Meta appears to have come to a consensus that when a user asks a question that is easily answered with Google that we should not close the question but rather answer it properly.

Embrace the non-Googlers

Easily googlable questions

How should we deal with Google questions?

However, what does this mean about asking questions that can be answered easily with a Google search?

More specifically, should we encourage or discourage such behavior, and how?

Here are some reasons that I can think of for each option.

Reasons to Encourage

"We're building a canonical archive of programming answers."

  • And that includes answers easily accessible via Google search

Sometimes we want to reference another answer to a question that hasn't been asked.

Sometimes we want to save others the Google searching we did by asking and answering the question ourselves. In this case, we did do the research, but the question by necessity must not show research effort.

Reasons to Discourage

Question quality may drop, as users will feel less incentive to Google their problem. This will also probably result in a greater quantity of questions as well that have already been answered on StackOverflow, as Google searches often result in StackOverflow answers.

  • 4
    Questions can be closed for lack of research. Can you give an example of a question that shows appropriate research without utilizing a search engine? – George Cummins Aug 2 '13 at 18:14
  • @GeorgeCummins My questions surrounding this discussion mostly come from this question. The answer did not exist on StackOverflow, but I had no way to add it to SO if the question was invalid. So what should I do? Should I not ask and answer the question - then the information remains unavailable on SO, or should I ask it, and then it gets closed? – Cory Klein Aug 2 '13 at 18:16
  • Easily "Google-able" questions tend to have already been asked and answered, and should be flagged if appropriate. – Mike D Aug 2 '13 at 18:18
  • @MikeD Should they be flagged as "not enough research" or as "duplicate"? What if they haven't been asked already, as was the case for this recent easily google-able question? – Cory Klein Aug 2 '13 at 18:19
  • @CoryKlein My point was essentially the same made by Steve P. If it doesn't exist and its a fair question, ask away. – Mike D Aug 2 '13 at 18:32
  • 4
    SO is NOT a documentation archive and it is not intended to be one. There is A LOT of documentation that is VERY easy to find and access on all sorts of topics. That question was put on hold and should be, it does not "need to be added" because as I posted on one of your other MANY threads you seem happy to spam the board on this same topic with, the answer to your question was clearly demonstrated and explained on multiple websites dedicated to vim and were found with a three word google search. – Jake Sellers Aug 2 '13 at 19:56
  • 6
    My take is (and has always been) that we are intended to be the resource people use from Google. If that means replacing hard to follow documentation with easy to follow examples, so be it. I do get a little frustrated when we try to point to the documentation instead of making the documentation easier to follow. – George Stocker Aug 2 '13 at 19:58
  • @JakeSellers I'm sorry if I've been spamming on this topic. I did open one other thread on a related (although different) topic. Is this not a separate question? Or are there other posts I made that I am not aware of? – Cory Klein Aug 2 '13 at 20:01
  • 5
    @GeorgeStocker thats just it, thats not the purpose of SO. Questions in the format of, "I'm clueless and didn't put any effort into learning or trying, how do I do this?" are bad questions. The general accepted format is, "I read this and tried this and can't make it work or don't understand, please help." – Jake Sellers Aug 2 '13 at 20:06
  • 6
    @GeorgeStocker I'm pretty sure it isn't our job to re-document every crappily documented API out there. If you're actually stuck with a specific problem, and the answer incidentally involves elaborating or more clearly articulating something from the docs, that's fine. But if a canonical resource on something already exists, I personally can see no reason to have contrived supplementary posts on SO retreading that ground just for the heck of directing traffic here. – user200500 Aug 2 '13 at 20:22
  • 1
    @Asad: «I'm pretty sure it isn't our job to re-document every crappily documented API out there.» Clarifying or expanding upon crappy documentation, based on experience, in response to specific programming difficulties sounds like exactly the job for SO. "Q: The docs {tell me that|don't really explain} flimflam and I tried it but got the wrong results, wtf?" "A: I've run into this; the docs should say that you need to double-flam <code sample>" That's a solid piece of reference material for the archive that's going to be valuable to lots of people. – jscs Aug 3 '13 at 1:48
  • I don't think that's covered by "easily Googleable", though. – jscs Aug 3 '13 at 1:49
  • @Asad: Yup, I was pretty sure we're in agreement; I'm completely on board with your last sentence. The one I quoted struck me as contradictory to it, though. – jscs Aug 3 '13 at 1:58
  • @Josh I can see why it might read that way; what I meant was that you shouldn't just arbitrarily post self answered questions in a bid to rewrite documentation (regardless how crappy you think it is). Eg. "how do I attach an event handler in jQuery" -> "you attach handlers using so and so methods, see doc" doesn't make for great Stack Overflow content. If you're genuinely having trouble with something and the answer turns out to be couched in the documentation, that's fine, because the fact that you were having trouble already indicates this might be a problem other programmers will encounter. – user200500 Aug 3 '13 at 2:03

