What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 130 Stack Exchange communities.

I came across this question in the review queue. Sometimes I come across questions of similar form. My thought is to vote to close as "too-broad" because:

  • Question is very broad, there are many, many ways to perform that task.
  • No code was posted, no specific coding problem was identified.

But, and this is in general, I always get tripped up when I see that these questions have a reasonable accepted answer: I would have voted it too-broad immediately had I seen the question pop up in the new questions stream; but when I see that the answerer seemed to find the question clear enough (and his answer was accepted and deemed satisfactory), I end up going back and thinking "if it was clear enough for the answerer to provide a specific answer, doesn't that mean it wasn't too broad after all?"

How should I handle questions like this?

And, in general, how should I handle questions that, alone, seem too broad, but when answered, seem specific? Should I ignore the answers when considering closing a question, or do they retroactively change the quality of the question?

share|improve this question
    
From comments below: The reason I view it as too broad is, if I answered I would have mentioned using existing components (like the accepted answer), custom drawing, third-party libs, etc. as options, with pros/cons. I see it as "too broad" because I have a tendency to list many/all possible options in my answers, and the answer I imagined myself writing was very long -- so to me it was "too broad". Other people answer differently, it wouldn't have even occurred to me to only mention a single specific option among many, but in reality, that's all the OP really needed. –  Jason C Feb 27 at 20:58
2  
For what it's worth, I don't necessarily consider presenting and contrasting multiple solutions to be a problem - if there's enough information in the question to expect that a high-level explanation will suffice. I tend to think of Too Broad more in terms of, "I haven't even decided which language I'm using yet - how do I render video?" –  Shog9 Feb 27 at 23:08
1  
@JasonC when I see a poor question before I commit to answering it I like to think what are the next 5,10 readers going to do... Is it likely that the question will receive 5 votes and get closed? If so I do my part by downvoting and voting to close even though I may personally like the question or would like to answer it. –  user221081 Mar 6 at 8:34

4 Answers 4

up vote 17 down vote accepted

Questions don't get any less broad when they're answered... But your perception of their breadth might. And that's ok!

"Too broad" means that the question asks for more than we can reasonably provide. A textbook answer, or answers to multiple, discrete questions. I really like jmac's summary of these questions:

If the asker identifies where they want to start, and where they want to end, but there are way too many pieces to fill in, then it is Too Broad:

too many missing pieces in the puzzle

When you're viewing an unanswered question and suspect it is too broad, you must make this call based on your interpretation of what the asker is looking for and your knowledge of the topic - there's nothing else to go on. If you're reasonably familiar with the topic, you stand a good chance of doing so accurately... Still, it is very easy to be wrong!

But when you're viewing an answered question, you have additional data: is the answer (or answers) complete? Do they satisfy the asker and stand a reasonable chance of being helpful to others facing the same problem? If so, it's a safe bet that the question is not too broad!

Note that you can always edit the question to better reflect what the asker was really looking for, at least in situations where you're able to determine this... And an accepted answer offers an excellent hint at what that was.

share|improve this answer
    
"But when you're viewing an answered question, you have additional data: is the answer (or answers) complete?" - With that in mind, I think it would be useful to have the accepted answer (if there is one) be displayed under the question. I made a post here to try and revive that request. It'd be a nice convenience instead of having to click through, review, then go back. –  Jason C Mar 2 at 6:48
    
"you can always edit the question to better reflect what the asker was really looking for" -- nope, not always. Sometimes existing answers essentially lock question from editing –  gnat Mar 2 at 17:01
2  
If the choice is between invalidating some answers and deleting all answers, editing is still a good option, @gnat. –  Shog9 Mar 2 at 17:10
    
@gnat, Too broad or not, if a question is not a good question, and it is reflected in the answer quality, I will edit the question and leave a comment on the existing answers explaining that I have edited the question and they may want to revisit their answers. On The Workplace, despite having done this many times, I have not gotten a poor reception from the answerers or the community, and that helps make our site a better resource in the long-run. –  jmac Mar 6 at 8:57
    
@jmac this works smooth when answers are new and of so-so value / quality ("fastest gun"). Requesting such answers to adapt to change doesn't cause much opposition of community / avid answerers. But if it happens that question change invalidates some great answer, I'll probably be the first to request rollback –  gnat Mar 6 at 11:11
    
@gnat, if it gets a great answer, it isn't too broad (or a bad question to begin with). Great answers can turn a mediocre question in to a good one -- if the question was absolute rubbish, it wouldn't get a great answer in the first place. I'm not advocating editing questions that are good to become great, I'm advocating editing questions that would be closed to at least be answerable. Obviously answer quality comes in to play there too. –  jmac Mar 6 at 11:36
    
@jmac this ain't so: there is even a feature specifically targeting unsalvageable poor questions with great answers: historical-lock. This feature is widely used at SO and Programmers; in comparison at Workplace it seems to be less in demand (since question quality norms were rather strictly enforced there from day one) but I wouldn't be surprised if it eventually gets used there too –  gnat Mar 6 at 13:30

If the only possible (quality) answers that you can imagine would be quite lengthy, and are well beyond what could be expected to fit into an SO answer, then the question would be too broad.

