cats. I'm a uni student so it would be really inconvenient to have a pet.
oh so the cats are other people's
I have 3
@WheatWizard You should get yourself a guinea pig or something. Guinea pigs are low-maintenence.
4:42 AM
well, i mean, they're not totally mine
@Pavel My lease has a no pets policy. Its not really about the maintenance.
I'm also likely to move frequently and that is not very good for pets.
Q: Why do suspended users have a reputation of one?

damodarI have some questions regarding suspended users: Why do suspended users have a reputation of one? According to the default reputation of a new user, he/she will have reputation of one and he/she has the privilege of posting questions. When comparing this, this is getting violated when compared ...

@quartata I'm talking in terms of semantics
4:56 AM
well you mentioned longjmp
GCC only uses that if it has no choice (incompatible ABI) or it's optimizing for speed
what you describe is much closer to how zero cost exceptions work
@Downgoat Does PPCG v2 work on Windows
wait nvm there's no bat
@ASCII-only ???? do you mean vsl
PPCG v2 is on web browser
Er sorry. I meant optimizing for size
Not optimizing for speed. Zero cost exceptions means no performance hit if an exception is never thrown
@Downgoat server
because look like very many thing to do so i might do PR
@ASCII-only oh, yeah, you need linux beacause redis doesnt work on windows
and redis is used for a lot of things
2 hours later…
7:25 AM
Is the plural of regex "regices"?
Asking for a friend
regexes or regexen according to Wiktionary
I tend to use regexes, but I have no idea if that's the proper term or even if there is one
regexen sounds weird
Sounds weird +1. Wonder who ever thought of that
Sounds like a bad forced German plural
Eh ¯\_(ツ)_/¯ thanks either way
8:01 AM
Regex regex regex. Now it doesn't sound like a real word.
Agree that regices is like a natural word though
A: List of bounties with no deadline

wastl100 reputation for a quine in Reflections A quine in Reflections would be really nice, but of course you may not use the quine function at (0|0). (That would not be a proper quine.) Please post your solution on Golf you a quine for great good!

In learncpp.com, I am reading that when you initialize an fixed size array, it's length has to be a compile time constant. But on sites like codechef or codeforces, I can even create fixed length arrays with taking it's length as an input in C++. Why ?
@AlexKChen Nonstandard GNU extension.
Search "VLA" for more information.
8:45 AM
I just read the indefinite bounty question, "200 for a segfault in Actually". I wonder if @Mego knew there is an extremely easy solution involving input from stdin, or was just "lucky" to add that prohibition.
8:56 AM
Ah never mind, Actually is Python 3 so the bug doesn't work.
9:42 AM
A: Sandbox for Proposed Challenges

SanchisesMinimal dot matrix adressing code-golf grid binary Consider a typical dot-matrix LED module. The LEDs are addressed in rows and columns. A LED will light up iff the voltage on the row R it's on is high (1), and the column C it's on is set to low (0). For example, 1 | 0 1 1 0 0 | 0 0 0 0 0 ...

10:07 AM
:/ There's no zypper plugin for oh-my-fish
Anyone here use oh-my-fish?
10:25 AM
Never mind, I had to copy oh-my-zsh's manually >_>
11:08 AM
@AlexKChen Even then, to avoid segfault I suggest allocate memory dynamically. Usingstd::vector.
11:43 AM
Q: There, I fixed it (with rope)

Kevin CruijssenRelated: Rather similar (but much easier) challenge horizontally: There I fixed it (with tape) Challenge: Given a string only containing upper- and/or lowercase letters (whichever you prefer) and new-lines, put rope vertically to fix it. We do this by checking the difference of two adjacent let...

12:06 PM
What is the time complexity of declaring int array[len]; in C++ ? Is it O(1) or O(len) ?
(Note that I'm not initializing it)
I'm no C++ expert but I would assume O(1). It's just a pointer move isn't it?
Declaring? Well isn't the space allocated at compile time
so O(0) :P
Yeah, I was just gonna add that :)
Big-O doesn't really apply to compile time
No I'm using gcc so len is taken as input
Q: Time complexity of declaring an array in C++

Alex K ChenWhat's the time complexity of the following code in C++: (note that I'm using gcc, so len is taken as an input from the user) int array[len]; \\array is uninitialized Is it O(1) or O(len) ? I am a bit confused.

