The Nineteenth Byte

well, I was going to try something new: recycling old code. you may choose any code block on PPCG (that is actually code, and not test cases or something), scramble it (i.e. form an arbitrary anagram), and that anagram must print 2015 in some language. shortest answer (i.e. shortest snippet) wins. I was also considering to add some other restriction to make it harder.

Joe Z.

ah

are you sure that none of the 2014 entries would be candidates?

I could imagine one of them might be rearrangeable in some way

especially the ones that use bases

Martin Büttner

I'd need to check, but even so, they aren't necessarily the shortest

Joe Z.

true

mine is more traditional (I suppose), but yours is a nice idea too

Martin Büttner

will it be restricted source again?

Joe Z.

no, not this time

I can tell you that much, at least :P

Sp3000

1:05 AM

CJam sounds like if you have KF and can dump the rest...

feersum

the real contest is scraping the site for short code blocks I guess

Joe Z.

yeah

Martin Büttner

@Sp3000 yeah, but then you still need to find a short one that contains those two and allows you to dump the rest

FryAmTheEggman

^KF is likely to be in Pyth code somewhere ;p

Martin Büttner

@feersum that's true

Joe Z.

1:06 AM

DiBBCp was genius, for ASCII-only

Sp3000

1

A: The 9 Billion Names of God

Pyth, 34 characters Kf<T"n"GJKFbJI>lb9Bb~Jm+bdfXVb*Y3K Explanation: Kf<T"n"G K = list of letters in the alphabet before n. JK J = copy of K FbJ For b in J: I>lb9B If length of b > 9: break b print(b) ~J J+= ~Jm+bd J+...

Pyth code indeed

I don't know enough CJam for dumping though

Martin Büttner

but people didn't mind scraping SO for the handful of duct tape coding challenges we had

KF"fTnGJKbJI>lb9b~Jm+bdfXVbYK"B3<*

Sp3000

That works :P it's a start

Joe Z.

also, the contest won't just be about making code that prints 2015 this time around

I'm not going to do that again

Martin Büttner

debugging scripts called from .qmail files is bloody annoying

@JoeZ. graphical output? ;)

hm, maybe if I require answers to be in the same language they're recycling? That'll be annoying with polyglots again though.

@feersum I'm also pretty sure it won't take longer than 24 hours for someone to write a snippet using the SE API which finds the shortest 100 (or so) code blocks on the site.

I posted the draft... the bot should post it here soon

New Sandboxed Posts

1:30 AM

0

A: Sandbox for Proposed Challenges

Recycling for the New Year code-golfnumberrestricted-source A good resolution for 2015 is living more sustainably, and we'll start by recycling. Recycling code, that is! Your task is to print the number 2015 to STDOUT, with or without trailing newline. However, you have to so by recycling some...

Sp3000

@MartinBüttner I'm guessing output should be the same for any input? Just checking because dumping q or l in CJam might be hard

Martin Büttner

yes, the submissions shouldn't take any input

(I mean they can read STDIN, but they mustn't crash if it's empty)

Sp3000

... so using q/l is allowed then? If it doesn't crash

Martin Büttner

clarified

well it's not

l crashes I think

not sure about q

yeah, q works once

l doesn't

Sp3000

Well basically I'm asking because codegolf.stackexchange.com/a/26833/21487

Martin Büttner

1:39 AM

I don't think that works

how would you get rid of the STDIN contents?

Sp3000

Well I was just showing why I was asking :P

Had you said input from STDIN is always empty or something it'd have been viable

Martin Büttner

true

also added a note about using the current date

otherwise an answer from last year can be reused

Rainbolt

Team PPCG assemble. I have a challenge for you.

Martin Büttner

hm, I'm growing more fond of the idea to require the source to be valid in the same language as the answer

@Rainbolt you should sandbox it

Rainbolt

Create some data in Excel and then hide the columns. Create a line chart using that data. Get the value to appear on the line chart when you hover over it.

