« first day (1362 days earlier)      last day (3476 days later) » 

12:58 AM
Code golf idea: you know when you want stack some books and it's really annoying one book is shorter but wider than another, so no matter how you order them, there's a bit of the top book hanging over the lower one? How about you get as input a list of book dimensions and you're to sort them so as to minimise the total overhanging area throughout the stack?
 
1) Rotate stack of books until spines are perpendicular to the surface 2) There is no step 2
I like it, though ;)
 
 
2 hours later…
2:58 AM
Have we had an Ackermann function golf yet?
Searching cg.sx gives nothing, even though it feels like a classic.
(In all fairness, I just found a short J answer and I really wanna share.)
 
 
2 hours later…
4:45 AM
1
Q: The Ackermann function

algorithmsharkThe Ackermann function is notable for being the one of the simplest examples of a total, computable function that isn't primitive recursive. We will use the definition of A(m,n) taking in two nonnegative integers where A(0,n) = n+1 A(m,0) = A(m-1,1) A(m,n) = A(m-1,A(m,n-1)) You may implement ...

Now we do.
I'll drop my J/k/q in tomorrow.
 
 
1 hour later…
6:15 AM
0
A: Sandbox for Proposed Challenges

Beta DecayComputer Logic Puzzles code-golf Challenge Your program will supplied with some code in the language of your choice, the output of that code and the input (if any). The thing is that code will have two lines removed and replaced with comments. Your aim is to write a program which works out wha...

 
 
1 hour later…
7:18 AM
@algorithmshark Not as such, but there are a lot of Ackermann answers to various broad "Calculate something big" questions
@user23013 codegolf.stackexchange.com/q/32015/194 ? I might have an idea.
 
7:43 AM
Yes
 
 
2 hours later…
9:45 AM
@Geobits As in, don't stack but put them next to each other? Even then I like to sort them by size :D
 
 
1 hour later…
10:50 AM
Deadfish's spec's gone all of the place...
 
0
A: Sandbox for Proposed Challenges

Martin BüttnerBook Stack Sort When stacking books you usually want to put the largest ones at the bottom and the smallest ones at the top. However, my latent OCD makes me feel very uneasy if I've got two books where one is shorter but wider than other (or vice-versa). No matter which order I place them in, th...

 
11:29 AM
can someone else please do me a favour and google ".net regex syntax" to check if my balancing groups post on SO really is on the first page for that? :D (might just be personalised search results or something)
 
Followed by RegexHero
 
Nope, no stack overflow
 
kk, I thought so
@Sp3000 for me it's between msdn and regexhero
 
 
2 hours later…
1:14 PM
Code snippet for the Korean flag question, nice
 
yup
three answers have them
 
Too bad my turtle one doesn't seem to work well with Skulpt :/
 
How well sandboxed are the stack snippets? Could we make a KotH where people fight with their snippets right there in the browser?
 
@MartinBüttner Would be cool, but I don't think we could make them communicate.
 
It'd be a good way to run a KOTH at least, since you could provide something like that for people to test their code
I guess OP could always just update the snippet to have more bots as submissions come in, to bypass communicating
 
1:26 PM
On second thought, my Ackermann question may have been better as a Rosetta-stone-style challenge.
 
@algorithmshark agreed
 
Is it too late to burn everything and try again?
 
Personally, I just wanted to see someone go nuts with multithreading and microoptimsations everywhere, or something :P
 
@algorithmshark I think so
You can add a per-language leaderboard to the question to encourage more people submitting in usually-not-used-for-golfing languages
 
Ackermann in FRACTRAN :D
 
1:29 PM
I wanted code-golf and not fastest-code because IMO fastest-code on Ackermann looks like special-casing each of m=0..3 and then relying on a hyperexponent for m>3.
Cool once in the first language, and then again in the fastest language, but that's it.
@MartinBüttner Now there's a thought.
Oh, also, @MartinBüttner: I didn't even realize that some languages would have restrictive recursion limits.
I think that it'd be okay to reconfig your language to fix that. Should I update to say all configuration costs no chars, or only the recursion limit can be redefined, or what?
 
I suppose a sentence on the recursion limit should be enough.
 
 
1 hour later…
3:05 PM
0
Q: Do we need the [function] tag?

professorfishI've noticed that 20 questions are tagged [function]. Looking at the questions themselves, they don't seem to have anything drastically different from other questions. The tag wiki reads: For challenges that ask for a function or involve functions in some way. (Also known as methods or subro...

