The Nineteenth Byte

ASCII-only

12:00 AM

@Pavel still, there's a comment on every line

@Pavel this is clearly much better

Pavel

@ASCII-only That's nothing, golflangs get comments for every character. :P

ASCII-only

@Pavel those are explanations

except for regexes

Pavel

@ASCII-only Those comments show literally what the code is doing, not explaining how this demonstrates async/await, which is the point.

ASCII-only

@Pavel >_> it wasn't serious, i was trying to make the comments redundant

Adám

@Pavel APL equivalent.

Pavel

12:07 AM

@Adám Nice

Adám

@Pavel This uses time slicing of a single CPU thread. The equivalent using actual CPU threads requires the isolates' mechanism. Unfortunately, we've not activated the glyphs for that. Once we do so, it'd mean replacing & with ∥ (parallel — get it?) and removing ⎕TSYNC. Meanwhile there are cover operators to use instead.

Pavel

@Adám Does ∥ do anything else currently?

Adám

@Pavel No. ¤ and ∥ are just reserved for "isolate" and "parallel". They currently give syntax errors.

xfix

"This question has more than 500 answers already."

i just wanted to shorten Golang Hello, world solution from 61 to 48 bytes...

so many solutions for hello world already

Adám

@Pavel The covers (poly-fills, if you want) for ¤ and ∥ are called ø and ll.

xfix

12:16 AM

oh wait, i realized a mistake

ASCII-only

@xfix :|

link to TIO pls

xfix

which wasn't obvious running a script manually, print function for some reason writes to STDERR, which is completely undocumented

Adám

CMC: π⁶ (pi^6)

xfix

that said, for your bounty, four bytes can be saved easily

package main
import."fmt"
func main(){Print("Hello, World!")}

Pavel

@Adám Mathematica: π^6

xfix

12:18 AM

by using dot imports

tio.run/##S8///…

the optimized version is this:

Adám

@Pavel Heh, on WolframAlpha you can just enter π⁶ :-)

ASCII-only

:| charcoal doesn't have arbitrary precision pow yet

xfix

package main
func main(){print("Hello, World!")}

but that outputs to STDERR

Pavel

@Adám That works in Mathematica too, but it's still stored internally as ^.

mudkip201

@Adám Pyt: π⁶

xfix

12:20 AM

π⁶ is a valid Perl 6 expression by the way, not that it matters

it wasn't back when i started using this avatar, but now it is

Pavel

Oh, you mean with the literal character for the superscript ⁶. That also works.

xfix

i mean, a superscript 6

xfix@hack:~$ perl6 -e 'say π⁶'
961.389193575304

mudkip201

neat

Adám

Wow, APL is so verbose on this one: 6*⍨○1

ASCII-only

@xfix that's superscript

xfix

12:22 AM

right, corrected

i sorta take blame for this working in Perl 6, but i mean, it makes programs shorter so whatever

mudkip201

does subscript do anything in Perl?

does π₆ do anything

xfix

i don't think so

ASCII-only

no

xfix

however, you can use ₆ as a number

mudkip201

... well then

xfix

12:25 AM

$ perl6 -e 'say ₆'
6

not too interesting

but if a challenge disallows using ASCII digits, well

PhiNotPi

That's not a Perl 5 thing, is it?

xfix

no, it's Perl 6 only

it's pretty much full on Unicode language, trying to support Unicode as well as possible

Adám

Too bad. Subscripts could conveniently have been, well, subscripts (i.e. indexing).

xfix

actually, i sort of agree, it may be an interesting idea to have a support for this

totallyhuman

Haskell, 4 bytes: pi^6

Downgoat

12:28 AM

@ASCII-only I assume that's the ritual where the goats are given lifetime supply of tin can from ellen

ASCII-only

@Downgoat that must be a very weird ritual O_o

Anonymous

@Downgoat ...oh, you're still here. Something must've gone wrong with the ritual.

xfix

also, this is a quite obscure feature, but whatever

$ perl6 -e 'say 2²⁺²'
16

totallyhuman

CMP: if i wanted to call the nth function in golf code, would it be more efficient to use superscript numbers or just push the number and have a command call the TOSth function

PhiNotPi

Hour 4 Post-Election: They still believe I'm one of them.

