The Nineteenth Byte

Martin Büttner

00:07

@Dennis No that's fine. I just mean tailored towards those particular strings.

Dennis

00:29

@MartinBüttner OK, thanks.

I'll try it tomorrow if my idea makes any sense.

feersum

01:52

Does anyone know an easy tool to draw a graph (gives some vertices/edges)?

Sp3000

Dot?

Here's a guide with examples: graphviz.org/pdf/dotguide.pdf, download from graphviz.org

Thomas Kwa

Hey everyone, I was wondering if codegolf.stackexchange.com/a/50543/39328 should be community wiki or not, since I improved on the routine I found.

Nathan Merrill

02:34

@ThomasKwa I'd recommend making a meta post.

The comments on this post: meta.codegolf.stackexchange.com/questions/2155/… has a comment mentioning "how far is too far"

but that point is never fleshed out

why must chat be dead?

New Main Posts

03:16

-1

Q: An Easy Index Puzzle!

Difficulty: Easy Try and implement a function that takes the following input and gives the following output: INPUT: X Coordinate (integer) Y Coordinate (integer) OUTPUT: An index (integer) Where the output depends on tracing the input in a circular fashion around the origin. The index c...

math

Thomas Kwa

04:05

Hmm. Complicating the issue of whether to post as community wiki, I guess, is that there are standard routines in every language. On almost every quine challenge I see the words "modification of the standard quine" or something of that nature.

xnor

05:18

aargh

how on earth did someone do this in 42 chars of Python 2: golf.shinh.org/p.rb?permutater ?

it prints all permutations of the input string in lexicographic order relative to the input order

i can't see how to get anywhere close

feersum

ooh looks like fun

Sparr

@xnor neat site. thanks for the link

xnor

usually, there a bunch of solutions that are within 1 or 2 chars of the best

so it's really weird how on this one there's two exceptional solutions

Sp3000

What do you have so far?

Oh wow, 72 nice

xnor

unfortunately, I didn't save my code and the site doesn't either

but it's the natural thing with itertools

Sp3000

05:21

Ah right

xnor

usually, i think itertools can't win

Sp3000

Do they have to be in lexicographical order?

xnor

but unlike here, you have to write a program rather than a function

it compares the output directly as text for the test cases

feersum

one thing to remember is that on anarchy golf, all problems are kolmogorov-complexity

xnor

(which also makes me wonder about shenanigans of optimizing to test cases)

feersum

05:22

it only has to work fro those 2 inputs

xnor

unfortunately, hoge is not sorted

the scary thing is that i wasn't even to write code that lists the rotations of the input in 42 chars

Sp3000

Damn there's not even any hidden test cases, it's as feersum says :(

But it does have to be in the order given

It'd be funny if they managed to get it right with, say, import random

xnor

in theory you could guess which of the inputs you have with id(0)

the same two golfers also found super-short solutions in other languaes

Sp3000

Ruby 20 bytes o_O

xnor

how many bytes does it even take to read input and print in Ruby?

feersum

05:38

perhaps the next great filesystem exploit has been discovered

Sparr

does ruby have command line parameters like perl to supply default behavior?

feersum

in python for example, there is some trick that you should read multiline input by opening a file '../fd0' or something

Sparr

perl golfers often count "-p" as just one or two characters, despite it putting a 30-character loop-and-print around your code

feersum

there you have to use #!perl though it is often worthwile

Sp3000

05:51

Based on the front page (search for "file") and also the comments by leonid in this blog post, I think at some point it was possible to write files with one submission, and read them later with another submission

feersum

oh this is a quite old problem

xnor

ah, so it was an exploit

that makes me feel better

Sp3000

I'm not sure if that's definitely what happened here, but it seems likely

xnor

06:08

well, i'm going to claim that we win!