 
Anyone know the nth term rule of a sequence of binary sequence?
 
@BetaDecay what?
 
For the sequence 1, 10, 11, 100, 101, 110, 111... what's the nth term rule?
 
3:25 PM
nth term or nth digit? (confused)
 
Nth term. For example, the nth term rule of the sequence 2, 4, 8, 16, 32 is 2^n
 
Oh, so you mean reading the binary sequence as decimal numbers?
 
Yeah
 
1 * (n % 2) + 10 * (floor(n/2) % 2) + 100 * (floor(n/4) % 2) + ... ? Maybe
 
Yeah... How would you find how many times you add it?
 
3:34 PM
up to infinity :P
otherwise something like log n / log 2 should give you an estimate
 
After a while floor(n/2^k) is always 0
 
Wait, what's k?
 
the iterator of the sum
 
Ahhh
 
tbh it'd probably be easier to just use bitwise operations
@MartinBüttner I tried to use Skulpt to update my Korean flag answer... I don't think it went too well :/
Either way, for some reason the interpreters that require external content don't seem to work in my Chrome, but it does in FF
 
3:38 PM
@Sp3000 I saw it... I'd upvote again ;)
 
Hmm so it works fine on your comp? Hmmm weird...
 
yeah, chrome on windows
 
Well it's not my extensions... ah well. At least it works on FF
 
 
3 hours later…
6:58 PM
@MartinBüttner Re your flag, I converted the post to a comment, but now it looks like arse.
 
@ChrisJester-Young hm, ideally it should be a comment on this answer with the solution in a pastebin link
 
@MartinBüttner That would be very ideal, but the mod tools don't offer a direct way to do it. So I'd have to do it manually. Which is fine, but then the comment will have the wrong sender.
 
@Doorknob: Please note that I take your deleting my answer in the Jump-the-Digit challenge extremely personally. If I'd known that a bit of satire would upset you so thoroughly that you'd unilaterally eradicate my response, 270 rep, and dozens of comments, I'm a reasonable man and I'd have deleted the entry myself.
As it stands, you couldn't be bothered to ask me or even notify me here, you unilaterally wiped out an answer that the broader community clearly considers acceptable, and you totally invalidated the three hours of work I spent writing the code.
Your action was heavy-handed, grossly disproportionate to the magnitude of the "offense", and I resent it regardless of whether you were acting in good faith. That is all, sir. - Miffed COTO
 
/me looks at COTO's post to see what all that is about.
@COTO Also, like, the proper venue for this is probably via a moderator flag.
@COTO To be honest, for what it's worth, I'd have nuked your entry too. (And I have nuked even higher-scoring non-entries.) I think I'm the person who wrote the "all entries must be serious contender" rule, so I follow it more tightly than others might.
 
7:14 PM
I wish searching "jump digit" produced something worth looking at. I can't follow this conversation :(
 
@ChrisJester-Young Thanks, Chris. I don't want to drag this out. I'm not going to undelete the answer or flag a moderator. My answer technically wasn't in compliance with the topic rules anyway.
 
@COTO That's fine. Just letting you know that I stand by Doorknob's decision in this instance.
 
Ah, I see. This reminds me of various KOTH entries named Emo*.
 
@Geobits Hahahaha.
 
7:22 PM
@ChrisJester-Young My ire stems more from the circumstances of how he did it. Like a bloody bull in a china shop. Give me ten minutes and I would've placed a "serious" Java solution in plain sight in the same answer as the functional-but-obviously-not-serious-enough version. Problem resolved.
 
@COTO Don't post two answers in the same post. Just post your serious answer separately.
Seriously, your joke post is "damaged" and probably cannot be salvaged, even if you 100% replace it with a "good" solution.
You should not have received any rep for your joke answer, and piggybacking a serious answer on top of that is not right.
6
 
@ChrisJester-Young I know, which is why I'm not undeleting it. Having said that, there can't possibly be a serious Java solution. The best possible Java solution I found was 156 characters long, didn't handle dashes, and was utterly unremarkable except for possibly the brazen idiocy of submitting Java code to a code golf challenge. Hence the satire.
As for "you should not have received any rep for...", many big name contributors here have assured me that upvotes and downvotes are entirely at the discretion of the voters and are governed by no rules whatsoever, hence "should" has no meaning with respect to rep earned through voting.
 
