The Nineteenth Byte

like... even in c

using a linked list like betseg would make it easy if done right

hmm. should i be preparing utility functions for the upcoming aoc

1:32 AM

@ASCII-only I'd just use Python: Batteries Included™

python is terrible

also not available online

ouch

~~i should really put more effort into making a good programming language~~

same

~~as in better than python~~

1:36 AM

I have yet to make one better than brainfuck

Define "better than python".

unrelated: some of the names on the AoC global leaderboard sound familiar

@Bubbler more easy to use than python's so called "easy to use"

@ASCII-only there's a private leaderboard for ppcg folk

I forget who set it up

more powerful than python, a way more massive standard library, implicit imports in certain modes

@Skidsdev betseg, but no i mean the global leaderboard

@ASCII-only define more powerful

1:38 AM

Python is way too powerful already, no?

how about not powerful enough >_>

fastest growing language 3 years in a row, 2nd most popular language on both GitHub and stackoverflow

like you need a million builtins to do code jam with it

@Skidsdev that's just because it's relatively easy to learn

not the easiest to learn, but more people already know it than for other powerful languages so more people teach it imo

@Skidsdev I guess the most popular one is JS for both

charcoal is in python. why? because dlosc was better at python than other languages. although i was an idiot and never gave him a chance to contribute >_>

1:40 AM

@Bubbler Actually I think Java is still the most popular on SO

@Skidsdev ...ew

SO why

@ASCII-only agreed

^

that's to be expected though. Java, Python and C/C++/Haskell are the languages most commonly taught in uni

but seriously? everyone. pls use C# instead of Java already

also, wow. hyperneutrino and doorknob are in the top 100

unihedron as well, and betaveros (and petertseng, kinda) a sounds familiar. cameron aavik used to be on TNB

I can't imagine many unis teach Haskell, it's pretty out there

More probably teach assembly

1:43 AM

@Pavel nah haskell is only for compiler courses

also you'll be a little lucky to find a uni teaching python, they like outdated curriculums...

nope

@Skidsdev Oh they teach Python alright

well, not where i live

You just have to be lucky enough to get Py3

1:45 AM

The most weird thing to me is that many unis teach C++.

@feersum what's weird about that?

Very inappropriate for beginners.

many would argue otherwise

Straight C is much more educational.

right but they want to fit object oriented stuff into the curriculum too

"C with classes"

1:46 AM

@Skidsdev In addition to what?

@ASCII-only Oh you know what, I just had the thought that eventually compiler courses will start migrating towards Rust

@Pavel why?

Isn't Rust a giant hodgepodge of features like C++?

remember, they aim for simplicity, not performance

if anything, perl would be a better choice

@feersum in addition to teaching a commonly used language, probably teaching some core low-level fundamentals depending on the course

1:48 AM

It has the type inference features of Haskell, and also really cool lifetime tracking for compile-time memory management

both perl and haskell's syntax/paradigm make lexing asy iirc

and it's "fast"

@Pavel but the syntax/builtins are very, very, very different

Rust isn't easy to parse so much as it just has a really fucking cool compiler

@Skidsdev C++ does not force you to learn anything low-level (on the surface...). C makes the students actually have to understand pointers.

1:49 AM

@Skidsdev "fast" doesn't matter here, otherwise unis would all be using PyPy

@Pavel nah i mean it's easy to write a parser in haskell

especially considering template haskell is a thing

@feersum doesn't force, no, but it's a hell of a lot more low level than say C# or Java

@Skidsdev kinda but not really

C# has all the same low-level features as C

@Skidsdev It would be more accurate to say that it allows to mix different levels.

you can go more low level, but really all you need to remember for C++ is new goes with delete and boom, you've turned C++ into a high level language. in fact (IIRC) you don't even need new and delete. possibly

1:51 AM

I mean you can always go more low level

C? Nah ASM. ASM? Nah Straight Machine Code. Machine code? Nah just flip the bits on the disk platter

@ASCII-only Huh? That's how you turn it into C.

@feersum wait what :|

crap ignore me then

ok. what are utility functions i'll need for aoc

I mean you also need to stop using classes entirely to turn it into C

except pods

C has structs

You can even stick function pointers in 'em

OOP C is totaly possible

1:55 AM

it might be hacky, and look a little weird, and be a little clunky, but yeah. basic OOP is possible

the problem with this theoretical language idea i have that is intended to be more powerful than Python: idek if it will actually work. in fact, i don't even know how to start on the parser >_>

@Skidsdev Well, there was a time where programmers were literally 'flipping the bits on the disk platter' ... except they were called punch cards. :p

and back then a "disk platter" was just a fancy type of tableware if anything at all

Got to look at the insides of a punchcard machine once

Apparantly they detect the holes in the cards mechanically with magnetic particles rather than optically like a scantron

But then comes ... the 5MB hard-drive \o/

El'endia Starman

@Pavel I would have never guessed that. Actually, I don't think I even thought much about how punch cards work.

