The Nineteenth Byte

TheNumberOne

00:02

@orlp The [ on the left can be recomputed from the rest of the expression, so we can remove those.

orlp

@TheNumberOne how?

how do you distinguish [+[+]] from [++]

TheNumberOne

[+[+]] -> +[+]

[++] -> ++]

afk again

The ] on the right can also be recomputed from the middle, unless it matches one of the [ on the far left.

orlp

no, you're leaving something out here

do you do anything special for .[<+]. for example

it has no bracket on either the beginning or the end

does that mean you do base 8 everywhere?

Sp3000

.[<+]. doesn't change

orlp

well, that's not true

your explanation is flawed then

Sp3000

00:12

How so?

orlp

because the [ doesn't get encoded

Sp3000

Why not? As I said, octal except the first digit

So [ is still a possibility after the first

orlp

well, that's not correct

give me one sec

wait now this is weird

TheNumberOne

[ is encoded using the octal digit 0

orlp

give me a couple thousand more seconds

TheNumberOne

00:16

ideone.com/A5agkY

orlp

oh wait

I think that's why it disappeared

leading 0s get nuked

TheNumberOne

^ First person to notice that.

Other than me.

orlp

ok I think I'm starting to understand your answer

TheNumberOne

That's why the translation can't start with a [. My encoding simply does not allow it.

orlp

any non-balanced brackets simply automatically get balanced

so you can have arbitrary balance in the middle

TheNumberOne

00:19

Yes

The hard part is distinguishing [+[+]], [[+[+]]], and [[[+[+]]]] from each other.

They are encoded as +[+], +[+]], and +[+]]] respectively.

+[+] is encoded as +[+

Dennis

WTH? github.com/mihirk/NSFW

orlp

@TheNumberOne I had a different idea

TheNumberOne

What is it?

orlp

encode into base-7, not counting [

when you encounter [, push onto a stack how many symbols you've processed so far

when you encounter ], encode a 6 in base-7 followed by a a base-N number containing stack.pop(), where N is the number of symbols processed so far

it should be trivial to see that this is a dense enumeration

TheNumberOne

That fails in the case where nothing follows the 6.

orlp

00:28

@TheNumberOne there always follows something after the 6

TheNumberOne

Decode 34 (decimal)

orlp

34 % 7 == 6

number of symbols processed so far = 0

base-0 number containing 0 (trivial)

[]

TheNumberOne

Decode 42

orlp

42 % 7 == 0

+

42 // 7 = 6

number of symbols processed so far = 1

base 1 number follows

6 // 7 = 0

[+]

TheNumberOne

Decode 6

orlp

00:32

[]

don't even need to think about it

remember I wasn't done with 34 yet

TheNumberOne

Oh :), finish then.

orlp

34 // 7 = 4

4 would be like ,

so [],

TheNumberOne

Do you decode the number left to right or right to left?

orlp

don't know

it's tricky to implement right

actually I think it needs one modification

[ [ ] ]

consider that

you could see it like this

[   ]
   [    ]

which is obviously not right

but the encoding right now could encode it like this, thus it's not dense

Mitch Schwartz

00:58

ah cool, nobody noticed the easy byte that i thought of at dinner, so 116 now

c+=(c<0<6<v)-(d>0==v) -> c+=c<0<6<v;c-=d>0==v

like i said, i didn't look into how / why it works, i've just been focusing on making programs that are equivalent to the previous one

Mitch Schwartz

01:18

@TheNumberOne not sure if you meant first person to notice that leading 0s get nuked, but if so, that's not quite true, and i don't think it's hard to notice. i mentioned a consequence of it here chat.stackexchange.com/transcript/message/23717192#23717192

"if we don't strip the initial '[' we would get the same d"

TheNumberOne

good point, you noticed it too.

I forgot about that. A lot of your optimizations were based on that.

@orlp Decode 19

@orlp 19 should uncover a problem with your decoding algorithm.

orlp

19 % 7 == 5

.

19 // 7 = 2

.-

weee

TheNumberOne

Hm... not what I intended.

orlp

:D

TheNumberOne

01:36

@orlp Decode 97

orlp

97 % 7 == 6

base 0 number

[]

97 // 7 == 13

13 % 7 == 6

base 1 number

13 // 7 = 1

yes, 1 % 1 == 1

Mitch Schwartz

ok i have a better idea of how @TheNumberOne 's encoding system works, very cool

orlp

[[]]

(0 % 1 == 0 would form [][])

TheNumberOne

@orlp Wouldn't it be a base 2 number?

orlp

@TheNumberOne no, we only processed one symbol so far

remember, we ignored [ for 'symbols so far'

but it definitely has some off-by-one errors, and one flawed nesting

but I believe that can be solved

TheNumberOne

01:40

@orlp Oh, I forgot that.