"We're building a canonical archive of programming answers."

And that includes answers easily accessible via Google search


For trivial questions easily answered by a canonical resource like the language's manual itself, we are not there to build an obstacle to Googlers, leading them to a (possibly incorrect) SO question instead of the correct resource.

My rule of thumb is: if entering the question title into Google results in an unambiguously correct answer among the first three to five results (it doesn't matter which number you use here - usually, it's on #1 anyway), the question needs to be destroyed, as its existence makes the Internet a slightly worse place.

  • 3
    What if the question is answered in page 24 sub-section 13 of the manual? rtfm is a terribly inaccessible method to a great number of learners. – Cory Klein Aug 2 '13 at 19:46
  • 1
    @Cory can you make a real-world example? – Pekka Aug 2 '13 at 19:46
  • To find the answer to this question, a user would have to already know about :h and they would have to know that selecting text in vim is known as "visual mode". And yet the question is easily Google-able, but none of the Google results are the manual itself. – Cory Klein Aug 2 '13 at 19:48
  • 12
    I find it particularly annoying when the #1 Google result takes me to SO and the top answer is a link to the #2 Google result (or a rehash of it). We don't need to get in the middle of that. – Bill the Lizard Aug 2 '13 at 19:49
  • 6
    I pretty much completely disagree with this. I very often would trust a highly upvoted SO answer that echoes the manual, because it also contains the value of an intelligent person saying, "yes, that's right, no, you don't need to look somewhere else in the manual, and yes, you are thinking of using the right solution to this problem." Manuals don't generally say "wait wait wait, you're thinking about this all wrong." People do. – djechlin Aug 2 '13 at 19:50
  • 1
    @Cory I did say if entering the question title into Google results in an unambiguously correct answer which in your example may not the case. I'd let that one stand on the grounds that it isn't answered that well anywhere else. – Pekka Aug 2 '13 at 19:52
  • 9
    @djechlin I see your point,but it is this kind of questions that are turning SO into a pointless rep factory where people ask trivial questions like How do I check if a PHP variable is set? and ten people answer isset() and earn points for it - while a Google search for the same term already points to the right manual page. I don't disagree with the idea that getting an answer from a human being well-versed in the technology is great, but it too often doesn't work that way in practice. – Pekka Aug 2 '13 at 19:53
  • 4
    @djechlin as said, I think these kinds of questions have taken the fun out of SO by drowning anything else, and usually attracting garbage answers by the dozen. It's the reason why I hardly answer questions any more. When a question is "how do I do xyz?" with no context, there is rarely room for subtlety. What we do by answering that stuff is teach people to come to SO with every problem they may encounter instead of doing some research first - which doesn't help them develop into actual, Internet-literate developers. – Pekka Aug 2 '13 at 19:57
  • 1
    @djechlin In addition to Pekka's point, the mere existence of an SO answer is not sufficient to ensure that it is correct. Sure, a highly-voted answer carries more weight than one with a net score of 2, but does an incoming Google user understand our voting process enough to recognize the difference? In the case of low-traffic questions with marginally-correct answers, we may be doing more harm than good, but they will rise to the top of the Google results just the same. – George Cummins Aug 2 '13 at 19:58
  • 4
    So people should regurgitate the man pages for people that can't use a man page? Are you serious? – Jake Sellers Aug 2 '13 at 19:59
  • 1
    @JakeSellers it's weird to ask me if I'm serious about something I didn't say. – djechlin Aug 2 '13 at 20:00
  • 1
    @Pekka웃 right... a bad question shouldn't be asked anyway, but it's not bad because it's easily googleable. – djechlin Aug 2 '13 at 20:08
  • 1
    @djechlin in my experience, most easily Googleable questions are bad ones. – Pekka Aug 2 '13 at 20:12
  • 3
    @CoryKlein If the top answer to a question is a link to the official documentation, that usually indicates that both the question and the answer are bad. Note that I'm only talking about the worst case where the answer is literally a link to or a copy/paste from a readily available source. (I've left my share of "Why didn't you just read the first paragraph of the Wikipedia article?"-style answers in the past, so I'm as guilty of this as anyone.) – Bill the Lizard Aug 2 '13 at 20:13
  • 1
    @JakeSellers Similarly, programming has a rather steep learning curve, which probably accounts for the creation of StackOverflow, and why so many questions can now be answered with a simple Google search. – Cory Klein Aug 2 '13 at 20:29

