The Nineteenth Byte

5:00 AM

Also, how in the world do we not have a Euclid's golf yet? I could have sworn we did, but if so my search-fu is failing.

Dennis

@AlexA. codegolf.stackexchange.com/…

Alex A.

@Dennis I stand humbly corrected.

Geobits

I love that the first result for that is Write a code golf problem in which Java wins :D

Reto Koradi

@Dennis Just using the word Java in the answer does not count. ;) One of the first search results: codegolf.stackexchange.com/questions/51871/…

Geobits

I'd also think Processing doesn't count as "Java", but maybe I'm just a purist :)

Alex A.

5:05 AM

I thought Processing was its own language?

Geobits

It is to my mind. A few users have tried to convince me to use it in my Java answers (esp on gfx challenges) to "golf it down", so I'm not sure everyone agrees.

Dennis

Haha, I'm the user with the most accepted "Java" answers to code golf challenges.

Geobits

That's just because you mention CJam's java interpreter too often :P

Dennis

3 in CJam, 1 in Bash. whistles

Zach Gates

@Geobits I never know how to phrase it :/

El'endia Starman

5:09 AM

@ZachGates In the only challenge I've posted so far, I just said that the input can be taken through any method desired (except hard-coding).

Geobits

If you don't have a specific input method that you want to enforce, you can use this list from meta. It's fine to be more specific, I just tend to participate more often in more open ones (and I assume at least some others are like-minded).

Alex A.

^

El'endia Starman

I definitely prefer making input and output whatever is convenient to the user because then the focus is more on the actual problem/puzzle/etc.

Reto Koradi

You have to be somewhat careful, depending on what you're trying to achieve. I've seen challenges where the input was wide open, and people found very creative ways of defining the input that simplified the problem for them. Which might be fine if you want to recognize the creativity. But if you wanted the focus on the problem itself, it could be undesired.

El'endia Starman

@RetoKoradi Examples? I want to form an opinion on that.

Zach Gates

5:12 AM

I just made an edit. I think it improves it.

Geobits

That's true. For many inputs it's no issue (like a simple list of ints, etc), but some cases are hard to spot in advance.

Reto Koradi

@El'endiaStarman I was afraid you would ask for examples. ;) Let me see if I can remember one...

feersum

A favorite hypothetical example is to use unary input on a challenge to add two numbers

Geobits

There was the hypothetical in chat the other day: "My language only takes integer input via a list of prime factors" or something like that :D

Alex A.

In ascending order, no doubt :P

Dennis

5:15 AM

I'd like to read the integers i and j as i j]$~_{2$1$:G1$1$md}h]7>4/"=( * )+ "N+S/f.+G, please.

feersum

That sort of thing shows up with specifications like "Instead of A and B, you may use any two distinct strings..."

Zach Gates

@Dennis Honestly, I'm not sure what that means..

El'endia Starman

Pretty sure that's a whole program. Makes it trivial to write the "real" program that actually executes it.

Dennis

If input may be taken in any form, I'd like it followed by the code I need to solve the problem. ;)

Alex A.

@ZachGates You're not the only one.

Geobits

5:18 AM

2

A: A Euclidean Algorithm

CJam, 46 43 bytes q~]$~_{2$1$:G1$1$md}h]7>4/"=( * )+ "S/f.+G Try it online in the CJam interpreter.

Reto Koradi

@El'endiaStarman Ok, this is not an extreme example, but it might somewhat illustrate where I was going. Look at aditsu's answer here: codegolf.stackexchange.com/questions/54105/…. While the problem is specified as having 3 inputs, the answer defines the input format as the 3 strings concatenated into a single string.

Which was very clever, and completely in the spirit of code golfing. But maybe not what the OP (which happens to be @Dennis) had in mind.

feersum

43 bytes for Euclidean algorithm ???

Zach Gates

@Dennis It seems like you're pointing out a flaw in my phrasing, but I can't be sure

feersum

I think that's how long it is in C.

Zach Gates

@Dennis OH. Now I see what you're saying..

Reto Koradi

5:21 AM

@feersum Most of it is probably for generating the step-by-step output. Just producing the final result would be very little code.

Zach Gates

No, that isn't allowed, haha

Dennis

@ZachGates I've seen people embed code in their "input format", so I usually try to be explicit in my challenges. Something like You may read the integers as such, an array of integers or a string representation thereof..

feersum

lol, I didn't actually read the challenge.

2

Zach Gates

@Dennis Fixed it. (I think)

El'endia Starman

@RetoKoradi Hmm. I feel like that doesn't quite count. The spec ends up defining b and B in such a way that they are one character each. Thus, given bnB, it is entirely possible to extract the three numbers from it. So I think that is a valid input method.