Would [+] be encoded as 6?

Never mind, above you decode 42 into [+]

orlp

6 encodes []

you already asked that :)

let me demonstrate the inverse

r = 0

[ ignore, push 0 on stack

TheNumberOne

How would you encode +[] vs [+]

orlp

+ -> 0

r = 7*0 + 0 = 0

hrm

yeah, here I have some issues

TheNumberOne

@orlp 8 would be 7, correct?

orlp

?

I mean, 0 digits are falling off

@TheNumberOne oops

TheNumberOne

01:46

@orlp That was my biggest problem with figuring out an encoding.

One way to fix this could be to start with 1

orlp

@TheNumberOne empty string vs +

TheNumberOne

@orlp "" -> 1, "+" -> 7

orlp

@TheNumberOne what would 2 be?

TheNumberOne

No program would translate into a 2.

Mitch Schwartz

got another byte, d>0==v -> d>1>v

if d is nonzero at that point, it has to be at least 8

because of the previous d=d*8+v

TheNumberOne

01:52

works, I didn't expect it to though.

ah, good explanation

Mitch Schwartz

thanks

TheNumberOne

02:18

@orlp You need a program where leading 0's don't change your program.

Alex A.

03:36

@Doorknob Is this possible in Snowman? (Possible in a reasonable amount of code)

@Dennis Are you in here?

Dennis

Nope. :P

Alex A.

Oh okay. Sorry, carry on.

Dennis

Just kidding.

Alex A.

I know.

Dennis

(Me, not the language.)

Alex A.

03:42

/you kidding

Would you jam out a CJam solution in 5 minutes for this?

@Calvin'sHobbies Hi

Dennis

Sure.

Calvin's Hobbies

o.o

Alex A.

Sure as in yes you'd want to? (I'm not asking you to. :P)

@Calvin'sHobbies ?_?

Looking for feedback on the challenge because I'm bored and I just want to post it. :P

Dennis

Done.

Alex A.

Done what?

Dennis

03:50

Done jamming.

Alex A.

:)

I think something happened to your link

Dennis

) doesn't usually get URL encoded, but it messes up the Markdown.

Alex A.

Ah okay

Nice work!

Dennis

Nice challenge. Brute force becomes rather slow for higher values of n.

Alex A.

Yeah, I noticed running that link.

Should I impose some kind of restriction?

I figured I'd let people do whatever.

Dennis

03:53

You might want to mention that n will be non-negative.

Alex A.

Good idea

Done

Dennis

I also don't understand the without leading zeros remark.

Alex A.

So for example 013 is not distinct from 13.

You'd count that as one.

Is there a better way to say that?

Dennis

Number of distinct primes should have you covered.

That also explains that multiplicities are not counted.

Alex A.

Better?

Dennis

03:56

Perfect.

Alex A.

:D

Is it interesting enough?

That's probably the most import question to ask.

Dennis

I think it's entertaining.

It would be a lot more difficult with a time limit.

That's not necessarily a good thing.

Alex A.

I feel like in order to impose a restriction like that I would need to have a reference implementation so that I know of a way to do better than brute force. (Currently I don't)

orlp

I can't golf it any further

def N(s):
 r=p=0;a=[0]
 for c in s.strip(']'):
  j='[<>,.-+]'.find(c);r=r*8+j;p+=r<1
  if j%7<1:a+=[-~a[-1]-j%5]
 return-~r*8**min(a[p:])

@TheNumberOne

Dennis

@AlexA. True. With and without the restriction, you have two entirely different challenges. I like it as it is.

Doorknob

04:01

@AlexA. Sure. Now, if you want a reasonable amount of code and for it to complete before the heat-death of the universe then... ;)

Alex A.

@Dennis Thank you so much for your input, I really appreciate it!

@Doorknob Do you have any input on the challenge? Would it be fun to do in Snowman?

Doorknob

Well, then it depends on how you define "reasonable" I guess. I'm sure it'd be possible to be efficient and under, say, 1kb.

Alex A.

:P

Doorknob

It would definitely be a challenge!

Dennis

@AlexA. No problem. This is your second challenge, right?

