« first day (1458 days earlier)      last day (3403 days later) » 

12:20 AM
Speaking of matrices, what do you guys think of this?
 
12:35 AM
0
A: Sandbox for Proposed Challenges

YpnypnResponsible governance king-of-the-hill Background For years, the Federal government of the United States has been in chaos. The Presidency and Congress are under different parties, and there are no compromises in sight. Instead of passing a budget or improving the tax code, the politicians are...

 
 
2 hours later…
2:53 AM
@Lembik Ahaha no thanks, I don't want to solve a cubic
 
 
1 hour later…
4:13 AM
@feersum On the word chain question, what did you mean by "Isn't that a Hamiltonian path?"
 
a Hamiltonian path is one that visits each node once, no?
an Eulerian path visits each edge
 
If you treat groups of three chars as the vertices, and words as the edges
Then it turns into Eulerian
 
fair enough
can go either way since all the edges are directed, I guess
 
 
5 hours later…
8:52 AM
@MartinBüttner About Jigsaw, if I have
print("""
      """.count(" "))
Is there a whole row of spaces at the end of the first line which would be counted?
 
9:33 AM
Yes
 
Interesting...
(Also, you're up early :D)
 
9:53 AM
@MartinBüttner can you shave off 2 bytes ?
 
Off what?
 
binary counter quine
 
Probably more
 
10:13 AM
Oh, you answered it too, that's why you asked. I'm not sure I can be bothered for a "golf-off" today. ;)
 
heh, busy ?
 
@Sp3000 I had a lecture at 9, so if got up at 7 and did some golfing over breakfast.
@Optimizer yeah I should get some project work done, and I also want to post the jigsaw challenge, which will probably cost some time
 
oh, I got a CnR idea
cops post some hint to an existing answer. like 2 sample input output, language and a range of byte size (50-100) and the robbers find that answer ..
 
I'm not sure it'll be interesting if the cops don't do much programming themselves
 
both cops and robbers dont do programming.. its more about finding a good code from the code-stack
 
10:23 AM
It's questionable whether that's a programming challenge.
Sounds like an interesting game for chat though to breed more Peter Taylors you know the existing challenges so well that they recognise duplicates in their sleep. ;)
 
10:40 AM
damn this user23013
 
Every time XD
 
okay, I think I'll just not golf mine at all :D
time for jigsaw puzzles...
any better ideas for a first post than $> and << on one line and 1 on another?
 
You can do Golunar, which would start the challenge off with an int
Oh you probably want to space things out for the first answer, right?
 
10:55 AM
maybe print numbers 1 through N?
 
@Sp3000 yep
forget what I said
still, I think it will be quite tough to do this in 10 characters.
 
print (
             1
   )
? :P
 
well, still $><<1
 
Replace 1 with 0 ** 0?
Not sure what you're going for :P
 
@Sp3000 I'm not sure that works in ruby
amazingly it does
 
10:59 AM
I don't even know ruby and I can do this in less than 10 characters
p *(1..N)
 
I don't think this will new-line separate them
but print an array
 
it does
I tried it
 
oh, right, it gets deconstructed into the argument list of p
I'm not sure it makes it any more interesting though. sometimes you'll have to set up a loop in a new language, which can be really annoying, or you can reuse most of the previous prints and just add your own (depending on how well your language can deal with the existing blocks)
with printing a new number it's more likely that you actually have to get rid of previous code
 
I don't understand this comment
How does doing operations which are strictly a subset make it more likely that there will be incompatible code?
 
because printing 1..N is the same as printing 1..N-1 and then printing N
whereas printing N does not include printing N-1
 
11:05 AM
print 1 works in more languages than print 1..N
 
but if there's print 1; print 2; print 3 you can just find another language which uses the same print and append print 4
 
you might argue that it's too hard; it couldn't be done in Python
I'm completely mystified about what point you're trying to show
 
I'm not going to start with code that prints a range from 1 to 1
I'm going to start with code that prints 1
so the next person can just find another language where that's valid syntax and append code that prints 2
and that goes on until we run out of such languages
at that point, people will start having to form loops, which will be pretty hard in some languages with 10 new characters
and I'm not sure either half of the challenge is very nice if it plays out like this
 
ok, too hard makes sense
 
Wait... sorry I'm confused. "append code that prints 2"? So it's not rearranging the code pieces so far and adding chars from the box to print the number of your answer?
(not sure what the 1..N is about)
 
11:18 AM
@Sp3000 this is if the task was "print all numbers from 1 to N"
 
Oh, right. Thanks.
 