old meme alert

Downgoat

12:31 AM

@xfix @ASCII-only we should add this to VSL

mudkip201

@totallyhuman I'd do the latter. Honestly, probably is less work to implement. But if you're looking to actually make the code the shortest possible, then maybe the former is better

totallyhuman

shouldn't be too hard, i already have the parser to work for it

Downgoat

also using things like ≠ for !=

ASCII-only

@Downgoat yes

@Downgoat and times and divide and other math symbols and unicode fractions

totallyhuman

@mudkip201 yeah, it'd be golfier but is it worth it to lose 10 characters is the question

mudkip201

12:33 AM

@totallyhuman I'd say probably not. I mean, if you find you don't need 10 characters, then you can always go back and make that change later

xfix

it's actually crazy how much there is special support for Unicode in Perl 6

ASCII-only

@totallyhuman no. definitely not. unless your language is based around calling functions, in which case definitely

totallyhuman

mm ok

xfix

like Perl 6 had « and » since beginning, then there was ｢ and ｣ for raw strings, later Unicode set operations were added, i made a module adding tau and τ constants, somebody noticed it and decided to add it straight to the compiler along with π and 𝑒, and then quite crazy things were added that i don't even know ordering of

totallyhuman

like using the atom emoji for atomic things

Adám

12:37 AM

@xfix Is there a summery list somewhere?

xfix

this i know was very late after all this stuff

Neil

@orlp I would have written 3P = 1 + P + P^2 which is readily rearranged into (P-1)^2 = 0 of course.

xfix

docs.perl6.org/language/unicode_ascii

but it's not even complete

like you can use Unicode fractions for instance

also see codegolf.stackexchange.com/a/112149/3103

totallyhuman

docs.perl6.org/type/atomicint

okay first of all, why do perl docs look so childish

xfix

yeah, the docs are rather poor, and it's mostly a reason why Perl 6 isn't really used in practice

and there is a lot of stuff in a language, it's probably the biggest language ever created

bigger than C++? yes

moonheart08

12:41 AM

How to hurt someone's brain:
Hide a user defined literal deep in your codebase

xfix

(standard library that is)

moonheart08

I've been examing the codebase of a friend, and a user defined literal hurt my head

xfix

i actually wonder, is there a language with more features built-in into the language than Perl 6

moonheart08

probably not

Doesn't Perl 6 have an entire syntax parsing functionality?

xfix

yes, there are grammars

in fact, Perl 6 language itself is defined as a grammar, allowing for modifications

moonheart08

12:44 AM

Oh god. Perl 6 is a self describing meta language, then

More like an esolang than a real language

brb waiting for people to start golfing with it

xfix

i mean, i did a lot of code golfing in Perl 6

ASCII-only

@moonheart08 People have done that for ages

xfix

although grammar modifications aren't all that interesting, they are quite verbose

making them unviable for code golf

ASCII-only

What makes you think any language hasn't been used

moonheart08

i haven't seen any answers, so :P

xfix

12:46 AM

at this point i'm surprised that somebody actually used a programming language i made just as a puzzle for myself whether it's even Turing complete

and published on esolangs because i felt like documenting it somewhere

moonheart08

Which language?

ASCII-only

@xfix wait what language

xfix

GPRX 3000

i actually found a single solution in it on codegolf SE

pretty much a Turing tarpit, except i don't know if it's even Turing complete

still didn't figure it out

Anonymous

I use PHP 5.4 at work, and it's awful :(

2

ASCII-only

I'm surprised there's someone using the language I (co-)wrote even though I don't update it for 2+ months every year because of summer holidays

xfix

12:48 AM

like i cannot prove it's not Turing complete, but i cannot prove it's Turing complete either

ASCII-only

>_> why did ~~people~~ dzaima have to start writing ascii-art languages only after Charcoal was basically done

xfix

esolangs.org/wiki/GPRX_3000

essentially a stack based machine limited to 3 elements

however, the elements are infinite precision integers, so you could in theory store memory here

and it cycles

Pavel

@Mego Why do they use outdated PHP?

Anonymous

@xfix You can do conditional loops via arithmetic and jumps, and you have arbitrary-precision ints, so it should be TC

Anonymous

@Pavel Legacy code

moonheart08

12:51 AM

@xfix That's neat. I'm going to look over it

Anonymous

Updating to PHP 7 will break a ton of stuff, and we don't have the dev resources to fix it

moonheart08

Pretty sure it's possible to fake a bitshift..

