46

After seeing how to make xkcd style graphs in Mathematica, LaTeX, R, Matlab, and now looking for one in Python, I thought that this is too much.

I am now eagerly awaiting for "xkcd style graphs in C++", "xkcd style graphs in C#", "xkcd style graphs in Haskell", and eventually, who knows, "xkcd style graphs in bra*nfuck".

Do we really want all these questions? Shouldn't people just reuse the algorithms from other languages, implement them in their favourite language, and if trouble arises, post questions about the specific trouble that arose? These questions give me a big "can you do this task for me?" feeling.

Is it really appropriate to keep asking this over and over again?

Improve this question
  • 2
    I felt that the first four are valid since all of those are for plotting and different. I voted to close the python as a dupe since it simply uses a matlab-like library. Other non-plotting "languages" should also be closed as dupes. – Mooing Duck Oct 3 '12 at 23:16
  • 22
    Over on Christianity, we tried and failed to think of a way to ask the question. So there is some hope we won't have a copy for each site. – Jon Ericson Oct 3 '12 at 23:20
  • 8
    These questions give me a big "Hey, this was big on the multi-collider, let's just see how many badges this will get me" feeling – sehe Oct 3 '12 at 23:23
  • 2
    @sehe: Excepting the Mathematica version, which is the prior art the rest of them are based upon, of course. ;) – Jon Ericson Oct 3 '12 at 23:25
  • @JonEricson Derp. Note "Hey, this was big on the multi-collider" (rather indicative of prior-art situation) – sehe Oct 3 '12 at 23:27
  • Also note that the last two were asked by the same person, who, looking through his history, has only one other python related post. And approximately 1/4 of his reputation is from those two questions. – ughoavgfhw Oct 3 '12 at 23:43
  • 2
    I don't care about who posted them; I'm concerned about the (IMO poor) content of the questions. I just want to know what the community thinks I should do when I see the next question like these: burninate on first sight or not? – R. Martinho Fernandes Oct 3 '12 at 23:45
  • 4
    @R. Martinho Fernandes: The Mathematica question (and answers) are exceptionally good. It appears totally worth the freakishly huge traffic. It's also on a site that is orders of magnitude smaller than SO. Each subsequent version is less interesting and less useful for building traffic. So, I would continue to vote to close. (I provided the final vote on the Python version, as it happens.) – Jon Ericson Oct 4 '12 at 0:30
  • 23
    +1 for bra*nfuck alone. That particular i always offended me. – Dennis Oct 4 '12 at 3:04
  • 5
    @JonEricson How can I draw an xkcd style cross? – Lorem Ipsum Oct 4 '12 at 4:22
  • 2
    I don't know if I've ever seen a better example of "don't ask a question unless you yourself are actually facing the problem" – Michael Mrozek Oct 4 '12 at 4:37
  • @yoda You sheep, come back to the herd – Dr. belisarius Oct 4 '12 at 5:08
21

If the question shows even a first attempt at something or is in the middle of a hurdle that needs that push, it's fine to leave it alone.

If it's bare bones and there to have the OP licking their thumbs counting off the number of badges and upvotes filling their gold plated dentures, then apply the same rules as you would any half-attempt at a question. Vote or flag to close as per your level of reputation and see it stay open for a few months yet before it is referenced down the line as some other user does the same and wonders why "da rulz" are not consistent.

Improve this answer
12

What the me-too questions haven't appreciated is that the original poster on Mathematica genuinely wanted to make an xkcd-style picture as a response to a friend. A slightly frivolous problem, maybe, but it was a genuine problem he faced and that he wanted to solve. I am willing to bet he had no idea that the question would go viral the way it did - a testament to the quality of the answers. Mathematica.SE has had blockbuster questions before that had 10,000-plus views in short order, but 100,000+ views on a site that usually gets around 2000 visits/day was hugely surprising.

With the exception of the TeX one, the copycat questions didn't have either the genuine need or a solid attempt in the question. I agree that the TeX.SE question stay open and the others ones stay closed.