Even if it is a basic question, I see no reason why we should discourage the asking of it, unless it already appears in the SO database. The goal is for people to be able to google and get pointed to SO for good advice, which is impossible if the question has never been asked.

On that note, there is a big difference between a basic question and a question that shows that you literally know nothing about the topic.

  • 4
    People started downvoting after @Pekka's response, but I ask you to consider this: "For trivial questions easily answered by a canonical resource like the language's manual itself, we are not there to build an obstacle to Googlers" I'm pretty sure that almost every question that I've answered on SO was the direct result of me reading the documentation. The issue is that a lot of people don't understand the documentation and need it to be interpreted for them. TBH, it just sounds elitist and rude to say that people shouldn't need the documentation explained to them. – Steve P. Aug 2 '13 at 19:59
  • Personally, I often link to the documentation with an explanation. It seems like you're complaining more about the answers than the questions. Additionally, I don't think that multiple answers to the same question is a bad thing, quite the opposite, in fact. Multiple ways of addressing something are important and good. Of course, if sources are exact duplicates, then it's bad, but that's a product of the answerer, not the question. – Steve P. Aug 2 '13 at 20:01
  • 2
    And obviously bad questions are bad questions, but the fact that there are answers readily available online outside of SO shouldn't mean that those questions no longer have a place in SO. That just doesn't make sense to me. If we use that logic, then most questions don't belong here. – Steve P. Aug 2 '13 at 20:11
  • Yeah, that is the point I was trying to make - documentation is quite hard to read. Analogously, Many of my packages/modules are nearly direct wrappers to third party ones, but when my teammates use my modules, they are given the guarantee that I've done the research that "yes, this is how to use it," and I'm a good person to prod about things working funny. – djechlin Aug 2 '13 at 20:12
  • 2
    @SteveP. This is an issue that splits the community down the middle, so there's no sense in getting all teary eyed about downvotes. You have more upvotes than downvotes, so I'm not sure why this upsets you. Also, not every disagreement will be articulated in comments to your answer. I downvoted your answer. Here's what I feel about this issue. I'm not going to paste that comment on every answer I downvote or upvote though. – user200500 Aug 2 '13 at 20:33
  • @Asad Fair enough, sometimes I just get frustrated when I receive a bunch of downvotes without any explanation. I understand that the explanation is: "I don't agree with you," but that's not exactly rich in meaning. The point is to discuss, by simply downvoting it seems more like reddit, not SO. – Steve P. Aug 2 '13 at 20:37
  • The issue is that a lot of people don't understand the documentation and need it to be interpreted for them. that's not the issue here. We are talking about "how to do xyz?" questions whose answers are trivial to Google – Pekka Aug 2 '13 at 20:39
  • @SteveP. no one is debating that you should be able to find good advice on SO, some people, including myself believe asking a question simply because it hasn't already been asked is a bad reason to ask a question, especially when it is a currently well documented subject. – Jake Sellers Aug 2 '13 at 20:41
  • 2
    @Pekka웃 I think the issue is a bit broader, at least in the sense the OP was asking, because there's questions where the the answer can't be reduced entirely to "see this page in the docs? yeah, do that." As far as those questions are concerned, I totally agree with you. But when you're looking at questions where searching the docs thoroughly enough and making the right connections would help you solve your own problem, I still think those questions are fair to ask here. You have to use your judgement, which is why I don't think coming up with a one size fits all consensus here would work. – user200500 Aug 2 '13 at 20:42
  • @Jake Sellers I guess my issue is that just because something is currently well documented doesn't mean that everyone is going to understand it. For example, there are tons of extremely well written books on mathematics. That doesn't mean that everyone who reads them will understand them. Some people need a little bit of handholding, especially beginners/people that don't really have good CS intuition. – Steve P. Aug 2 '13 at 20:52
  • @Asad I agree. There is not a clear cut answer to this question. Nonetheless, I still have an issue with the conclusion of Pekka's answer. Firstly, "unambiguously correct" again precludes the possibility that people won't understand even after reading, and secondly, I think it's a good thing to have multiple sources of information. Again, if all an answer does is point to a resource, then it's a garbage answer...It's not that I think that it's always yes, as can be seen in the bottom of my answer. Whereas Pekka seems to suggest (and most agree) that it's never ok. – Steve P. Aug 2 '13 at 21:03