Peter Taylor

@trichoplax Did you know that some of Perl's syntactic design decisions were taken to allow a friend of Larry Wall to write poetry in it? perlmonks.org/?node_id=451207

2

xnor

07:13

i'm currently looking at: golf.shinh.org/p.rb?DozzFizzBuzz

since it lacks a description, it's a generalization of FizzBuzz where each prime has its own four-letter words

and among the primes, the first letter cycles through 'DFBMPQD'

and the second letter through 'oiuea''

and then 'zz'

each prime-word appears once, so 4 is Dozz, not DozzDozz

Sp3000

Trying to beat 120?

xnor

just got it!

Sp3000

:P nice

xnor

still 3 more days until it closes though, might not hold

isaacg

I had an idea for a challenge, as follows: For each programming language, find the shortest string in that language that can be output by a program shorter than the string. Ties are broken lexicographically.

How does it sound?

xnor

07:20

hmm

lemme think about Python

Sp3000

Like, print id?

xnor

-10 has length 3, and can be done as ~9

oh, probably can't REPL

isaacg

Right, no repl

xnor

@Sp3000 that's a great one!

this seems way easier for languages with automatic output

isaacg

Sure, in Pyth it's T

I want to emphasize the language by language aspect, not overall length.

xnor

07:23

so a Rosetta Stone?

it will be trivial for lots of languages

wait, it's the output length that matters

so print id is probably too long

isaacg

Perhaps

Probably is.

xnor

ok, this is more interesting that i thought at first

print'1'*12

Sp3000

print 1e7

isaacg

I want to emphasize language by language winners - not rosetta stone, which is 'do this task in many langauges.'

Is there an existing tag for that?

xnor

@Sp3000 for the first time in golfing history, we're happy e gives a float

Sp3000

07:25

Indeed :P

isaacg

If it didn't, 1e9 would be just as good

xnor

@isaacg in theory, all challenges are compared language-by-language, depending on who you ask

isaacg

Sure. OK, I'll right up a proposal for the sandbox.

xnor

for golf.shinh.org/p.rb?DozzFizzBuzz , it's funny how the run-time (which is publically displayed) spikes up for my answer and the 120

Sp3000

Alternative 10: print{.5}. Would lose on the tiebreak, but print-1e6 would be better anyway

Actually, are we counting newlines?

isaacg

07:29

Hm. I'd say newlines don't count.

The "length of an output" is actual length -1 if it ends in a newline, or actual length if it doesn't.

xnor

i think that might lead to exploits in languages that can quote multiple empty lines

isaacg

Or in benefits to running the program on a windows machine.

Sp3000

Broken lexicographically by program or by output?

(also I'm not sure about the fastest-gun-in-the-west problem with T/A/etc.)

isaacg

I'd say by output.

Sp3000

Then CJam would probably win I think

isaacg

07:36

I'd hope fastest gun would be solved by winner being most languages won.

Each individual language won gets you 1 point. Most points wins.

Invalidated at my discrecion if it doesn't look golfed.

Sp3000

Hmm well the problem is, say you post T for Pyth. Then I come along, late to the party and start on my answer. I'd probably refrain from posting T even though it's obvious because it feels like copying?

isaacg

Right.

But, you can post other language answers.

Sp3000

Doesn't really stop someone getting a major headstart by posting, say, 10 obvious answers and making everybody else feel like they can't post them

(It's a cool idea, but it kinda has the same problems as golf practice IMO)

isaacg

I could just say '1 answer per day' or something.

Sp3000

Hmm that's an interesting idea...

Katenkyo

07:43

@isaacg Even for people joining after 5 days?

isaacg

Maybe just no more than 1 answer on the first day.

Katenkyo

Or will it be something like maxAnswer per Post = days since start

isaacg

Hm, @Katenkyo I like that the best.

Katenkyo

It would not limit new challengers more than first days ones, and still be 1 answer/day FGOTW

@isaacg Nice :)

Katenkyo

08:19

Controller will soon be over for my koth

Finally !

