The Nineteenth Byte

Uriel

00:01

@HyperNeutrino APL: {,/(⊢,⍵+1-⊢)¨⍳⍵÷2}

HyperNeutrino

Nice

Now time to read through that to make sense of it because I'm trying to learn APL lol

ATaco

Yay, atleast it beats APL!~

Uriel

@HyperNeutrino it takes the range of half the input, and for every item gives the item and the input + 1 - the item (7 = 8 + 1 - 2)

@ATaco note APL is not a golfing lang

HyperNeutrino

ah ok thanks

Pavel

@Uriel What are you talking about

HyperNeutrino

00:15

gtg o/

Pavel

It totally is

What sane language looks like {,/(⊢,⍵+1-⊢)¨⍳⍵÷2}

ASCII-only

@Adám JS: c=>'A1 H9 IAG NOR UNC WHO XS'.match(c+`\S+`)+''||c

@Adám Wait that doesn't work, JS: ('A1 H9 IAG NOR UNC WHO XS'.match(c+`\S+`)||c)[0]

Uriel

@Pavel I believe the guy who coined the term golfing was not even born when the first APL line was written

ASCII-only

@Adám JS: c=>c+({A:'1',H:'9',N:'OR',U:'NC',W:'HO',X:'S'}[c]||'')

Pavel

@Uriel I'm sure when Hero of Alexandria invented the steam engine he was already thinking of it's use for powering trains.

Uriel

00:32

@Pavel you can ask @Adám how different is APL nowadays

ATaco

Just because it wasn't designed to golf, doesn't mean it's not golfy.

Uriel

@ATaco oh, surely it is

Maatouq

hi
on this code
x,y,z;
double R(){return--y?!z?y=R(),R(z=1):x*R():x;}
main(){
scanf("%d%d",&x,&y);
printf("%f\n",R());
}
i couldn't understand how double recursive function call works
y=R(),R(z=1) in one statement
it's a solution for this problem
https://codegolf.stackexchange.com/questions/5562/implement-hyperexponentiation-tetration-without-the-use-of

ASCII-only

@Maatouq please format ty

@Maatouq Basically it evaluates y=R(), then returns R(z=1) I think

ATaco

x,y,z;
double R(){return--y?!z?y=R(),R(z=1):x*R():x;}
main(){
	scanf("%d%d",&x,&y);
	printf("%f\n",R());
}

(Boop)

Pavel

00:38

How does that work, R isn't defined to take an int parameter.

So the call to R(z=1) can't work

ATaco

C just forgets the passed parameter, the writer is using it to golf.

Pavel

Oh right, I forgot C functions don't have defined parameters

El'endia Starman

01:04

@HyperNeutrino 95 messages, apparently.

You're also #11 in terms of posting the most ':P' messages relative to all of your messages.

Pavel

01:19

@El'endiaStarman Is there a way to search by regex? I want to do a search for .*ಠ_+ಠ.*

ATaco

@El'endiaStarman C'mon, I thought you knew SQL :P

New Main Posts

0

Q: Lights out in more than 2 dimensions

What this isn't A rehash of another challenge posted for solving Light's Out. What this is Light's out in 3D... or 4D, or 5D, or even 1D; you get the picture. So, what are the rules? Input You can choose to format your input however you wish. One example could be: number of dimensions, siz...

El'endia Starman

@Pavel PostgreSQL has three pattern matching functions. postgresql.org/docs/9.3/static/functions-matching.html

@ATaco That's the query to use when you want a ':P' anywhere in the message. The ones I linked look for messages that are literally just ':P'.

4 hours ago, by HyperNeutrino

I wonder how many messages of mine contain :P and only :P

ATaco

Well isn't that better fitting? Posting just an emoticon is noise!

Oh... ;-;

Pavel

ppcg.starmaninnovations.com/tnbde/#ldXdIQCHIR

^ Most frequent Kannadian face users

HyperNeutrino

01:25

@El'endiaStarman huh I'm surprised I'm not higher up :P But thanks for the stats! :D

ATaco

That thing has a name?

