The Nineteenth Byte

Netscape's source is now available :)

auto playing video ^

gahh

Boooooo

6:51 PM

-1 (to cnet.com) autoplaying video trend sucks ._.

betseg

Linux kernel's source is now available :)

?

that is docs

7:07 PM

interesting, so I once wrote in a programming language that wasn't allowed according to the rules, but I was hoping nobody would care enough to report that

and now it was proven to be Turing complete because it's apparently possible to implement Rule 110 in it

those rules have been abolished for a long time :D

@xfix Technically, languages don't need to be turing complete for you to answer with them

Yeah

just find prime number and cat

iirc

CMP: what percentage of your votes are on questions? (Check electorate badge progress )

Where do I find that?

^^

7:10 PM

i don't even know if checking for number primality in that language would be practical, as it doesn't have a number type

57% over here :P

the language in question being "ed"

ed is fine

@DJMcMayhem Profile > Activity > Votes Cast > By Type > Divide the count of "Question" votes by "Answer" + "Question" votes

Giuseppe

@Christopher 34% -- and sorry for the ping earlier, muscle memory took over :P

7:11 PM

What if I already have electorate?

@Giuseppe all gud

@DJMcMayhem do what Mr Xcoder did

with the question like "The world ends in ed" i just had to write it in ed

and the language was easy enough to implement in it

About 1/3rd

@DJMcMayhem you are the only person with the badge

@DJMcMayhem not bad

That's not true

7:12 PM

Roughly 31.1567164179

> awarded 66 times

huh

idk why it showed me only u

I am about half way there

Is it because you went here?

mbomb007

@MartinEnder chat.stackexchange.com/rooms/41525/retina is frozen again

@Dennis I really hate it when people say something and then someone creates a monster

@DJMcMayhem lol yeah

7:14 PM

@mbomb007 Want me to thaw it?

ninja'd

badges are interesting, some of them seem like really hard to get

martin is not even active in the room

and he already unfroze

for instance, good luck getting Tumbleweed on a site with barely any "questions"

@xfix some of them are never gonna be given here

@Christopher he hasn't recently talked, but that doesn't mean he doesn't have the tab open (he's here)

7:15 PM

true

Tenasious and unsung hero are never gonna be given

i think i was first person to get Reversal on this site, and even now it seems like a rare badge

but yeah, i don't see Tenacious be ever given

that's in the group of the unfortunate users' badges

yeah...

it's like, this isn't QA site, so accepted doesn't really mean much

to get that, you need to write shortest programs, but somehow not get upvotes, which is just not going to happen

i suppose a strategy would be to make "tips" threads for obscure programming languages, accept your own answers, and hope nobody cares about that programming language

that's probably the most practical way to get Tenacious here, but even then, seems almost impossible to do

yeah

I tried to get it once

never got accepted answer except on 1 answer :/

Dennis

@Christopher Anything you say may be used against you or to create an esolang.

15

7:23 PM

@Dennis Just when I thought you were the serious mod :P

what does that have to do with being a mod o_o

He actually reminds me of my dad

My dad seems like a serious person until you get to know him

although chances are that even if you were to make a thread like "Tips for golfing in Malbolge", chances are it would get upvotes on answers, although not sure about that

True

Tips for golfing in dodos

So question about text compression loophole. If a challenge says "answërs in characters" can't I just post an answer with a 1 char byte that has all the text encoded in it?

listening to music while working and a new versions of Stacy's Girl came on. Called Stacy's mom.

i think you need to use an encoding that was defined before the question was asked (and you cannot use encodings like MetaGolfScript)

7:29 PM

@Christopher "a 1 char byte" ? :)

@xfix well create an esolang that takes the binary byte value and splits it into 8 bit segments then turns each 8 bit segment back into a byte

@xfix no, that was abolished too, I think a year ago :)

@EriktheOutgolfer time to gulf

Gulf: Golfing but with lame a cheaty tactics that are allowed

i mean, encodings, not languages?

codepages?

but the binary file is the same regardless of the codepage

7:31 PM

@Christopher That's not really possible. UTF-8 bytes don't get longer than 4 bytes

although i'm sorta glad that this rule about languages was removed, because there is no real issue with using a programming language newer than a challenge if it wasn't specifically designed for it

Plus that's exactly why we say bytes instead of chars

@DJMcMayhem who said UTF

if somebody wants to use Java 9 in an old challenge, well...

Get on my gulfing level

7:31 PM

Point me to the code-page where the bytes to "Never gonna give you up" is contained in a single character.

@DJMcMayhem utf8 can support up to 6, but >4 are never used

@ngn you might be confusing that with UTF-16?