By the way, I have a doubt on it... Should I put friendly fire on or not...?

Lembik

08:43

chat is moving to New York??

Katenkyo

@Lembik How that?

Optimizer

@Lembik Will we all be moved along with it too ?

Lembik

a notice came up saying it was moving from OR to NY

Katenkyo

09:00

@Lembik Where is that notice? Oo

randomra

I could swear the new CH challenge was asked with hexa digits before but can't find it

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

code-golf Shortest String with a Shorter Program In this challenge, you will find a program whose output is longer than the program itself. The winning program will be the one which gives the shortest output, while still being shorter than that output. Scoring: Each language will be considered...

Sp3000

09:18

@isaacg Can we edit a post to golf a program?

isaacg

Did you post a comment?

And give it a few days?

Sp3000

No I mean for the shortest string challenge

We can't post new answers until the challenge has been up for more days, but can we edit existing ones

isaacg

See most recent edit

Sp3000

Does "total number of submissions" mean that if someone else gets optimal, we can ditch that language and replace it with another?

isaacg

Hm. That's the current implication, but I'm not sure if I want that to be the case

What do you think?

Benefit: Reduces feel-bad for people whose anser is beaten.

I think it's better overall.

Katenkyo

09:26

I should try to golf one day :3

Sp3000

I was wondering whether disallowing it would make people think about their submissions more before posting

isaacg

@Sp3000 It incentivizes perfection on the first try, which I think is somewhat unfortunate.

Sp3000

Hmm well I'd be fine either way, just thought you might want to specify more clearly :)

isaacg

OK, will do.

Sp3000

btw there's better than T in Pyth :P

Optimizer

09:34

@Sp3000 is there ?

isaacg

Oh?

Optimizer

^ lang creator not sure

Sp3000

Makes it beat the CJam solution I had in mind, actually

Optimizer

A ?

B ?

Sp3000

No, for CJam it'd be W

Optimizer

09:36

why ?

Sp3000

"The second tiebreaker is the lexicographical ordering on the output. "

Optimizer

oh

so H ?

Sp3000

There's still one better :P

(also H is actually worse than T)

Optimizer

its not listed online then

isaacg

Well, Y and [ and ] and ( all work in pyth, but T beats them all.

Optimizer

09:39

Y is also worse than T

rcrmn

@isaacg regarding the rosetta-stone tag, I believe it's purpose is to mark the questions in which the winner is the one with the most languages used in their solution - you are just restricting which solutions everyone can use, which are the ones that are the shortest for each language

isaacg

That's one way to interpret it, I suppose. I feel like it's sufficiently diffferent from existing challenges with that tag that it shouldn't use it.

Sp3000

@isaacg Why not (?

isaacg

@Sp3000 Oh, right, I forgot that ( was so early lexicographically

rcrmn

I don't think it's that different from the others, you just add a restriction to the winning condition of typical rosetta-stone challenges

By merging code-golf with rosetta-stone

Optimizer

09:42

W in CJam

someone would consider A or B at first, but not that :D

randomra

right, tried to beat the pyth T in Cjam

isaacg

So, ( feels like the winner so far.

Sp3000

Nope, got better

Optimizer

@isaacg you should transfer ownership of Pyth to Sp3000 if you don't know the better one now :P

Sp3000

This one's not in Pyth :P

Optimizer

09:46

@Sp3000 then there is no point. right ? its intra language

Sp3000

I thought we were talking about overall winner

Optimizer

overall winner is the winner of most intra lang ones :P

isaacg

The real winner is going to be some language with a 1 character output with no input.

Like B****fuck .

Optimizer

GS : p

isaacg

What is GS on empty input?

Optimizer

09:48

nothing

Sp3000

@Optimizer So close :P

Optimizer

@Sp3000 still no ?