@COTO Well, bogus answers would (at best) be community wiki'd or converted to comment or (at worst) be deleted, none of which will confer any rep for you, regardless of any upvotes.
 
As a Java golfer, I find that comment utterly ridiculous.
3
 
@COTO If 156 characters is the shortest Java solution then that's a serious entry. As usual, code golf isn't about getting the green check mark but squeezing as many bytes out of your language of choice as possible.
 
7:35 PM
It's also about having fun, providing the community with creative answers, and making the best of a bad situati--- oh, nope. Sorry. It wasn't "seriously" golfed, so who gives a care what the community thinks and let's ignore the unique circumstances of the question, into the junk bin it goes.
 
It wasn't "not seriously golfed", it was purposely the exact opposite of golfed to get a laugh.
 
@COTO Just so you know, this site was created because Stack Overflow did not tolerate code golf questions (despite it being "fun" and popular with the community). So just because something is popular does not make it a go.
I've posted a number of golf questions on SO back in the early days, all of which have been closed precisely because it's not tolerated on SO.
160
Q: Stack overflow code golf

Chris Jester-YoungTo commemorate the public launch of Stack Overflow, what's the shortest code to cause a stack overflow? Any language welcome. ETA: Just to be clear on this question, seeing as I'm an occasional Scheme user: tail-call "recursion" is really iteration, and any solution which can be converted to an ...

^ Quite a popular question, as you can see. But still no go. Understandably.
 
@COTO I still don't think you're in a position to find any reaction to your answer offensive. You were lucky that the OP himself found it funny and thought about making it a popcon. But technically, you're answer was taking the piss, ridiculing the question and implying that it's too boring to golf, so you did something more fun. Another question author might have found that offensive.
 
@ChrisJester-Young Answer me this: did they unceremoniously delete your question and wipe out all the associated rep?
 
@COTO They did for many code golf questions on SO.
 
7:40 PM
@COTO My question was community wiki'd, so same diff re rep.
Also what @MartinBüttner said.
 
Haha, one of those links was one of my questions. :-P
(The decision tree one.)
I should try to migrate some of those to this site, if that's at all feasible.
I miss LiraNuna's challenges. He wrote good challenges. (Re his lasers challenge that got deleted.)
 
@MartinBüttner If the OP had deleted the answer, or made any indication whatsoever that he considered it offensive, I would have deleted. As I said, it's not my intent to upset people. As it stands, (s)he didn't, and I didn't.
@MartinBüttner And the deletions no-doubt cheesed so many people off that they created a whole new SE board for them.
 
@COTO The questions that Martin linked to have all been nuked by a mod with no consultation with the OP.
 
I believe the deletions happened after PPCG was created.
 
7:44 PM
Yes, the deletions happened last month, even.
 
in fact, they were deleted a month ago
 
ninja'd
 
Jinx jinx!
@COTO Like, I didn't even know my decision tree code golf got nuked until Martin mentioned it just now. :-P
No, I didn't get any rep-drop notifications. Why? Because it was community wiki to begin with, which means no rep for anybody on that thread.
So deleting it won't affect rep for anyone.
 
@ChrisJester-Young There are tens of millions of active users on SO. There's maybe 50 here. Take the Gini coefficient on the rep distribution for the past year and you'd get pretty darn close to zero. It's not unreasonable to make a one-to-one "please edit or delete" request here. That's the bare minimum of courtesy I'd extend as a mod of a small board, at any rate.
 
7:51 PM
@COTO Why air dirty laundry publicly? (Note that there is no means to privately contact users except for "mod messages", which are for serious breaches of site rules.)
So far in the nearly 4 years this site has been around, only 5 mod messages have been written. It's very rarely used.
 
Also, there were comments on the post for a few hours before deletion, clearly stating that it didn't meet the rules. Just because they weren't posted by diamonds doesn't make them invalid.
 
100% agree.
 
Does anyone know how to get the UTC creation time of the original post (not the latest edot)?
 
@Calvin'sHobbies Yes, just click on the "latest edit" time for an edit history.
 
@Calvin'sHobbies If you look at the revision history, you can scroll to version one and hover over the "9 hours ago" or "3 days ago" to get a timestamp.
 