2:01 AM

@Pavel did optical detection technology exist back then though

@ASCII-only Invented by IBM in 1937

@Arnauld Did you hear about the 16TB HDD being announced recently

Vaguely. I didn't read the articles.

Me neither

did you guys hear about the atomic hard drive IBM made?

I can barely fill 1 or 2 TB.

2:06 AM

it uses only a single atom to store a single bit of information, making it the most space efficient data storage in existance

ignoring the fact that it needs an observatory-sized electron microscope to read/write said bits

@Arnauld and then there are some with 1tb of music :P

@Pavel Which, even at its invention, was used for the sole purpose of grading tests automatically.

I have 2TB of games

@Skidsdev *a STM? wait no, microscopes wouldn't be able to read that

and 1.5TB of audio clips, music samples, sound libraries and VST plugins

2:08 AM

du /?

@Skidsdev you could make it more efficient by using a smaller atom :P

it uses the orientation of a single Holmium atom’s magnetic field

@ASCII-only yeah but modern hard drives utilize roughly 100,000 atoms to store a single bit (According to IBM)

also they chose holmium due to its powerful magnetic field created by “many unpaired electrons.”

@Skidsdev yeah of course. we don't have the technology to read more finely

@Skidsdev they chose one that was more easy to make work, not a smaller one

IBM figures that it can store the entire iTunes catalog (all 35 million tracks) onto a disk the size of a credit card by using this technique.

just from the surface area?

2:12 AM

It currently requires a liquid nitrogen-cooled tunnelling electron microscope operating in a vacuum to work

which by the way, is easily one of the most sci-fi movie phrases I've ever heard

as in like... 2050 sci-fi instead of 2500 sci-fi?

Meh, needs a "superconducting"

and a "quantum"?

"iridium"

The system uses atoms of holmium seated atop a magnesium oxide surface, which keeps the atom's magnetic poles stable -- even in the presence of other magnets.

So no drive distruction by powerful magnet

We Adeptus Mechanicus now, boys. Make sure to say your holy litanies before uttering "Play Despacito"

2:30 AM

Since when is physical storage space an actual limitation?

surely the rest of the computer is far more limiting than the amount of room that hard drive takes

A limit to what?

@NathanMerrill there are standards around hard drive size, so making it smaller allows one to fit more storage in the same standardized size

@Veskah Praise Tzeench

@NathanMerrill Also I'm pretty sure this tech isn't for consumer drives, it's for highly specialized massive-scale data storage

@Skidsdev Filthy Heretic

Right, for consumers, it's about price, not physical limitations

@Veskah jokes jokes, I'm all about the Greater Good

2:33 AM

most consumers aren't buying high-density drives because they don't need them

however, I don't see why massive-scale data storage needs it either

@Skidsdev Filthy Blueberry :^)

@Veskah Filthy Clanker

because the land is surely cheaper than all of the equipment needed to store in such a small area

Big data ain't a misnomer. And there's only so much rack space.

At least before you have to buy another rack

I agree, it takes up more room. But once again, is it going to be cheaper to buy an acre of land and more racks, or a highly-sophisticated state-of-the-art storage device?

It's not about room, it's about price

2:39 AM

Yes, early adapters have always paid a premium, no one was surprised.

Yeah, but I don't think the price will ever compare (at least, in our era). I may be wrong, but that's my prediction

You're probably right, we're pretty good at putting magnetic sand on a platter. However, those with bottomless purses might have good reason to grab these

it's IBM

they did it to prove they can

in it's current form it'll likely have zero practical application in any industry

because it'll have abysmal IO speeds too

Right. It's really cool. I just don't think it's practical :)

And I also thing there another application for tiny storage moving forward into Internet of Things

Bleh, IoT

2:44 AM

but the reader needs to be just as small as the actual storage

SoC armed with flash memory are pretty good these days

@NathanMerrill I mean we have NAND storage for that stuff

ignoring the global shortage of NAND memory

Don't worry, the companies who make it totally won't gouge the price in the meantime.

It's different than last time :^)

Yeah, but say in 20 years, could we fit a computer in a shoelace? When computers get really tiny, I could see that having interesting use cases

Depending on what you're defining computer as, we're already there.