Destructible Lemon

maybe i should try and prove tcness

I think you could prove it tc in the way you prove a minsky machine tc, except it has multiplication and stuff builtin

except, can it be checked as easily as in a minsky machine?

ASCII-only

@moonheart08 rx2* or something like that?

moonheart08

Maybe. If you can perform a binary rotation, you can easily access the infinite memory as a 'tape' of 1 and 0

from there you can make a boolfuck conversion.

Pavel

12:55 AM

@Mego Did PHP 7 really break that much stuff?

quartata

@Pavel oh god yes

moonheart08

But yea, it's almost efinitely turing complete. I'd be suprised if it's not

Destructible Lemon

i have an idea

quartata

@xfix has moarvm gotten any faster

xfix

i honestly don't know, i didn't write bigger programs in Perl 6

moonheart08

12:57 AM

Does ELVM have JIT, or is it just an interpreter?

quartata

@moonheart08 neither

it's AOT

like normal LLVM

Destructible Lemon

you can check if somthing is divisble by 2, with the modulus. this allows you to use a power of 2 counter

moonheart08

oh

Destructible Lemon

through a similar mechanism, you can probably check if something is divisible by 3, but you have to have a case for modulus being 1 or 2, not both of them

Pavel

@quartata As far as I can tell, yes, but it's still crap.

Destructible Lemon

12:58 AM

etc.

quartata

@Pavel !remindme 10 years

Dennis

@moonheart08 elc is a compiler, eli is an interpreter. No clue if eli is AOT, JIT, ATL, or whatever.

Destructible Lemon

the only issue might be overlapping parts of the destinations with the whole: jumping by values rather than exactly conditionally, but i think that can be worked around by arbitrarily long jumps or something

actually maybe the issue of the registers always being in the same order

ASCII-only

> Probing output cache...

xfix

i'm fairly certain with 4 registers it is Turing complete, it's not with 2 registers

but 3 registers, i have no idea

moonheart08

1:00 AM

3 cell brainfuck, @xfix

Anonymous

@Dennis No idea what ATL means, but eli is purely interpreted, statement-by-statement

moonheart08

3 cell brainfuck is turing complete, it has no negative numbers and 3 infinite registers

Neil

@ASCII-only Ｍ⁵↑ＵＯχ*Ｍ⁵↘Ｔ³¦³# and then change the up arrow to left, right or down to see what you should get

Destructible Lemon

@moonheart08 but that doesn't have the issue of needing a register for jumps and stuff

Dennis

@Mego Always Too Late

moonheart08

1:01 AM

Good point

Anonymous

Lol, never heard of that one

ASCII-only

atl.com.au

moonheart08

I mean, it can't be indeterministic, so that's not an issue. It's either turing or it's not

ASCII-only

> ATL Translation and Interpreting

Dennis

@Mego Take any programming language that takes 10+ seconds to run Hello World. That's ATL compilation.

Pavel

1:02 AM

@Dennis Is that a joke or a real thing

@Dennis Ah, so just Kotlin and Ceylon

Dennis

And F#.

ASCII-only

and .NET with VS and a terrible computer

quartata

F# and Kotlin are plenty fast

ASCII-only

~~and depending on your luck (or lack thererof), Jelly~~ wait it's not compiled

quartata

you can't include compilation time

xfix

1:03 AM

i wonder how slow those compilers are

quartata

@Pavel that includes compilation

Anonymous

@Dennis So Python + Sympy?

xfix

like Rust compiler is well known for being slow, but Hello World takes 1 second to compile

quartata

obviously that's slow, both have very complicated type inference systems

Pavel

@quartata We're literally comparing compilation methods

Dennis

1:04 AM

@quartata Compilation time is precisely what I'm talking about.