Sp3000

`

Optimizer

@Sp3000 its the same output

Sp3000

Backtick is before p

Optimizer

09:49

wait .. tie breaker involves code too ?

isaacg

Second tiebreaker

If outputs are identical

Optimizer

lol

Sp3000

Maybe isaacg should emphasise the tiebreakers more :P

Optimizer

I havent read the question yet :P

Sp3000

D:

randomra

09:50

@isaacg length of code doesn't matter? code_aa is better than code_b ?

isaacg

@randomra You're right, I should make that length of code, then lexi on code.

Optimizer

so its like day 1 - everyone will be winning in their languages.

no day 2

or more like, day 2 will be finding new esolangs

randomra

@Optimizer (removed) before I could make fun of that.. :P

@Optimizer exploring esolangs is always fun

Optimizer

so finally 5 tie breakers ?

isaacg

4 I think.

primary is output length, then output non-ASCII, lexi output, code length, lexi code.

Then time, I suppose.

randomra

09:55

beating a posted language could score double so you wouldn't always just search for an unused language

Optimizer

@isaacg that is 5

randomra

might be less fun though

Sp3000

I don't think that'd work with the current rules of editing posts if you're beaten

Optimizer

if you are beaten, even if you copy the same code, u posted it later..

isaacg

Perhaps an extra point for beating a solution that has been up for at least 2 days?

Optimizer

09:57

thats hard to track

Sp3000

I mean if you switch languages, you'd have to make sure people keep the original submissions up

isaacg

No, I can just look at edit histories.

Optimizer

switch languages as in ?

Sp3000

You could, but I'd imagine that would be even harder to track

isaacg

Yeah, it seems like more trouble than it's worth. I think improving on someone else's solution will likely be its own reward without needing to be built into the scoring.

New Main Posts

12:06

-2

Q: Rotation algorithm in PHP

Joel and Henry are playing a game. They have N balls numbered from 0 to N−1. Henry asks Joel to reverse the position of the balls, i.e., to change the order from say, 0,1,2,3 to 3,2,1,0. He further asks Joel to reverse the position of the balls N times, each time starting from one position furth...

php algorithm-advice

trichoplax

12:21

@PeterTaylor Thanks for the perl poetry link :)

New Meta Posts

0

Q: Does PPCG.se need a distinct icon?

Currently, we have a semi-default StackExchange icon, with letters in a sky-blue speech bubble. Can we, and should we, have something more distinctive?

discussion feature-request

VisualMelon

12:39

@Sparr did I hear something about your asexual bitstring family trees not being solvable simply with a minimum spanning tree of ln(1 / probably of string generation)? After drawing some pretty pictures and learning logs again, I can't help but think it is just as simple as building said spanning tree, and picking any string as the root node

(I'd love a "yes, you're missing something", or "no, it's that simple" response ;) - please don't tell me what I might be missing if you are confident that I am)

trichoplax

I guess you need to ask yourself whether the probability will be the same if you choose a different node as the root

(vague enough?)

VisualMelon

I have... I am uncertain about whether my thinking breaks down because there are more than one sets of steps to build such a tree

(i.e. I am finding the probability of building a tree any way, but I don't consider that there may be more than one ways)

... funny how being forced to explain things makes you realize how wrong you are, thanks @trichoplax ;)

trichoplax

:) Explaining is one of the best ways to learn

2

VisualMelon

that it is

Katenkyo

@trichoplax Words of wisdom

VisualMelon

12:47

@Sparr (feel free to ignore my last communique)

trichoplax

@Katenkyo Thanks :)

Agawa001

explaining is a way to expose off your mind failure

VisualMelon

(or to just realize there is something you meant to check but forgot to, let's not be too hard on my mind, it's got other things to think about than trees at the moment)

trichoplax

It was still your mind that identified the problem...