HyperNeutrino

huh apparently i'm seventh for ಠ_ಠ (watch this message push me to sixth)

Pavel

Well ಠ is a letter of the Kanadian language

ASCII-only

@ATaco It's called the look of disapproval

ATaco

I wonder if you could graph the usage over time to pinpoint exactly when the TacoScript for it was released.

ASCII-only

01:27

@ATaco Yes you can

Pavel

Probably, but that sounds too fancy

ASCII-only

knowyourmeme.com/memes/_-look-of-disapproval

Pavel

I don't know that much SQL/JS

ASCII-only

You could possibly ask El'endia

Downgoat

@Pavel did someone say SQL/JS :P

@ATaco userscript-request: add user-friendly AM/PM timestamps to chat

ATaco

01:30

(Screams in local timezone)

El'endia Starman

There are a bunch of line graphs on the TNBDE useful queries page. You can modify one of those.

Downgoat

@El'endiaStarman speaking of TNBDE, I don't mean to sound pushy or anything, but is there a reason this PR has't been merged?

El'endia Starman

@Downgoat Yeah, laziness. :P

Christopher 2EZ 4RTZ

@Dennis @MartinEnder @DJMcMayham @Hyper Neutrino @Ascii-Only @ATaco @Downgoat @FunkyComputerMan any of you care if I mention you (in a positive way) for helping me on this website and coding in general? (common app college essay)

ATaco

Sure, go for it.

Pavel

01:39

@Christopher2EZ4RTZ WheatWizard is now @FunkyComputerMan. Also, you added a space.

Christopher 2EZ 4RTZ

Opps

That edit just pinged a bunch of people

Funky Computer Man

I don't mind

Pavel

Also in HyperNeutrino's name.

Christopher 2EZ 4RTZ

Oh well

It still pings

Pavel

Yep.

Downgoat

01:40

@Christopher2EZ4RTZ yeah, that's fine

Christopher 2EZ 4RTZ

I just wanted to check since I don't want any issues

It is a pretty damn good essay already

ATaco

Scream my name to the heavens as the coder who brought you light!

5

Or something like that I suppose.

Christopher 2EZ 4RTZ

253 words about PPCG on my 623 word essay

@ATaco brb screaming

Pavel

CMC: Package Hack into an RPM and put the result on COPR.

ASCII-only

@Christopher2EZ4RTZ I don't mind

reads chat

Haha everyone agreed with a different wording

Pavel

01:46

@ATaco I have pull requests for changing ChatJax delimiter to $$ and for fixing the /o_o bug, please respond.

Christopher 2EZ 4RTZ

Not funky and you @ASCII-only

ASCII-only

@Christopher2EZ4RTZ oh btw you misspelt DJ's name

Christopher 2EZ 4RTZ

@DJMcMayhem The super ping list

@ASCII-only rip

Can I get a few short snippets (working answers) of crazy esoteric code?

ASCII-only

@Christopher2EZ4RTZ Anything on PPCG :P

Downgoat

@Christopher2EZ4RTZ I would say to add JSF or @ConorO'Brien obfuscated JS if this is for your apps

Pavel

01:55

@Christopher2EZ4RTZ Here:

ṡ2EÐḟFs2ḣ€1S€ṡ3M€Fċ2
0;⁸;0Ç

Christopher 2EZ 4RTZ

@Pavel what does that even do

ATaco

It makes me cry for starters.

Christopher 2EZ 4RTZ

Now if only Dennis would step down so I could become mod (that would look great on my essay)

Pavel

1

A: How many peaks in my mountain range?

Jelly, 27 bytes ṡ2EÐḟFs2ḣ€1S€ṡ3M€Fċ2 0;⁸;0Ç

Question: Is codegolf.stackexchange.com/a/143241/60042 valid IO? I understand the need for a C submission to take the size of an array as a parameter, but the output format seems odd to me.

