In an effort to more fully understand the gestalt of Stack Overflow, I offer this question for perusal:


Since the question has since been deleted, here it is quoted inline:

Title: i want a dialog box in case of error

I am developing an application for iphone/android and i am using webapp-net. I have some fields in my application, what I want to have is this: when input is entered and its incorrect(so the app will realizes that its not a correct input immediately) I want a popup window to show up with a message that indicates the input needs to be changed. I dont want a new window or anything like, a simple dialog box with an error message will do the job. and when the user touches the textfield area to make a new input I want that dialog box to be gone. So far I couldnt find a good example to do this. If you need any more information i would gladly give it to you.

Now, it seems pretty clear to me what the OP wants: they want to know how to perform form validation, and don't have the necessary background to figure out how to approach it.

There are a number of ways to respond to this question; my own comment (which BTW is a re-rendering of someone's misguided and obfuscated LMGTFY link, which I deleted) is just one way; it links to a Google search that will help him research some strategies. A simple code example posted as an answer is another way.

But the usual response to these kinds of questions is "Show us what you've done so far," and "Show us some code," which is how the other commenter responded. This turns the question from a request for guidance into a troubleshooting problem ("What's wrong with my code?").

Here's the thing, though: as a software professional, I find this sort of code troubleshooting completely uninteresting. It is the very definition of too localized: This question is unlikely to ever help any future visitors; it is only relevant to a small geographic area, a specific moment in time, or an extraordinarily narrow situation that is not generally applicable to the worldwide audience of the internet.

And it's not what the OP is asking.

With that, here are my questions:

  1. Is this question off-topic, and if so, why?
  2. Are these kinds of "beginner questions" welcome here, and if not, why not? Are we justified in closing such questions? Which close reason do we use?
  3. Is this question "Very Low Quality?" (it was moderator-flagged as such) Why?
  4. Does compelling people to show code turn us into a "coding support site," and if not, why not?
  • 4
    It seems that he thinks SO is a free flowing discussion board from his line, "[if] you need any more information i would gladly give it to you.".
    – user7116
    Sep 22, 2011 at 19:07
  • @sixlettervariables: Perhaps. But that's what comments and question edits are for: clarification of the question.
    – user102937
    Sep 22, 2011 at 19:09
  • whats wrong with LMGTFY, if it answers the questions? Sep 22, 2011 at 21:13
  • 6
    @VJo: Everything. Sep 22, 2011 at 21:20
  • 3
    If questions like this are allowed, I'd like to know how to ignore them. Sep 23, 2011 at 3:54
  • 1
    Here's a similar question/situation I ran across today: stackoverflow.com/questions/7519946/…
    – jscs
    Sep 23, 2011 at 6:51
  • @AndrewGrimm That's a great feature idea. It would be great to be able to set a filter to show only questions with a score above a certain threshold. e.g. Show me questions with a score of at least 2.
    – Mike Hay
    Sep 23, 2011 at 20:36
  • 2
    @Mike: How are you going to vote a question up or down if you can't see it in the first place? Leaving others to vet your questions for you is just laziness.
    – user102937
    Sep 23, 2011 at 20:40
  • @RobertHarvey Isn't laziness the driver behind all great inventions? ;)
    – Mike Hay
    Sep 23, 2011 at 20:43
  • @Mike: Heh, there's already a search option for that anyway: stackoverflow.com/search
    – user102937
    Sep 23, 2011 at 20:44
  • 2
  • 1
    Is it really better to provide the link instead of the LMGTFY? It seems it would have been nicer and beneficial to say "I think you are looking for jquery validation - is that correct?". I dunno, maybe it is just me but the LMGTFY and links to google search in general seem condescending. Sep 27, 2011 at 18:31

5 Answers 5


Based on the OP's last sentence, "[if] you need any more information i would gladly give it to you," I don't think they understand how SO works. Based on their prior question, it seems obvious they are looking for code samples/examples. Regardless,

  1. Is this question off-topic?

    No, it is certainly about programming and is not some code golf nonsense.

  2. Are these kinds of "beginner questions" welcome here?

    Certainly, albeit with some stipulations I've given below.