VisualMelon

that's the spirit, I'd have certainly got round to it eventually without trichoplax's help, I'm sure, maybe, perhaps

trichoplax

12:59

and I think (hope!) Agawa001 was talking about minds in general, not yours in particular ;)

Agawa001

i havent read before

i smell a star flood

someone isnt ready to explain , when he is learning, he just listen

Rainbolt

@Agawa001 You haven't read before? I feel bad for you.

Agawa001

@trichoplax im not stupid to talk trash about someone whom i owe nothing , if i do , there should be dozen of understandable reasons which push me to

and honestlt honestly , lots of idiocies are starred in this room

trichoplax

@Agawa001 I understand. I was saying that your statement about the mind was about minds in general, so VisualMelon would know it wasn't personal. I didn't think you were being confrontational

VisualMelon

no offense was taken, @Agawa001, just having fun

(what trichoplax said)

Agawa001

13:12

taking confrontations in chat too personnal , is like suiciding for losing videogame , im responding to ur last accusation , about me undersestimating peoples minds

trichoplax

@Agawa001 yes, anyone can star anything for any reason, so sometimes stars show something insightful, and other times just something odd or funny...

@Agawa001 I've read back and I can't see anyone accusing you of underestimating people's minds. If it looks like someone did, let us know in case it wasn't meant that way and we can clarify.

VisualMelon

my apologies, that wasn't intended as an accusation, my family lives in a constant state of sarcastic vitriol, when playing poker, I refer to myself in the third person as "idiot" and everyone knows what's going on, this is how I live my life. Sorry if it wasn't clear I was just having fun

(sarcasm turns out to not help in group projects, which is something I'm yet to quite get used to)

Geobits

Pfft, just sounds like the wrong group.

:P

Agawa001

someone starred my comment for purpose

trichoplax

It was a true comment - I'm not surprised someone starred it

Geobits

13:17

Yes, but for what purpose? I can't tell at all.

Agawa001

(starred as another idiocy)

4

VisualMelon

@Geobits yeah, not wishing to offend, it's the foreign students who just aren't used to how us British nuts speak that get confused by it, and it's to be expected, as any difference in culture can create problems without any ill-intent

trichoplax

Even within Britain there's a wide range of levels of sarcasm. Some people talk to complete strangers with viciously cutting sarcasm and the strangers don't always understand that it is meant as a way of connecting

Geobits

@VisualMelon I had a bit of a problem with that when I first moved to Japan, both speaking and listening. It's really easy to take a second language too literally :D

VisualMelon

yeah, as far as I'm concerned, when I feel comfortable with calling someone a "moron" and say things like "that work you've produced is just fantastic, we definitely don't need to re-write this section because you've clearly had a lot of time to work on it", then I know I'm comfortable with working with them ;)

Katenkyo

13:22

@Geobits I agree, that's, sometimes, the problem I have with English ^^

Geobits

Don't worry, native English speakers are often pretty bad at it, too, especially in written form ;)

VisualMelon

oh, yes, you need the intonation and rhy smile

Rainbolt

@Geobits So true. My boyfriend is a high school teacher, and I am totally unable to help him grade papers because I can't even read their handwriting.

trichoplax

I have the same problem even in spoken English, as I don't change my intonation or facial expression when being sarcastic

Katenkyo

Haha, I think it's the same everywhere !

Rainbolt

13:25

I used to have facial expressions, and then I started playing Magic: the Gathering

Geobits

@trichoplax I've learned to, depending on surroundings. My mom, for instance, *really needs the clues, or she just doesn't catch it at all.

trichoplax

I always thought that was half the fun of sarcasm

VisualMelon

yeah, and sometimes you just have to look serious and lay things on thick or it just isn't funny

Katenkyo

But I think it is much more hard to know that an english sentence is sarcastic in written form than in my mother language (French,France, not Canadian).

Geobits

It is half the fun, but sometimes you need someone to get it ;)

Vioz-

13:26

Hey y'all :)

Katenkyo

@Vioz- Hi

trichoplax

Hey @Vioz-

Vioz-

What's everyone up to?