ATaco

Yep, C languages can take inputs as Pointers, and may take the size of an array as a parameter.

Pavel

02:03

I'm not talking about that

I mean the output

It requires a 4th input, an array that it populates.

ATaco

(And may output through those pointers)

It seems valid to me.

Pavel

I know that you can output by returning a modified version of the array, but this is different and to me it feels like the correct way to do that would be to return an int pointer.

Also, what is this "dynamic array" thing? I haven't heard of it. (Note that I am not a C person)

Christopher 2EZ 4RTZ

@Pavel lol C and dynamic

@Downgoat are you smart?

Downgoat

no

chat.stackexchange.com/…

Christopher 2EZ 4RTZ

ok @ATaco are you smarter then him?

@Downgoat chat.stackexchange.com/…

ATaco

02:09

Apparently

chat.stackexchange.com/…

Christopher 2EZ 4RTZ

I want a 2nd opinion on my paper and want to know if you would read it XD

You just have to read it and say what you think (I have english teacher for grammar and such)

Downgoat

if your paper is on CS i can maybe be of use if it is on anything else I am useless

Dennis

@Christopher2EZ4RTZ Sure, mention away. :)

Christopher 2EZ 4RTZ

@Dennis ty

@Downgoat onetimesecret.com/private/gp29aqg1eabut3nq5or3kv1wyck2v54

Give it a shot

Downgoat

@Christopher2EZ4RTZ this is wrong link

you have to give the other one

it should be in yellow like /secret/randomhexdigits

Christopher 2EZ 4RTZ

02:13

I am an idiot

@Downgoat that makes me saying it once XD

onetimesecret.com/secret/6wbi2wxy9l6p6j427pb427dr5hfi6s9

Downgoat

its' -> its in the KahnAcadamy sentences

Destructible Lemon

@Downgoat muphry's law?

Downgoat

"Processing.js that was modified to make it simpler" -> "Processing.js, a language designed to make programming easier than usual." or something sentence kinda confusing at first

@DestructibleLemon yse

HyperNeutrino

@Christopher2EZ4RTZ Yeah I'd be fine with that, though I don't recall having done anything to help? huh

Downgoat

also :D someone else who learned how to code w/ Processing.js

"Looking back I realize", missing comma after back

StackOverflow and StackExchange have spaces

I recommend defining code golf too

HyperNeutrino

02:19

(doesn't include spaces)

Downgoat

> One challenge only allowed characters that were not banned by the last answer and each answer banned another character. Quickly code became messes until somebody won using only the number eight 2.13 × 10705 times to print the text “‘Round up the usual suspects.’ - Capt. Louis Renault” (number 32 on the top 95 movies quotes of all time).

Challenger5

I remember reading a meta question asking whether golfers can make the challenge harder for themselves by adding additional stipulations, and the consensus was that they could. But I can't find the post. Does anybody know what it was?

Downgoat

I think you should re-word this or make it less specific. It's confusing to me

ATaco

ComputerCraft did a lot to teach me how to code.

Downgoat

though I think otherwise is 10/10 essay, really shows your passion for programming

HyperNeutrino

02:22

wait link to essay

Christopher 2EZ 4RTZ

@HyperNeutrino onetimesecret.com/secret/fv7ccqimeah5akx0ocnidjh980kpij8

HyperNeutrino

ah ok thanks

Christopher 2EZ 4RTZ

@Downgoat thanks for the help

I will add it all tmr

Downgoat is a good goat. pats on head

HyperNeutrino

How should I give feedback for minor/major grammatical corrections / etc?

Downgoat

Yes math textbook, when I am driving I obviously solve quadratic equations to determine my stopping distance to the next car so I will not crash. Obviously using graphic utility will not cause that.

