The Nineteenth Byte

@ETHproductions ah the stats say no

@ETHproductions how old are you? (If you don't mind my asking)

Rainbolt

A 25 year old block of sand, I am.

We all treat each other equal, I like that

@ETHproductions i thought you had to be at least 13 to be on SE .___.

12:00 AM

@noɥʇʎPʎzɐɹC Who do you think answered 16-20? :P

@Rainbolt 25 years, I've been in the Minecraft desert.

"I still remember the first block of sand"

@ETHproductions <=13 is only 2 points

@noɥʇʎPʎzɐɹC I'm talking about the strawpoll

@ETHproductions ye, strawpoll says that

@noɥʇʎPʎzɐɹC yeah, techinically SE doesn't allow kids <13

12:01 AM

I'm like

Just over

So yeah

@DJMcMayhem 16-20

PPCG Ages with most rep

Dennis

@ETHproductions I'm 32.

I personally am 13 (turning 14) soon.

@Dennis Someone older than 20 joins the conversation ":D

12:02 AM

looks like martin is 27 /s

@EᴀsᴛᴇʀʟʏIʀᴋ How do you know

it has the most rep on that query

@Qwerp-Derp because its age with most rep

shit ninja

and martin is the only >100k user on PPCG

ninja'ed

@EᴀsᴛᴇʀʟʏIʀᴋ yay, I'm no longer a fugitive of the law! :P

12:03 AM

Ah

COPPA is kinda stupid

13 -> 12

meh, it's well meaning

@EᴀsᴛᴇʀʟʏIʀᴋ is ninja cat confirmed

@ETHproductions I'm older than 20 too. Which is also true if you take "20 too" as 22. :P

@Downgoat What if you divided it by the number of users with that age to get the average rep per person?

12:04 AM

how u know dis

I'm younger than 20.

shhh i am ninja cat nuke ^

Just computed a large adjacency list by hand and bit into a tomato that tasted like blood. I've had better afternoons

4

@DanTheMan data.stackexchange.com/codegolf/query/545010

hahaha

@quartata are you sure that's a tomato

12:06 AM

@Downgoat "Avergae rep"

@EᴀsᴛᴇʀʟʏIʀᴋ SOPA was well meaning too

@Qwerp-Derp DAMN IT

not getting it to that

i haet speling

You should rename your name to Downgaot

12:06 AM

^^

i changed name to tpyogaot but people got annoyed

(I dont blame them)

@noɥʇʎPʎzɐɹC no it wasn't

@Downgoat Just Downgaot will be fine

ok i should be downgaot

To all non-US PPCGers, the highly-anticipated first presidential debate between Hillary Clinton and Donald Trump is due to start in an hour. You can watch here or probably here.

12:08 AM

action music starts

@HelkaHomba And to US PPCGers.

@HelkaHomba s/anticipated/feared/

Is that to non-USers because the USers are sick of them already? :P

I can't even watch, it kind of makes my stomach sick

THERE'S GONNA BE GUNS!

12:08 AM

@HelkaHomba I'll just wait for the summary. :P

AND BLOOD!

Does everyone here think that the US is doomed however the population votes (either Hillary or Trump)?

one choice also dooms the planet with the US too though

@El'endiaStarman "In summary, the audience got third-degree burns and the entire interview was ← All messages moved to Trash."

@DanTheMan I think he assumes most of us already know.

12:09 AM

@Qwerp-Derp Mildly doomed with Hillary, absolutely doomed with Trump.

5

^

@El'endiaStarman Very true

@Qwerp-Derp Nah, the US is only doomed if the population votes Trump. No matter how much you dislike Hillary, she isn't a xenophobe or a birther.

Verry verry true

But she plays the "woman" card too many times

12:10 AM

@DanTheMan amen

The woman card?

"Oh, I'm Hillary Clinton, vote for me, I'm going to be the next female president"

Trump: GET YER FRONT ROW ROW SEATS FOR WW3!

0

A: Build a nest

Brainfuck, 157 bytes -[+[+<]>>+]<+++.[->+>+<<]>>++>>,[>-[<->-----]+<+++[-<+<<.>>>>-<]>[-<<-[->+<]<<<[-]>>>>[-<+<<<+>>>>]<<.>>+<[>-]>[-<+<<.>>>>]<<>>]<,]<<<<[>.>.<<[-]]>>>+[-<.>]+ Reads an integer in binary on stdin, and outputs the nested list on stdin. Requires 0 to be input as the empty stri...