This random SoC I found is 4x4x0.65 mm

nxp.com/docs/en/data-sheet/MC9S08SU16DS.pdf

PDF warning, I guess

3:02 AM

That's not bad. All you really need to add to that is wireless communication and wireless charging. Because, do you really want to plug in your shoes at night :)

Draco18s

4:20 AM

@ASCII-only Oh, sure. My esolang is just weird enough that having it eval is super weird. Multiple simultaneously executed instruction pointers is madness enough without having each one of those IPs also be executing subworld copies of another program with its OWN multiple simultaneously executed instruction pointers.

so when it comes to C, what's the general consensus here on GCC vs Clang?

One starts with a G, the other one starts with a C

And you are writing in C so the choice is obvious

Draco18s

That, or you're trying to commit cuicide.

@Skidsdev they're competing with each other, so tbh not much difference. but generally people prefer gcc i think?

@Draco18s so... kinda like charcoal's lambda functions (yes, functions, not eval)

For code golf, gcc is a clear winner though. clang requires return statements.

4:29 AM

@ASCII-only fair. I've found competing arguments and benchmarks for both, some showing GCC is better, some showing Clang is better

Draco18s

@ASCII-only Only familiar with Charcoal in that "I've seen other people post with it"

@Dennis while that is good to know, I'm not at the point of golfing in C just yet

oh yeah. gcc probably has more extensions

gcc comes standard with pretty much every Linux distro, so I've been using that thus far

@Draco18s well. charcoal functions are a half-assed feature. i doubt anyone has used them on PPCG before. ever

@Skidsdev :|

4:30 AM

I was just wondering if it was actually worth considering clang

@Skidsdev nope

never

@Skidsdev Clang has more helpful error messages and GCC has better optimizations

I prefer gcc, but I must admit that clang has much better error messages.

ninja'd

so use gcc until I run into a tricky error, then compile with clang. Gotcha \s

huh. gcc really needs to up their debug game i guess

Anonymous

4:34 AM

IIRC clang does better in benchmarks, but GCC has more extensions and has better support and availability.

Anonymous

And supposedly it's much easier to write an LLVM backend than a GCC backend

yeah, because llvm is modular iirc

it's why there is such an insane number of llvm backends

Riker

5:11 AM

John McAfee

@officialmcafee, House McAfee

If you believe that anyone, at any time, for any reason, ever, gives money away to total strangers, then you deserve to get scammed.

25.6k tweets, 879k followers, following 13.3k users

@quartata new person for you to aspire to be ^

@Skidsdev my personal experience tends to be that clang is much stricter, erroring on things that are warnings in gcc

can't really say that's a general thing though

wait when did day 7 come out

oh just now

6 days after day 1, I presume

@Riker i already did this

apologies for the lack of content over the last couple hours...i just released a new version of my Antivirus Software and as usual the trolls spiked ny tomato juice and put me on a plain to Honolulu

Tweeted by quartatertots on June 24, 2018 at 2:51 AM

@ASCII-only llvm frontends are easy. backends suck

@Pavel have you seen sun.misc.Unsafe

5:28 AM

@quartata Yes and I don't think it quite counts

no of course not just having a giggle

i wonder how that gets compiled down in gcj

I heard they were going to remove it in java 9 at one point

No idea if they actually went through with that

@Riker in case you didnt know mcafee is actually 100% serious. he really is that insane and terrible

Today's part 2 was frustrating as hell

Downgoat

@Riker how much coke does this man do

5:39 AM

@GoodTipLeaver @CryptoIsse @J34512063 @missmayn @RationalMale I don't do cocaine. It's a yesterday drug. If I were to do uppers, I would choose one of the hundreds of Chinese research chemicals of the Cathinone class. Legal, and make's Cocaine feel like weak coffee

Tweeted by officialmcafee on November 13, 2018 at 2:09 AM in reply to GoodTipLeaver

6:03 AM

ugh

who boasts about doing weak coffee

my grandma probably tried it

@quartata In case anyone's wondering, Cathinone is a schedule 1 drug and illegal in most of the world, including the US

mcafee does not care for your laws

~~he is a fugitive from justice~~

6:29 AM

goddamnit

doing AoC on mobile is hard

and one small mistap can lead to all progress lost :|

6:46 AM

also :/ :/ day 7 part 2 is too hard

idk where i'm messing up. i can tell that i did, just i'm bad at debugging js

[generic JS is bad joke]

wait nvm

I messed it up four times, and I had to wait for 5 minutes (it happens after 4 wrong answers) 3 times.

lol, that's what's happening to me now

i probably should stop submitting solutions i know are wrong

I made some pretty dumb mistakes too

6:58 AM

see, this is why we need a better language

At first, I only had 4 workers, and I forgot to take into account the last step's time...