0
Q: The Jigsaw Code Puzzle

Martin Büttner This is an answer-dependent challenge! The order of answers matters, and your exact task depends on the last answer that was posted. You might want to sort the answers by oldest. Let's piece together some code! Here are 1920 random (printable ASCII) characters. Think of them as a big heap of...

@Sp3000 thx
 
@MartinBüttner That's weird, the snippet's not recognising my comment
 
it takes a few seconds
@feersum you should mention that this assumes the Python interpreter (with NUMERIC_OUTPUT = True)
 
ok
 
11:29 AM
and as far as I can tell you didn't remove any pieces
or did you replace them with spaces?
 
replaced with spaces
 
That's what I was going to do too :P
 
Entering the answer really sucks; the box is too narrow for the width of one line and I can't find any way to make it wider
probably there will be many answers with misalignment, missing pieces etc.
 
Man, I'm trying to find out where feersum took the two chars from XD
 
there's one next to the letters 'mn'
Why doesn't the stack snipper do this? :P
 
11:36 AM
@feersum I tend to not actually code in the answer text box
okay, the stack snippet displays the languages now. I should check for duplicate languages, but humans will be better for that anyway, because of different spellings and versions
 
I just wrote myself a Python script to check XD
 
@feersum because finding code blocks in the answer body is haaaaaard
 
I'm pretty sure the new answer's invalid...
 
no time to read the whole question, I just have to print N , the answer number, using code out of the remaining boggle board, right ?
 
11:51 AM
it's not a boggle board
 
jigsaw board ?
 
you have to execute the entire board in order
 
you can rearrange blocks of characters in the existing code, and you can add characters from the remaining ones
and then you run the entire board as your code
and it should print N
 
0
Q: Can numeric input be in the form of a character code?

Martin BüttnerThis answer just prompted a question, which I've wondered about a few times when answering challenges in esoteric languages: For, some esoteric language, reading characters as character codes is the only means of input, and interesting golfing of numeric challenges is near impossible with decima...

 
@MartinBüttner Oh thanks, @ meta question
Been wondering myself
@MartinBüttner Also, should it be mentioned that the |s don't actually count as part of the code?
 
11:58 AM
I believe it's mentioned somewhere
if you can think of a better place to mention it, feel free to edit
search the page for the word "never"
 
Ah, found it
I was doing a search for the char itself
 
0
Q: What tag should we use for answer-dependent challenges?

Martin BüttnerBy my count, we've had at least 5 "answer-dependent" challenges now, starting with Evolution of "Hello, World!". I think it's time we started tagging these, so we can put some general information in the tag wiki, and it's immediately apparent that a challenge is of this type. So, here's another ...

 
@MartinBüttner Nice work with the snippet. Though reading the rules I'm getting major deja-vu ;)
 
@Calvin'sHobbies In fact, I copied most of the code from my previous answer-dependent challenge (for which, I admit, I reused a lot of your very good wording ;))
 
And I just stole Calvin's tag suggestion, see how this works?
 
12:04 PM
lol
 
(Don't worry, we love you Calvin :) )
 
@Calvin'sHobbies btw, I've been trying for weeks to come up with such a challenge based on the idea of Chinese Whispers, but to no avail... maybe that theme inspires you to something? ;)
I've been thinking about either processing the last answer's output, or basing your code on the last answer's output (instead of its source), but I haven't been able to come up with something decent.
 
Hmm, I might tinker with that idea
Though hopefully tomorrow I'll be posting another new type of experimental challenge, so stay tuned
 
oha
you might wanna wait until showcase your language drops out of the HNQ... you're drowning out all the other PPCG challenges ;)
 
@MartinBüttner The snippet in your jigsaw challenge seems to give the leaderboard in reversed order (lowest score first).
 