7:54 PM
Look at the time for revision 1.
Jinx!
 
@Calvin'sHobbies Hover over the time posted
 
Ah, thanks!
 
(like, the one in the little plaque with your user name)
 
I guess so. So, to summarize what you gentlemen are saying: content being unceremoniously wiped out, without warning, for minor infractions, and in spite of community feedback (i.e. voting), is commonplace and them's just the breaks?
 
@COTO "Minor" infraction is perhaps a projection on your part.
Off-topic stuff gets deleted all the time.
Heck, I do it a lot myself, and I do much less moderation these days than Doorknob does.
 
7:56 PM
and "in spite of community feedback" is irrelevant
2
this isn't reddit
 
@COTO Look, I don't like deleting stuff much. But yours seems like a clear case of "An answer that is not an answer should be deleted" to me.
 
Correct me if I'm wrong, but the only way moderators get to be moderators is because the community voted them into the position, no?
 
@COTO Right now, because the site is still in beta, mods are appointed "pro tempore" by the Community Team.
 
I don't believe beta sites have elections.
 
that's a different community
 
7:58 PM
@COTO I will promise you, if the site graduated and came to an election, I'm sure Doorknob and Martin will both get elected.
 
I'm pretty sure that the majority of votes on your question was not by active PPCG people... anyone with the association bonus can upvote... but only people who are active here can downvote. this skews the "but it was popular" thing even more.
 
I have that much esteem in their position in the community and in their levelheadedness.
 
@ChrisJester-Young wait what? :D I'm sure you meant Peter ;)
 
Also, you'll note that the accepted answer to your linked question on meta.SE is "I may be in the minority, but I say no. You should downvote it or leave a comment explaining why it breaks the rules." and cites a meta.SE rule stating "Answers that are answers but are factually wrong or are actively bad (say promote SQL Injection) should get downvoted, but not deleted."
 
Right. I wrote that, and I'm saying that your answer was not an answer.
 
8:00 PM
@COTO If you read the entire post, he was referring to answers that accidentally break the rules, not those that deliberately break the rules.
 
You skipped over that part of the cited meta.SE post.
 
@MartinBüttner Lol. I'm sure Peter will get in too, but I've interacted with him less than with you or Doorknob.
so it's harder for me to vouch for him openly. :-P
 
@Geobits You mean the part that says "If you delete answers that don't meet the spec, you're simply inviting more answers that break the same rules, because nobody can tell it's already been tried and shot down. Showing people that we downvote might discourage that."?
But... whatever.
No satire allowed.
Got it.
 
@Geobits while you're here, did you have a look at the book sort proposal?
 
8:05 PM
@COTO No, your post met the spec (I assume, I haven't run it). It just wasn't an answer to a code golf, since it wasn't golfed. At all. On purpose.
@MartinBüttner Yea, I didn't see aything glaringly wrong with it :)
 
any opinion on which optimality criterion to choose?
 
I like area, personally.
 
hmmm
I'm starting to think that by that criterion it might actually always be optimal to sort by area
okay, no I think I've got a counter example
it requires some weird book dimensions though :D
 
I was just trying to draw one, what have you got?
 
5x5, 1x24, 5x4... this is the order by area which has an overhanging area if 19+16=35... if you swap the top two you get 20
the difference is pretty large, so this does happen for less extreme dimensions, too.
 
8:17 PM
Makes sense, just make sure you include an example of that in your test cases.
 
@ChrisJester-Young You actually can. (I'm at school right now; I'll elaborate at home)
@COTO Response coming soon (see above message).
 
@Doorknob Okay, I'll just nuke the comment and let you explain (or just do it). :-)
0
A: Jumblers vs Rebuilders: Coding with Tetris Bricks

feersumCrack of xnor's 340-area answer No cryptographic mumbo-jumbo for me. Code: 10000001111111222 22388892357889356 69968999352143221 80867227952780811 a=23333333444444 b=71122368066667 c=47948316823798 d=79254432699451 496857235742 579159386011 606961806231 pow(a,b,c)-d 626562...

Now undeleted.
 
@ChrisJester-Young There's no need for the conversion any more, if the author just reposts it in the other question.
which he just did
 
8:32 PM
@MartinBüttner Uh okay, what should I do?
 