You have the thanks of many tired developers if you solve the challenge. And you win.

Martin Büttner

1:47 AM

I almost would have written "I don't really Excel"... sorry Rainbolt, I don't really use MS Excel.

FryAmTheEggman

-1 for not using a pun ;)

2

Rainbolt

It is sad that shortly after I fix this problem, we'll be moving to strictly web-based reports. Goodbye Excel, hello modern world.

But this does not absolve me of fixing the problem :(

Sp3000

You know, if I get this submission working this could be a pretty hilarious solution :P (might save it for when it comes out)

Martin Büttner

@Sp3000 "this submission"?

Sp3000

For recycle

Martin Büttner

1:51 AM

but which one?

is that a secret? ^^

Sp3000

Saving it for when you post the question ;D

Martin Büttner

lol k

@JoeZ. are you still willing to tell me about your challenge in private? :P

feersum

whenever I try to make a graph in Excel it's PLEASE JUST KILL ME NOW

Sp3000

Yep it works, now to wait for new years :P

Martin Büttner

how long is it? does it ruin the challenge? :P

you seem so happy about it.. any rule abuse? :D

Sp3000

1:57 AM

Shorter than the one above, is in the same language as it originally was. I'm just happy because of the irony - I can tell you in secret but I don't know how

FryAmTheEggman

^ Quick everyone spam martin >:D

Sp3000

Oh right, email - that thing

Martin Büttner

@FryAmTheEggman I can (and will) delete that address any time

@Sp3000 have you got it?

feersum

that's why the spam needs to be quick!

Martin Büttner

hope you've got it, it was getting close to the 2 minutes ^^

Sp3000

1:59 AM

Yup got it

Sent

Martin Büttner

thanks

anyone know how I pass the STDIN of a bash script on to a program called within that bash script?

Sp3000

But yeah, re: email I think I like the same language restriction. Makes it more interesting

esp. in Python and such with keywords - maybe a per-language scoreboard?

Martin Büttner

yeah, I'll include one

@Sp3000 updated spec, is that part clear?

well, I should really sleep

good night everyone

Sp3000

2:18 AM

Valid syntax eh, interesting :P

Yeah I think that works, just make sure the original code runs without crashing in the language you chose, right?

(But yeah, night :P)

Rainbolt

2:40 AM

I crumbled and asked a real question: superuser.com/q/853064/399241

Joe Z.

3:44 AM

@MartinBüttner I'd have to find a channel for that first

xnor

anyone know what version python starting having enumerate return an iterator?

or, did it ever return a list?

alternatively, is there a shorter way to convert an enumerate object to a list without the six chars of list() in Python 2?

starred assignment doesn't exist, unfortunately

FryAmTheEggman

I think you could use []?

xnor

how do you mean?

FryAmTheEggman

Nope, I was wrong

Hrm, I don't think you can get shorter than list... what are you enumerating?

xnor

a variant of feersum's minsweeper answer: codegolf.stackexchange.com/a/42344/20260

except I'm not wrapping the enumerate in a dict

FryAmTheEggman

3:59 AM

hrmm.. do you use range or len on what you are enumerating?

You might be able to use zip

xnor

you mean zip(range(len(L)),L)?

FryAmTheEggman

Yeah

xnor

that works but is long

FryAmTheEggman

Guessing you tried already ;)

xnor

longer than list(enumerate(L))

FryAmTheEggman

4:00 AM

Yeah that's why I asked if you used range/len elsewhere

xnor

nope

feersum

4:22 AM

R=range requires 3 uses of ranges to save bytes

FryAmTheEggman

Not if it lets you use a different algorithm. It saves an additional 2 bytes in this case, which means it only needed 2 (not that it matters ;p )

Peter Taylor

8:41 AM

Must dash but, Martin, what on Earth is the ellipse in your hat image? It looks nothing like the dimensions of the ellipse in the SVG.