Dennis

5:24 AM

To be fair, that's exactly what I had in mind when I wrote the challenge.

Geobits

I would agree if it didn't explicitly say "Accept three strings as input:"

Dennis

@Geobits You may read the three strings in any convenient order, as such, an array of strings, a string representation thereof, concatenated or separated by single-character delimiters of your choice.

Geobits

Ah, I missed that. Was expecting input rules in the input section :)

I've seen better examples anyway, but they're really hard to search for. is:answer input form is a pain to sift through.

Reto Koradi

@El'endiaStarman It's absolutely valid, and apparently what Dennis was looking for in this case. My general point is that if you leave the input rules too wide open, you risk that people will do something that is not what you wanted.

@Geobits Yes, it's not the ideal example. But the only one that came to mind at the moment.

El'endia Starman

@RetoKoradi Definitely true. I just don't know how much of a worry it really is. I'll just be safe from here on out though and do something like Dennis' method to disallow code/etc.

Dennis

5:28 AM

This might be a good example:

7

A: Find the odd character out in a pattern

Bash Perl, 231 229 218 178 164 166 138 106 74 bytes /^(((.*).*)\2+)\3$/;$_.=$1x2;$.--,die$+[1]if/^(.*)(.)(.*) .*\1(?!\2).\3/ The script requires using the -n switch, which accounts for two of the bytes. The idea of appending two copies of all full repetitions of the pattern has been taken fro...

You must output the (zero-based, starting from the top left) coordinates of the odd one out. For example, in the above input, the corresponding output is 4,2. You may also output 4 2, or "4""2", or even [[4],[2]], or any other format, as long as you can tell what the output is supposed to be.

Zach Gates

I'm off to bed. Have to be up at 6 /:

Dennis

I chose 4 at script.pl line 1, <> line 2..

El'endia Starman

...you output the answer by throwing an error?

That's clever. And definitely something I would disallow.

Alex A.

Well, in R, stop() is shorter than return()...

Geobits

I've been tempted to System.exit(n) several times rather than print to STDOUT.

Dennis

5:33 AM

@El'endiaStarman We now have defaults for I/O that forbid using STDERR for output, so it wouldn't work anymore.

Alex A.

What if I made a language where all output was to STDERR?

Could I never use it here?

Dennis

There are language like that.

Alex A.

:O

Which, and why?

Geobits

I think if you did that we'd find a new disparaging tag to star.

Alex A.

I... don't know what you're talking about. >.>

Reto Koradi

5:35 AM

There are languages that don't write to either. I think? JavaScript, I figure?

Dennis

4

A: The Programming Language Quiz

Gnuplot, 51 bytes, cracked by Sp3000 set timestamp "off"; f(x)="Hello, World! print f(1) Edit: I just realized that it printed to STDERR. But arguably that's the closest alternative.

Alex A.

Damnit, someone else starred the tag now. :/

Dennis

Tag?

Alex A.

See starboard.

Reto Koradi

Yes, you win the star competition! Are you proud?

Dennis

5:37 AM

fights urge to pin tag

Alex A.

clearly-i-have-no-idea-what-i-am-doing

Geobits

Don't fight it man :D

Alex A.

@Dennis ಠ_ಠ

Geobits

False-eyelash-guy makes another appearance.

Alex A.

Quoi

Geobits

5:43 AM

Quit Using Obscure Initialisms?

3

Alex A.

Can't stop, won't stop ಠ_ಠ ಠ_ಠ ಠ_ಠ

PEACE

Geobits

People Exist As Creepy Emoji

3

Aww, we pushed the tag off the board. Now it really does need to be pinned :P

El'endia Starman

@Geobits It really does annoy me that there's not a scroll bar or functionality over there.

Geobits

Yea, on my laptop it's not so bad, but on my desktop I have 3-4 inches of dead space where more stuff could be.

Dennis

Done. :P

For some reason, I had to click unpin this message to pin it.

Geobits

5:50 AM

With that accomplished, I think I can go to sleep in peace.

Beta Decay

6:08 AM

Hey @MartinBüttner did you find anything wrong with FizzBuzz? :)

trichoplax

7:04 AM

This seems potentially promising:

5

A: Are Stack Snippets a good enough reason to increase the 30k post limit?

This seems to not be a restriction of the database. On Code Review, it seems the restriction has been doubled, as per @ChrisJester-Young♦'s comment: I just upped the MaxBodySize limit, so you should be able to have the whole code in the question body. For Code Review, I believe it's importan...