I'll leave a comment
 
Looks like I came back at just the right time!
 
for?
heated discussions? ;)
 
Done before you typed that lol
 
8:35 PM
Thanks for removing it before I had to do it myself. :-P
 
I almost said something inappropriate in response.
 
How is October 13 still on the transcript? Don't you guys say anything funny or have any kind of strong opinion that is exactly opposite to mine when I am not here?
 
@Rainbolt Until this afternoon it went back to 10 Oct
Also I think the last week was generally very quiet in chat.
man, I was just going to reply :D
 
8:46 PM
mods reading deleted posts...
 
@MartinBüttner Yeah, it's the Streisand effect here. :-P
 
For example, there is absolutely zero point to deleting messages in Teacher's Lounge, except to stir up Streisand effect. :-P
 
9:01 PM
@ChrisJester-Young I guess it's too late now, but to do the thing you were trying to do: Edit the answer to change it to a pastebin link (or whatever), mod -> convert to comment, paste a link to the thing you want it to be a comment on, submit.
 
@Doorknob Interesting, thanks!
 
I've added a few test cases now
0
A: Sandbox for Proposed Challenges

Martin BüttnerBook Stack Sort code-golfarray-manipulationsorting When stacking books you usually want to put the largest ones at the bottom and the smallest ones at the top. However, my latent OCD makes me feel very uneasy if I've got two books where one is shorter but wider than the other. No matter which o...

It's quite tough to come up with ones, where I'm certain of their correctness without writing a solution myself, but they do cover all the relevant cases: identical elements, 0-overlap solutions, ambiguous solutions, solutions that can be sorted by total area, and solutions that can't.
If no one spots any problems, I'll probably post it in an hour or two.
 
@COTO First of all, this was by no means a unilateral action. The post had several "not an answer" flags by various community members, and other users clearly voiced their objection in the comments. Secondly, the score of a post and how many votes it has is mostly if not completed irrelevant when I decide how to handle a flag or a problematic question/answer. (Think a la code-trolling.) cont...
 
... (cont.) Finally, deletion isn't really an outstanding thing on Stack Exchange; we're not afraid to maintain site quality, and removing objectionable content (soft-deleting, not completely nuking, mind you) isn't all that exceptional of an occurrence. We can doubtlessly both agree that your answer was not, in fact, a golfed solution to the challenge presented, and while it made me laugh, "joke" or "just for fun" answers are more in the style of Reddit / forum-like sites, rather than SE.
 
9:14 PM
oh speaking of code trolling
I have a great idea for a challenge
@COTO you said you weren't around when code trolling was a thing. this is actually relevant to the discussion, you might want to read the discussions about it meta.codegolf.stackexchange.com/q/1514/8478. Code trolling was literally intended to get answers like yours. And it was really popular - with people from other SE sites who found through the HNQ. And the discussion revolved a lot around fun & popularity vs quality standards. So it might be an interesting read.
 
9:29 PM
@MartinBüttner I don't even know what "code trolling" is. Maybe we should start with that. Any kind of satirical answer, I take it, regardless of popularity, relevance, the effort needed to produce the code, and quality? Believe it or not, there isn't a single link to a "code-trolling" question on your meta link, and clicking the tag leads to a dead end. So what exactly is and isn't code trolling?
Scratch that. I found a link.
 
@COTO Code trolling doesn't exist any more. It was a type of underhanded popularity contest which asked for trolling answers. Most questions were in the form of a really bad give-me-teh-codez SO question and answers were supposed to take the piss.
 
I think this was the first and "canonical" code-troll, which is why it's locked rather than deleted.
 
So... the first one I visited has more answer upvotes than the 80 most recently asked questions combined. I think it also has more answers than the 80 most recently asked questions too. And it was... somehow... diminishing the quality of the board?
 
OP of codegolf.stackexchange.com/q/40150/194 just edited to answer one question, so I presume the reason he hasn't edited to answer mine is that he doesn't understand it or doesn't want to answer it. Does anyone have suggestions for alternative phrasings which might make it clearer what the problem is and why it needs fixing?
@COTO It was the direct cause of more utter crap being posted in one week than the entire content that has been posted in the last month.
 
I guess so. Although considering PPCG gets maybe 1-2 new questions per day, that's not exactly a feat.
But if it was out of control, I can understand that.
 
9:50 PM
One of our top code-trollers also posted this wonderful question: skeptics.stackexchange.com/q/7639
Correction: Victor is not a top troll. He is the grandfather of code trolling.
 