  3. Is this question "Very Low Quality"?

    No it was not worthy of the flag; it was salvageable with an edit (nice work Mike Hay). However, they do not seem to enjoy capitalizing "i" very often, nor did they have a well formatted question.

  4. Does compelling people to show code turn us into a "coding support site"?

    To be honest, I thought that's the definition of StackOverflow. Programmers working on fixing actual problems other programmers face. Some questions are going to be way more monotonous than others. Not everybody has neat, awesome problems, myself included.

    I vote to close almost every question that doesn't meet the criteria, "practical, answerable questions based on actual problems that you face."

As for beginner questions, there is something to be said for providing some rambling problematic code or at least a mock-up of what you'd like to happen. Any effort is better than no effort. I think this question in particular is a run of the mill "show me the code" question which isn't likely going to benefit the community in the long run. I would vote to close this question as Too Localized.

Can I give a concrete answer as to how much of something related to code I need to see to hold off on a vote to close? Not really, but anything is better than nothing.

  • Thanks. "Too localized" is debatable. Data validation is of interest to almost every programmer.
    – user102937
    Sep 22, 2011 at 19:34
  • I agree, but the question could be made more generic/useful to escape the realm of Too Localized. With Mike and Jim's help it likely won't need to be closed.
    – user7116
    Sep 22, 2011 at 19:37
  • 3
    Data validation is of interest to almost every programmer. Validating this specific piece of data is not. Sep 27, 2011 at 17:30

Are code troubleshooting questions really what we want to encourage?

Yes, that's a pretty big part of Stack Overflow. We should keep answering those questions.

But the question you link to as example is completely off the grid. It is neither a troubleshooting question nor is it about code nor is it about an actual problem. The OP is asking on how to achieve something he should do on his own, or pay a programmer to do it for him. He's outlining behavior and now wants us to tell him how to do it. These questions have always been downvoted and closed. I fail to see the difference between that question and give me the codez.

As to your comment, if I remember right linking against search machines was discouraged for a good reason. While I understand your motivation, I think we should stick to that rule. We should not encourage such behavior by even throwing the smallest breadcrumbs at the people. Additionally, we can't draw the line between Google-Questions and and those who are not deserving such a smelly answer.

Edit: To explain this 'hostile' approach, I'd like to quote two things:

The Downvote-Tooltip:

This question does not show any research effort; it is unclear or not useful.

The Not-a-real-question close reason description:

It's difficult to tell what is being asked here. This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form.

Both fit on this question:

  • It lacks any research
  • It lacks crucial details (what is invalid text? What method is used to identify invalid text?)
  • It is most likely open ended, with additional trouble shooting and extending of the given example ("Can you please make the box move more to the right?")
  • 1
    Seems like an awfully hostile approach.
    – user102937
    Sep 22, 2011 at 20:00
  • @RobertHarvey: May be...but I'd like to point at the Downvote-Tooltip and the Not a real question close-reason description. Both fit very good on this question in my opinion. You can of course ask the user for clarification, or write him a complete example on how to achieve that. But as you already asked, should we encourage that? If you meet a question which should not be encouraged, downvote and close it. At least that's the approach I'm using, can very well be that I'm wrong. Sep 22, 2011 at 20:19
  • 1
    Regarding the "Not enough research effort" issue, it's impossible to perform a Google Search if you don't even know what you are looking for (The OP didn't use the word "validation" in his question). The real issue is whether or not we want beginner questions. People come up with all sorts of rationalizations for closing these questions, but it essentially amounts to "I don't like this question, because it is too basic."
    – user102937
    Sep 22, 2011 at 22:03
  • 1
    @RobertHarvey: Note: General Reference Close Reason. Anyway, if he can't even word his question (missing the possibility of dictionaries and translators, if his mother language is not English), do you really want to throw a block of explaining text and code at him, and hope that it will help him? Also, I'm able to get results from entering "javascript check if textbox is correct", there are even examples using the word "validation" popping up in between and showing inlined error messages. Sep 23, 2011 at 6:44
  • It's been my experience, in real life, that usually, people who cannot even get started on a task take a very long time to help. That said, he was able to at least state his goal. It may take a lot of guidance, several comments and requests for clarification, and it make take a long time for this guy to get an answer, but ... why not leave it open for those who have the patience to help guys like this?
    – Mike Hay
    Sep 23, 2011 at 20:24
  • The value of a poor question like this is that the next guy who comes along and uses the same (misguided) starting point as the OP can find this (if he searches first, debatable) and benefit from any answers or help that were offered.
    – Mike Hay
    Sep 23, 2011 at 20:29
  • @Mike: My usual response to such people is to guide the OP to a tutorial or book, to point them in the right direction. The problem is not so much closing these kinds of questions; I really don't have a problem with that. What I take issue with is asking the OP to "show their work" when it is apparent that there is none to show. I believe asking such things can obfuscate the actual problem, and steer questions into this highly-localized troubleshooting mode. But I don't think I've made that argument adequately here, judging from the answers posted so far.
    – user102937
    Sep 23, 2011 at 20:50
  • In short, I wish the community would make more effort to adjust their answers to the OP's level of understanding, rather than making excuses to get rid of a question because it is "too simple."
    – user102937
    Sep 23, 2011 at 20:54
  • @RobertHarvey: Yes, I understood your question completely different to what you said now. The FAQ talks about a specific programming problem. So maybe we should define if those questions are a specific programming problem (in my eyes they are not). Sep 24, 2011 at 18:02
  • "Yes, that's a pretty big part of Stack Overflow." It shouldn't be. Sep 27, 2011 at 17:31
  • @TomalakGeret'kal: Of course not those! But there are legitimate versions out there. Sep 27, 2011 at 17:57
  • @PaddedCell: "low quality" and "too localised" are orthogonal. Admittedly that meta Q&A seems to be about both at the same time by now, but really it's supposed to talk just about "too localised". You can have a well-formatted question that's a request for troubleshooting and that is still what I am "fighting against". There are plenty of forums/message boards/IRC channels for that kind of thing. Sep 27, 2011 at 23:37

In this case, I don't think that code would improve the question.

Code is useful for "I am having this specific error in a specific portion of my code". It's not useful for "I have no idea what I'm doing."

In general, the solution to "I don't know how to program" or "I don't know the basics of web development like field validation" is to go experiment and read tutorials and whatever else. Telling someone to go somewhere else and learn isn't a good answer, and if it's the only possible answer then that makes it a bad question. SO isn't built to host tutorials.

I don't think catering to questions this basic makes for a website that experts want to participate in. I might close as Not A Real Question but at minimum the downvote tooltip says "This question does not show any research effort; it is unclear or not useful", and it's only not unclear.

Oh, and I don't think the other commenter was looking to troubleshoot. It seems clear to me that he is looking for context in order to provide better guidance tailored to how the OP was creating his web form.

  • Thanks. Note that "Not a Real Question" doesn't really fit: "This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form." The question is none of those things.
    – user102937
    Sep 22, 2011 at 19:31
  • @RobertHarvey Yes. General Reference would be closer, but on sites without it NARQ is generally used (from what I've seen). It could also be Not Constructive, since it's not, although the description doesn't necessarily fit either.
    – user154510
    Sep 22, 2011 at 19:35
  • Too localized doesn't fit either, this is a common requirement. The ideal close reason is exact duplicate. Sep 22, 2011 at 20:04

Complete text of question.

Title: i want a dialog box in case of error

I am developing an application for iphone/android and i am using webapp-net. I have some fields in my application, what I want to have is this: when input is entered and its incorrect(so the app will realizes that its not a correct input immediately) I want a popup window to show up with a message that indicates the input needs to be changed. I dont want a new window or anything like, a simple dialog box with an error message will do the job. and when the user touches the textfield area to make a new input I want that dialog box to be gone. So far I couldnt find a good example to do this. If you need any more information i would gladly give it to you.

I don't think this is a good question and I don't think it belongs here. Even with edits -- which are a waste in this case.

Primarily, it fails the what have you tried to make this work? test. Forget showing code, how about showing us some effort! Like it says in the sidebar of the /ask page:

Provide details. Share your research.