Conor O'Brien

02:27

@Christopher2EZ4RTZ "and I would like to thank (sniff) all of these wonderful people, comic book characters, hosts, weird computer men, tacos, and others (sniff) (sniff) for helping me create programs abominable to good programming practices"

10

HyperNeutrino

how to calculate whether or not you will run into the car in front of you in 3 easy steps:

1. Encounter a situation where there is a car in front of you
2. Press the brakes suitably to avoid crashing
3. If this fails, then you have crashed into the car

@ConorO'Brien don't forget the subatomic particle :( nobody ever notices neutrinos

Conor O'Brien

Notices what

HyperNeutrino

>.>

DJMcMayhem

@Christopher2EZ4RTZ I would not mind at all! I'd be honored :)

HyperNeutrino

I am still wondering what I have done to help anyone though... I mean besides JHT I guess

anyway gtg o/

Challenger5

02:35

@ConorO'Brien Why are there so many PPCG characters that are named after food? Wheat (that's not really a food, but still...), lemons, watermelons, tacos, etc.

I haven't tried a neutrino, but I've been told they're delicious

Conor O'Brien

@Challenger5 lack of creativity? there's a lot of food-related golflangs too

Dennis

There are?

Challenger5

@ConorO'Brien I can only think of Jelly and CJam right now

Conor O'Brien

maybe not golflangs

ppcg langs

jelly cjam cheddar confiture (obscure fork of jelly with only like 2 commands) sesos (technically)

maybe not that many but its certainly a trend

Dennis

Wait, confiture?

Challenger5

02:38

Is it like a tacit tarpit?

github.com/ConorOBrien-Foxx/confiture

Conor O'Brien

confiture

it's not even TC iirc

I think it has addition and maybe something else

ATaco

TacO

Dennis

Update README.md, THIIIIING, THIIIIING2, alright, let's get rolling! sounds legit.

Conor O'Brien

this except I forgot to try

but yeah the idea was to make a turing tarpit based on the evaluation style of jelly

then school started IIRC

Challenger5

02:56

Thinking about it, it could be completely bitwise, with vectorizing NAND

Conor O'Brien

I already have an esolang sorta like that

github.com/ConorOBrien-Foxx/NTFJ

ATaco

Of course, you can't talk bitwise langauges without talking bout Wise

Conor O'Brien

we were until you mentioned it :P

Downgoat

03:22

dumb questio: is floor(a) + floor(b) == floor(a + b)

Destructible Lemon

I don't think so

1.6 and 1.6

Downgoat

ah yes duh >_<

Destructible Lemon

it's ok

I still want to make a koth

ATaco

But is round(a) + round(b) == round(a + b)?

El'endia Starman

Depends on how you round. I think.

ATaco

03:29

The answer is no, 1.3 and 1.3

El'endia Starman

Yeah, I just realized that.

ATaco

(Where round is floor(n+0.5))

El'endia Starman

I was thinking of potential edge cases due to the IEEE standard (I think) that numbers that are exactly halfway between integers get rounded to the nearest even integer.

ATaco

Floating point numbers are bullshit.

El'endia Starman

How might you implement them better?

Challenger5

03:31

Not have NaN, negative 0, and all that other stuff.

quartata

Those have purposes

Very well defined ones at that

Challenger5

Explain to me what the purpose of NaN is as opposed to Infinity.

ATaco

NaN may not have a sign, also, 1/0 isn't infinity, as much as javascript argues.

quartata

Errors and 0/0

ATaco

It's also a nice Null number.

Challenger5

03:35

Given that floating point handling is usually implemented at a very low level, I would have expected most invalid results to just return 0, and trust the user to deal with it (after all, this is how integer overflow and /0 is handled)

Destructible Lemon

@El'endiaStarman 😡 this is not what they taught us at school

quartata

@Challenger5 and if the result is 0?

Also signaling NaNs are much better for error handling

ATaco

@El'endiaStarman I can't but that doesn't mean they're good.

