The Nineteenth Byte

because I can never be bothered to implement a flood fill in CJam... and you already have your implementation which you can just reuse from previous challengees

Optimizer

its not mine

its Peter's converted to cjam by user23013

its just a universal fact now, so no point in giving credit in each answer :D

and when did things become propriety here ? :P

Martin Büttner

they aren't, but I'm not really interested in reusing it

Optimizer

its not like every flood filling is same

you are basically arguing that, you have your own mapping method, which you can reuse

{ ... }%

Martin Büttner

no I'm arguing that I don't enjoy implementing flood-fill based algorithms in CJam :P

Optimizer

that is okay.

but your main reasoning for that was that there is a standard way to floodfill , so its not a challenge for you :P

oh, hey, we are at 3008 questions

Martin Büttner

5:21 PM

or 2995, ignoring the locked ones

Optimizer

(per 4 years)

Martin Büttner

(which are all code trolling I think, so it's totally cool to disregard them)

@Optimizer actually, a third of those were asked within the last year

(if not even more)

Optimizer

so we just need 5 more tips questions ?

Martin Büttner

actually, a third was asked within the last 8 months

yeah, closer to 50% in the last 12 months

Optimizer

5:37 PM

are you gonna fine grain that stats even further ? "yeah, 54.32% in last 11 months and 28 days"

Martin Büttner

no, I just went to page 33 and 50 of the questions list and checked when those questions were posted (30 per page)

lol, +6/-5

Geobits

6.41% in the last 55 days. That's much better than our average of 3.70% per 55 days.

Optimizer

+6 -5 wher e?

Martin Büttner

linky

Optimizer

oh, votes

i was text searching

Rainbolt

6:11 PM

Huh. We got a new Geobits. I hope the old Geobits doesn't notice.

I guess we always knew there were two of them. Otherwise it would be Geobit.

Optimizer

so 8 of them will make Geobyte ?

this can go on

Rainbolt

I just found out that a nibble is 4 bits.

Martin Büttner

I'm amazed how long this article is

Optimizer

Geonibble will come first then

Martin Büttner

"semi-nibble" is just ridiculous

should be "quit" :D

Geobits

6:25 PM

For the record, the name Geobits came from Geobot. The bits/bytes thing is just a happy accident.

isaacg

@mbomb007 To answer your question from here: meta.codegolf.stackexchange.com/questions/1797/… There isn't really a lambda syntax. Functions are not first class in Pyth, they cannot be assigned to variables or passed around. Also, range's step parameter cannot be used, nor can slice's. To replicate that functionality, use %<int><seq>

Optimizer

I wish I would get a bronze badge everytime my inbox reaches 10

and solver and gold

and hopefully, platinum

feersum

a badge of inbox non-opening?

"Self-Restraint"

Geobits

That seems counterproductive.

Martin Büttner

@feersum Mysticial would get a platinum badge for that ;)

Rainbolt

6:31 PM

Oooo what is a platinum badge?

Martin Büttner

not a thing

aditsu

it's worthless compared to an adamantium, unobtainium or eludium badge

Martin Büttner

mysticial would get all of them

Rainbolt

I think we need to convert to Wood, Stone, Iron, and Diamond badges.

aditsu

not to mention elerium-115

Optimizer

6:35 PM

paper badge ?

rock paper scissors badge, get any two and 1 gets removed

get all three at same time for an unstable equilibrium

Geobits

If I have two rocks and get a paper, do I still have one rock?

Optimizer

no, count does not matter

just the badge

so 0 rocks ... rock ?

Geobits

If I have paper, can I visit profiles of rock-carriers to delete their badges? Please say yes.

aditsu

only if you also bring a Spock with you

Geobits

Of course, accidentally clicking into a scissor-profile would be disastrous.

aditsu

6:43 PM

so, is it better to get the number of occurrences of an item in an array, or get an array of indices of those occurrences?

Geobits

In general? I'd say the count is more commonly used. Which would be less golfy if I had to do it the "hard way"?

Rainbolt

7:14 PM

I wonder why Optimizer's recently starred message is taking up two lines on the transcript

Geobits

It's not for me?

Rainbolt

Refreshed the page. Still looks like this:

Geobits

Weird. Mine isn't like that.