@SteveP. Response too long for comment and I'll be better able to write clearly here:

I realize this may be splitting hairs a bit, but I'm not trying to say "noob" questions are bad.

To use your example, there are lots of math books, but even upper level questions, perhaps,

How do I calculate a three dimensional arc length?

would still be a bad question. Is it difficult, high level math? Yes (IMO), but that doesn't mean its a good question. Maybe your text book sucks, maybe your professor sucks, maybe you're bad at math, it doesn't matter, that's not the point.

The mission statement of SO is to ask and answer very focused questions on specific subjects or problems, as I understand it. A better question would look something like,

I'm working on this arc length problem: [arc length problem here] and I keep getting it wrong. I've read this and I'm following this example here but I'm clearly doing something wrong, help much appreciated.

Now, that's not to say that if a question doesn't match that format, its bad. Questions such as this and this are maybe borderline questions (the first is actually locked) that ended up creating incredibly helpful threads.

BUT, they still distinguish themselves, in my opinion, in the fact that what they are asking is not easily googleable. For example, searching google for a question similar to the second one is liable to confuse you. When should I use java instead of Ruby for a web app? According to Oracle its probably every single time! Not very helpful, I'd much rather have an expert answer on SO.

So its a blurry line I end up drawing, but its STILL a line, and "how do I select text" falls under that line in my opinion.

