Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 155 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

I've noticed lately that there are questions on Stack Overflow being closed as "not a real question" which are to an expert in the subject clearly "real questions" -- unambiguous, not rhetorical, not overly broad, on topic, with straightforward answers. But the subject of the question is sufficiently obscure that even a reasonably knowledgable person might not "get" what the question is about.

Two recent examples that come to mind are:

Why doesn't C# support the return of references?

Here we have a question that was voted down to I think -6 before Bill The Lizard was kind enough to resuscitate it; it is an excellent question and went on to be upvoted pretty highly.

An example from today is:

Is there attached property in C# itself?

Again, same thing. The question has been closed as "not real" despite the fact that it is on topic and has an unambiguous answer (namely, "no, but we tried").

(It has since been reopened, though of course it is likely that attention brought to it here was a factor.)

The fact that these are both in a sense "the same" question -- namely "does C# support this obscure feature? If not, why not?" -- is I think not particularly relevant to the more general issue. (*) The more general issue here is that both these questions are quite good questions about a topic so obscure that five smart people agreed that it was "not a real question".

I perceive no particular barrier to getting answers to ordinary, boring questions about common scenarios on this site. I think that there is huge value in this site's capability to provide answers to the difficult, obscure, interesting questions. Closing questions about obscure topics early, before experts can actually see them, seems to be working against that value.

Is this more general issue a widespread problem that anyone cares about?

If it is, are the existing mechanisms (voting to reopen, calling in a moderator, asking questions like this on meta, and so on) sufficient to address it?

If they are not, what mechanisms could be improved to address it?

(*) I tend to push back on questions of this form because the answers to them are typically all the same, namely "nice idea but we have higher priorities". These two are exceptions to that rule in that they are both features that we thought were good enough that we actually implemented them before we cut them.

share|improve this question
The fix for your second example is to edit the question so that it makes sense. If nobody reading the question can figure it out, it's going to get closed because nobody can figure it out. – Robert Harvey Jun 24 '11 at 19:07
The first question probably irritated people because it was grammatically incorrect, even after an edit by a well-meaning community member. It's difficult to be sympathetic when the OP can't even compose a complete sentence. – Robert Harvey Jun 24 '11 at 19:11
up vote 10 down vote accepted

I think the system is well adapted to handle this, and that it is not a problem. I've seen this kind of "This seems entirely nonsensical" thing happen on other sites. On Skeptics, a question was hastily closed as "Not A Real Question", when those who first observed it noted that it seemed a preposterous claim that no one believed. It ended up mostly as quickly reopened after it was shown to indeed be a pretty well-known claim. Admittedly, the only visible actors in that particular sequence are moderators, but there was actually a hefty discussion between multiple users on the subject.

Which I think is a strong point towards the system - fact is, with as many experts as we have, there often is someone in our number who can identify what is a truly off-handle question, and what is just something that is obscure. In your second example, it was even reopened by 5 normal users, not requiring moderator intervention.

It definitely is quite disconcerting to see a legitimate question closed because of ignorance, but fact is, there do exist questions which fall along the same lines that are indeed "Not a Real Question". You mention yourself that these two are certainly exceptions among their kind. Culling problem questions helps keep the question quality high, especially when it comes to really quick questions (like the roots of those you cite) that are extremely short and have little information to them. It isn't difficult to imagine how someone might feel that it just comes so out of left field that it is not a real question to them.

As such, the presentation of an obscure question helps best when it contains pertinent data to show that it is legitimately a question. Failing that and ending in closure, we can afford a margin of error because we possess the tools to reverse any erroneous closures of this kind, however. We may be tough on enforcing our quality standards, but we're also tough on making sure our decisions are right. The community rarely backs off on reopening a truly inappropriate closure. I think that this overall exerts an emphasis on desirable quality in our questions, and yet retains the ability to salvage even the most obscure of elements.

share|improve this answer
+1, but without Eric these questions are not constructive. Those of us without intimate knowledge of the language/framework designer's whims have no way of accurately answering "why does language X have/not have feature Y". – user7116 Jun 24 '11 at 18:52
+1 especially for the third paragraph and «these two are certainly exceptions among their kind.» «The presentation of an obscure question helps best when it contains pertinent data...» – Josh Caswell Jun 24 '11 at 19:09
Thanks for your thoughts Grace, I appreciate it. – Eric Lippert Jun 24 '11 at 21:50

This certainly is an issue: A question that doesn't manage to make sense upon a first quick scan is definitely in danger of getting closed.

Whether the existing mechanisms are sufficient, I can't say for sure - it's very hard to tell how much good content gets unfairly closed this way. However, all the cases I've seen worked out fine, and if the OP is able to make a good case on Meta without letting too much of their anger show through, the chances of getting a good question reopened are excellent.

Very effective methods to help an OP who asks a misunderstood high-level question include:

  • Adding a comment saying that this is a good question. If it comes from a respected member of the community, it will stop people from pulling the close trigger too quickly.

  • Trying and editing the question to make it clearer.