El'endia Starman

@DestructibleLemon I was taught that rounding rule in school a couple times. I don't normally use it though; I just round up.

Challenger5

@quartata Well, checked overflow is also good for error handling, but it's not done that way. Floating-point is going to be implemented at a very low level, so it needs to be simple

It has too many special cases, and the special cases it defines aren't even mathematically true

quartata

03:40

I mean errors in user land where a float is to be returned

Also almost all architectures set a flag on overflow

ATaco

Also, Javascript uses Doubles, not Floats.

IIRC, Floats don't handle Infinity & NaN.

quartata

They do

Also "keeping it as mathematically accurate to our arbitrary definitions of R as possible" is an argument for 0/0 = NaN

Challenger5

imo it's an argument for anything/0 = NaN

ATaco

According to the scarce mathematics I know, anything/0 should equal NaN.

Challenger5

But in IEEE 754, positive/0 = ∞, negative/0 = -∞, and 0/0 = NaN

Destructible Lemon

03:47

anything/0 = 0.5

Challenger5

¿!?

quartata

The idea is that for all x != 0 lim x -> 0+ 1 / x = infinity

only exception is 0/0 since that's indeterminate

Challenger5

@quartata Yes but approaching the limit is different depending on whether you approach 0 from the positive or negative side

quartata

It's more convenient than an exception at low level at any rate, plus kinda fits with the approximate nature of floats

Challenger5

And approaching from one side or the other is a bit arbitrary

quartata

03:50

That number could be close to 0 but result in Infinity

@Challenger5 Other side would be for signed zero is all

Challenger5

Sorry, but I don't understand what you mean by that.

quartata

note the + by the 0

That's a one sided limit

Other side is lim x -> 0- 1/x = negative infinity which is what negative zero is for

sorry for no latex

Destructible Lemon

0*infinity = 0?

ASCII-only

I don't know how to write math help :/

quartata

oh

Challenger5

03:56

But why should approaching one direction or the other be dependent on whether the numerator is positive or negative?

quartata

I wrote that wrong

I meant for x to be in the numerator

for all x > 0 : lim y -> 0+ x/y = Infinity and lim y -> 0- x/y = -Infinity

Challenger5

In IEEE 754, (1 / 0 == Infinity) but (-1 / 0 == -Infinity)

Destructible Lemon

noooooo

quartata

That's a good point

But at this point they kind of have to enforce consistency

ASCII-only

TIL Chef esolang guy is Australian

quartata

03:58

Since the sign of the denominator makes a difference

ASCII-only

And also SROMG (and other related webcomics) guy too

Challenger5

(-1 / -0 == Infinity). WHY????

quartata

Negative number divided by very tiny negative number

It checks out

Funky Computer Man

@quartata That's not true

It is only true for positive x

quartata

Can't see what you're replying to

Oh

Funky Computer Man

04:00

> for all x > 0 : lim y -> 0+ x/y = Infinity and lim y -> 0- x/y = -Infinity

Pavel

@Challenger5 What about that doesn't make sense

quartata

x > 0 is right there I changed that after I realized the clause was dependent on the sign of the numerator and I didn't want to type it out

I'm on mobile guys cut me some slack

Challenger5

@Pavel The fact that (-1 / 0 == Infinity) but (-1 / -0 == -Infinity).

Funky Computer Man

Oh I missed the change

Challenger5

If you told a mathematician, "in programming dividing by zero gives a result that is either positive or negative infinity, depending on the sign of the numerator and the sign of the zero," what would they think about you?

quartata

04:04

They'd understand it if I defined it in terms of a new algebra and explained why that's convenient practically

Along with the limits above

Floats are inherently approximations of real numbers. Sacrifices must be made

Challenger5

I still don't see why signed zero and signed infinity are convenient practically.

quartata

Well signed infinity is most definitely a thing

As for signed zero:

Challenger5

Signed infinity might be a thing, but it's definitely not what you get from dividing by 0.

quartata

Signed zero

Signed zero is zero with an associated sign. In ordinary arithmetic, the number 0 does not have a sign, so that −0, +0 and 0 are identical. However, in computing, some number representations allow for the existence of two zeros, often denoted by −0 (negative zero) and +0 (positive zero), regarded as equal by the numerical comparison operations but with possible different behaviors in particular operations. This occurs in the sign and magnitude and ones' complement signed number representations for integers, and in most floating-point number representations. The number 0 is usually encoded as +0...