  • 1
    I completely agree with you. Tbh, it drives me crazy when people ask questions like that. I've been a teaching assistant at a top university for the past five years. Despite being surrounded by very intelligent people, you'd be surprised how many of them desperately need handholding, at least in the beginning (mostly people coming from a non-engineering background). I guess I just feel bad for beginners who just want an answer. Yes, they need to learn how to research things for themselves, but that can be pretty intimidating if you're a complete neophyte. – Steve P. Aug 2 '13 at 21:17
  • 5
    @Jake Sellers - so say you don't know how to calculate a three dimensional arc length, and then you figure it out with the help of your professor, text-book, or Google. At that point, you want to share your knowledge on SO with a self posted Q&A. Is it wrong to phrase your question as "How do I calculate a three dimensional arc length?". To what point do you need to feign like you're asking a question that you're stumped with, when you actually know the answer? – Cory Klein Aug 2 '13 at 21:49
  • @CoryKlein Perhaps this is the real crux of the discussion, in my opinion, that is just not at all what SO is for. You should not be asking questions simply to add data to SO like its a big data base. – Jake Sellers Aug 2 '13 at 22:19
  • @JakeSellers What is your opinion on It's OK to Ask and Answer Your Own Questions? – Cory Klein Aug 2 '13 at 22:21
  • 2
    @CoryKlein answering your own question is great, asking a question you already know the answer to just to add info to the site is completely different. You don't need to post every nugget if knowledge you possess, just be happy you learned something. In the future if you see someone else has a similar problem you have an an excellent answer for them. If everyone just started posting every possible question they could think of and knew the answer to in an attempt to preemptively answer everything, the site would go the hell real quick. – Jake Sellers Aug 2 '13 at 22:25
  • 2
    @JakeSellers The link above specifically says "if you have a question that you already know the answer to... and you'd like to document it in public". Also, when you click "Ask Question", there is a checkbox at the bottom saying "Answer your own question", with a link to the above article. – Cory Klein Aug 2 '13 at 22:29
  • 2
    Yes and if you read the rest and the examples I think we can agree that basic text editor usage doesn't fall under that category. "No good info or buried under bad info" doesn't not apply to your question. If you spend days or weeks banging your head against the wall on a hard to solve problem or obscure bug, them by all means help us out. – Jake Sellers Aug 2 '13 at 22:32
  • @JakeSellers Ah, I see where you are coming from! That reading of the article makes sense, and I wasn't able to see it from that viewpoint before. Thanks for pointing that out, and thank you for humoring me in all this drawn out discussion. You may not believe it, but I really do believe in SO, and want to contribute. :) – Cory Klein Aug 2 '13 at 22:38
  • @CoryKlein Yes, but the article fails to qualify that advice with the parenthetical "(provided it isn't already sufficiently documented in public)". I read that as being implicit, but if the advice is literally to dump the contents of your brain onto SO then I can't say it makes much sense. – user200500 Aug 2 '13 at 22:39
  • @Asad I agree with that reading now. This Meta discussion has been really helpful for me to understand the purpose of SO better. – Cory Klein Aug 2 '13 at 22:49

Yes, if the question is not already in SO, and provided you have done your research. If the question is not on SO and you found it by googling, then please self-answer.

  • I will always trust a well-upvoted SO answer over even official documentation. The reason is that the answerer is not just repeating the doc, but also asserting that I'm using the right technology, thinking about the problem the right way, and not missing some tricky subtlety elsewhere in the doc. Manuals can't say "Wait, you should be using this method." Answerers can, and many upvotes mean there likely isn't some catch.
  • "Official" manuals are not optimized for Q&A. It still takes me 30 seconds to parse through and find the right answer for many of them. Even in the Javadoc, which I generally regard highly, I often have to read general class and interface documentation before the documentation on the method makes sense. This is often several minutes of effort when my question could have been answered in several seconds.
  • SO has a uniform format. cplusplus.com, Javadoc, php.net are already, dev.mysql.com already constitute four totally different repos I have to know how to read just to get info. I'm largely a Java and Node programmer but once in a while need to do something quick and dirty in, say, Ruby, in which I'm a total noob. And in those situations I already know how to use SO, but not ruby's documentation, so I will prefer answers on SO.

That being said, bad questions shouldn't be asked. It's just not being easily Google-able that makes them bad. "How do I concatenate two strings in Java?" (with no other info) would be a bad question that should be closed as off topic (or duplicate, but that aside). The user needs to provide context so the benefits I described are possible, and so that future users can realize they had a similar problem.

You must log in to answer this question.

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