@AlexKChen >_> oh in a function
@AlexKChen well of course it's O(1) still
you're not doing anything to the memory
@AlexKChen :/ that's not how big-o notation works
12:22 PM
@ASCII-only Why not ? I am new to big-O notation, can you explain ?
@AlexKChen wikipedia is a thing you know >_>
@AlexKChen you remove all constants in big o notation
if it depends linearly on length then it's just O(n)
@ASCII-only So where is the constant inside my big O notation ?
well len = n, isn't it ?
@AlexKChen sizeof(Foo) is a constant...
@AlexKChen and why would it be sizeof(Foo) at all
oh ok
ok ok
Also note: you shouldn't use VLA for arrays that are too large otherwise you'll run out of stack space
12:53 PM
@AlexKChen Although it's just pseudocode, I think //instead of \\ should be used for comment.
Also, use 4-spaces indentation for code block formatting. I think it should be used whenever the code is not inline.
1:04 PM
Also... why do you have the association bonus?
@user202729 He does?
@ASCII-only @AlexKChen .
@user202729 how do you get association bonus if you're only on one site
@ASCII-only Yes you can. But I don't think that's enough rep for assoc bonus.
hang on what is this magic
@AlexKChen why is your network rep 279
1:09 PM
Probably deleted account on some site.
Q: Inverse Gamma Function

Luis felipe De jesus MunozInverse Gamma Function The Gamma Function is an extension of the factorial function, with its argument shifted down by 1, to real and complex numbers. But, Can it be inverted? Your task is, given a positive real number, find the inverse gamma function. The result must be accurate to 4 signi...

@user202729 yeah but who said he had assoc bonus
Try searching for Alex K Chen on PPCG.
@ASCII-only "he"? ...
And why isn't it listed...?
1:11 PM
Because you can hide your accounts if you want to.
... Technically someone can search for all of them, there are only ~170 sites. But can you have different information on different sites?
Oh yeah >_> I was on that page recently IIRC
It's less effective if you use the same name and avatar.
reads answer on Alex's post hang on object arrays aren't set to null?
Of course, it's unintialized memory with unknown value.
1:15 PM
@user202729 Yes?
@user202729 "Save changes just for this community" is a thing, you know
@AlexKChen (BTW, saying "I have no idea" on SO feels like downvote magnet)
NMP deleted, I didn't even suggest the sandbox.
@feersum 1) Yes, I know about Python 2's bug. 2) Yes, Actually is Python 3, so it's avoided.
Why the restriction?
Because, if it's possible, it's possible without STDIN
What if there's another eval bug? :P
1:30 PM
@feersum Eval can't be used anyway.
It's possible, but all uses of eval are disallowed
What Python bug are we talking about?
I mean ast.literal_eval.
eval('(' * 999999)
(This bug also works with literal_eval).
Can't use that, either. It's only used in input with the -i flag. TIO doesn't use the -i flag, and there's no input allowed.
How does it normally do input?
1:32 PM
With regular eval
Oh wait, no, it always uses literal_eval
I forgot that I changed that
That's also not a segfault, but a stack overflow - a regular Python error
That's because the bug doesn't work with exec.
1:35 PM
Segfaults look like this: tio.run/##K6gsycjPM/7/…
@feersum exec isn't used at any point during that.
I'm referring to the exec mode of compile().
Perhaps instead of you trying to explain and me continuing to not know what you mean, you could post an example?
I already wrote the program that segfaults above. What else do you want to see?
1:40 PM
But eval('(' * 999999) doesn't segfault. Same error.
Oh sorry, I remembered it wrong. It's eval('()' * 999999).
There we go
Got a segfault locally
Though that doesn't say much, because my local install of 2.7 always segfaults on exit
1:44 PM
Part of the reason I moved Actually to Python 3 was because I was tired of coredumps continually accumulating
@Dennis It's a known bug with Python 2 on Cygwin
One more reason to migrate to Python 3.
Exactly the Cygwin maintainers' philosphy
I'm constantly amazed by the fact that it's still being used outside of PPCG.
"it"? Python 2?
1:46 PM
Migrating to Python 3 is a waste of time if you've got a large codebase.
We use PHP 5.4 on Ubuntu 14.04 at work :(
All versions of PHP are the wrong version.
You are very correct
Unfortunately, our use of PHP was due to a poor decision in the past, and now it's too much work to change it
1:48 PM
Oh, also, the PHP installation is single-threaded, and we run a decently-large website on it
Anyone know why OpenSUSE on WSL can't install mariadb? Like it works but when I try starting the server it says no such service mysql
Probably for the best, I mean would you really trust PHP to implement threading correctly?
@ASCII-only mysqld is probably the server
@feersum The multithreaded version just uses FFI to use libpthreads
Uh, that's hardly the only step to making the language thread-compatible.
Got to make sure there is no shared globals modified in the standard library, dependence on thread-hostile C functions, etc.
1:52 PM
Oof yeah, pretty sure the standard library abuses $_SERVER and $_SESSION a lot
does mysqld run the server as mysql
also: does this look right
and does anyone know how to run WSL headless
We're not tech support :)
It's easier to just assume nothing works right on WSL, and be pleasantly surprised when something does work
Yeah, I don't know why you'd bother with that instead of using a real VM.
2:20 PM
Q: There, I broke it (with scissors)