12:10 PM
@Zgarb oh, thank you
oh, already fixed
 
Well, that was fast.
 
ugh, how do I verify that unlambda answer...
 
do you need an online tester?
 
that would be perfect... otherwise I guess I'll have to download one of the others
 
you can abuse golf.shinh.org
submit it to a random question
and it will show the output
 
12:16 PM
Wow, they actually have unlambda
 
okay it works
@feersum thank you
 
You're welcome
 
I'm surprised the first two answers are in esoteric languages straight away
 
normal ones use too many characters...
 
I would have kept Prelude until much later, because it just ignores unknown characters.
 
12:19 PM
I'm waiting for someone to post print lol
 
then someone else can snatch it
 
@Sp3000 so you can answer in Python?
 
So I can use the n for ><>
 
I'm pretty sure if someone adds print it will be because they answer in python :P
@Sp3000 oh nice :D
 
(Not going for shortest answer here, just something interesting)
 
12:31 PM
ugh, I forgot 3 lines of the box of pieces in my first answer
will fix up all answers in a minute
 
told you that would happen ;)
 
12:42 PM
@feersum not a stack snippet, but now there's a CJam script to check which characters were removed
 
1:04 PM
is there a way to check other people's reviews of edits so I could get the hang of it?
 
@MartinBüttner this is just mine
 
hm, I can see everyone's reviews there. maybe that requires some additional privileges.
 
Nobody was doing it so I added it myself :(
 
1:20 PM
I think I've learnt that the first line and the first column are really important :P
 
@Sp3000 does the offical fish.py work with python 3.4? fish.py -c '1n;' seems to create infinite loop
 
It should - that's what I'm using
@randomra You don't need the quotes
Bonus points if you place 5\; as your three chars :D
 
@Sp3000 ahh, thanks, help tricked me
 
grc
@Sp3000 why do you want 5\;?
 
1:35 PM
Just thought it'd be more fun that a straight 5n; :P Never mind me
 
grc
so you don't have some awesome plan then :(
 
I was talking about if ><> was the next answer :P
Although speaking of awesome plans, Java sounds fun here
 
grc
is the grid big enough?
 
@grc I'm sure it is, but I'm not sure whether it'll be too busy by the time there are enough characters for finding a Java answer in there
does anyone have a TI-Basic calculator to verify the newest answer?
 
It is, but nobody's going to bother putting static down just for Java :P
 
1:40 PM
or know an online interpreter?
@Sp3000 well if that's the only thing that's missing, then someone might ;)
 
class public void main?
 
maybe if we get some stray letters ;)
I'm sure main will find its way in there
 
System.out.println? :P
 
minimal Java can be done like enum A{a;{System.out.print("N");}} as long as errors are OK
 
we should make Martin's jigsaw puzzle's rules part as a template for answer dependent puzzles
 
1:49 PM
@feersum they generally are, as long as STDOUT is correct, I think
 
... they generally are?
 
for full programs, I think so
so not for function-only answers, where defining the function throws an exception you would have to catch or something
 
so now we are stuck until someone finds an old calculator where that TI BASIC program works ? :/
 
I have a TI calc but don't know how to make a backtick
I can't rule out the possibility that there is a way...as it does have foreign languages with accented characters and I don't know how to make those either
 
grc
is Timtech around?
 
1:53 PM
@TimTech Are you around ?
 
Here's a Timtech answer which may/may not help
 
nice, managed to golf down that Prelude answer from 32 to 20 by using more voices... proof that combining prelude with cjam may allow for better golfing :D
 
I thought Timtech answers are traditionally impossible to verify
3
 
The answer has backticks
 
I managed to find a character resembling one here: tibasicdev.wikidot.com/83lgfont#60
still no idea how to type it
 
2:00 PM
were calculators created before Copy Paste was ?
 
nice, now prelude is beating all but CJam and J on that question :)
 
when the loop starts, it starts for all lines/voices ?
 
yes, there's only one instruction pointer, which points at full columns
clarified
 
Is the Prelude answer pushing a char to act as the int input?
 
yes... well it's pushing the input char's code point as an integer
 
2:06 PM
kk, if you're doing that I might do that for ><> then :D
 
stawp u guys
is there any other 2d golf lang ?
 
@Optimizer Befunge, PATH, Rail, Manufactoria, Marbelous...
 