haha

Protip: get all the rules right, then try the sample input with exact same conditions

i forgot about the last step's time too

oh wait

... you can use the sample

i did that on codejam (because you have to) but forgot to do it here

goddamnit

wrong again

I found that I was getting +1 than expected at the very first tick

Damn off-by-one errors

7:01 AM

i'm getting it right for the example tho

wait

i'm using 4 workers as well

lol XD

lul

wait no

oh wait. i didn't try the example with 2 workers

finally

darn off-by-one errors

my insanely bad code is now on hackmd

wonder if i should use a different language every day from now on

> 7 stars

> behind anon with 3.5 stars

yay me

That's just a problem with private leaderboard

well i can change ordering to sane ordering

@dzaima why 6.5 stars

> changes ordering to global score

Mmm

7:16 AM

0

:P

im #1 in local score :P

i have to settle for ordering by stars

@ASCII-only all i could manage before running to the bus

For star ordering, you can take the top by waiting in front of your PC

The problem is that you have to do it every day to keep the top

7:20 AM

@Bubbler waiting?

until the next problem opens

well. same with every ordering

like... i have 7 stars but am below average in all but star ordering

after a week or so youll be above average

its coz you started late

i've done that for the past 2 days but i only have 10-15 mins before i have to go :/

> you'll be average after a week or so

that's not for a while :P

@JoKing any progress on leap years?

7:23 AM

i did an O(n) solution for day 5 while in maths class trying to keep nice notes and it was my peak for aoc

@betseg peak?

greatest moment and probably wont do anything as good as that

again, for aoc

greatest moment?

@ASCII-only No, i haven't looked at it much

doesn't look like we can get much further with current approach though. hmm

7:26 AM

I probably should have recorded my programs for aoc... I just deleted them afterwards

Yeah

@ASCII-only every other challenge was something like O(n!^n!) and with time on hand so yeah id consider doing O(n) with no time a good thing™

@JoKing nah not like aoc is anywhere near as impressive as code jam

@betseg wait. wat

only slight exaggeration

@betseg mine are like... O(n^2) max. well maybe n^3 for more stringy ones

7:39 AM

@JoKing hmm. i'm trying the second form of "Junctions can also be used to terminate sequences with side effects" but have no idea how to translate it fully, it can potentially be 25 bytes

@ASCII-only I don't think that one's applicable. the condition stuff ends up on different numbers than the printing

yeah :/ was seeing if there was any way to short cicruit

New Main Posts

7:55 AM

1

Q: Number that can eat itself

Given a positive integer, output a truthy/falsy value as to whether the number can eat itself. Rules Leftmost is the head, rightmost is the tail If the head is greater than or equal to the tail, the head eats the tail and the new head becomes their sum. If sum>=10 then the head is replaced by...

code-golf

8:09 AM

also. this is just sad

someone on code-golf.io pls learn lisp >_>

@ASCII-only I know it's irrelevant, but what does your avatar really depict? Is it a bird? Or two mountains separated by two valleys?

:|

i did it really badly didn't i

what's it supposed to be?

here

i assumed it was a face, not specifically a cat's

8:22 AM

hmm. maybe i should make it a 5x5 ASCII art or something :P

@ASCII-only I see. One less question to take to the grave.

some people have no idea that ^^ is eyes

and not a double caret :P

surprising fact

@FrownyFrog I don't get why they are eyes.

8:24 AM

by convention

this js leap years is tough too :/

lisp though, i'm the only one there >_>

@FrownyFrog What's the convention for the mouth?

it's the closest ascii equivalent to -^

@Arjun it's a convention, not the convention

what's a convention for the mouth?

@ASCII-only this face would (probably?) be ^u^ in ascii

o/O/u is also common for the eyes and v/w/_ for the mouth iirc

A: Sandbox for Proposed Challenges

9:03 AM

for aoc 7b why can't worker 2 start E as soon as it is finished with F?

New Sandboxed Posts

0

Compressed RSA keys cops-and-robbers number-theory primes Silly cops-and-robbers idea. As computers get more powerful, RSA keys have grown longer and longer to maintain security. This makes it difficult to: hand out your encryption key on business cards recite it over the telephone quickly s...

@Cowsquack Step D must be finished before step E can begin.

this helps for js in leap year

@JoKing ooh now it makes sense

Neil

10:49 AM

@Pavel Thunderbird's MIME code is written in OOP C. People are scared of it :-(

A: Sandbox for Proposed Challenges

11:45 AM

blogs.windows.com/windowsexperience/2018/12/06/…

edge as an electron app

Quintec

12:38 PM

@ASCII-only I tend to use ^.^ becuase it’s more recognizable as a face and looks cuter

New Sandboxed Posts

1:08 PM

0

Not a challenge, but not certain of the wording so trying it out here: Tips for golfing in JSLint tips javascript code-golf What general tips do you have for golfing in JavaScript restricted to satisfying JSLint with default options? I'm looking for ideas that can be applied to code golf p...