If it turns out, at some later point in time, that someone else managed to find a way to effectively answer the question in a reasonably concise manor, then you were simply wrong about the scope of the question. It happens sometimes; you think a problem is harder than it is, or weren't aware of some mechanism that greatly simplifies the problem. This is why we require 5 close votes, and not just one, and why there is a means to reopen a closed question.

This specific example is a bit beyond my area of expertise, but it seems to me like the existing answer is an acceptable answer (the OP thought so, as it's accepted), so I wouldn't say that the question is too broad. While it's entirely reasonable for you to have though so before seeing the answer, it would appear that you were mistaken. It happens.

What you should be on the lookout for is answers to excessively broad questions that are just the tip of the iceberg. People like answering really broad questions; almost as much as people like asking them. Lots of people try to write answers, but when the question really is very broad you just can't give a quality, comprehensive answer. You end up giving people just a small start, and leaving them with 99% of the work to still figure out after reading your answer. People also like to post answers that just link to blog articles, or suggest a book (or a chapter in a book) or some site tutorial that will comprehensively cover the question. Answers like these are exactly why broad questions should be closed; these aren't the type of answers that SE wants to have.

I don't feel (although, granted, I'm not familiar enough with the subject material to say conclusively) that your example answer fits what I described above. While it doesn't do everything for the author, it seems to be proving a sufficient solution, rather than "just a start".

share|improve this answer

If the OP accepted a specific answer, it could be that although the wording of the question was broad, the OP actually had something specific in mind. In some of these cases I have edited the question to change its emphasis. But whether this is worthwhile or possible without radically changing the question is something I do on a case by case basis. Changing the title of the question tends to be safer than changing its body.

In general, however, I do not think the existence of an accepted answer indicates that a question is a good question.

share|improve this answer
1  
+1 and just to add to your last sentence - when OP accepts an answer it doesn't mean nothing but "he is happy with the answer given". It doesn't mean the answer is good, best possible etc. that is usually indicated by the score by community. It is not necessarily good to answer poor/broad questions as if 5 people come around and decide the question is off-topic it might get closed and deleted altogether therefore what we do not encourage people to do it's to answer poor questions. Bad questions should be closed and deleted. –  user221081 Mar 6 at 8:29
    
+1 very good insight! –  Jason C Mar 6 at 8:31
    
I may be in the minority, but I am a strong advocate of aggressive edits to questions. If you can keep the spirit of the question and get useful answers for the asker, change as much needed to do that and make it a good question. If you can do that with just a title change, great! But an open question will always get better answers than a closed question. Focus on creating a good resource where possible, and in my experience it works out in the end. –  jmac Mar 6 at 9:00

Use "Too Broad" for Big List questions, or questions that would require the better part of a book chapter to answer. The question you cited is neither Big List, nor Book Chapter.

Example of a Big List question:

What are all of the common Javascript libraries?

Example of a Book Chapter question:

What are all of the capabilities of [this javascript library], and how are they used?

I'm guessing you voted "too broad" because the question is icanhazcodez. But that doesn't necessarily make a question too broad.

share|improve this answer
3  
My guess is that Jason was imagining the solution as something that's doing a whole bunch of custom drawing, rather than something as simple as using some existing font. That's a reasonable reason to think that the question is too broad. –  Servy Feb 27 at 20:38
1  
It's a good thing the OP included a picture, otherwise the question would have been indecipherable. –  Robert Harvey Feb 27 at 20:40
1  
@Servy That's an accurate guess. Had I answered that question; I would have mentioned using existing GUI components (like the accepted answer), custom drawing, and third-party libraries as options, with pros/cons. For me, I interpret questions like that as "too broad" because I have a tendency to list many/all possible options in my answers, and the answer I imagined myself writing was very long -- so to me it was "too broad". Other people answer differently, it wouldn't have occurred to me to only mention a single specific option among many, but in reality, that's all the OP really needed. –  Jason C Feb 27 at 20:55

You must log in to answer this question.

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