VisualMelon

definitely not talking about sarcasm

Geobits

Starring nonsense and discussing the finer points of sarcasm, which is totally unrelated to the starred nonsense.

VisualMelon

13:29

naturally, how could anyone suspect such a thing?

trichoplax

Trying to simulate the rivulets that form on the inside of a glass of smoothie/yoghurt as simply as possible

Rainbolt

Trying to understand Telerik reporting tools

Geobits

A "perfectly smooth" glass?

trichoplax

@Geobits I don't have any of those, but I'm modelling one nonetheless...

Modelling makes it sound like I'm doing something intelligent. I'm playing with pixels...

Rainbolt

Anyone here attend Visual Studio Live ever?

Geobits

13:32

@trichoplax I have one you can borrow; I keep them out in the barn with my spherical cow.

trichoplax

Are they stored under vacuum?

Geobits

Well, near-vacuum. It's a space barn, so...

trichoplax

You have a perfectly smooth glass but only a near-vacuum??

Geobits

Budget cuts suck, man.

trichoplax

That must be awful

I hope you're not getting vacuum ablation on your spherical cow

Katenkyo

13:34

Wow, my "Hi" got starred...

Geobits

It's supposed to be a perfect vacuum, but we couldn't afford the monitoring software license any more. Turned out okay, it couldn't seem to monitor the vacuum without disturbing it in some way.

Katenkyo

It's so revolutionary that it got a star

trichoplax

@Katenkyo Yeah, we used to get such a flurry of 'Hi's that they ended up getting starred

Geobits

But no, the cow is okay. The sheep ableated long ago, though.

trichoplax

@Geobits You're getting dangerously close to being starred

Katenkyo

13:36

@trichoplax seems logic, it does have sense

I think we could star his whole story !

Geobits

:D

Please not the whole thing.

I don't want to bump off your game-changing 'hi' :P

trichoplax

Yeah, 'cos it would end up out of order on the starboard and then it would make no sense...

Geobits

God forbid something I say makes no sense. That would be terrible.

Katenkyo

@trichoplax A senseless starboard would be scary !

trichoplax

The starboard should always be a source of wisdom and contemplation

@Geobits I had less problem reading "make" but I honestly can't say which is "correct"

Geobits

13:40

I know... English is stupid, but I think it's right, even though the other sounded better to me.

trichoplax

I guess there are some implied words in there somewhere that would make "makes" make sense

Katenkyo

The stupidity of a language reflect the widsom of its users :)

trichoplax

I can only apologise for my last sentence

Katenkyo

@trichoplax No need, it was beautiful

Geobits

^

trichoplax

13:41

Aw thank you

I only seem to end up spattered all over the starboard when I least expect it

Geobits

Nobody expects the... Starrish Exposition...? (I'll see myself out now)

4

Katenkyo

@Geobits Haha, it is starred with the typo :)

Geobits

I know not what you mean. If I type something, it is what I meant to type. Anything you think might have been a typo is merely me changing my mind later.

This goes for all sit you ate shuns.

trichoplax

4

Hey! No metastarring!

Katenkyo

@trichoplax Meta-star is life, Meta-star is love

Compass

13:47

Hey all! Is anyone here a Java superstar? I had a quick crazy question.

Rainbolt

I wish imgur links would onebox when hovered over in the transcript.

4

Katenkyo

Will go for a quick pause in some secs, but i can give it a try

(if i can't answer before the pause, i'll after)

Geobits

I use Java, but I don't call myself a superstar. What's the question?

Rainbolt

I'm not a Java superstar, but I am a Java guy.

trichoplax

@Rainbolt Raise a feature request - I'll upvote it

Compass

13:48

So, I have here a comparator

it compares a bunch of longs, with the following scenario

if a = b, returns 0, if a > b, returns 1, if a < b, returns 0.