Alex A.

04:02

Yep

First was World Big Dosa

Doorknob

@AlexA. I haven't had lots of free time recently, though if I get the chance I might give it a shot.

Alex A.

@Doorknob <3

Dennis

I still have to write an explanation for that one. I kept thinking that I would find a shorter way without converting the entire code to unprintables, but eventually forgot about it altogether.

Blame cops and robbers.

Doorknob

Nice challenge, too! :)

Alex A.

What, mine? If so, thanks! :D

Doorknob

04:05

Yep, sounds fun (especially in snowman :P)

Alex A.

Every time I try doing anything in Snowman I get so confused that I give up and use a different language. I have yet to have the patience to learn it well.

Doorknob

I have to go to sleep now though, because school. I might be able to try a Snowman solution on the weekend. Just be prepared to be whacked with a block of unintelligible gibberish. :P

Alex A.

I'm always prepared for that.

Goodnight!

Btw, has been posted.

New Main Posts

0

Q: Fond Memories of Past Primes

Consider a prime number p, written in base 10. The memory of p is defined as the number of distinct primes less than p that are contained as substrings of p. Challenge Given a non-negative integer n as input, find the smallest prime p such that p has memory n. That is, find the smallest prime w...

Dennis

OK, I managed to achieve a significant speedup. 3.5s for the last test case.

Alex A.

04:09

Ooh, nice!

Oh, should the title not be title case?

Fond Memories of Past Primes --> Fond memories of past primes ?

Looking at the current question list I guess the latter is more common

Dennis

It's a title, so title case should be ok.

Alex A.

I remember at one point someone edited a challenge to remove the title case. Might have been Beta.

Sp3000

Doesn't really matter, both are fine

Dennis

Find the first prime with memory p should be lowercase, but Fond Memories of Past Primes looks better like this imho.

Alex A.

Okay cool. Thanks :)

Dennis

04:23

I'm a knucklehead. I posted the older, slower, longer version.

Alex A.

It happens.

Dennis

It's remarkable that the sequence is not increasing.

Alex A.

Right?

Sp3000

@Dennis It's kind of like "smallest number with exactly n divisors" - the exactly kinda screws things over :P

Dennis

OK, maybe remarkable is a string word. It's counter-intuitive though.

Alex A.

04:30

@Sp3000 Do you like my second challenge?

Sp3000

It's okay :) I wouldn't say "like" though because anything primes in Python is insta +20 bytes

:P

Alex A.

You could use something other than Python ;)

Sp3000

:/

Alex A.

Or not

Dennis

Btw, this was my first code golf since August 24. The PLQ leaves time for nothing else...

Sp3000

04:32

You don't say :P I've pretty much given up on PLQ though

Alex A.

Wow, seriously! That's an entire 3 days.

Dennis

@AlexA. Yup. And my third since August 17.

@Sp3000 I try to get loose, but I can't. :P

Alex A.

Martin has created an addiction.

Dennis

1. I'm browsing esolangs to crack answers. 2. I find a language I really like and post a new answer.

It's a vicious cycle.

Alex A.

Sounds fun though.

Dennis

04:34

Well, let's say Martin made it worse. :P

Alex A.

Haha

Sp3000

I see 3279 byte Wordys and I fling my arms up in the air and go "Nope."

Reto Koradi

@Dennis Yes, I was worried about you when I posted a CJam answer on a challenge before you. ;)

Alex A.

Dennis

Haha.

Wordy was a fun one. I initially wanted to make a Malbolge/Wordy polyglot, but I'm not sure it would have stayed under 30k characters.

Sp3000

04:37

No fun for the crackers though :P I never touched it

Dennis

Well, when I wrote it, you had just cracked my Malbolge/Wordfuck polyglot, so I made it extra special. :P

Still, over 50% needle.

Sp3000

"needle"

Alex A.

As opposed to haystack?

Dennis

That's one big haystack. — Sp3000 Aug 20 at 5:51

I guess that depends on the size of the needle. — Dennis Aug 20 at 5:53

Alex A.

Found the needle

Dennis

04:51

sigh 5 minutes to write an answer, 15 minutes to make it Foo(l)proof.

Alex A.

^ story of the PLQ

So it looks like our name probably won't be changing, at least not anytime soon.

22 for no change, 13 for highest voted replacement name

Dennis

If those are the options, I prefer no change.

Alex A.

You mean those two? Because you posted an option yourself.

Dennis

Those two, yes.

Alex A.