Rainbolt

My stars are no longer anonymous :(

Geobits

When you hover, does the little arrow (normally to the right of his name) show up on the bottom line?

It looks like it's just at the breaking length.

Martin Büttner

7:18 PM

how did I miss this question when it was asked codegolf.stackexchange.com/a/47044/8478 ... I guess because it wasn't really well thought through, but it's such a nice challenge for Mathematica...

FUZxxl

7:31 PM

I believe this question is not answerable.

First, by the theorem of Rice you can always construct a function that behaves like another function but is different in its construction.

So the return value cannot tell whether the function is the right one.

And neither can external state.

For instance f(f(x)) and f(id(f(x))) are always indistinguishable (where id(x) is the identity function), even if f(x) modifies global state.

and even there; how can we distinguish a = f(x); f(a); from f(f(a));?

I hate questions that are only solvable in certain languages because it requires certain kinds of reflection.

Martin Büttner

@FUZxxl that's the puzzle

sure it's not possible in all languages, but no challenge is

TheBestOne

@FUZxxl I can't solve it in HQ9+ :(

Martin Büttner

that's because HQ9+ is not a programming language.

2

FUZxxl

Well, good puzzles should be solvable in all languages that have basic IO capabilities and that are turing complete.

Martin Büttner

@FUZxxl anyway, the point of the challenge is that you should find a way to determine if f was a parameter, which most definitely involves not looking at the evaluated result. in fact, the last example is actually one where the evaluated result has no trace of the value of f any more.

@FUZxxl that's really ambitious. most challenges I'd classify as programming "puzzles" are usually very specific to working around the usual semantics of a handful of languages.

if you can point me to a counterexample, I'd be really interested.

FUZxxl

7:46 PM

counterexample for what?

TheBestOne

@FUZxxl In Cjam, try drawing a smiley face to the screen. All output to stdout or stderr is ignored.

FUZxxl

Then that is a shitty challenge.

Martin Büttner

@FUZxxl to my previous statement. I'd like to see a good programming puzzle that makes sense in all turing complete languages with basic IO capabilities.

Geobits

All graphical (non-ASCII) output is bad?

Martin Büttner

(programming puzzle in the sense of programming-puzzle)

FUZxxl

7:48 PM

@Geobits If it's allowed to generate an image file (e.g. a PNM file) instead, that's fine.

@MartinBüttner Any arithmetic challenge for instance.

TheBestOne

@MartinBüttner Print "Hello World" to stdout.

FUZxxl

where you need to compute certain functions.

Martin Büttner

guys

"programming puzzle"

not "challenge"

something like this

FUZxxl

I'm sorry. Didn't see that.

Geobits

@FUZxxl I disagree. That changes the challenge considerably.

FUZxxl

7:49 PM

@Geobits In what way?

Martin Büttner

actually

an even simpler example than graphical output is file system manipulation

is any challenge that requires the creation of files shitty?

FUZxxl

Oh yeah, in CJam you could probably solve this by generating an image file and then invoking an image viewer.

@MartinBüttner Challenges should try to provide alternative ways of IO for language that don't provide file system access capabilities.

Martin Büttner

well some challenges aren't about the I/O but actually about dealing with a directory tree

I'm referring to those

like codegolf.stackexchange.com/questions/43201/a-skittish-program

Rainbolt

So is this challenge that requires finding pixels in a PNG image poor? codegolf.stackexchange.com/q/37041/18487

Martin Büttner

or codegolf.stackexchange.com/questions/44640/h-tree-directories

Geobits

7:52 PM

@FUZxxl On most systems there is overhead to display (just as in your CJam->file->viewer example) that isn't present with just image creation.

Martin Büttner

@Rainbolt that plug... still hoping someone will find the easter egg...

Rainbolt

I suppose I could have loaded the entire PNG image into memory and then passed it? I don't know how this would help languages that can't parse a file (unless they have a deep understanding of how PNG files are stored in memory)

Martin Büttner

@Rainbolt I guess the challenge could be more inclusive though if you allowed people to read the file from STDIN

FUZxxl

@Rainbolt It would be better if the challenge allowed solutions to specify what image file format they want to read.

Anybody can read ASCII PPM.

Rainbolt

Do all image formats have a concept of pixels?