Pavel

@Dennis I'm getting 3 seconds for F# on TIO

quartata

I see you've drunk the clang kool-aid then

@xfix what are the registers initialized to

Dennis

The Mono one is fast. I meant .NET Core.

xfix

0

Pavel

Ah

xfix

1:05 AM

but it doesn't matter because you can set at beginning whatever value you like

as there are integer constants

quartata

oh, I didn't notice that

ASCII-only

@Neil Thanks (kinda distracted right now so it'll take a while to get fixed :/)

quartata

it's TC

Οurous

this is an irritating bug. I was trying to see how long a loop would fit in the heap size for TIO, and instead discovered that multiple-digit integers don't work properly. Change it to a 9 and it works fine

xfix

1x2x3x

sets register values to 1, 2, 3

ASCII-only

1:06 AM

@Οurous :

@Οurous excuse me. pls add wiki ty

quartata

use one counter as scratch for multiplications

one for doing jumps

and one for the actual memory (one-counter multiplicative)

pretty sure it checks out

Οurous

@ASCII-only once it has enough functionality to actually be useful for anything yeah

ASCII-only

@Οurous ok so how to do 9 + 9

quartata

@Οurous ha, is this written in clean

Οurous

@quartata yeah

quartata

1:08 AM

dirty in clean

ASCII-only

also what is all this mirror/clockwise stuff

Destructible Lemon

@quartata the only issue is retrieving the stuff, due to how calculations only go one way

Οurous

@ASCII-only 9 9⭧+. Or later on, 9 9∑

@ASCII-only It's 2D.

ASCII-only

> later on

Οurous

@ASCII-only "in active development". It didn't have loops three days ago

quartata

1:09 AM

yeah pretty sure using one of them for the state register works

Οurous

Well on the plus side it's just a floating point issue.

._.

ASCII-only

@Οurous wait wat

Οurous

@ASCII-only yeah that's what I thought, it's not an accuracy issue, it just only manifests when the underlying value is a Real

Pavel

1:28 AM

@Οurous You're using -- for short flags. - is for short flags, -- is for long flags. e.g. -f vs --force.

Also, -ns is the same as -n -s

ASCII-only

1:58 AM

:| using a terrible uni computer atm

backspace gets stuck so often

it's really annoying

Οurous

@Pavel I'm actually using - for flags to the interpreter, and -- for flags to the runtime of the language itself, but I'll take heed of your wisdom. Is there a manual for this kind of thing? Because I couldn't find one. And believe me I've looked.

Anonymous

@Οurous gnu.org/software/libc/manual/html_node/Argument-Syntax.html

Οurous

@Mego Well crap... there is a manual. Thanks.

Anonymous

Credit to this SO answer for pointing the way

Dennis

There's a manual for everything. The question always is how many people can be ~~tricked into~~ convinced of following it.

Anonymous

2:12 AM

@Dennis So my mod manual is in the mail?

Dennis

6 to 8 weeks

moonheart08

@Mego No, it's on [obscure website from the 1990s]

It gives full details on how to moderate a under construction site

Anonymous

@moonheart08 Got it, going to Geocities

Anonymous

@Dennis Somehow I knew this would be your response :P

moonheart08

Dennis is hardcoded, you can find his responses by reading his sourcecode. gl getting it tho

it has a gigantic license agreement that traps you in a "friendship" and/or "weirdo" contract

It can potentially get you locked in a "arrested" contract

Anonymous

2:16 AM

@moonheart08 Oh, I thought he was written in Jelly

moonheart08

Nope, he's written in FORTRAN, with some weird macro based language intermixed

no-one really understands the macro language

mudkip201

@moonheart08 Except for him, ofc

moonheart08

true

Dennis

Heh, I still have no clue why the Fortran Hello World needs print'("Hello, World!")'.

Sorry if that killed the mood for robotification.

Ørjan Johansen

@quartata I think this fits very nice with Collatz functions

Anonymous

2:28 AM

@Dennis Nah, I'm just pressing all of the new buttons to see what they do

Dennis

Careful, some buttons can only be pressed once.

Anonymous

nukes own account oops

quartata

@Mego have you found this button yet:

in Tavern on the Meta on Meta Stack Exchange Chat, Aug 26 '10 at 18:14, by Shog9

Mods can also come into your house at night and leave your milk out. Some mod abilities aren't talked about much.

I think it's under tools?

or is it on the profile

Anonymous

By "pressing buttons", I mean "cautiously looking at all the buttons and deciding not to push them in case I break something"

Anonymous

@quartata Found it. Put your milk back in your fridge tomorrow morning.

Dennis

2:38 AM

If only there was a button for that.

Riker

@Mego merge-completed :D

Anonymous

@Riker Huh what?

Dennis

@mınxomaτ Got mine today. It's seriously incredible.

Somewhat related: DHL On Demand Delivery emails are almost indistinguishable from spam.