@COTO it gets 3.8 ;) (we're on a bit of a recession though, it used to be above 5)
@COTO You're still arguing by its popularity. That does not make it a good challenge. The main problem with code trolling was that all questions were like "write code that can do whatever you want as long you can somehow claim it's related to the topic I've given you"
 
@COTO The good code-trolling challenges were sparse. They took the fall with the terrible challenges, which were plentiful. The real problem with them was that, by definition, everything labelled code-trolling was automatically on topic and designed to be ugly.
 
@Rainbolt I saw that one. And the "multiply without + or *" one. They're popular, full of wonderfully creative answers. The most popular answers seem to be the most creative and on-topic ones. I guess I just don't see the problem. Unless there were like 3+ of them per day.
 
I don't know if there were 3+ per day, but we definitely got one like "Print your mom." every day.
 
OK. Yeah. That would be annoying.
 
9:59 PM
Let me put it this way: we had at least four exact duplicate questions copied and pasted from SO to here and labelled code-trolling.
And that's only the ones I personally saw.
To make matters worse, sometimes the author would direct the original poster here, to show them how much "fun" we were having at their expense.
Code trolling isn't really gone. All along it was just a popularity contest in disguise.
 
@Rainbolt Not really...
 
@Timtech Oh? The most popular answer didn't get accepted as the answer?
 
@Rainbolt Right
 
@Timtech Well, I disagree.
 
@Rainbolt I think I argued back then that most code trolling questions were just (bad) underhanded pop-cons, and we even retagged one or two of them.
 
10:05 PM
They were fun though
:)
 
Is this stuff documented anywhere remotely visible? I skimmed the site rules when I first came on, and there's nothing about "code trolling" anywhere. Even the "satirical answer = code trolling = not an answer" rule is subsumed by the broader "off-topic" rule, which I didn't even consider since as far as I was concerned, whether or not something is satirical has nominally nothing to do with whether it's off-topic. "I like turtles." in a question about how to write a loop is "off topic".
 
No, it isn't documented.
 
@COTO in regard to your answer, Peter quoted the relevant phrase from the help centre, I think
 
Yes, the "off-topic" page of the help center, which I specifically didn't read because I thought I knew what "off-topic" meant. :D
 
cya
 
10:07 PM
well where else should we put it :P ... it's the one site-specific page of the help centre
 
Oh, I guess @TimTech was right. From the old tag wiki: "For the winning criteria would normally be , but this will not be implied."
I guess it just turned out that way 99.9% of the time, so I assumed that's how it was.
Can anyone else see the wiki history?
Because I can't
 
let me check
in other words, "yes"
 
I started to edit the tag months ago, and my edit is still hanging out there, half written. That's how I found it.
Maybe if I discard my edit I'll be able to see the revision history.
Ah, there it is.
Is Stack Exchange going super slow for anyone else?
 
@MartinBüttner Putting it in codegolf.stackexchange.com/help/deleted-answers would help. That's one of the 6 questions that comes up on the main page of the help center. It lists several examples of off-topic answers, all of which I would consider to be "obviously" off-topic.
@Rainbolt It's chugging for me too.
 
@COTO I believe that one cannot be customised for PPCG only.
@Rainbolt I was just about to ask
 
10:13 PM
Guess that means it's time to go home :)
 
Well, it figures that my post popular answer ever--by far--wasn't long for this world. Live and learn. :(
 
@COTO Save them for popularity contests. ;)
 
0
Q: Bug on home page?

TimtechThe home page is blank for me. It is the same with meta. Anyone else with ideas/solutions/problems?

 
0
Q: Bug on home page?

TimtechThe home page is blank for me. It is the same with meta. Anyone else with ideas/solutions/problems?

 
Oh by the way, we missed the deadline :3
23
Q: Let's commit to graduate

TimtechIt's already been 1,082 days in beta, and I'd like to see our site graduating sometime in the near future. James Khoury proposed aiming for the 1,337th day since beta started, and I like that idea, especially since it's only in 255 days. I'd suggest increasing the voting going around so our us...