HyperNeutrinoChallenge Given a string describing a cutting rule and another string, cut parts out of the second string using the rule described by the first string. Both strings will consist of letters a-z or A-Z, whichever one you choose (they don't have to be represented the same way). The way the second ...

2:33 PM
fixed sorry
stack exchange should have git-like merge conflicts >:D
And M. Ender just got Reversal too on the same question I think
Currently I'm getting 36 bytes (Jelly) for We end up where we started. Anyone want to solve that too?
35 bytes.
2:58 PM
A: Sandbox for Proposed Challenges

ZgarbSwap halves to test divisors code-golf number-theory decision-problem source-layout Four integer sequences In this challenge, you will test four different properties of a positive integer, given by the following sequences. A positive integer N is perfect, if the sum of proper divisors of N...

3:08 PM
CMP: Should I learn Bash?
That depends. Which shell do you know now?
... Windows cmd. None on Linux.
I vote yes then.
(technically I can use Python with os.system calls)
You probably should.
3:12 PM
ϕ and φ are considered the same variable name in python
Yes, they are both phi.
but they're two different characters
I just realized something: refi64 made Hy!
3:21 PM
@user202729 no reason for X
@EriktheOutgolfer Yes there are. Char can't be passed as command line arguments.
(just try it and you will see)
Make it a function.
... yes, I was just going to say that.
But that's more of a flaw of Jelly (not allowing cmdline char).
@user202729 yes, learn at least the basics from a practical point of view
Then doesn't work.
(can I just use ?)
3:25 PM
o_O I just got Popular Question for my challenge about Markov processes and it's only been a day
@user202729 Yes, you can.
Looks like I can. But that's just because I was fortunate, generally it's impossible.
So, 33 bytes.
The value of isn't reset when you call a link. You can use it in monadic link you call, but not in dyadic ones.
I need to read the Jelly source
@totallyhuman it probably does unicode normalisation
@ngn Considering ϕ and φ to be the same variable name, yet a and а to be different, is a weird choice imo.
Q: One hundred passwords

YpnypnI have one hundred vaults, each one within another. Each vault has a password, consisting of a single number. 95 43 81 89 56 89 57 67 7 45 34 34 78 88 14 40 81 23 26 78 46 8 96 11 28 3 74 6 23 89 54 37 6 99 45 1 45 87 80 12 92 20 49 72 9 92 ...

I wonder if there's some sort of pattern to those numbers
If a way to express the sequence as an equation can be found, the solution for a bunch of various languages becomes possible
But the restriction... I doubt Java can do much with that many bytes.
3:39 PM
that first one looks more like an italicized psi than a phi
@Dennis both phis are greek, greeks consider them to be two forms of the same letter; the second a is cyrillic, it's got nothing to do with the latin alphabet; unicode is indeed a mess
interface A{static void main(String[]a){System.out.print( That's too long...
is markov convergence even possible with float precision
@Dennis I wonder what romans would say about ascii :) what?! u and v are different?
I'm more concerned about visible differences.
3:43 PM
33 so far. Stax is 24.
those two letters are definitely not canonically equivalent
As ÐĿ is "loop until repeat" (unlike Stack Cats {...}), and <dyad>¿ is "loop until dyad apply to adjacent is true", neither can be used here.
Yet another 33 bytes alternative, again. So many alternatives without improvement.
@totallyhuman What is horrifying?
so many 0s
3:51 PM
i can't do this challenge, machine numbers are bad ;-;
@HyperNeutrino can i assume numbers are equal after a certain amount of decimal digits
"this challenge"?
@totallyhuman No.
Q: Convergence of a Markov Process