And then there's Trump

12:10 AM

that took a while :P

@DanTheMan no, instead she's a crook and a liar

Who just is... no

@orlp O_O wat

I think the voting turnout is going to be like 10%

@DJMcMayhem To be fair, Trump is also those things, and he's Trump.

7

12:11 AM

Vote trump, no ww3. And ww4.

@Geobits I agree 100%

Yeah let's drop the politics and admire orlp's answer

2

It's like choosing between cyanide and arsenic.

I still plan on voting for Hillary though, if I remember to figure out how to absentee vote in time. Bernie did succeed in pulling the party's platform more over to the left than it would have been otherwise, and apparently her chosen running mate is a good guy.

@Geobits He also comes with a habit of denying science for paycheck for Big Oil Co™

12:12 AM

Who uses what IDE

@quartata I can't fully appreciate it, not being a bird and all.

@DJMcMayhem So the Trump supporters are like hippies who think arsenic is "all natural, yo"?

I heard somebody compare Hillary to broccoli and Trump to salmonella. You may or may not like brocolli, but you'd have to be insane to want salmonella.

IDK

I don't particularly feel like getting into a fistfight over Hillary

12:12 AM

@Geobits @ AlexA. pls help us appreciate orlp's post

@Qwerp-Derp that's not at all what I'm saying

1 min ago, by quartata

Yeah let's drop the politics and admire orlp's answer

@DJMcMayhem Yeah, IK, I was just trying to find an idiom out of that (TBH, I don't even know what I'm doing)

@DanTheMan idk, i feel like Trump doesn't actually believe anything he's doing, he just started his campaign for the lulz, and its actually working and hes just capitalizing on it as much as he can, so we basically have a 50/50 chance on his actual policies, as opposed to hillary who we know kinda sucks

but then again, if he believes in even 10% of what he says...

strawpoll.me/11310788

12:14 AM

Yeah, the most consistent thing about Trump is that he's inconsistent.

btw

@noɥʇʎPʎzɐɹC No Atom?

@Maltysen lol, he's gonna impose weird laws, the ones that are made up in worldbuilding se

Lol at the thought that you'd vote for somebody because they're probably only trolling us all.

can we count brainfuck as 3 bits per instruction?

Steven H.

12:15 AM

Are there any meta posts about answering questions using interpreter forks of existing languages?

rounding up to the byte?

@Qwerp-Derp dammit, should've done write in

because honestly that's all it needs

Aug 10 at 22:22, by flawr

Please include the title of the poll when posting a link.

@orlp Yes! i have been thinking about that, actually

12:15 AM

@orlp No but there is a variant that works like that

@orlp I think that's been answered as "no" in the past.

And what quartata said too.

@StevenH. I'm sure there are. IIRC it follows the usual "if it was before the challenge" rules.

IntelliJ 'cause it's good!

@Blue Hi, Blue!

because honestly

Meet Yellow!

12:16 AM

it would be trivial to write a brainfuck interpreter

that could interpret my program at 3 bits per symbol

@noɥʇʎPʎzɐɹC Does PyCharm count as IntelliJ based?

^^^^

@Qwerp-Derp Yes, of course

Well yeah, but you could do that for basically any ASCII language to drop it to seven bits, too. It's just not interesting imo.

Steven H.

Isn't that what Sesos was intended to be? Compressed BF?

@Qwerp-Derp same core interface

12:17 AM

Oh yeh Sesos does do that

JetBrains gives pro licenses for free to students

"jif" or "gif"?

@orlp There is a variant that does that

en.wikipedia.org/wiki/Beef_brain . (also known as sesos)

Hell, I only need 18 characters for complete Java using /u0123 codes. Maybe I should write a bit-jamming "interpreter" for that :(

12:17 AM

@Qwerp-Derp i had a jif v gif challenge... I wonder what happened with that

Hmmm

@Geobits Really only 16, since you could drop the /u part and reinsert it.

I say "gif," pronounced like "jif"

Should I do a compression thang with Logicode?

By the way am I the only one who notices how orlp talks in half sentences? It kind of makes me hold my breath in anticipation of the next bit of the thought

12:18 AM

I'm sure I've notic

ed it before.

here it comes

@Qwerp-Derp In all honestly, no. Short of a nuclear world war, I think America will pull through for 4 years. But I'm not voting for the candidate who has shown no desire to understand the nuances of avoiding that scenario.