Maybe we need a site specific feature request on our own Meta?

feersum

This is pathetic. You should be grateful for the opportunity to golf your oversized posts! :P

2

trichoplax

I know Calvin's Hobbies had to golf both the code and the spec for some of the Stack Snippets KotHs. I'm half considering writing a KotH and then posting the controller as a golf before then using the result to post a KotH

feersum

Everyone loves a KOTH in which a single round takes 100 years to run.

trichoplax

7:19 AM

"The leaderboard will be updated shortly after the heat death of the universe"

El'endia Starman

7:43 AM

That awesome feeling when your (my) Befunge answer is just under half the size of its competitor.

New Meta Posts

0

Q: Should we increase our post character limit to allow substantial stack snippets?

Calvin's Hobbies posted on Mother Meta about increasing the character limit for posts across the Stack Exchange network. This answer shows that it is possible to increase the limit from the standard 30,000 characters for individual SE sites, and this has already been doubled for Code Review. To ...

orlp

8:23 AM

@MartinBüttner I wasn't aware of that rule, and I'm certain it's not always valid.

For example it's common practice to post decent but not fully fleshed out approaches to hard code-challenges

feersum

EmoGolf, it's the new new thing

orlp

I could delete said answer, but I'm not certain if it's hurting anyone or anything. I just found the simple stack based parsing method pretty.

Justin

8:41 AM

I'm working on my CS lab (C++). It works in windows, but not on g++. Apparently g++ doesn't allow me to call function(std::istringstream{ "some string" }); when function expects a std::basic_istream<char> &. I have no idea what is going wrong here...

trichoplax

@orlp If you golf it, you can still leave the readable version in the answer...

feersum

@Justin Did you pass the -std=c++11 switch?

Justin

@feersum Oh, let me double check.

orlp

@trichoplax I already posted 2 golfed answers to that question

trichoplax

@orlp Not questioning your effort - I think the only problem is the example it sets. I can see that code being displayed either alongside its golfed form, or edited into another answer for the same language, to say "I like this pretty approach but it turned out not to be the shortest"

orlp

8:46 AM

@trichoplax I'll do that

Justin

@feersum Yes, I did. g++ tells me this (cropped): ^

no matching function for call to ‘lexer::Lexer::Lex(std::istringstream)’
lexer_test.cpp:43:35: note: candidate is:
  lexer.Lex(istringstream{ string });\
In file included from lexer_test.cpp:1:0:
lexer.hpp:27:8: note: void lexer::Lexer::Lex(std::basic_istream<char>&)
   void Lex(std::basic_istream<char> & inputStream);
        ^
lexer.hpp:27:8: note:   no known conversion for argument 1 from ‘std::istringstream {aka std::basic_istringstream<char>}’ to ‘std::basic_istream<char>&’

Wonder if it has to do with the reference.

orlp

@Justin you're passing a temporary into a function taking a non-const reference

Justin

@orlp I didn't realize that wasn't legal...

Why does the Visual C++ compiler have to hide these problems! Ugh.

orlp

@Justin because it sucks

using gcc on windows

@Justin github.com/orlp/dev-on-windows/wiki/Installing-GCC

Justin

I should figure out how to do that with visual studio. But now I just want to finish my lab and sleep (it's 3 AM here. been working on this lab for a really long time today)

orlp

8:54 AM

@Justin I don't think that's really possible

Justin

@orlp Using gcc with Visual Studio? Then I'd have to find another C++ IDE...

Maybe the one by JetBrains would work.

orlp

@Justin vim :P

Justin

@orlp I avoided vim because I need tools like a debugger, code hints, IntelliSense (whatever that's called, where the comments are looked up on mouseover or something similar). I don't know how to configure vim.

orlp

@Justin and today will be the day where you start cursing the world for not developing good tools like such as standalone applications

Justin

@orlp What would make it not a temporary? Would doing this work: std::istringstream stream{"value"}; function(stream);?

orlp

8:58 AM

@Justin yes

Justin

Okay. Sorry for the dumb questions. I've spend the last 12 hours straight working on this lab.

orlp

although I'm not certain why you need a non-const reference to an istringstream

feersum

Reading from a stream changes its state

Justin

@orlp It could just be that I'm coding in bad style. But I want to have an input stream, which I read from.

And I wanted it generic so that both an ifstream and an istringstream would work (for testing purposes).

orlp

@feersum I guess, yeah

I generally only pass output streams around, though

Justin

9:01 AM

I might change it to just take a string as input.

@orlp Hooray! It compiled and ran now.

Justin

9:18 AM

Night

Martin Büttner

10:05 AM

@BetaDecay having a look at fizzbuzz now

I'm not a fan of the comma separation. I think newlines are canonical for fizzbuzz. I'd say it should either be that or a flexible delimiter (comma, space, newline).

requiring newline-separation (\n or \r\n) and allowing an optional trailing newline seems cleanest

Transcript for