Q: What are our standard-defining posts?

@Pavel Not Cathinone, chemicals of the Cathinone class ("bath salts"). Bupropion, e.g. is sold as an antidepressant.

New Meta Posts

1:59 PM

0

What are the posts on meta that specify rules or standards every user should know?

discussion faq-proposed

Adám

2:21 PM

@Quintec It is also valid and meaningful APL.

3:05 PM

I finihsed aoc 7b, it was extremely awful

i got a nice head start in the first 5 days but it seems like ill lose the first spot

yeah I had trouble with AoC 3a and just sorta gave up :P

3:29 PM

@Quintec I'm more of a fan of ^-^

Quintec

3:49 PM

@betseg Ay, I'm representing hard

Alion

4:37 PM

@betseg What is this?

@Alion PPCG's advent of code leaderboard on a browser extension i found

firefox chrome

5:16 PM

I thought it was a historical listing of visible chat stars. Which also sounds rather interesting

@NathanMerrill You could use TNBDE

Uh, anyone know SQL?

Quintec

5:29 PM

Yay gold ranking on halite.io :D (Team PPCG step up your game :P)

@Pavel got some working knowledge. What do you need?

Blue

Is anyone good at regex? I'm trying to write a regex that doesn't match strings that contain ( unless it's preceeded by \ So \( matches but ( doesn't.

markdown.

\\\( ?

^([^(]|\\\()*$?

Blue

That looks like it still doesn't catch the (

I'm effectively trying to implement backslash escaping

5:40 PM

@J.Sallé "A historical listing of visible chat stars."

anti-matches of (?<!\\)\(?

@Blue an example where it doesn't work?

Generic form: /^(AB|[^B])*$/ but both \ and ( have to be escaped.

(so, what dzaima said)

alternative answer: ^(\\.|[^(])*$

Blue

Oh, it appears to work, though the online tester was odd

5:46 PM

@Pavel in which order? Most to least starred?

I guess

30 mins ago, by Nathan Merrill

I thought it was a historical listing of visible chat stars. Which also sounds rather interesting

@Cowsquack or matches of ^(?!.*(?<!\\)\()

@Pavel is this it?

I limited to 50 results because query times are limited to 10s, but I think even 150 results would still run under that

@Pavel I more interpreted as a history of the stars visible on the starboard per-day or something, and i don't think tnbde isn't able to do that

6:02 PM

Wait.. he wasn't always called Leaky Nun?

@Skidsdev I'm getting a 404 error from that link

@J.Sallé ooohh yeah I forgot people with low enough rep can't see deleted posts

HyperNeutrino

6:51 PM

lol I was running through SO LQ review queues and tried commenting on a post

turns out it was an audit and i got pranked ;-;

Q: P Pr Pre Pref Prefi Prefix Prefixe Prefixes

Does that count as success or fail?

HyperNeutrino

well, the comment failed and it said "this is an audit" so I just gave it a Recommend Deletion and passed :P

because the post was already deleted + locked lol

Oliver

7:27 PM

Is there a golfy way to use a variable inside of a JS regex match?

Example: s.match`users\/([^\/]*)\/oliver"` -> I want to replace oliver with a variable.

New Main Posts

1

Given some finite list, return a list of all its prefixes, including an empty list, in ascending order of their length. (Basically implementing the Haskell function inits.) Details The input list contains numbers (or another type if more convenient). The output must be a list of lists. The s...

code-golf array-manipulation function functional-programming

@FrownyFrog Why did you delete those entries? codegolf.stackexchange.com/posts/177147/revisions

@flawr they don't include the empty first array

Ah I see

ran into the same problem in both apl and canvas too :|

7:36 PM

@FrownyFrog it might be worth adding the reason to the edit summary in a case like that

i wonder what percentage of answers will be shorter if written without the starting element

@dzaima start a new challenge? :D

AdmBorkBork

Stupid PowerShell

13

@flawr too duplicate-y

@AdmBorkBork :O

AdmBorkBork

7:50 PM

What? It's perfectly logical to love something you sometimes think is stupid

Arrays in PowerShell suck

@AdmBorkBork and still I hate myself

AdmBorkBork

Aww, don't say that. :)

:D

is it reasonable to require that the submission is the only function in the program (besides the main one) and that the stack is empty on its calling?