They're not golfing languages though :P
 
oh golfing lang
never mind...
 
yeah . golfing
I dont want 100x100 bytes of rail tracks in my code
 
2:08 PM
then what do you mean by "other" 2d golfing lang? :P
I don't know any :D
 
There's the currently non-existent Golfish :P
 
as well as the currently non-existent Big Band
 
prelude and fish are pretty concise
 
@Optimizer prelude is good for like 3 things
it's definitely not a golfing language
 
Big Band is not the right name
PreJam is so much better
 
2:09 PM
The thing about Prelude/><> is that ints/strings are the only thing they can really do
There's no concepts of arrays, for one thing
 
@Optimizer dude... it's a jammin' big band
(big not referring to the code, but the instruction set)
 
still no
 
@Optimizer and Befunge is pretty similar to ><>
 
what ?
 
wtf brain
 
2:09 PM
:D
 
Ack, I can't beat Prelude with ><> because I need to print newlines
@randomra To save on whitespace, you can bend the program around a bit - and don't forget that ><> wraps around!
 
2:30 PM
I don't like snippets in meta at all
 
snippets in meta?
 
stack snippets
 
@Optimizer since that TI-Basic answer has an upvote, we can't just delete vote it. I'm considering fixing it.
@NathanMerrill then post an answer ;)
 
Isn't that what Peter Taylor is arguing?
I did upvote his :P
 
@NathanMerrill he still suggested having a list of them on meta. and I'd still include the current version of the snippet in those answers, so you can actually try out the snippet right there.
I thought his main point was a) having them in only one place on meta and b) outsourcing bug reports and feature requests to GitHub or similar
 
2:33 PM
@MartinBüttner u n me downvote
-1
 
well if the only issue is the wrong count and not removed characters, that can be fixed easily
just because no one seems to know about the backtick doesn't mean it isn't possible.
 
so lets say correctness cannot be verified at all.
what then ?
 
hm good question.
I'd probably ask the OP to prove correctness then or voluntarily remove it. but deleting someone's answer because I can't tell if it's right doesn't seem nice.
can anyone else chime in with an opinion here?
@Sp3000 ?
@feersum ?
 
if you are making it mandatory for the correctness to be verified, you should have put a time limit from the beginning ..
 
Well based on the Timtech answer I posted early, we know at least that backticks are valid syntax...
 
2:39 PM
the question is about the delay in verification of answer
 
well then... I'm sure there's some way to load code into a TI-Basic interpreter without typing it into a calculator ^^
 
So if the colons really skip everything I wouldn't be surprised. But I'm just not sure whether said backticks need to be matched
 
that is a very crucial part to answer dependent challenges
 
@MartinBüttner how much duplication of snippets is there currently?
 
okay, here is what I'm gonna do... I'll fix the technical issues, and I'll give him the benefit of the doubt for the backticks. if it turns out to be invalid, he can still remove it so he doesn't get a score for it.
@NathanMerrill I think CH and I are pretty much the only users of snippets so far (with 2 or 3 people having used the Levenshtein distance snippet). I keep reusing my own snippets from previous questions. A list on meta would hopefully encourage more people to use the snippets.
 
2:42 PM
@PeterTaylor I think a simple solution is to have the snakes have no thickness. Just be a line that can be rotated at integer positions.What do you think?
 
Hmm I get the feeling TI-Basic just ignores backticks individually, from reading Timtech's answers
Hmm no thickness solves a few issues, so that'd be a step forward
 
let the games continue.
 
@randomra btw I'm not sure how you feel about the output, but printing newlines is easier than printing spaces
 
wow that looks like quite a KotH for a new user and not going through the sandbox:
0
Q: KOTH: Warring Towns

ThraxIntroduction The game takes is strongly inspired by Manu's Game of Town. It takes place in a world of swords and magic. The king who ruled the whole continent just died, and the lords of the many city-states are now fighting over territory. You are one of these lords and your goal is to conquer ...

 
Trying to look for pitfalls now, but it's a bit to take in.
 
2:58 PM
Basically you have 3 different types of warriors, each basically representing Rock, Paper, or Scissors
the rest are all management of the town
and peons
the necromancers are cool though, as they can resurrect
 
@Optimizer I fixed the answer btw, if you were waiting for that.
 