UTF-16 is where every character is two bytes

@DJMcMayhem CRS-1

UTF-8 has 6 byte encodings, although because max Unicode character is U+10FFFF, it just won't happen

7:33 PM

@EriktheOutgolfer i'm not ^

Christopher's running Sh*t

cl.cam.ac.uk/~mgk25/ucs/utf-8-history.txt

@DJMcMayhem no, that's UCS-2

it supported characters up to 7FFFFFFF, but that turned out to be not really all that useful

@Christopher That's almost the exact same thing as this:

Do this, and if somebody tries to cheat, let everyone know and watch as it (hopefully) gets downvoted into oblivion. — Esolanging Fruit Jun 14 '17 at 17:26

7:37 PM

@DJMcMayhem lol

Not for here :P

which is a fair policy :)

if you really define a language that defines 0 bytes to solve that given challenge, then sure, you may have won the code golf, but prepare for downvotes

dzaima

I like how on this site, even if you find a loophole, it's usually pointless to utilize it, usually even bad for you

I would have no qualms about straight up deleting an answer that is that cheap and low-effort.

@DJMcMayhem :P

@EriktheOutgolfer Thanks for the edit.

7:42 PM

Wouldn't removing the cheap and low-effort answers get rid of most jelly answers? Heck most golfing answers for that.

No... Posting answer in golfing languages != low effort. (I wish I could caps lock punctuation)

No? How is a powerful well-designed golfing language even remotely comparable to inventing a new language (or encoding) that only does exactly what the challenge asks for?

</joke>

what about a real language which just happens to do something at 0 bytes?

^

7:46 PM

That's better, but I find it annoying when the 0-byte behavior is an edge-case designed to solve one particular challenge. I think it's much cooler when that behavior is fundamental to how the language works

i once had a stupid idea of making a language in which empty program prints an entire alphabet, just in case somebody would post a challenge like that

but even if i did succeed at that, honestly, it still would be cheating

it's rather boring, but whatever, github.com/xfix/AlphaBeta/blob/master/alphabeta.pl

honestly it's just a bad golfing language

also "e - eval current file name" - what was i thinking back then...

@xfix or you could watch the sandbox and create such languages just before the challenges are published :)

fair enough

@ngn I think that "using languages intended for a specific challenge" is in fact a standard loophole

@EriktheOutgolfer that's not an objective criterion

@EriktheOutgolfer technically, I could claim I made that language for my cat's birthday and for this challenge

7:55 PM

@ngn well, here it is

@EriktheOutgolfer I'm not disputing that it's a "standard loophole", I'm only saying that it's hard to check

honestly i don't see a problem with HQ9+

because HQ9+ is allowed, but what could it solve?

it was a rather clever joke language, and not really designed to win code golf challenges

and arguing it shouldn't be allowed because "not Turing-complete" just let people to make languages like HQ9+B

it's trivial to add Turing completeness to any programming language, so requiring that didn't really stop anything

are quines technically kolmogorov complexity implementations, and hence a non-language could participate in the quine challenge?

8:01 PM

no, every program has a different output, so it's not kolmogorov-complexity, it's just quine

that doesn't go against the definition of kolomgorov though

> kolmogorov-complexity questions effectively crowd-source the approximation of it.

codegolf.stackexchange.com/tags/kolmogorov-complexity/info

this suggests that quine challenges aren't kolmogorov-complexity

> In the context of code-golf, Kolmogorov complexity is a challenge to compress a specified string or dataset, so a post should be tagged kolmogorov-complexity iff the bulk of the challenge is to produce a string or a subset of a dataset which is given in the question.

caird coinheringaahing

I think quine is a subset of kc

except that you're not trying to make your code as short as possible to output an arbitrary string, but your own source code, which imposes an additional restriction making it not kolmogorov-complexity

8:05 PM

@totallyhuman Kolmogorov complexity is the length of the shortest program that computes a given function in a given language

the term is used incorrectly in PPCG's tag

@ngn But the goal is to find that shortest length in a given language

@EriktheOutgolfer but additional restrictions have never made challenges not KC...

to be fair, i don't see a problem with allowing "non-languages" whatever that means (although something that could be reasonably considered to be source code would need to be required, for instance for rot13 "language", the source code is a file that will be "encrypted" output, which pretty much makes it useless for ROT13 challenges)

and it's not that you can cheat at "quine" challenges by saying cat language

but additional restrictions make the code not necessarily be the shortest possible anymore

i mean, you can, but so you can specify the language to be PHP and do the same

8:10 PM

@EriktheOutgolfer hasn't mattered before

@DJMcMayhem most (all?) KC challenges on PPCG don't specify what language they must be solved in

i could find more examples but i think everyone's seen examples

Yes, but the challenge is still to find the shortest program per language

It's just that different users and different answers use different languages.

Neil

@Christopher 16%

'dang

Happy Pi Day!