Martin Büttner

10:21 AM

@PeterTaylor ummm, I didn't participate in that challenge

@JoeZ. I've got a temporary email address, which I can give you if we're both in chat at the same time

Peter Taylor

Ah, sorry. I saw that you were making queries about integer dimensions, then I saw that you'd edited the question and an image had appeared, and I didn't check whether there had been multiple edits.

Martin Büttner

@Sp3000 well it could still crash with a runtime error

Peter Taylor

@Optimizer, what on Earth is the ellipse in the hat JPEG? It looks nothing like the dimensions of the one in the SVG.

Martin Büttner

@PeterTaylor no I only removed an obsolete sentence

ProgramFOX

@Howard Congratulations on getting 20k!

Martin Büttner

10:25 AM

@PeterTaylor @Optimizer also, the corner radii look closer to 10 than 20

Peter Taylor

Talking about corner radii makes a nonsense of the original "an exact construction as appearing below", which is still immediately above the SVG. The SVG mainly uses Bézier curves, not circular arcs.

Makes me wish I hadn't withdrawn the vote to close as unclear until the spec was finalised, because in effect I've written an answer and the spec has changed in the meantime.

Optimizer

The SVG draws a bezier curve only for the two dark grey blocks, for rest all , its a normal border radius

@PeterTaylor which one are you talking about ?

Peter Taylor

The big red one which is cropped at the bottom

Optimizer

@MartinBüttner that is definitely not 10. Look at one of the dimensions, its 8 and the border radius is not a full quadrant of a circle

@PeterTaylor that is a circle representing the arc in dark grey color

Peter Taylor

The SVG draws Bézier curves for every single curve apart from one ellipse (the top of the middle grey bit) and two circles (one red, one yellow). And that one ellipse is much much smaller than the red outline in your JPEG.

Optimizer

10:38 AM

Obviously SVG draws a bezier curve because the complete rectangle is a single curve

but the question is that is that bezeir curve not a circle ?

let me draw that in CSS and tell.

Peter Taylor

The path spec supports elliptical arcs, but this particular SVG doesn't use them.

Read the SVG. The only instances of the character a are in attribute names, element names, RGB values, and the transform.

Optimizer

10:54 AM

@PeterTaylor let them be bezeir, that does not mean they are not a circle. See this pen. - codepen.io/scrapmac/pen/yyOmWW

And yes, the 20 is too high. I blame late night spec writing.

Peter Taylor

It does mean they're not a circle. It's impossible to create a perfect circle using non-rational Bézier curves.

Optimizer

Look at that codepen. All three div are perfectly overlapping the curves

Martin Büttner

@PeterTaylor for the recycling spec draft in the sandbox... do you think "syntactically valid" could be ambiguous/ill-defined for some dynamic/esoteric languages?

Peter Taylor

@Optimizer It's an approximation. math.stackexchange.com/a/873589/5676

Martin Büttner

11:14 AM

@Sp3000 now with the same-language rule I was considering to allow any code block (even if it's test cases or something), but I'm afraid there'll be an example output like 0512 in some code block on the site.

Sp3000

Hmm I think anything that was code or meant to be code is fine

Test cases don't seem like they'd end well

Martin Büttner

probably

Sp3000

You could do a *2 penalty for using a different language block?

Or something

Martin Büttner

hm, no, the simpler the better, I think (specwise)

Sp3000

Ahaha k. But yeah I'm trying to look into example output blocks atm

Peter Taylor

11:27 AM

@MartinBüttner If you're worried, state that it may not produce compile-time or run-time errors of any kind.

Sp3000

Hmm different questions format cases differently - some separate input/output, per case, and some have everything in one big code block

Martin Büttner

@PeterTaylor I was thinking that would run into some form of halting problem.

Peter Taylor

Why? You're expecting all submissions to halt, surely?

Martin Büttner

but the original source doesn't necessarily halt

(I'm talking about errors in the original code block, not the submission itself)

Peter Taylor

11:42 AM

Ah. "You must be able to show that the original source block is syntactically valid in your language. In the best case you can do this by showing that it runs without compile-time or run-time errors."