FUZxxl

7:53 PM

@MartinBüttner Well, that challenge is hard to alter to be more inclusive. It's still interesting.

Geobits

I just don't think that "every challenge needs to be feasible in 100% of languages" should be a rule.

6

Try to make it as inclusive as you can, but don't sweat the small stuff.

FUZxxl

@Rainbolt I was referring to raster image format.

It's just unfair to exclude languages that don't have a PNG decoder in their standard library.

@Geobits Indeed. Good policy.

Martin Büttner

that being said... I am genuinely interested in a good programming puzzle (in the sense of the corresponding tag) which isn't limited to at least one language paradigm

Rainbolt

@FUZxxl Why is that unfair?

TheBestOne

I personally like King-Of-The-Hills.

Martin Büttner

7:56 PM

@Rainbolt maybe not unfair, but probably unnecessary

FUZxxl

@Rainbolt Because a PNG decoder is quite hard to write.

Geobits

I think the most usual violater is [restricted-source] anyway. Many of those are downright impossible in many languages :P

Rainbolt

How can you even be unfair to a language? It doesn't have feelings.

Martin Büttner

@Geobits and source layout

@Rainbolt :D

FUZxxl

And unless the task is specifically about decoding PNGs, the language shouldn't require people to implement something that is not really a core part of the challenge.

Rainbolt

7:56 PM

@FUZxxl Then don't write one. Solve the challenge in one of hundreds of languages that can decode PNG images.

This isn't like a "GolfScript or you lose" situation. You have hundreds of options.

And parsing a PNG file is, to me, a real life task.

FUZxxl

@Rainbolt Well, I have some right to complain. Per policy, posing the same challenge but without the PNG parsing would be too little difference to constitute a new challenge.

TheBestOne

@Rainbolt Haven't you heard of Draco or Dylan.

FUZxxl

@Rainbolt I agree with your last statement. If the task is to parse a PNG, that's a good challenge.

If the task however is not directly related to PNG parsing, requiring people to parse PNGs is a shitty prerequisite.

Rainbolt

What if you want to level the playing field so that all inputs and outputs for all participants are exactly the same. The only thing unique to each submission is what it does to that input in order to produce an output.

In other words, all things held constant except for the submission itself.

Geobits

I skip many a [code-golf] simply because it isn't "short" (obviously subjective) in my favorite language. I don't think it's unfair that array or stack based languages have huge advantages on some, even though it's undeniably true.

FUZxxl

8:00 PM

@Rainbolt That's a noble goal but you should choose file formats that are easy to parse and to generate unless generating the file format is a direct part of the challenge.

Geobits

It just means I either need to use/learn one of those languages or be content skipping some challenges.

Martin Büttner

@Geobits Hm, that's one thing, but I agree with FUZxxl that a good challenge should isolate the core of the challenge as much as possible.

FUZxxl

@Rainbolt Your challenge wouldn't loose anything from using PNM as an image format.

The only thing that's lost is the advantage for languages that implement PNG de/encoding.

Rainbolt

Why limit it to PNG and PNM?

Geobits

@MartinBüttner I agree with that in principle. Sometimes the specifics get in the way ;)

Rainbolt

8:01 PM

Where do you stop?

FUZxxl

OTOH if the challenge forbade people to use an existing PNG de/encoder, that would be fine, too.

Martin Büttner

@Rainbolt At all publicly specified raster image formats.

Rainbolt

Where is a list of publicly specified?

FUZxxl

@Rainbolt At a format that's easy to parse for almost all languages.

Martin Büttner

@Rainbolt where is a list of all programming languages?

obviously, anyone who chooses an obscure format has to prove that he didn't make it up.

Rainbolt

8:02 PM

@MartinBüttner Where is the spec for engima?

FUZxxl

@Rainbolt As usual; any format that doesn't exploit a standard loophole and was specified before the beginning of the challenge.

Rainbolt

You say "as usual" as though that has always applied to image formats.

Geobits

For example, parsing input isn't the "core" of many challenges, but it's still a pain in the ass (score-wise) in Java sometimes. Only recently have people started loosening requirements on that (function args, etc).

FUZxxl

@Rainbolt That was meant as in “as for programming languages”

Martin Büttner

@Rainbolt I believe it is

(at least for graphical output)

Rainbolt

8:05 PM

I remember now why I didn't use multiple image formats. I converted my image to another format and some fuzzy areas showed up.