 
10:24 PM
Sep 26 at 10:45, by Martin Büttner
today is the day Timtech wanted us to graduate:
 
@MartinBüttner <3
 
The byte of days passed overflowed after 255 ;p
Can't you just hardcode the output for this question? ;p :
16
Q: Short Deadfish Numbers

MegaTomDeadfish is one of the best known non Turing-complete programming languages. It has only one accumulator (which starts at 0) to store data, and only four commands: i - Increment the accumulator s - Square the accumulator d - Decrement the accumulator o - Output the accumulator A Deadfish progr...

 
Sure but if you look at the existing answers, that won't be shorter than calculating them.
 
Unless...
Binary :)
Each letter is 2 bits?
That's like 4.3k total score
Which is a lot :(
 
10:38 PM
I think I even tried packing it in Unicode
didn't help
 
Okay :)
 
10:59 PM
0
Q: Book Stack Sort

Martin BüttnerWhen stacking books you usually want to put the largest ones at the bottom and the smallest ones on top. However, my latent OCD makes me feel very uneasy if I've got two books where one is shorter but wider than the other. No matter which order I place them in, the top book will extend beyond the...

 
cya
 
later
@xnor thanks for catching that
 
sure
are you going to restrict away brute-forcey solutions?
 
yes, I think that would be fairly boring
 
yeah, it's an annoying pitfall for optimization problems
i totally agree with a restriction
 
11:13 PM
unfortunately I don't know the actual complexity of the problem
 
hmm
it has optimal substructure so there should be a dynamic program
 
I'm thinking about just requiring polynomial time solutions. that should be possible.
 
actually, i'm not sure
do you have one?
 
@xnor it doesn't have optimal substructure for each subset
otherwise you could just sort by total area I think
 
ok, yes, the dynamic program would have to remember the subset and the top element
and there's exponentially many subsets
but n*2^n at least beats n!
you might be able to embed hampath/travelling salesman though
 
11:16 PM
I could ask for exponential complexity with fixed exponent
 
like 2^n?
 
fixed base, of course
yes
or 10^n
 
hmm
 
or I could turn it into with the tie breaker being code length
 
i guess that technically rules out n!, though barely
 
11:19 PM
@PeterTaylor does meta.codegolf.stackexchange.com/a/2366/8478 seem to be equivalent to some NP-hard problem?
 
i'm failing to reduce hampath to this because you can't embed an arbitrary graph; if edges ab and bc have weight 0, so must ac
 
It might be equivalent to Hamiltonian path in some given metric, but I'm not sure which metric and that doesn't guarantee that it would be NP-complete. I was just about to go to bed, and I'm already half asleep, so that's as much of an answer as I'm willing to commit to right now.
 
@PeterTaylor Thanks for looking at it anyway.
@xnor I think if you have that case, you can probably collapse those vertices into one, removing edges pointing to intermediate vertices. Could it be possible to reduce Hamiltonian path to it after that step?
 
i see
then you're left with an instance where no book is strictly bigger than another
at that point though the optimum is probably just to order the books by thinness (length-width)
oh, never mind, you're only removing triples where a>b>c (> meaning full covers)
not just when a>b
 
11:36 PM
yes... I could have phrased that more clearly
this problem is surprisingly hard
 
well, if we can't come with something, that's a good sign for a complexity challenge with golf tiebreak
 
even then it could be quite tough to prove optimality and complexity of submissions
 
i think it's reasonable to answer from answerers
test cases would help with the understanding that simply being right on the test-cases doesn't replace a proof of optimality
 
@xnor what do you mean by that?
 
by what?
 
11:42 PM
hover over my post or click the little arrow on the left ;)
 
i know, i was hoping you'd say specifically what part was unclear
i mean that you'd include test cases
but i think that a solution giving the right answer on all test cases shouldn't be sufficient
 
but that's not the post I was replying to ^^
 
rather, answerers would still be required to give a proof of optimality
 
oh, my mistake :-P
i thought the arrow was to the whole block
i meant "reasonable to expect from answerers"
 
11:44 PM
ah I see
 
it's the unfortunate type of thing where i expect more work correlates with fewer upvotes
but i think those who'd do this type of challenge are dedicated enough to not care
i'm sad COTO's hilarious java satire was deleted
i understand it had to be done because of the rules
but that was a lot of work and a lot of cleverness
if code trolling produces gems like this instead of drivel, i'd be all for it
 
I've edited in a draft for a fastest-algorithm scoring blurb...
 

« first day (1362 days earlier)      last day (3476 days later) »