  • Providing a serious answer. If a question has a sensible answer - especially from a respected member - it usually has an increased chance of surviving.

share|improve this answer
Aye much to the final point. Many a time, it is an excellent answer that is the savior to the question - even truly poor questions that are poor in their own right, they can be saved by an excellent answer. – Grace Note Jun 24 '11 at 18:20
Guilty on the first point for the former question. That question was already going down when I answered it and a comment on the question might have prevented it from being closed in the first place. – Rick Sladkey Jun 24 '11 at 19:50

The question might not have been closed if the author had invested a bit more effort. The more it looks like a lazy one-liner the more likely it is to get closed. It's an appropriate auxiliary metric for gauging question quality. And quite frankly I find the 50 upvotes unwarranted, regardless of how clever the actual topic is.

With the answer at hand it's also obvious that the question was only ever answerable by the language designers and/or the vendors marketing department. And that IMHO legitimates the NARQ votes to some extend. "Is there a special reason?" gave the first question a subjective slant as well.

I'm wondering if it is a widespread and real problem that valid questions get closed due to complexity and difficulty however.

share|improve this answer
+1: Clever topics are not the same as clever questions. – Lightness Races in Orbit Jun 27 '11 at 22:03

I think these two highlight an interesting feature of StackOverflow, namely your expertise and intimate knowledge being available to answer these questions. If you weren't available, both questions would merely be us lusers pontificating on subjects we have no real way to answer.

I don't vote that way in or anymore, but usually any "Why doesn't Language X have a feature from Language Y" is Not a Real QuestionNot Constructive because the answers can only come from designers, or someone privy to the whims of the designers.

share|improve this answer
Though I certainly appreciate your point -- I find it personally irksome when someone pretends that they know what my colleagues were thinking and gets it wrong -- we are trying to be more transparent about these sorts of design issues. The annotated specification, blogs, forums, and so on, are all ways that we disseminate information about design choices into the world. Product designers do not necessarily have a personal monopoly on information about their design decisions. I've certainly noticed Jon, and others, giving really good answers about design decisions. – Eric Lippert Jun 24 '11 at 19:11
@Eric: I can imagine. I just have a tough time leaving open a question that basically is, "Short answer: No. Long answer: see the spec or wait for Eric." – user7116 Jun 24 '11 at 19:14
@Eric: I'd also like to note that my answer may read like I feel you guys hold back, which isn't the case (I'm a fan of the volumes of information from the .Net/C# teams). Just that there are a lot of things even the really sharp folks would have to guess at. – user7116 Jun 24 '11 at 19:19
Sure, I take your point (though I note that anyone who had happened to read and remember my blog article on extension properties would have been able to answer the second question reasonably well). I don't want to dwell on the specifics of these two questions overmuch; I'm more interested in the general case of questions where it is not apparent that there's a "real" question there because the subject is obscure. I don't have a broad enough perspective on the site to know if this sort of thing happens a lot, and if it does, if it is a problem not addressed by existing mechanisms. – Eric Lippert Jun 24 '11 at 19:28
It's not a stellar answer but I did try to answer the former question and even got six votes before Eric came along and the Eric-effect kicked in. Having industry insiders around helps but questions aren't pointless without them. – Rick Sladkey Jun 24 '11 at 19:57
@Eric: That's more than just something to note, really. Just because a question can only be answered by an insider, doesn't mean an insider needs to answer it personally on SO. On a few occasions I've gone digging through blog posts and mail archives and such looking for the motivation behind design decisions in order to answer a question--in fact, faced with either of the questions you linked to, my first step in answering would have been pointing a web search at your blog. – McCannot Jun 24 '11 at 20:21
I think the obscurity is exactly what makes it NARQ and what turns it into one when you answer it. If someone asks "which room in Kate's house is bigger, a or b?" the general response will be "who knows and who cares?" But if I write a lovely long answer that tells you all about my house design priorities, maybe people now care and learn from the answer. I think all you can do is come to meta when you see one of these closed, and ask for reopen votes so you can spin straw into gold. – Kate Gregory Jun 25 '11 at 1:06
@Kate: b is bigger, obviously. – user7116 Jun 25 '11 at 1:13

I think the answers you provided are perfectly clear albeit suffering from some grammatical/formatting mistakes in the beginning. The answer to your question is that some people are too busy to take the time to read and understand the question before they are flagged. There is a balance between giving people too much leeway on questions and not enough. It seems we have not found that yet in some people - and it seems lately with those on SO.

share|improve this answer

I won't write a full answer as I think that everything has been covered, but I do want to add my two pennies.

I've noticed lately that there are questions on Stack Overflow being closed as "not a real question" which are to an expert in the subject clearly "real questions" -- unambiguous, not rhetorical, not overly broad, on topic, with straightforward answers. But the subject of the question is sufficiently obscure that even a reasonably knowledgable person might not "get" what the question is about.

Two recent examples that come to mind are:

Why doesn't C# support the return of references?

Here we have a question that was voted down to I think -6 before Bill The Lizard was kind enough to resuscitate it; it is an excellent question and went on to be upvoted pretty highly.

Looking at the question, I'd guess that there were a spate of initial close votes because, on a site where low-quality, poorly-researched, misstated and vague questions are becoming somewhat rife, it's become almost a reflex action for some to react in this manner for very short questions perhaps without really considering the question's merits.

share|improve this answer
(OK, to be fair, mario pretty much covered this already.) – Lightness Races in Orbit Jun 27 '11 at 22:03

You must log in to answer this question.

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