Martin Büttner

@Geobits which, I think, is a good development

Rainbolt

Solid red pixels became red blobs with shades of red surrounding what used to be a solid red pixel.

Martin Büttner

(unless the challenge is about input parsing)

FUZxxl

@Rainbolt Was the other format JPEG? JPEG isn't loss-less. That's an unsuitable image format.

Rainbolt

But it is raster, and it is public

Martin Büttner

8:05 PM

sorry, I forgot a word

Geobits

@MartinBüttner Sure, but nobody ever complained about it or said it was "unfair". It was more like "suck it up and don't use Java" :P

FUZxxl

@Rainbolt It's not loss-less though.

I think a good rule-of-thumb is: “If your challenge mandates a complex file-format, using builtin functionality to parse and create files of this format should be prohibited.”

Geobits

Okay, so maybe some people complained :D

FUZxxl

That's a large difference.

Martin Büttner

I don't think "we used to have it worse and no one complained back then" arguments have much merit, sorry

FUZxxl

8:07 PM

@MartinBüttner Indeed.

Geobits

Oh, I agree. My :P face must not have gone through.

Rainbolt

@FUZxxl I do not like your rule of thumb. If the challenge requires a complex file format, it should be up to the challenge author to allow or disallow using built in tools. By default, it should be allowed.

Geobits

Like I said earlier: Make it as inclusive as you (reasonably) can, without stressing over it.

FUZxxl

@Rainbolt It's my rule of thumb for making challenges. The intent is that you should not mandate complex file formats if you don't want people to implement de/encoding that file format.

Rainbolt

Even if it's your rule of thumb, you still have to explicitly prohibit using built in tools.

FUZxxl

8:09 PM

@Rainbolt Indeed. That's what the rule-of-thumb says.

Martin Büttner

@Geobits I can agree with that

(and I think "publicly specified, lossless raster format" is definitely the realm of "reasonable", but luckily "reasonable" is subjective)

Geobits

@MartinBüttner No, you stress over it. See "random" for example ;)

Rainbolt

This really sounds like "It's my rule of thumb but my point was that I would never need to use my own rule of thumb because I would never do that." which really makes it just a rule of your pinky toe.

Martin Büttner

@Geobits nah, people stressed me about... I didn't stress :D

for the first challenge I just tried to make it nice and simple and got a lot of sh*t for it

Geobits

I noticed. It started on my answer IIRC :D

So, you know... you're welcome.

Martin Büttner

8:12 PM

thanks ;)

FUZxxl

Anybody interested in a Paeth transformation?

Martin Büttner

can't find a wikipedia article. what is this?

Rainbolt

@FUZxxl gist.github.com/fuzxxl/3451755 linky

(Spoiler alert)

FUZxxl

@Rainbolt Ja, that's mine.

who where did that come from?

Rainbolt

I know. I'm always watching your GitHub.

FUZxxl

8:18 PM

Portable Network Graphics

Portable Network Graphics (PNG), also known as PNG's Not Gif, is a raster graphics file format that supports lossless data compression. PNG was created as an improved, non-patented replacement for Graphics Interchange Format (GIF), and is the most used lossless image compression format on the Internet. PNG supports palette-based images (with palettes of 24-bit RGB or 32-bit RGBA colors), grayscale images (with or without alpha channel), and full-color non-palette-based RGB[A] images (with or without alpha channel). PNG was designed for transferring images on the Internet, not for professional-quality...

Martin Büttner

@FUZxxl it's the first hit on google

FUZxxl

oh wow

Martin Büttner

correction

it's the first two hits on google

FUZxxl

Yeah, I wrote that for the QTV project.

QTV is a loss-less video codec for synthetic image data

like that kind of data usually created by computer images.

but that's not important.

randomra

8:35 PM

Had a great avatar idea:

2

Geobits

Griffin

2.9k 16 25

I think his is white, not transparent. Not easy to tell, though.