Not a fan of Code Sport, eh?

It's not my favorite.

Dennis

04:56

Nope. If it was Code Sports, I could get behind it.

Alex A.

HOORAY.

Somebody agrees with me.

The lack of the s bothers me quite a bit.

Dennis

It's BE. But since this is a US site...

Alex A.

I guess in British English they say "sport," but in American English it's usually "sports." This makes it more amusing then that "Code Sport" was suggested by an American.

I just suggested the s

Dennis

I just upvoted your suggestion. :P

Alex A.

@Dennis I originally though you were American because you use American English.

Dennis

05:00

OK, just posted a 17 byte PLQ answer. After the 0 and 3 byte answers, I think this is the shortest yet.

Alex A.

:D

Dennis

@AlexA. AE is simply the lingua franca of the internet, which has taught me 90% of the English I know.

Browsing sites like this one, you find a lot more AE than BE.

Alex A.

I feel like there's a large presence of BE here though.

More than other places I've been on the Internet.

Dennis

Presence? Sure. But AE still dominates.

Alex A.

Yeah, true.

Dennis

05:10

Btw, for a German, the complete lack of correlation between orthography and pronunciation is quite disturbing.

Alex A.

What do you mean?

Dennis

There are no ghotis in German. Or Spanish. Or any other language I know.

Reto Koradi

Probably that you can't tell how something is pronounced from the spelling. Or vice versa.

Alex A.

Ghotis...?

Dennis

Ghoti

Ghoti is a creative respelling of the word fish, used to illustrate irregularities in English spelling. == Explanation == The word is intended to be pronounced in the same way (/ˈfɪʃ/), using these sounds: gh, pronounced [f] as in tough [tʌf]; o, pronounced [ɪ] as in women [ˈwɪmɪn]; and ti, pronounced [ʃ] as in nation [ˈneɪʃən]. However, linguists have pointed out that the location of the letters in the constructed word is inconsistent with how those letters would be pronounced in those placements, and that the expected pronunciation in English would be "goaty". For instance, the letters ...

Alex A.

05:14

Oh I see what you mean.

Yes.

Dennis

Alex A.

Haha excellent and relevant!

Dennis

For a German, there should be an I in team.

Alex A.

Why not "teem"?

Would that be acceptable in German? (I don't know what "i" sounds like in German.)

Reto Koradi

To be fair, German has its own set of difficulties. Like using 3 different articles for objects, where there's absolutely no logic behind which one is used for what. Why is a fork female, and a spoon male?

Alex A.

05:22

Seems like it should be the other way around.

Just saying.

Dennis

The letter I is pronounced ɪ in German and Spanish. ee would be a long e.

Yes, the articles are nuts.

Alex A.

And don't you capitalize all nouns?

Dennis

So are the tenses. Simple past, in particular.

Yes, that's correct.

English capitalizes almost nothing. But it's I, not i. Never got that.

Alex A.

I'm glad I studied Japanese and French rather than German. While Japanese comes with the little circle of hell that is kanji (borrowed Chinese characters), it's otherwise pretty simple.

@Dennis Self-importance? :P

Dennis

We had to choose between Latin and French in high school. I chose Latin.

Alex A.

05:29

In high school I got to choose between French, Spanish, and Japanese. Then I took an intensive French course in college.

Why Latin?

Dennis

Why is it an option or why did I choose it?

Alex A.

Why did you choose it?

With French at least you can communicate with living people...

Dennis

I'm not particularly fond of the French language. I would have chosen pretty much any other option.

Alex A.

It's not fantastic. What about it don't you like?

I chose it in college because it was offered at a convenient time. Japanese or Persian would have interfered with my work schedule.

Dennis

For starters, I don't seem to be able to pronounce it properly.

Alex A.

05:38

Nobody pronounces French properly.

Dennis

Don't tell the French.

Alex A.

I won't.

Reto Koradi

I really haven't used French since school.

Forgot almost all about it.

Dennis

I think I forgot Latin in a year.

Alex A.

Haha. I've forgotten everything.

Languages, math, everything.

Dennis

05:43

I was pretty good at it, but nothing stuck.

I still know a little math. :P

Reto Koradi

I still sort of understand it when people around me talk French. But I wouldn't even try speaking it. I use English with the French people. Even though I learned French first.

Alex A.

@Dennis As a recent math PhD graduate I would hope you would.

@RetoKoradi Oh English isn't your first language?

Reto Koradi

No, my 3rd or 4th, depending on how you count.