@MartinBüttner can I flip a puzzle piece ?
 
@Optimizer what does that mean?
 
:" becomes ":
 
3:06 PM
or
"
:
 
see the third invalid example
 
then I'll wait :D
 
...
why? :D
 
I'm guessing Optimizer wants to do Pyth and comment out the rest of line 1 with a single quote :P
 
it's not like you're losing points from answering with something else now though.
@Optimizer if that was really your plan, then it's your loss now :P
 
3:14 PM
D: Fry stuck the 6 with :"
Wait... did he move the whole of line 1 down?
 
well his assumptions are wrong
the previous online compiler had a bug. The python version and the current online compiler, both support multiline
 
The Python interpreter doesn't do multiline though
 
how are you giving the source code?
 
I'm pretty sure it was isaacg himself who first told me that everything after the first line is ignored
@Optimizer what input?
 
py -3 pyth.py input.txt is how I always command line it
 
3:17 PM
Oops, i thought blocks could be joined by spaces :/
 
then the whole board is a single block ?
:D
 
Nov 7 '14 at 2:42, by isaacg
@Sp3000 You misunderstand me. Pyth enforces one-line-ness, newline is a comment character
 
Thought so, that's what I remembered too
 
@Optimizer Yeah its pretty dumb now that I think more about it :P
 
@FryAmTheEggman if you fix the move count I'll verify it
 
3:23 PM
I have a better answer
 
oh okay
let me know when it's undeleted
 
Ok, should be up.
 
@FryAmTheEggman You moved 2
(the exclamation mark)
 
Couldn't get the ` to appear properly in the explanation :/
I thought that was a block with the "?
 
It's on its own at the end of line 1
 
3:28 PM
Oh shoot
Its fine in the other place
doesn't do anything my bad, ill edit
 
@FryAmTheEggman you need to double the surrounding backticks, not the ones inside :P
 
Markdown pls :/
 
@FryAmTheEggman I get an invalid syntax :/
# converts to while True:, but the preceding spaces put parens around it
 
Oh shoot, you're right :/
 
The best I can do is 1 move, 2 add
 
3:31 PM
That fixes it I think
 
Yep that seems to work
 
gahhhh
I wanted do do an add-2 golfscript answer after this :D
now I can't ^^
or maybe I can?
 
Yeah cant you use }
 
hm, at least the online interpreter seems to stop execution at the #, instead of commenting out only one line
ah no, of course, the problem is the "
I could move 1 and add 2
I didn't need to add }
an unmatched " has the same effect
 
Grabs popcorn for snacking while watching Martin scheme
Oh, you're done
 
3:44 PM
lol
oh, there was bug in the snippet... it still multiplied the scores, instead of adding them :D
@Sp3000 I'm glad you like the smiley but would you mind verifying that it's correct? :D
 
I'm trying to work out what the colon does in Golfscript if it's followed by a quote
 
@Sp3000 you can actually assign values to strings it seems
 
@MartinBüttner yeah, same question.
 
7:"a""a"
 
try 5:"a" "a"
@Sp3000 nice one :D
 
3:48 PM
XD
Hell, didn't know you could do that, that's insane
Not very golfy, but it gives you an unlimited number of variable names to work with
 
liek you would ever golf something having unlimited number of variables ;)
 
Lemme go post that code-golf question where input is 1001 different variables...
 
god, adding and removing these | will take up all the time for me , and someone else will post before ..
and it will keep repeating
 
@Sp3000 GS variable name aren't single-character anyway
 
Martin posted a CJam script in the OP
Well, the removing | part anyway
@MartinBüttner Oh, they're not? :o
 
3:50 PM
@Sp3000 oh, good point. adding might be nice as well.
@Sp3000 nope
 
Too used to Pyth/CJam
You should put the :"a""a" in the explanation
It's funky
 
Is it possible to create a cyclical chain of duplicates?
A dupe of B, B dupe of C, C dupe of A?
 
@Rainbolt I don't know. probably is. I can't imagine cycle detection being feasible on the graph that is stackoverflow
@Sp3000 done
 

« first day (1458 days earlier)      last day (3403 days later) »