8:15 PM

@DJMcMayhem I understand that. So, strictly speaking it's multiple Kolmogorov complexity problems in one.

i might raise this on meta, since this needs some kind of rule either way

Actually, again speaking strictly, every golfing challenge is about coming as close as possible to the Kolmogorov complexity for a certain problem.

wait what

what was removed and what was added?

One space was added

first of all how did he edit with so low rep

8:22 PM

maybe he made the edit and then bountied some

Q: Euler-Poincaré-Characteristic of Polyhedra

hm, looks like it

flawr

9:02 PM

@Riker The top comment chains are great: reddit.com/r/YouShouldKnow/comments/84f8xz/…

ETHproductions

9:13 PM

> And instead of landing to pick up more passengers you could just fire people up in the air with a giant trebuchet and catch them as you fly along.

caird coinheringaahing

Seems legit

New Main Posts

9:26 PM

1

Given a triangulation of the surface of a polyhedron p, calculate its Euler-Poincaré-Characteristic χ(p) = V-E+F, where V is the number of vertices, E the number of edges and F the number of faces. Details The vertices are enumerated as 1,2,...,V. The triangulation is given as a list, where eac...

code-golf math geometry polynomials topology

CMC: figure out what 2**rip refers to (1 word)

@Riker ?elcric a fo aerA

mudkip201

9:51 PM

circumference?

nah, it's aera

it says 2**rip, not 2pir

mudkip201

oh derp

it's backwards

the whole thing

pir**2

Οurous

What's the double times supposed to do?

caird coinheringaahing

Power

betseg

@Οurous Try it online!

Οurous

9:57 PM

Ahh.

Pavel

10:35 PM

I'm surprised there's not a /usr/bin/exec.

Like on one hand, there's no way it could actually work, but on the other hand, the rest of the bash builtins have a /usr/bin equivalent

Even /usr/bin/alias exists

@DJMcMayhem 'area' but yea

also +1 ty for not LTR

*not using a RTL override

"area" backwards is "aera"

flawr

"area" backwards is "ɐǝɹɐ"

10:57 PM

backwards, not upside down :p

@totallyhuman nope, he lost about 700 rep due to a user being removed

another 100 lost rom a bounty + a couple small reversals

found this today

@EriktheOutgolfer you sir need to turn your head weird directions

11:13 PM

@Riker you can tell how old this is by looking at the DTD

lol dear god

> html 4.01... 1999

(from google)

has anyone here implemented automatic memory management using ref counting before

I have some Qs for VSL regarding it

@Riker sounds about right

@Downgoat fire away

dumb idea: .dotcom TLD

@quartata so when ever in VSL you allocate an object we create an alloc() call with allocates the object T* and a ref-counting struct in the form of: { i32 0, T* o }. That said, I'm not sure when to deallocate it

because LLVM will allocate the object to the stack (which I wouldn't want) if I make exprs not increment/decr RC

so do I do it everytime an object is assigned to an identifier? escapes the scope?

flawr

11:16 PM

@Riker don't effin onebox! (not meant for you, but the website)xkcd.com/1963 (but the xkcd link is meant for you)

@Downgoat no you want it on the stack

well ok

@quartata every object should be allocated on the stack?

because that would mean they are deallocated at the end of the scope

no not the object itself

what you borrow from it

I don't know a lick of LLVM but when you borrow from the pointer you want to have it in whatever wrapper you need so that when it goes off the stack you can call a destructor and decrement the count by 1

@quartata what do you mean by 'borrowing' from an object.

@flawr LOL

11:18 PM

getting a new reference to it

(at the edit and post :)

Adám

@Riker Wouldn't the last page be 63 of U+10FFFF?

so really it should be like {i32 count, T obj}, and that's on the heap

then when you get a reference &obj you wrap it with something with a destructor

in the destructor you then decrement the refcount

it's RAII

@Adám no clue

@flawr this is dark but fascinating

5

(also be sure you do the increment/decrement atomically)

11:20 PM

so if I were to do f(a: Obj()) and f is defined as func f(a: Obj) { return a }

I'm assuming that's not pass-by-value

yes, so we'd create incr ref, pass it to f which will incr again, f will decr at end of func, and the callee will decr at the end of its func?

yes

and when the refcount hits 0....

kaboom goes object

trying to go through the original boost::smart_ptr code to show what it looks like in C++ but it's kinda confusing because they have inline assembly for the atomic increment/decrement part

not sure why they didn't abstract just that part out rather than having one file repeated the pointer logic per architecture....

oh well

11:27 PM

would you be able to observe the difference between a by-value object and an immutable object (without language reflection or anything)

yes

if the object stored a pointer to itself (with a shallow copy constructor)

then you could tell if it was copied or not, the pointer would refer to the original object