What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 131 Stack Exchange communities.

Should we pose questions that we know are difficult to answer? (But can be answered.)


For closing posts on the Stack Exchange network, the reasoning often is:

This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form.

Where is the line of what's reasonable in the case of a difficult problem?


Update: If a question wasn't clear, then of course, it cannot be answered without due clarification. A call for improvement is probably what the closing statement above is mostly aimed at — especially the part: "in its current form".

However, for example, "overly broad" can have different meanings. Modifying CBredlow's words below, it may be easier for one person to answer a question that another deems as overly broad; and vice versa. So "overly broad" can simply mean difficult for some.

A philosopher could approach questions differently than a scientist.

The same goes for "ambiguous" — the definition of the word when it comes to language-use is: "open to more than one interpretation". Something like that could be difficult to answer for some, but not for others.

share|improve this question
    
We should ask good questions. That does not necessarily mean difficult. –  Bart Nov 28 '12 at 13:31
1  
You got the emphasis wrong there, it's 'cannot be reasonably answered in its current form.'. E.g. if you ask a difficult question that can be reasonably answered by providing more information, then that's great! Your question here, as it stands, is rather vague, please improve it's form. –  Martijn Pieters Nov 28 '12 at 13:38
    
There is a difference between a difficult question and a difficult-to-answer question. –  BoltClock's a Unicorn Nov 28 '12 at 13:50
    
Ah, interesting, where do you see the difference? –  Baumr Nov 28 '12 at 13:51
1  
That's a difficult question :P –  BoltClock's a Unicorn Nov 28 '12 at 13:53
    
LOL! Wise guy... –  Baumr Nov 28 '12 at 13:53
1  
"Nothing wrong with asking difficult questions, but if they are going to be as vague as this one, please don't." - I don't think this is a vague question. It's technically the opposite - direct, concise and straightforward. Furthermore I think it's a good question. How many times have you not asked spmething anywhere - not just here - because you felt it too noobish/silly/whatever? I have(n't) many times, I'll admit it –  skopp Nov 28 '12 at 15:16
    
@skopp Please look at the initial revision of this which is what that comment refers to. –  Bart Nov 28 '12 at 15:27
    
@bart Ah, okay. I misread the... context/tone (for lack of a better word right now). My bad - still quite a noob around the stack network. –  skopp Nov 30 '12 at 13:21
    
@skopp No problem. The comment by now is missing its context. Easy to confuse. –  Bart Nov 30 '12 at 13:22

3 Answers 3

up vote 3 down vote accepted

"Cannot be reasonably answered" can refer to several things. If my question is

I have written this standard "Hello World" program, but it keeps crashing on me. What might cause that?

it cannot be reasonably answered because I did not include enough information. In this case the utter lack of code is the problem. Its current form does now allow us to answer it besides a random guess. Update it and it might be good to go.

Were I to ask

Why did the creators of platform/library X decide to forego on functionality Y

that might not be reasonably answerable because we would require one of the people involved to respond. This can especially be so in the case of legacy APIs or libraries. Granted, closure of such cases can lead (and has lead) to extensive discussion on those questions and here on Meta.

Cannot be reasonably answered does not imply that difficult questions are not allowed. By all means, if it's a good question, ask it.

share|improve this answer
    
Interesting, so — considering your second example — one should take into account the availability of people on SO/SE? –  Baumr Nov 28 '12 at 14:18
2  
Well, it's a difficult one. Sometimes questions are so out there that an answer really can't be expected. And closing those seems fair. Though some argue that just because a user is not here (yet) the question is not by definition unanswerable. It's a bit a case-by-case decision. Just so I'm clear, I do not mean that questions about tools/libraries/etc. which have a small community presence on SO should be closed because of that. I'm talking about those requiring unlikely first hand knowledge and often don't go beyond a trivia status. –  Bart Nov 28 '12 at 14:25
    
Ah, interesting. Whether or not questions should 'hold out for the future', so to speak, was going to be my next question... so I guess that's a no? –  Baumr Nov 28 '12 at 14:36
    
Note that the second example does not require people to be SO users to answer! In cases where functionality Y seems like such an incredibly obvious thing to include in library X that its absence is conspicuous, there's a good chance one of the people involved has given the reason elsewhere, typically on a blog or mailing list. –  McCannot Nov 28 '12 at 15:46
1  
@McCannot Which makes these cases not clear-cut and causes a lot of debate if they end up being closed. It's a tricky one. –  Bart Nov 28 '12 at 15:48
1  
Well, usually they are clear-cut to the kind of people who read the relevant blogs/mailing lists/&c. and are thus capable of providing such answers, and entirely unclear to everyone else. On niche tags this can produce ire due to a perception of questions being closed by unqualified people, but what else can be done? Ah, well. –  McCannot Nov 28 '12 at 16:01
1  
@McCannot :) Which is exactly why I included the example. I'm not 100% clear on a stance one way or the other. All I'm saying is that closures like that may happen. Sometimes imho justified, though often hotly debated. –  Bart Nov 28 '12 at 16:04
    
Interesting discussion, I think that cases of "questions being closed by unqualified people" is an important problem –  Baumr Nov 29 '12 at 22:32
    
@Baumr Hang around on Meta long enough and you'll see several discussions like that come by. One side says "closed by unqualified people" the other side says "I'm fully qualified. You just don't like the closure". Question reopened. Question closed. Locked. Unlocked. Endless fun to observe from the sidelines. Not much fun to participate in. –  Bart Nov 29 '12 at 22:35
    
@Bart, haha! Sounds fun... might stay out of meta then :P But isn't the safe choice to leave such things open? Surely, someone ultra-qualified might chime in one day (and this kinda hits at that 'holding out for users of the future' thing). –  Baumr Nov 29 '12 at 22:40
    
On such edge cases I simply don't have the definitive answer. I tend to stay away from them. ;) –  Bart Nov 29 '12 at 22:42

Where is the line of what's reasonable in the case of a difficult problem?

There is no line. All levels of difficulty are welcomed here -- at least on the high side.

You've misunderstood the quotation:

This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form.

This is not talking about a problem's difficulty. This is talking about a question's form and phrasing. It's the difference between asking a vague question like "my code is broked, why?" and asking a clear question like "here is my code, it is crashing on this specific line with these inputs. Why?" Or asking ridiculously broad questions like "how do operating systems work" versus asking "how does the round robin scheduler work under Windows for user-space programs?"

share|improve this answer
    
Thanks, I've made some changes to my question based on your comments –  Baumr Nov 28 '12 at 14:46

In short, yes questions that are difficult to answer should be posted, but since I saw an earlier question where one liners are frowned upon, I'll try to go into more detail.

If it's a difficult concept, many people can throw their two cents in, and over time, the answer gets improved so that people can come and understand this question. Not all questions should be "How do I do XYZ in ABC", where XYZ is some programming 1 question.

Also, what may be difficult for one person may be something really simple for someone else.

When I see that reason for closing a question, it's usually because someone didn't give any specifics about their problem, or the question isn't really a question.

share|improve this answer
    
"many people can throw their two cents in, and over time, the answer gets improved so that people can come and understand this question"...that to me sounds more like the sign of a problematic question than something positive –  Bart Nov 28 '12 at 13:33
1  
I look at it as a question gets asked for a certain language and is answered quickly, and the language changes over the next couple of years, someone comes along and can update it. There could also be more than one way to answer the question and more people can throw in their own approach. –  CBredlow Nov 28 '12 at 13:46

You must log in to answer this question.

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