4

So... anybody use emacs? anybody use vim? they could get together and... ya' know we could grab some popcorn and watch

just any sec now

boop

haha!

Not a criticism it's just amusing

Blue

12:19 AM

nano ftw

it's a chat room

@HelkaHomba oh ye, launching nukes does not need congressional approval.

I kinda like 'simulating' a conversation

@Blue IntelliJ better code completion

@HelkaHomba But I don't have high hopes for the US

12:19 AM

@Blue auuughhh why :P

if I wanted to write full paragraphs I'd use email or smt

A: Sandbox for Proposed Challenges

1

Gif - Jif, Jif - Gif code-golf The point of this challenge is to (not) settle debate on the pronunciation of "gif" The pronunciation of gif is debated and while it's supposed to be (and should be) pronounced jif, it's still commonly disputed. In this challenge you will be given a set of wo...

challenge-feedback-request

Let me onebox a physical tiger...

@orlp ytalk ftw, amirite?

anybody like progress bars, if they were low effort and low overhead? cause I got a package you'd like

*library

12:21 AM

@Maltysen I feel like "Amirite" should be an actual mineral name thingo

It is...

@Qwerp-Derp There are 26 new chat users in the past month.

It's just hard to google because of the slang/meme usage

so what do I do?

what variant is there that gives me a 3-bit-per-symbol encoding?

Team Fortress: The Force Awakens REMASTERED Trailer Parody

@Downgoat, @ConorO'Brien, @quartata, @DJMcMayhem

►

12:23 AM

So how is "orlp" supposed to be pronounced?

@DanTheMan as it's written

@DanTheMan I presume like "welp", but starting with an "or".

^

Sesos is what you are looking for

orlup?

12:23 AM

@orlp esolangs.org/wiki/Spoon

do you know earl grey tea?

half u?

well

orlp

instead of earl

it's orl

12:24 AM

"orulp" was what I was reading it as

and then with a p at the end

@orlp wait that's huffman, not base conversion

@Maltysen even better?

true

I was right!

12:24 AM

I pronounce it oo-roll-p

is there a meta post

on languages that store source files in bits?

like binary lambda calculus

@orlp IIRC someone posted an answer in that where they scored themselves in bits

but maybe there's an interpreter that does that

I think @LuisMendo asked about that a while back

for MATL

@Maltysen Quite a few answers have

Eminem vs. Thomas the dank engine - The real shady train

►

Beautiful

12:26 AM

I love all tomas the dank engine remixes

Blue

@Qwerp-Derp y u do dis

@Blue What do you mean, it's beautiful

@Qwerp-Derp These are the 26 users who started chatting in the last month.

1/10 would not watch again

Here is a compressed BF variant: esolangs.org/wiki/CompressedFuck

12:28 AM

> ['Marky Markov']

@Qwerp-Derp wait isn't it the real slim train?

he's older than a month i thought

> ['Newly Featured Questions']

hahahaha

Yup. Only because SE fixed that bug.

I know my next project

so what's everybody's best at in programming? (front-end, full stack, back-end, mobile, etc.)

12:29 AM

writing a good encoding for brainfuck

with arithmetic coding

@noɥʇʎPʎzɐɹC I've been doing a lot of full stack programming.

and an adaptive model

@El'endiaStarman can you filter by > 10 messages or perhaps by more than one day chatted?

@El'endiaStarman full stack, better at back end because OH GOD CSS

@DanTheMan "cal" actually needs to be replaced with "ble" so it's not as nifty as I though. Sorry :/

12:31 AM

@HelkaHomba Nonsensical?

Question: Suggestion for human-readlabe datetime literal?

I was thinking something like:

> January 1, 2016 00:00:00

I could also just have time literal like:

> 00h00m00s

Time zone please.

At least the option for one.

maybe like -8 January 1, 2016 00:00:00 matches January 1 w/ timezone UTC-8

i don't like that though :/

That's pretty unorthodox. Zone at the end would be better.

that means I'd have to reserve every single month though (well maybe not)

12:35 AM

iirc, snek used an ISO standard for these

link?

docs.python.org/2/library/datetime.html#datetime.date.isoformat

oh that's date not datetime

docs.python.org/2/library/…

alright

converted my answer to spoon

63 bytes

does that turn out to be better than base compression?

not in this case

12:38 AM

Only here would this make sense:

@Geobits Yes!

I have very few +s

> converted my answer to spoon

@Maltysen are you familiar with arithmetic coding?