Riker

@Mego you can nuke your ppcg design patch branch now

Anonymous

Oh

Riker

2:44 AM

I'm just highly satisfied with merging in under a minute on my computer

yay for github not delaying notifications

Anonymous

I noticed that issue earlier, but only had time to fix it now

Ørjan Johansen

@quartata xx0+x42x%xx999x*g lets you branch based on a modulus 42 with distance 999 between branch codes. Then a branch code can look like 42x/x555*x32+xx0g (to continue; to halt replace the last 0 by program length).

(Numbers made up, of course.)

I should add that to the wiki.

New Main Posts

0

Q: Print the numbers -1, 0, 1

This is a very simple challenge, and it is to print the numbers -1, 0, and 1. The following paragraph is credited to Print numbers from 1 to 10 Your output format can be whatever your language supports. This includes arbitrary separators (commas, semicolons, newlines, combinations of those, etc....

Οurous

2:59 AM

Windows Defender, I'm 100% sure that the heap debugger I'm trying to get running is not a virus.

Dennis

From behavior alone, debuggers are probably indistinguishable from malware.

Pavel

> Make useful heap debugger
> Secretly insert virus
> Everyone adds exception to antivirus for my program
> Virus runs while no one is suspicious because heap debuggers are supposed to act like that
> ???
> profit and/or arrest

Οurous

I mean it would be applicable if this particular one couldn't hook into applications, but just instead analyze raw memory dumps.

PhiNotPi

@Οurous [Insert story about how I got Perl quarantined once.]

Dennis

@robbie0630 Welcome!

robbie0630

3:08 AM

hi!

i was just curious if anyone could solve my challenge with 2 bytes or less

Dennis

M can print [-1, 0, 1] in just two bytes: -R

- is a shorthand for -1, R creates a range from -n to n for a negative argument n.

robbie0630

nice!

user202729

Uh... so who became mods?

Dennis

7 hours ago, by totallyhuman

Congratulations to DJMcMayhem and Mego, our new moderators!

Time to reboot my computer. See you in 10.

Anonymous

3rD should make [-1, 0, 1] in Actually, but I foolishly decided to overload Decrement with stddev

Ørjan Johansen

3:14 AM

@quartata wiki updated.

Zacharý

Mego's a mod? gulps

And we all saw DJ (the V guy) coming

Ørjan Johansen

(GPRX 3000 TC proof)

Anonymous

@Zacharý I'm just your friendly neighborhood penguin. Except now I have a nuke button, in addition to my steel-flippered boots.

Zacharý

You're just the un-punny guy.

Anonymous

I'm not against puns. I'm just against the use of puns to annoy people, rather than to enhance a conversation.

PhiNotPi

3:18 AM

Phi see what you did there.

Zacharý

Ah. But you ARE the one who kicks most for puns.

Anonymous

@Zacharý The identity of the RO who issues a kick is not public knowledge. The only way others know which RO did the kick is if the RO says so. Your accusation towards me is baseless and rude.

PhiNotPi

Random question: would people be interested in unfreezing The Palette? (chat room for PPCG's artists)

Zacharý

Well, I inferred from the fact that you are the common RO from most of my pun-based kicks. It wasn't baseless, but it was rude.

Anonymous

The point of kicking is to inform a user of unacceptable behavior and to give them a moment to think about their unacceptable behavior, so that it isn't repeated in the future. And, after the kick, it's expected that all parties move past the incident, just like suspensions. Singling out a RO for assumed kicks and holding a grudge is not moving past it, and not Being Nice.

Anonymous

3:25 AM

Also: kicks can be issued without being in the room. All you need is to make a simple POST request with an authorized user token. The presence or absence of ROs does not correlate with kicks issued, because neither browsing the transcript or kicking a user makes a user appear in the room.

HyperNeutrino

@Zacharý If you're basing it off of which RO is at the top of the list the most often when you get kicked, that's completely baseless because I can be in chat for 10 hours and just not speak and still show up at the bottom even if I'm actively observing.

Anonymous

@PhiNotPi Ooh, that's a button I haven't pushed yet. If there's interest, I'd be happy to unfreeze it for you.

PhiNotPi

@Mego Go ahead.

Unfreezing rooms is like 90% of the job of mods.

Anonymous

It's now unfrozen

PhiNotPi

The Pallete

For the artistic side of programmers. Both music and visual arts.

Dennis

3:28 AM

Nah, 90% is dealing with comment flags, 10% is elevening messages after the grace period.