Martin Büttner

thank you

Peter Taylor

12:04 PM

So, longer than JS...

Optimizer

@PeterTaylor cool

Peter Taylor

Added an outline image, so you can see what I mean about the ellipse.

Optimizer

Is it possible to approximately map it to a circle ?

Peter Taylor

12:36 PM

Not very well. If you fit a circle through the top of the ellipse and the points where it meets the other mid-grey bit then there's an error of about 3 pixels in my 50% (713-pixels wide) export.

Optimizer

I see. IMO, it would been harder to spec out an ellipse there.

Peter Taylor

^ This is what I don't understand. Yesterday you said

> the reference SVG is there, which is vector
> so I cannot see anyway which can invalidate any existing answers if they followed the spec of "exact construction"

but then you went on to add a second, contradictory, spec, and today you've made a series of comments which suggest you don't actually understand the SVG spec. `<ellipse cy="43" cx="101" rx="53" ry="27.4" fill="#424143"></ellipse>` isn't exactly hard to convert to a bounding box.

Optimizer

12:53 PM

I mostly meant on the lines of , there is a vector graphic file present which can be used to get proportions etc.

not that it should define what types of curves to be used

3 px error in a 713 px wide means less than 1 px in the original resolution of 192 by 192 pixels

Peter Taylor

This must be some meaning of "exact" with which I'm not familiar.

Optimizer

@grc your star still is smaller and not centered

grc

@Optimizer have you seen my comment?

Optimizer

Wait, font ?

grc

is that cheating?

Optimizer

1:04 PM

now I can see why its smaller and towards left

not really.

grc

yeah I'm using unicode for the star

Optimizer

I just did not notice that u are using a font there

can you move to any other unit than px in the 105px font size ?

grc

do you think a monospace font would work cross-browser?

Optimizer

I think percentage based will fit well

it should

I am on Firefox only. so cross-browser is not the case

but both my windows 7 and 8.1 laptops show the start like in the image i commented

grc

try the snippet now

I'm using '790% monospace'

Optimizer

1:08 PM

nope

exact same as before

grc

how about now?

Optimizer

are you using 105px arial or is the script not updated ?

grc

yeah it's arial

Sp3000

2:02 PM

(I hope we don't get a draw X wave of questions...)

2

Optimizer

2:15 PM

yantra is hard

undergroundmonorail

2:28 PM

hello

ProgramFOX

3:05 PM

Anyone who wants to play a chess game? en.lichess.org/tournament/OlYdufdK

(password = winterbash)

undergroundmonorail

can i play even if i'm terrible at chess?

ProgramFOX

Every SE member can join if they want.

undergroundmonorail

neat

oh god the signup captcha is a chess puzzle

1. i'm bad at these
2. i think computers are good at these

ProgramFOX

Yes, and sometimes they are difficult ones.

Optimizer

@grc can you see if the star in this renders correctly for you

grc

3:09 PM

@Optimizer not quite

Optimizer

screenshot please ?

grc

Optimizer

ok, so relying on font-size is not the right way

grc

does compatibility matter for code golf?

Optimizer

well, its not an answer if it only works on my computer

FryAmTheEggman

3:15 PM

^ I think if you can specify precisely when it will work, you are usually ok

chrome ?

Chrome 31, Windows

non-standard CSS properties.

but what's with that wiggle wiggle wiggle ?

Peter Taylor

Wiggle wiggle wiggle?

Optimizer

3:17 PM

its arms are wavy

the star

@grc can you refresh now and see ?

grc

it looks slightly better but not enough

Optimizer

I just want to know if the size is consistent using real life measurement scale millimeter

Peter Taylor

It's what I commented on grc's answer: the star in Arial Unicode MS has a different shape.

Optimizer

@PeterTaylor oh, so its not from my codepen ?