@HelkaHomba Yeah, I think the "cal"->"ble" link makes it too easy.

12:40 AM

@orlp no

is that different from huffman?

@Maltysen it's much better than huffman

oh

huffman can do 1 bit / symbol at best

arithmetic coding has no such restriction

so its like an analog version of huffman?

sort of ish

12:41 AM

brb wikipidiaing

in short you encode an entire message as a number in [0, 1)

let's say

your first symbol is 25% chance to be X, and 75% chance to be Y

then you say [0, 0.25) is X, [.25, 1) is Y

sorry, fixed

assuming the chances don't change

XX would be [0, 0.0625), XY [0.0625, 0.25), YX [0.25, 0.4375) and YY [0.4375, 1)

and you send the real number in binary

@orlp and when they do change, you do the same analysis on bigrams etc, i'm assuming?

0.1 in base 2 is 0.5 in base 10

@Downgoat New users within the last month sorted by the number of days they chatted.

@Maltysen what is a bigram?

12:44 AM

@orlp you actually measure the probabilities of XX, XY, YX, YY appearing

> There are 2 things that must be completely written in one line; i.e., without “return” key. What are they?

actually it gets better than that

@Maltysen but hold your horses

Downgaot

What does this even mean ._.

so we have this

XX would be [0, 0.0625), XY [0.0625, 0.25), YX [0.25, 0.4375) and YY [0.4375, 1)

0.5 = 0.1 in binary

so if I were to send the '1' bit, we already know it's YY

because 0.5 falls in [0.4375, 1)

12:46 AM

whoaaaaaaa

actually, those ranges should be inclusive

to notate ambiguity

if I were to send 0.01b = 0.25

it can be either XY or YX

the next bit being 0 or 1 would disambiguate that

in order to send XX we would need 1/16 or 0.0001b

but that's what a good coding scheme does

assign longer codes to rarer symbols

@Maltysen using this scheme you can have arbitrary probabilities

if X has a 0.0001% chance of occurring

I can encode it in the fraction of a bit

(if it does occur, it will result in a long code, but that is expected)

@Maltysen understood that so far?

@orlp wait how can there be different better schemes than just using the frequencies

@Maltysen what do you mean?

> that's what a good coding scheme does

@Maltysen yes, what is your question?

a low-probability symbol must be encoded using a long code

this is because of the pigeon hole principle

12:51 AM

@orlp oh wait, you were describing arithmetic encoding as "a good coding scheme"

english parsing error

all short codes are assigned to high probability symbols

srry

alright

so arithmetic coding can handle arbitrary probability

it reaches exactly the shannon limit (so it's perfect in that sense)

but the best part is

after every symbol you can change your probabilities at will

@orlp oh, so you know the probabilities for say the 17th character

@Downgaot New users within the last month sorted by number of messages.

12:53 AM

you can do for example

'if the last symbol encoded was X, the probabilities are X=33%, Y=67%'

Pinning tabs is useful

Conor O'Brien

@Downgaot you changed your name

@orlp that's essentially what I meant by bigrams

cool

@Maltysen well

it's not just bigrams

@ConorO'Brien :O He actually did it!

12:54 AM

@orlp well yes, n-grams

@Maltysen no, arbitrary systems

@orlp hmm?

'if the number of parenthesis emitted so far is balanced, probability of ) is 0'

ohhhhhh

'if the last pixel encoded in this image was red, give priority to red-ish colors'

Downgaot

12:55 AM

@ConorO'Brien i have been identifying as a gaot so I descided I'll just change it

as long as the decoder has the same amount of information as you do

so no look-aheads

mınxomaτ

debate (for folks outside US)

then you can arbitrarily decide the probabilites for the next symbol based on what has already been decoded

@Maltysen the beauty of an arithmetic coder is that it takes away ALL complexities about encoding

all you have to do is give it a list of symbols with their probabilities, and an actual symbol

and in the decoder just the list of probabilities

@mınxomaτ cnn.com works

you don't have to do base encoding, smart schemes, anything :)

12:57 AM

@mınxomaτ Nope, copyright claimed already

mınxomaτ

New link

the only disadvantage of an arithmetic coder is that it's relatively slow compared to huffman and co

implementing it isn't easy either

I oversimplified here

by pretending we have real numbers [a, b] to represent the range

in reality we have a fixed precision, and output bits as we find out

so if [a, b] < 0.5 we already output the '0' bit