Anonymous

@Dennis I thought it was 100% editing feed messages to make inane statements about butter :P

user202729

... which grace period?

Dennis

2 minutes for chat messages, 5 minutes for comments.

Ørjan Johansen

What's elevening?

Dennis

Mod abuse.

Conor O'Brien

3:32 AM

Apr 6 '16 at 23:03, by Dennis

♫ I have eleven problems, but mod abuse ain't one. ♫

user202729

(That was on the deleted meme page)

Zacharý

I thought it was just usage of mod powers

Dennis

@user202729 It was not. Unlike the other memes that were either self-explanatory or nobody would care about, this one never made it.

Anonymous

@Zacharý Well, any mod power use is jokingly referred to as "mod abuse" (with an optional "eleven"), because we don't actually have any history of actual mod abuse as far as I know :P

Zacharý

@Dennis You definitely made it on the meme page more than once.

Dennis

3:35 AM

I'm painfully aware of that. :P

Zacharý

One does not simply ... outgolf Dennis

Conor O'Brien

:o what if deleting the meme page was mod abuse so dennis could preserve his character /s

Pavel

Is there a way to undo a commit that doesn't involve git push -f

Dennis

You could delete the repo.

Pavel

@Dennis I've done just that before, actually.

Zacharý

3:38 AM

I have a totally unrelated question. Would revealing a post that has personal info be "doxing"? I want to know the meaning further.

DJMcMayhem

Yes, doxing means revealing personal info

Pavel

damn gitlab not allowing git push -f to master

Anonymous

@Zacharý "doxing" is typically revealing personal information with malicious intent. You shouldn't reveal personal info regardless, but you may not be treated as harshly if it's on accident.

Pavel

@Mego Did you stop being an RO when you became a mod or does it just not render blue and italics at the same time

Anonymous

@Pavel The second one

Dennis

3:44 AM

The latter. I'm also a RO.

Anonymous

Moderator status completely supersedes RO status - there is nothing a RO can do that a mod can't. So, it's not really needed to display RO status alongside mod status.

Anonymous

Also mods can add and remove ROs at will, even without being an RO, so it's extra pointless

HyperNeutrino

question: how do I run arbitrary programs locally without risk of my computer being broken by whoever submits the program

Anonymous

@HyperNeutrino Using a VM mitigates a lot of the risk - a program would have to exploit vulnerabilities in the VM to escape

HyperNeutrino

hm ok. I was considering using TryItOffline but I can't get docker to work and I also don't really know what that is anyway

Anonymous

3:46 AM

Using a Linux VM with SELinux or similar would make it even harder to exploit vulnerabilities

HyperNeutrino

hm ok

Anonymous

Docker is like a VM on steroids. Rather than being a usermode application that can do some hardware-accelerated stuff for speedup, it runs in a hypervisor, which is much closer to the OS, so you get even more speed

Pavel

@HyperNeutrino TIOffline is horribly outdated anyway

Anonymous

Also Docker instances are much faster to spin up than VMs

HyperNeutrino

oh okay

Dennis

3:47 AM

@Mego Well, a mod on that site, anyway.

Anonymous

@Dennis There are other sites?

7

HyperNeutrino

so which would you recommend then? basically i'm creating an automated contest grader

Pavel

So, for example, Riker can't make someone an RO here?

Anonymous

@Pavel Correct

Pavel

Or pin messages/change room description

Dennis

3:48 AM

I think he can do all three.

DJMcMayhem

I think so too

Just not on meta or SO sites

Pavel

So what can't a mod from another site do?

Dennis

Enter private rooms.

Anonymous

@Dennis Oh, you're right. I was looking at CHQ, which is on the MSE server :/

Anonymous

Interestingly, I can pin messages, cancel stars, and edit the room description of CHQ, but I can't add people as ROs

Pavel

3:50 AM

I just experienced the opposite of "Works on my machine!", we have a script that works on everyone's computer except that of the developer that wrote it.

Anonymous

Oh wait, yes I can. It just has to be someone who isn't already a mod

Anonymous

So yeah, mods can do literally everything on other sites except enter other sites' private rooms

ASCII-only

@Pavel how O_o

Dennis

@Pavel My favorite kind of bug is the one that goes away when you add a print statement for debugging.

Anonymous

@Dennis I call those Heisenbugs

Downgoat

3:58 AM