Martin Büttner

I think you should take another look at the screenshot

Geobits

I see... I guess that only works if you don't have any gold badges ;)

Martin Büttner

yeah

or you need to make platinum after all ;)

randomra

@Geobits for the last message you got your old avatar back

Martin Büttner

8:46 PM

must be a caching problem on your end

Geobits

Yea, mine shows right.

randomra

right, refresh solved it

TheBestOne

9:16 PM

I joined Meta just so I could test this. I rather liked it :)

Rainbolt

Can Microsoft Paint not save transparent images?

I can't seem to figure it out

Geobits

You bloomed!

Rainbolt

Even just opening a PNG and then resaving it loses the transparency

TheBestOne

@Rainbolt I liked the old one more :(

Geobits

@randomra All this avatar-changing is your fault, you know ;)

Rainbolt

9:19 PM

This one would be nicer if the background were blue like it was supposed to be.

The last transparent image I uploaded ended up with a blue background

TheBestOne

@Geobits I changed mine first :D

Geobits

Oh? I thought you'd had that one for a week or more.

Rainbolt

I wonder where Fedorabits is today.

Geobits

Over with Flowerbolt.

Maybe Daisybolt is better :P

Rainbolt

Or would it be more proper to say, "I wonder where Fedorabits are today?"

randomra

9:21 PM

@Geobits I just changed from default. You all did that too! :)

Geobits

Geobits is a single entity named as a collection, so use it as singular.

I assume Fedorabits is the same.

TheBestOne

Does trichoplax's avatar ever look like it's animated to you guys?

Rainbolt

Link?

TheBestOne

@Rainbolt Only in chat.

Lembik

hi

0

A: Sandbox for Proposed Challenges

Finding all-but-one matches This challenge is about writing code to solve the following problem. Given two strings A and B, your code should output the start and end indices of a substring of A with the following properties. The substring of A should also match some substring of B with up to ...

New Sandboxed Posts

9:46 PM

0

A: Sandbox for Proposed Challenges

Finding all-but-one matches This challenge is about writing code to solve the following problem. Given two strings A and B, your code should output the start and end indices of a substring of A with the following properties. The substring of A should also match some substring of B with up to ...

Rainbolt

10:50 PM

sigh Our CEO is really pushing for employee participation in this stupid "track your steps with a pedometer" program.

He pretty much labeled every employee who wasn't participating a "disengaged employee"

I do the street cleanup every other month. Seems like that would be enough to show that I am "engaged".

I also made bacon wrapped duck. If he makes me participate in this stupid health thing, he'll get no more bacon wrapped duck from me.

/rant

Martin Büttner

@Lembik that impatience

Geobits

The circling worms weren't doing it for me.

This feels much better

Doorknob

Who starred "hi"?

Are we going to start starring "hi" now?

3

randomra

@Rainbolt what is the goal? to walk a lot at work?

Geobits

@Rainbolt Better than a "track your steps with this implanted microchip", right? It's all about perspective :P

Martin Büttner

11:05 PM

@Geobits the blur makes my eyes hurt

Geobits

It was a hasty job. I can fix it up later when I have more time.

Martin Büttner

@Doorknob I even missed the point where we started saying "hi"

Geobits

It would be nice if there were vector-based assets somewhere on this site. The sprite is too small to enlarge to decent avatar size.

Doorknob

Sep 17 '14 at 16:46, by Geobits

I need a userscript that hides 'hi' messages if more than one are present within ten messages of each other.

(click link, scroll up)

Martin Büttner

(yeah, I've been there... I didn't mean that literally)

(in fact I starred that (Geobits') message)

Geobits

11:10 PM

Is there a 'hi' message? It must be hidden.

Martin Büttner

so guys... this dude was so nice to ask on meta what he should do about his answer instead of going the usual rookie root of either posting the invalid answer anyway or duplicating the question with a minor change:

2

Q: Answering a not-quite-duplicate

I haven't created a duplicate question... and I don't even want it to be necessarily reopened. I'd just like to be able to post an answer (either to it, or to the supposed "original" question) Obviously, as it almost always happens... duplicates aren't perfect duplicates, and in this case there'...

discussion

I think that deserves an answer, but honestly, I'm not sure what to write

Sp3000

11:26 PM

I'm not quite sure I get the question...

Geobits

I was trying to figure out how to answer that earlier. The honest answer (IMO) is "just move on", but I don't know how to word that. If the only difference in the new question is that eval is allowed...

Transcript for