Peter Taylor

It is. The two of you have the same problem.

Optimizer

3:22 PM

so its using arial in your browser.

it uses segoe for me

let me fix that

@PeterTaylor how does it look now ?

grc

oh well, I'm off to bed

good night

Optimizer

gn!

@ProgramFOX how did u get waffles on PPCG :O

Martin too

ProgramFOX

hahaha :D

The real question is: should I tell you?

Optimizer

I know :/

but does it work without suggesting an edit but directly editing it ?

ProgramFOX

Yes.

Optimizer

3:31 PM

nice

I was up late trying to figure out hairboat, but it was too late and by the time i woke up, someone figured it out. Although, he also did not get a eureka for it

Peter Taylor

@Optimizer Identical to before.

Optimizer

can you check which font is actually being used ?

Peter Taylor

Arial Unicode MS is being used for one glyph which isn't in the selected font, Times New Roman.

Optimizer

weird..

When I say Times, why does it choose Arial

your system does not have Times New Roman installed ?

Peter Taylor

It does, but (its version of) TNR doesn't include the star.

Optimizer

3:39 PM

oh!

is that okay enough limitation ?

Peter Taylor

What OS are you testing with that has a version of TNR with a star?

Optimizer

Windows 7 and 8.1 both work

oh, it falls back to Segeo

Segoe UI Symbol

Peter Taylor

I'm on Windows 7

Optimizer

so its not falling back to Segoe UI ?

it looks a bit tad weird for me in Chrome. But it still says Times new roman as being used.

Peter Taylor

No

Optimizer

3:58 PM

@PeterTaylor did you see the latest answer. SVG is down to 470 bytes.

I am testing via browser stack on various configurations..

yeah, Chrome has some issues on windows 7 and 8 in the size and shape of the star

Martin Büttner

4:27 PM

...

Optimizer

LOL :(

weird , I am not getting my selfie hat ..

feersum

@MartinBüttner I just checked the input data of polyominoes... and zero outputs are because the sum of lengths is odd

Martin Büttner

@feersum huh, that shouldn't have happened

or did are talking about some test data you generated yourself?

feersum

never mind, I think I missed coping the last number.

copying

Martin Büttner

3 digit numbers in CJam are so annoying... they're really long, but too short replace them by some computation

Rainbolt

4:39 PM

@MartinBüttner That seems to happen to you far more often than it happens to me

Martin Büttner

@Rainbolt well, I've got a few more (especially recent) questions and a lot more answers, so I'd expect anything voting related to happen to my posts more often ^^

FryAmTheEggman

Currently, Stack Overflow is that it summarizes the seven million or more questions, each supports the programmer of the month 44 million people. - Oh Google translate ;)

Also, how did martin's serial voting amount to 53? Was he serially up and down voted? o.0'

Lembik

hi all

Rainbolt

4:55 PM

How many points is a downvote?

Is it -3 or -2?

ProgramFOX

-2

Rainbolt

If I reversed a downvote into an upvote, you would then net 7 points.

ProgramFOX

If you compare with the reputation after the downvote, yes.

Rainbolt

So here's a theory: whomever serially voted on Martin's posts had already downvoted a few of them. So when it got reversed, he lost an odd number (53)

FryAmTheEggman

But does the algo take that into account? Wouldn't the downvote still be reversed?

Rainbolt

4:57 PM

I'm thinking he was downvoted slowly over time by some user. The downvotes were NOT serial.

And then one day the user upvoted all of Martin's posts.

Those upvotes WERE serial.

Lembik

I have a coding question

I would like to iterate over lists of length 12 where the elements come from [-1,0,1] and each list has exactly four 1s, and four -1s

feersum

does it involve inner products of 0s and 1s?

Lembik

is there an elegant way to do this?

@feersum almost :)

@feersum although I think the stereotyping is mean :) codegolf.stackexchange.com/users/9206/lembik?tab=questions

Optimizer

elegant ? on PPCG ?