@Challenger5 But it's what you get when you divide a number very close to 0, which is common in floating point math due to error

Do you really want 1/0 to throw some bulky high level exception and then have 1/0.0000000001 return Infinity anyways (or even worse a giant imprecise number near the end of the floats)

Downgoat

@TuxCopter how to do title case in french? is it like L'indice de la Qualité de L'air?

ATaco

04:11

@quartata Yes. :P

quartata

Then I don't know what to say, go ahead and make your own standard

It's not easy

ATaco

BRB rewriting modern computational mathematics.

Challenger5

I would still prefer if it just had NaN for "universal error" rather than having a bunch
of edge cases (±inf, NaN, etc.)

ATaco

(I just wish Lua had an easier way to detect NaNs)

quartata

But infinity isn't really an error

Sometimes infinity is legit

ATaco

04:15

Infinity is nice if you want to compare it to other numbers, which may sound strange, but it has its purpose.

quartata

Also floats don't really overflow, throwing an "error" is kind of a weird concept IMO

@ATaco Another good point. That's what math.huge is if I'm not mistaken

Challenger5

Maybe, but inf should be "the number greater than all other numbers" rather than "the number you sometimes get when you divide by 0" or "the number that is returned on overflow"

quartata

Well it is the former too

Also signed zero should never really be seen unless you're working at a super low level

Or doing something hella weird

It usually works itself out

ATaco

function getClosest(some_vector_table){
	max_distance = infinity;
	target = null;
	for(i=0; i < some_vector_table.length; i++){
		this_distance = sqrt(some_vector_table[i].x^2 + some_vector_table[i].y^2);
		if(this_distance < max_distance){
			max_distance = this_distance;
			target = some_vector_table[i];
		}
	}
	return target;
}

Yay, a practical use for infinity!

quartata

Lua does use math.huge right

I couldn't remember

ATaco

04:18

Yeah, math.huge is infinity.

Challenger5

@ATaco If that's JS, I don't think it works; Infinity is capitalized

ATaco

But it doesn't let you work with -0.

Also, it's just generic code, no particular language.

(And you probably shouldn't use sqrt for best speeds)

Challenger5

That reminds me of 0x5f3759df

ATaco

We don't need sqrt because a > b == a^2 > b^2

Challenger5

@FunkyComputerMan I am now using your amazing icon for my one of my school accounts.

Funky Computer Man

04:29

@Challenger5 The icon is actually taken from my schools information system which despite its appearance of having been forgotten in the 90s was just redesigned this year.

Destructible Lemon

is that... legal? /s

Funky Computer Man

to have a crappy website? yes

Challenger5

Redesigning the site, or propagating the icon?

Funky Computer Man

The website also has a rather large security issue I won't point out

Destructible Lemon

@Challenger5 well I know Dennis would not do an illegal thing

probably

Challenger5

04:31

@DestructibleLemon O RLY?

Funky Computer Man

Can C output to STDOUT without any preprocessor directives?

Destructible Lemon

@Challenger5 😧

Challenger5

@FunkyComputerMan Do you count compilers that #include <stdio.h> if you use IO functions without declaring them?

Funky Computer Man

No I'm talking about the C standard not any specific compilers

ATaco

I don't think it's part of the standard.

Actually, I don't think C has any native support for STDOUT.