Obviously, that's not how it's supposed to work

but this is legacy code and it's been in there and was not causing problems for some unknown reason

Geobits

eww

Compass

I've gone ahead and fixed it, but would a mergesort actually "break" if that were the comparator?

Rainbolt

You can shortcut this with return a - b; but please continue.

Compass

@Rainbolt like i said, not my idea >_<

Katenkyo

@Compass It shouldn't

Rainbolt

13:49

Doesn't seem like a Java question lol

VisualMelon

(or better, a.CompareTo(b), I can never remember which way round they are meant to be anyway)

Compass

well it just makes no sense for me since logically, it should work, even if its... not pretty...

Rainbolt

Seems like an algorithms question that anyone but me could answer

Geobits

So, is the question, can you mergesort if the only comparison you have is "strictly greater than"?

Compass

but it just didn't sort one of our things and I fixed it but I don't know why it just up and broke after not being detected for... 3 or 4 years

i guess that could be the question

VisualMelon

13:50

Merge sort is usually stable, so I think it would break things

Rainbolt

I think it depends on your implementation of mergesort

VisualMelon

(e.g. if it's less than, but the merge treats it as the same, but it's behind originally)

trichoplax

Are any of the other places it is used dependent on the sort being stable?

Katenkyo

@Compass If the mergesort is functionnal with this comparator, it means it only try to see if a>b (try to sort ascending), so it's not a problem to fix the typo on a<b

VisualMelon

I am good at terms

Compass

13:52

OK. I'll spend a bit more time trying to see if something else is broken.

Just don't want to mark it as fixed if it could potentially not be the comparator being nasty

Geobits

I'm not sure it would 'break' anything, I think it may just not sort the entire list in certain cases. Especially if the list starts off partially sorted, I could see that going unnoticed for a long time.

Compass

well, unnoticed until the "sorted" pages are not sorted anymoreeee

and it is only reproducible on certain environments

trichoplax

If fixing it sets off a cascade of other bugs that also need fixing, that will be annoying but probably a cascade that needs to happen sooner rather than later

Compass

same data on my dev server, locally sorts fine, real server, breaks, test server, works fine, same builds everywhere!

stabs eyes out

Geobits

Hopefully it's not your eyes you're stabbing. That will make the problem a lot harder to spot.

trichoplax

13:55

Compass: stabs eyes out

Rainbolt

Turns out this has already been requested:

14

Q: Adding image thumbnails/previews to star wall in chat

Server Fault chat has turned into make-fun-of-bad-questions-with-funny-pictures chat and we're totally fine with that. We do still have a lot of serious conversation, but the star wall is constantly littered with links to images. I'd like to propose one of three ways to improve starred images: ...

feature-request chat chat-stars

Upvote this four year old post so that it may continue to be ignored.

Always makes me sad when something I wanna do was suggested four years ago :(

Compass

This comparator logic is making me lose my grip on reality! Wheee

trichoplax

@Rainbolt Upvoted, but doesn't look likely :(

But on the bright side there's a userscript answer...

Rainbolt

I don't understand what the issue is. I'm not a smart guy, but I can make lists of numbers and run code and see what comes out.

Compass

@Rainbolt I can do that too

What happens though is that on different machines with the same data

Machine 1 sorts, Machine 2 doesn't sort, Machine 3 sorts.

trichoplax

13:58

Focus on Machine 2. It has the answers

Compass

I focused on it. It just... didn't sort.

Machine 1 and 2 shared the same database and Machine 1 sorted and Machine 2 didn't.

Rainbolt

The comparator you described to us doesn't depend on the machine

trichoplax

What is different between them?

Geobits

Are the machines using the same java? Different versions may have different mergesorting internals.

Compass

As far as I know, nothing. Same Java version. Just me building locally vs the server building and deploying.

I'll have to find out what the servers have java-wise.

Rainbolt

13:59

Write out the list at each level of the mergesort

Transcript for