HyperNeutrinoChallenge Given a left- or right-stochastic matrix where the limit as x approaches infinity of the matrix to the power of x approaches a matrix with all finite values, return the matrix to which the matrix converges. Basically, you want to keep multiplying the matrix by itself until the result n...

det(A)=1. Correct?
@totallyhuman no but if there are floating point "issues" i'll pretend i didn't notice
most of them do "m-multiply until equal" which never really works without floating point inaccuracies so I'm allowing that
@Dennis Why don't you use æ×`ÐL?
3:59 PM
Because it doesn't work.
(interesting behavior.)
i dunno why it ends up at 0
Floating point issues?
well yeah that's the generic term for every issue with this code :P
4:01 PM
Wait a minute the determinant isn't even 1....
So in that case det(B)==0.
I wonder why function(m){ while(any(m!=(m=m%*%m)))0 m} doesn't work. Numerical inaccuracies preventing the termination condition from triggering? — CodesInChaos 7 hours ago
[64th application](https://tio.run/##NcixDoIwFEDRna94JAytoYQEgglpv8HRoXYgWORhiw08FsO3W3Fwuid37NandS4O6ha3mdApKRlTih/lGSVWadPahMArrdfNZ28MV6SRnTjCtE9SDC93X9h/t5xb8GZHKbyJvsMZFIQFZ4IMWJpCU/NDSHbpyAKB1mVxzqEsapPD4ernxpj46QfXPdYo@hCiuFRf "Haskell – Try It Online") vs. [65th](https://tio.run/##NcixDoIwFEDRna94JAytoYQExYS03@DoUDsQLPKwxQYei@HbrTg43ZM7tMvTOhd7dYvrROiUlIwpxffyjBKrtGlsQuCV1svqszeGK9LADhxh3EYp@pe7z@y/G84teLOhFN5E3@IECsKME0EGLE2hPvFdSHZuyQKB1mVxzqEsjiaH3dXPtTHx0/WufSxRdCFEcam@ "Haskell – Try It Online")
@totallyhuman Can't you use a less exact type?
It will never converge that way.
wait less exact would be better?
Because it will repeat eventually.
(understand "better" as in "take less bytes" here)
ok lemme check if haskell has one
4:05 PM
But why...... doesn't ` work?
--> numerical error. Everything is the fault of numerical error.
(actually that's because the algorithm is not numerically stable)
So, the given matrix has all elements >= 0, sum of each row = 1. Does that say anything about the determinant?
(or is it?)
Volume of nD paralleloid... looks like it's true.
... why can't I prove it... it seems really trivial. Can anyone hint please?
4:25 PM
A: Sandbox for Proposed Challenges

simonalexander2005Background I saw this very interesting Venn Diagram on wikipedia: https://simple.wikipedia.org/wiki/Alphabet#/media/File:Venn_diagram_gr_la_ru.svg It shows the letters (physical shapes) in different alphabets that are common between Russian, Greek and Latin alphabets. The Challenge Given an i...

(But it won't work with the identity matrix)
So I want to answer a question with ALPHA (The WIP language I made earlier) because it fit perfectly with the question
Is this allowed under PPCG rules?
if the version of the language post dates the challenge, then the answer must be non-competiting
actually...didn't we do away with that?
@NathanMerrill Yes, new languages compete normally now.
(but if it specializes for the challenge, then people will definitely downvote it)
so, yeah. A resounding "yes" :)
4:28 PM
Yeah its post the question date
Ill just make it non-competing
That good?
no need to :)
the only reason we made that rule was to get rid of stuff like Metagolfscript
Some of the ascii stuff was added to the planned command list a little bit before I found the question
But you still can add a "language postdates challenge" note, if you feel like.
4:29 PM
Still ALPHA?
Its kinda frozen in development while caird is doing his exams
I fixed the readme and added some new commands for set 2
Maybe I can get a little bit of publicity and get some bug testers and forks to complete the interpreter
My Spring Break is now', I should probably work on Jagony some more.
Was it you that named it?
4:33 PM
Named what?
Not sure
Ive been wrestlin- I mean learning arduinoC
I have to collect my self for a couple hours before I can write good code in C++ right now
Quirks of learning a lang i guess
@totallyhuman Would you mind choosing a crack from TPLQ2 that you'd like me to accept? (Most impressive/difficult, or something like that)
sifted through the entire quark interpreter to find a bug, only to find out he added a ! into a single statement by mistake
Hey ... I feel you. After programing APL for a bit, my coding style in my other languages dropped significantly.
4:36 PM
@cairdcoinheringaahing oh sure
Well Ive not been affected by golfing as much as others
Its just you can get carried away with proposed "solutions" in c++ and write code that is hot garbage
Well, now i've fixed it, and QUARK now has a working truth machine \o/