Funky Computer Man

04:36

Ok that is unfortunate

Challenger5

But how can the preprocessor add features to a language? That doesn't make sense...

Is <stdio.h> written for every platform?

Funky Computer Man

Maybe I should actually ask a SO question (shudders)

ATaco

I don't know enough about C to answer that, unfortunately.

Challenger5

@FunkyComputerMan Nah, a question that useful would probably be closed.

Funky Computer Man

That seems like a rather unuseful question as a whole

you can usually use preprocessor directives.

Challenger5

04:41

But the explanation of why would probably go into a substantial amount of detail about platform-specific IO and how it interacts with the standard.

ATaco

C doesn't have explicit stdio.h definitions because it's designed to compile to a wide variety of architecture, some of which have no support for that form of IO.

Funky Computer Man

Ok I guess that makes sense

so stdio.h must not be written in C then

Challenger5

Not in standard-compliant C

ATaco

It's my assumption that it makes direct CPU calls, rather than just does C things.

Or, I suppose, OS calls.

Funky Computer Man

Ok this is all reather unfortunate

or fortunate

depends on how I look at it I guess

quartata

04:45

github.com/lattera/glibc/blob/master/libio/stdio.h

Not sure what you mean stdio.h isn't defined in C

ATaco

C is a rather beautiful language for something made out of spite and excitable devs.

Funky Computer Man

Oh?

Challenger5

Spite? What spite?

quartata

GCC is just lazy and only throws a warning if you use it without the prototypes from the header

It still gets linked in from libc so it doesn't care

ATaco

Spite for B, the interpreted language of choice prior.

Funky Computer Man

04:47

@quartata Do you know if that is compliant with the C standards?

quartata

Probably not but you can declare the prototype yourself

Just copy whatever hellish one it declares for printf

It'll resolve

Funky Computer Man

Ok. I don't think I'm good enough at C for this right now

I'll come back to this later

Anonymous

@Challenger5 No. stdio.h usually has all of the standard declarations, some of which are enabled or disabled by preprocessor macros based on the definitions exported by your system or the configuration of your compiler. It doesn't actually contain the code for those functions - that's in libc (or glibc, or whatever C runtime you use).

quartata

Look for extern blah blah blah printf(blah blah

Funky Computer Man

@quartata Oh yeah I found it, it just is, as you said, "hellish"

Anonymous

04:50

You can declare printf without including stdio.h, but you have to get it right (or close enough)

Anonymous

printf, no #include, no warnings

quartata

Close enough

Funky Computer Man

How does this even work

I'm very confused

quartata

That's a forward declaration in essence

Except instead of being in your code the function is in a library

glibc

Which is implicitly linked by default

ATaco

Better to explicitly call it an extern :P

Anonymous

04:52

@ATaco extern is pointless with a single compilation unit

ATaco

(Screams in C)

Funky Computer Man

@quartata Oh so if we did not link glibc it wouldn't print

Anonymous

@FunkyComputerMan If you didn't link glibc, it wouldn't do much of anything :P

quartata

It would compile but not link

Try it with -ffreestanding

This is all basically because C compiles to individual object files and then links all of the references to other files later

It's for saving memory

only need to compile one at a time

Funky Computer Man

@Mego Really how much of C is in the standard library?

quartata

04:56

Just functions really

Anonymous

@FunkyComputerMan Anything that you don't write yourself.

quartata

Like all the core language constructs except for variable arguments work freestanding

Funky Computer Man

Ok this is good

quartata

C++ though needs the standard library for exceptions and RTTI

Funky Computer Man

are there any types without the standard library?

quartata

04:57

Hm?

Funky Computer Man

Is the int type part of the standard library?

quartata

No. But technically uint8_t and what not aren't defined without an include but those are compile time macros

Are you writing a kernel or something

Funky Computer Man

Hm I seem to get a segfault when I try to remove the standard library

quartata

What code

Oh

That's because you need to bootstrap calling main I think?

Transcript for