Lembik

PPCG is beautiful AND elegant :)

what do you mean?

feersum

5:00 PM

elegance is one byte shorter and 100 times slower!

FryAmTheEggman

Oh, apparently if you vote too much on one person, each kind of vote you passed is reverted

Geobits

If each list has four 1, four -1, and four 0, isn't that just all permutations of [-1 -1 -1 -1 0 0 0 0 1 1 1 1]?

Optimizer

no, its 19 bytes shorter and 40000 times slower

Lembik

@Geobits yes it is

FryAmTheEggman

meta.stackexchange.com/a/126857 @how does the system detect voting fraud

Lembik

5:21 PM

@Geobits err.. no it isn't :)

you want [-1,-1] to be the same as [-1,-1]

I solved it in any case

Geobits

Ah, right, good point.

FryAmTheEggman

There are 34650 such lists, yeah?

Optimizer

so its unique permutations.

Lembik

right.. bpaste.net/show/c41ab450457a will do it

just involves some actual coding :)

Optimizer

12C4*8C4

FryAmTheEggman

5:26 PM

Funny, I used 12!/(4!^3) ;p

Optimizer

isn't first one (mine) more intuitive ?

FryAmTheEggman

Probably, but I was taught the way I did it... Both make sense, imo.

Optimizer

they are same

ofc.

FryAmTheEggman

Yours says pick from 12 seats the 4 spots for the ones, then pick from the 8 remaining seats 4 for the -1s

Optimizer

also, they haven't mentioned it anywhere, but Selfie hat requires and upvote. To prevent random self answers. Like other hats (read "Fear and Loathing") are quality driven hats

FryAmTheEggman

5:29 PM

Mine says arrange 12 unique things into 12 seats, then divide by the number of ways you can arrange 4 people into 4 seats 3 times

Optimizer

mine says pick all possible four locations, then pick all possible four locations from remaining ones and you are done

FryAmTheEggman

Yeah, I shouldn't have said ones or -1s but that's basically what I said ;P

Joe Z.

5:45 PM

@MartinBüttner I'm here now, but dunno if you are.

Optimizer

6:00 PM

@MartinBüttner What algorithm are you using there in the last sundays problem ?

Optimizer

6:16 PM

How can people be on 20 hats ?

Geobits

6:29 PM

I had 20+ last year, just haven't been bothered much about it this time around. Call me crabby this year I guess.

Optimizer

:|

Geobits

:D

Optimizer

what I meant was, that there is a hat which is awarded when you get 20 hats

get it now ?

FryAmTheEggman

Uh, SE server delay?

Optimizer

maybe

Geobits

6:33 PM

I'm now unsure of the question. I thought you were asking how people had managed to get that many hats in the short time it's been going.

Optimizer

I have 19 hats, so I could not possibly ask that

FryAmTheEggman

^ That seems a tad extreme ;p Anyway, he meant given that you get a hat for having 20 other hats, logically nobody should ever have exactly 20 hats

ProgramFOX

You only get the Aztec hat if you have 20 hats on one site. If you have, for example, 18 hats on site A and 2 different hats on site B (total 20 hats), you won't get the Aztec hat.

Optimizer

oh!

that is the answer I was looking for.

Geobits

Ninja'd by FOX ><

ProgramFOX

6:36 PM

lol

Optimizer

2 more 2 go

hairboat hat is impossible to get on PPCG

ProgramFOX

Indeed :(

Optimizer

we should ping her to post something here :P

she likes pancakes, so maybe a pancake based code-golf ? :P

FryAmTheEggman

Isn't there already a question on that?

30

Q: Coating Every Pancake

You have a stack of pancakes on a plate with a glob of syrup on top so thick it can't run down the sides. You won't be happy to eat until both faces of each pancake have at least touched the syrup, but right now only one face of the top pancake has. You know that the syrup will never soak throug...

code-golf

Optimizer

that one is weird. something involving creating the perfect pancake shape or spread

Transcript for