Jelly almost definitely beats this, but who cares, i made it work :P
@cairdcoinheringaahing this one would be nice
Time to finish this little bit up, push it, and submit my first answer in QUARK (Just a truth machine for the catalog :P)
@Zacharý same
4:41 PM
I published by first alpha answer
@J.Sallé I should really start golfing some more in APL, I just haven't found a challenge that I can do that hasn't been Adám'd nor ngn'd
It was a Hello World with semi-common chars
@Zacharý I feel you, bro. They're just 2quick4me
@J.Sallé Yep. Guess I'll wait until the student competition to get destroyed by a little kid program some useful APL.
What even is APL?
4:44 PM
A Programming Language
The programming language named "A Programming Language".
A programming language name.
(this can quickly gets confusing)
join us
I'm probably going to have to rewrite the entire JagonyObject class... :(
@Zacharý indeed, I can't wait to get utterly outgolfed by a 10 year old.
@totallyhuman Accepted :P
@J.Sallé Not outgolfed for the student competition.
It's based on good APL code, and he'll be 12 at that time. He was ~11 when he last competed.
4:49 PM
Anyway, can't wait :p
@J.Sallé @Zacharý OK, maybe I should just make a note of the challenges I'd want to solve, and hold off to give you guys a chance.
Nah, it's fine... for one challenge I wish TAO would come now
I never thought I'd say this, but can we have some more Low Quality Posts? I want something to review :/
Why do you need to rewrite the jagony object?
@Zacharý Which one? 17.0 has been given to some customers for alpha-testing already…
4:55 PM
@cairdcoinheringaahing one I get 20 more, I'll stop reviewing them :P
@Adám nah, I don't mind. If I feel like solving something you've already solved I'll just make a tradfn since it's what I know best :p
@Adám Block sorting or something like that. F
@DJMcMayhem posts a bunch of spam Can you stop reviewing them now? :P
spam != VLQ :P
@Adám btw 17.0 will have builtin sorting, right?
4:58 PM
@Zacharý my latest is array-friendly, it hasn't been Adám'd yet and it won't be ngn'd
@J.Sallé No, but it can grade everything, so (⊂⍋A)⌷A will sort any A without refs.
@ngn Nor answered in Jelly as of now...
@Adám ah, I see
I am officially overly proud over the fact QUARK can now do a truth machine
A: Implement a Truth-Machine

moonheart08QUARK, 11 This is 6.75 bytes in QUARK's encoding, but that's not finished yet, so I'll score this in UTF8 until then. i‶{p}⟲ As QUARK became capable of doing a truth machine (implementation wise) less than a minute ago (from when I started writing this answer), it obviously needs an explanati...

Do the tables in the snippet in the Hello world challenge overlap for anyone else? (I'm using FF 58.0.1 on Linux)
5:13 PM
QUESTION: Would i get issues if i judge my QUARK entry in the QUARK encoding (which has decoders in the interp), but i don't have the encoder ready yet?
I'd say yes
@NieDzejkob Yes, they overlap for me (Chrome, Windows)
@moonheart08 If the interpreter produces the intended behaviour when fed the specified binary file, then you can claim it's encoded in that encoding
Mk. So i just need to actually add the flag so the interpreter can read it :p
(It reads unencoded files by default, to make lives easier)
A: Sandbox for Proposed Challenges

DatOops, I blew it up (with numbers) Inspired by There, I fixed it (with rope) and There, I fixed it (with Tape) Input : abcdefgklmnopqrstuvwxyz Output : abcde1234321mnopqrstuvwxyz Why? Between the g and the k, it is missing hij. So i, being the char in the middle is the epicenter of the "expl...

5:32 PM
@user202729 Do I remember correctly that you succeeded in running the Esoteric IDE on Linux? Which version of Mono did you use?
I finally submitted an APL answer.

« first day (2595 days earlier)      last day (928 days later) »