(Full disclosure: I'm a moderator on Mathematica.SE)

Improve this answer
  • 3
    Frankly, I don't see how any of this is relevant. If someone asks a high quality question, who cares if it's a problem they actually faced or one they just made up? In this instance, the mathematica version is of no higher quality than any of the others: specifically, the OP was not in the midst of an attempt, and got stuck on specific code. They just asked for ideas, like all the other versions. – joran Oct 4 '12 at 15:23
  • 4
    @joran: The primary difference, as far as I'm concerned, is the community that underlies the various sites the question was asked in. For better or worse, Mathematica and TeX - LaTeX are much smaller communities than Stack Overflow. The standards of the former may or may not resemble the standards of the later. Further, it's increasingly less likely that subsequent versions had any intrinsic motivation to ask the question. It's certainly hard to see what asking for the same algorithm in a different language achieves besides some upvotes and attention. – Jon Ericson Oct 4 '12 at 16:50
  • 3
    @JonEricson Different standards on different sites is fine to a point, I guess, as long as the questions are being evaluated on their merits, not based on the perceived motivations of the asker, so I stand by my assertion that that should remain irrelevant. – joran Oct 4 '12 at 17:00
9

I would only like to point out that the repetitiveness of the questions is not the issue here. There are likely thousands of examples we could find of people asking to do basically the same task in different languages that no one bats an eye at. For instance, there are surely countless questions asking about how to accomplish basically the same task in every SQL variant imaginable.

Additionally, note that the spurt of XKCD graphing questions has stopped all on its own, which suggests that concerns about a wild proliferation of such questions is probably overblown. These things tend to die out on their own.

Let's try to focus our irritation on the quality of the question, not the fairness of which questions win the popularity lottery. As I noted in some comments, the fact that some may ask these questions in order to farm rep or badges may well be annoying, but that really should have no bearing on our judgement about the quality of the question.

In this case, I think the only really valid criticism one can level is that the OP in each case did not make any attempt themselves, and instead simply asked for ideas and code.

Personally, I agree that ideally more effort should be displayed when asking these sorts of things. But on the other hand I tend to be less anal about that when the question yields high quality answers in short order, and the resulting content is likely to be useful to significantly more people than just the OP. If I really feel the question displayed little effort, but produced some great answers, there's no need to close the question and risk losing content, I can simply down vote the question.

Improve this answer
  • 2
    The wave of me-too questions died out because they were being closed and weren't the rep magnets the askers had hoped. They didn't die out "on their own". The community provided the incentive. – Verbeia Oct 4 '12 at 20:04
  • @Verbeia Fair enough, that probably contributed. (I think they would have died out on their own anyway, though.) Still, there's really no need to be so cynical about people's motives for asking those questions. – joran Oct 4 '12 at 20:07
6

The XKCD questions have a lot going for them:

  • They're interesting
  • Each language implements it differently
  • They are non-trivial problems
  • They have a large following
  • They are actually problems people face (though mostly they're just really neat).

We may hate fun, but I feel like these questions have a serious application.

I was around for the great Hidden Features of X debacle, so I have a bit of perspective on this.

I've mentioned the positive side to the XKCD questions, but there is a dark underbelly: When does it stop? When is enough enough? The answer is, of course, it depends.

It depends on:

  • Community Sentiment
  • The value the individual question brings (as a commenter pointed out, there are diminishing returns)
  • Who is actually trying to solve a problem, and who just wants the reputation
  • Who wins the close war

Ultimately, these questions differ from the Hidden Features of X questions because they aren't opened ended. There is normally 1 or 2 ways to implement such a thing in a language. That may make them more applicable to Stack Overflow's mission, however annoying it may be to see them in the short term.

Improve this answer
  • 12
    Yes....but "what have they tried?" – Kev Oct 4 '12 at 0:29
  • 1
    Except for your second point, the law of diminishing returns kicks in for each copy-cat question. – Jon Ericson Oct 4 '12 at 0:43
  • Even your second point gets tenuous after the third or fourth question as some language implementations aren't going to vary that much – ChrisF♦ Oct 4 '12 at 11:34
  • I spent some time listing the positives, but I didn't list the negatives. I've gone ahead and listed what I thought the negatives to these questions might be (just so you wouldn't think I was all about them). – George Stocker Oct 4 '12 at 12:58
2

If someone would be able to define one single good version of the question as a "programming puzzle" on CodeGolf SE (if it would fit their FAQ of course), which allows any language to be used, we could then just point all new occurences to that one?

Improve this answer
  • 1
    Sadly, cross-site duplicates are not a thing. – John Oct 4 '12 at 18:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .