Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 153 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

Before stackoverflow (and currently, to a lesser extent) I helped out on several IRC channels. There have always been certain FAQs for each channel. Over the years there have been a variety of resources we use as authoritative references; SO has become a great site for many of them.

More often there are the non-FAQ questions. "Why is this particular site/code/application doing what it is doing, and how do I fix it?" IRC is, IMHO, perfect for this. Immediate, personal, tailored problem analysis and solution. Posting a blog entry about the solution would—most of the time—be of little use to anyone else.

I see SO as the opposite of IRC in this regard. While it's nice to get a community of rabid experts vying for the opportunity to fix your problem, it seems to me that there is little long-term value (either to SO or the 'net as a whole) to hosting questions and answers related to a very specific problem.

Some examples:

  • Is anyone but the OP going to benefit from this question, describing a problem with SVG pie-charts in Ruby where the problem was a mistake in degree-to-radian conversion?
  • A user with invalid HTML markup that needed a specific CSS hack to fix his layout?
  • This user has created a site that looks horrible in IE and wants help fixing it.

I realize that the world isn't black-and-white. It's possible to conjure up some almost-plausible scenarios where a future user might find the question after searching and get an answer. But I submit that by-and-large questions like this are a drain on SO.

If you could categorize and identify such questions (either absolutely or on a scale), do you think that they should be discouraged?

If so—if this is a worthy problem to investigate—do you have any suggestions on how to identify such questions, separating the wheat from the chaff?

share|improve this question
So you basically want to get rid of anything that's no FAQ-able? Well, then why have SO in the first place? I can just Google the FAQ's, I come here because I have very specific problems which need expertise, Google can't help me on that one. In fact I'd propose the opposite, close all "googlable" Questions. – Ivo Wetzel Dec 7 '10 at 13:33
@IvoWetzel I think there are more categories for questions than FAQ and not. Roughly (and broadly), there are perhaps: RTFM, FAQ, Common, Algorithm, Advice, One-Off. "Common" questions are the sort where SO has excelled for me: users posting specific backtraces or error messages and getting solutions. "Algorithm" are "what's the best way to do this?". "Advice" are "which library is the best?". If it was up to me (let's be glad it isn't) I would try to discourage or delete RTFM and One-Off questions, and minimize Advice questions. – Phrogz Dec 9 '10 at 4:09
I agree on that, everything that can be found in the docs, via Google or via 15 minutes of the usage of intelligence should be discouraged. But as I said the "Is anyone but the OP going to benefit..." point is stupid in my opinion, because the solution might not be helpful for you, but someone else might encounter the same obscure hard problem, and then it will be helpful. – Ivo Wetzel Dec 9 '10 at 5:01
up vote 1 down vote accepted

Information is information, and humans are spectacularly bad at anticipating the situations in which information will prove useful.

Sadly, as evidenced by this question, we're also spectacularly bad at learning from history.

share|improve this answer
So it is your belief that we cannot predict which questions are one-offs and which will be useful, so we should allow them all? I disagree, but fair enough. I wonder, though, if there might be a mechanism to prune them long after the fact. Can we use history itself to tell us which questions might ought to be purged? – Phrogz Dec 7 '10 at 5:12
@Phrogz: How exactly do you determine if someone googling by found the information useful? That knowledge is unavailable to us. If the information is useful once, assume it will be useful again. There is no other sane course. – Nicholas Knight Dec 7 '10 at 5:16

The only one-off questions I completely object to are of this form:

Q: I need to frozz my bobbit, but I can't seem to, even though I've zazzled the furmglobinator several times and checked my mutz levels repeatedly.

A: Ooops, my bad, I forgot to turn the power on.

Beyond the "oops, dumb mistake, never mind" type of one-off questions -- which should be deleted -- the criteria I use is simple:

Could I learn anything from this question?

This depends on how well the question is asked, and whether it provides meaningful background and context. How to ask covers this:

Be specific

If you ask a vague question, you'll get a vague answer. But if you give us details and context, we can provide a useful answer.

Make it relevant to others

We like to help as many people at a time as we can. Make it clear how your question is relevant to more people than just you, and more of us will be interested in your question and willing to look into it.

If a question is failing to teach me anything, it's generally because it ran afoul of one or both of these How to Ask rules, in my experience.

share|improve this answer
+1 I recently answered a question where someone had a problem with a very specific problem JS Graphing lib. So I took a look at the 11k lines of code and well, I fixed it in about half an hour, showed that the lib was poorly designed in places, helped him to get rid of this project blocking problem and I got a 500 bounty. Everyone won. Such questions are great, yeah they don't give you the 15+ upvoted that "jQuery how do I fade an element" would. I'd rather get rid of those ones, since these are just a google to docs away. – Ivo Wetzel Dec 7 '10 at 13:39
Wouldn't it make sense then to create another site which would be more like a mix between an irc-channel and stackoverflow. Where these "what does this error message mean?" or "why doesn't my use of this thing work as expected?" questions could be asked without being stuck in stackoverflow's database polluting the search results when looking for answers to more generally useful questions. I think this breeds laziness in that people immediately start reaching for their browser when faced with a problem instead of even trying to solve their problem by themselves first. – Sam Mar 3 '11 at 13:09
@sam you mean like ? – Jeff Atwood Mar 3 '11 at 13:22
Yeah I know SO has a chat, but I'm not the one needing it. I mainly ask my questions of that type on IRC. However it seems there are a lot of people asking these "help me debug this"-questions, that don't know about either SO's chat or IRC. – Sam Mar 3 '11 at 13:27

You must log in to answer this question.

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