Alex A.

:O

What was your first?

(and second, third, and fourth)

Reto Koradi

German. Or at least a variation of it. Then real German. Then French. Then English. We didn't start English until 9th grade.

Alex A.

05:47

Swiss German?

Dennis

German is neat. Travel 100 km in the wrong direction and you don't understand a word.

Alex A.

Haha

Dennis

Seriously, the regional dialects are insane.

Reto Koradi

Right, Swiss German.

I heard that there are places in Austria where people from one valley don't understand people from the next valley.

Alex A.

That wouldn't surprise me.

Brb have to restart my computer

Dennis

05:51

It starts with telling the time. Not even 20 km from home, people say three quarters 10 instead of a quarter to 10.

Well, that's it for today. Good night y'all!

Reto Koradi

Good night.

Alex A.

Goodnight!

Martin Büttner

07:49

@Dennis it's only consistent ;)

Martin Büttner

08:08

@Dennis apparently it's even worse in Belgium

feersum

09:10

I'm pretty sure my program does not output this..

imgur.com/vikybDy

Beta Decay

What is your program @feersum?

flawr

@@AlexA. Did you ask about the ebook reading device?

feersum

golf.shinh.org/p.rb?Joseph+and+the+Amazing+Technicolor+Balls

flawr

10:07

@AlexA. We were discussing about recursion in matlab using anonymous functions. I asked on SO, here the results: stackoverflow.com/questions/32237198/…

And I have to admit, that I did not expect an article about functional programming in Matlab that exclusively uses anonymous functions.

The code then looks even worse than Lisp.

Martin Büttner

10:41

the very first PPCG answer in Labyrinth:

1

A: Let's see some action!

Labyrinth, 394 386 bytes I proudly introduce... <}74}}:23}29}59}}}}}}}:111_}}}}:::::::23_}:111 ? @ :" }}_47}_95 3""""""""( _ : } _ } {=} 2 23_}29_ _ ; : \ 0 ; 3 +_( 3_" 60{ .{.{. "-_95:}}"" 2 0 ) 2 " _ _ { "" _ : 2 _ ."(; } 3...

Martin Büttner

10:54

okay, I think I'll be posting hello world in a bit

could someone give it another read?

16

A: Sandbox for Proposed Challenges

"Hello, World!" code-golfhello-world So... uh... this is a bit embarrassing. But we don't have a plain "Hello, World!" challenge yet (despite having 35 variants tagged with hello-world, and counting). While this is not the most interesting code golf in the common languages, finding the shortest...

Sp3000

11:10

(I still think it looks good, but I was waiting for someone else to respond :P)

Mitch Schwartz

@feersum is that reproducible by submitting the same code?

feersum

11:45

@MitchSchwartz no

ProgramFOX

@MartinBüttner Looks good to me.

Martin Büttner

I'll post it after lunch then

Martin Büttner

12:24

posted

New Main Posts

2

Q: "Hello, World!"

So... uh... this is a bit embarrassing. But we don't have a plain "Hello, World!" challenge yet (despite having 35 variants tagged with hello-world, and counting). While this is not the most interesting code golf in the common languages, finding the shortest solution in certain esolangs can be a ...

code-golf hello-world

3

georgeunix

12:55

@MartinBüttner I would have got that in the bag if it weren't for that meddling Stuck: codegolf.stackexchange.com/a/55441/43338

Martin Büttner

"This is not about finding the language with the shortest "Hello, World!" program. This is about finding the shortest "Hello, World!" program in every language. Therefore, I will not accept an answer."

(that said, /// would also be 13 bytes)

Geobits

Asked: 31 minutes ago. HNQ status: 2nd (so far)

Optimizer

PHP is more of a programming language than HTML ?

Sp3000

>>> crack('Hello, World!')
   Works in ///
   Works in dogless
   Works in itflabtijtslwi

georgeunix

@MartinBüttner Thanks

Sp3000

12:57

@MartinBüttner ^^ Add another to the list :P

New Main Posts

0

Q: Turn a three column file into a table

Given an input txt file with three columns, you must generate an output file where: The values on the first column of the input are going to generate the rows in the output. The values on the second column of the input are going to form the columns in the output. And the third column is going ...

java

Geobits

Works in English

georgeunix

Well, it says it must 'print to STDOUT'

HTML can't do that

@MartinBüttner as it's has the code-golf tag, I assumed it was a competition. Still, it can be xD

Transcript for