I do not accept questions that are brief meta-descriptions of what the author wants to accomplish without demonstrating any actual attempt to solve their own problem or sharing the research they did on the topic. These are just "gimme teh codez" in a different colored wrapper.

  • 3
    You didn't really address the root of my question, which is that, by insisting on showing code, are people turning the site into a "What's wrong with my codez" support forum? ZZZzzzz....
    – user102937
    Sep 24, 2011 at 6:39
  • 1
    you missed the point of my response entirely; I did NOT say show the code. I said SHOW YOUR RESEARCH AND SHOW SOME EFFORT. I care nothing for code. "So far, I can't find a good example to do this." Really? I don't believe you. Prove it. SHOW. SOME. EFFORT. Sep 24, 2011 at 8:20
  • 2
    No, I got what you said. But the OP was asked to show some code, not demonstrate their research effort. Bah, I should have picked a better example question; everyone's getting distracted because this one is too simple. WHERE'S MY GENERAL REFERENCE CLOSE REASON?!
    – user102937
    Sep 24, 2011 at 14:52
  • 8
    I don't care if other people requested code; what's necessary here is proof that the user did in fact research this problem, because honestly, I think he's full of crap. If you can't show your research, it's because you suck and didn't do any. Sep 24, 2011 at 21:19
  • @Jeff: Finally, we agree on something ;-) Sep 27, 2011 at 17:31

Code troubleshooting questions are a good fit for Stack Overflow, within reason. “Within reason” means that:

  • The code must not be too long. There's a post length limit, but if you come anywhere close to it, your question has far too much code. You (the asker) must make the effort of shrinking your code to a small example that exhibits the problem. This effort will make you understand the problem better (and sometimes will even let you solve it by yourself), and it is necessary if you expect people to answer.
  • The posted code must be complete. You'll get far better answers if people can try the code on their own machine and look at their compiler's warning messages or spend a couple of minutes with a debugger (but don't expect that they'll always do that!). Try to minimize dependencies (e.g. don't include libraries that aren't relevant to the problem).
  • You must post an explanation of the desired behavior of the program, and the behavior you observe, and it must be clear how the behaviors differ. Posting the code with no explanation may get your question automatically rejected, and even if you pass through the filter, a question that does not have a clear explanation of both the desired behavior and the problematic actual behavior is likely to be closed as “not a real question”.

In moderate quantities, “why doesn't this code work” can be a fun puzzle. In large quantities, “why doesn't this code work” is debugging hell.

But not all questions are about code troubleshooting. At the other end of the spectrum, there are times where you don't even know where to start. You have a high-level description of your data, now you're looking for a suitable data structure. You know what you want to do with your data, now you need an algorithm. You're entering a new subfield of programming and you need to learn the techniques that are specific to this subfield. You're confronted with a new, complex piece of hardware or software and you're looking for a good way to become familiar with it. All these questions are about programming, and (provided they meet the general Stack Exchange guidelines, such as asking “practical, answerable questions”, avoiding extended discussion and so on) they are perfectly suitable for Stack Overflow.

The question that sparked this is clearly a beginner's question. This doesn't make it any less legitimate. Stack Overflow isn't only for those programmers who've already seen it all and done it all; the Stack Overflow FAQ has nothing to say about the level of expertise required to participate on the site (unlike some other Stack Exchange site, which are not the topic at hand here).

Just because a question is basic doesn't mean it's simple. There have been two main benchmarks proposed to define questions that are too simple: questions whose answer is easily found by a web search (Google questions), or questions whose answer is easily found in a common reference (Wikipedia questions). (See the discussions regarding a general reference close reason.) This question is not too simple by either definition. It's about input validation — but the asker didn't know the word, which made it difficult to search for good references. Even with the word, the now obvious Wikipedia article is not much help: it's very abstract, there's nothing in here that helps you validate form data in Javascript.

Sure, if you're a seasoned web programmer, this is elementary to you. Well, it's not elementary to everyone. Imagine you've just hired a junior developer, or perhaps an intern. You will have to guide him through many things that seem easy to you. This question is one of those things. And it's not so easy to write a good answer, is it? Not everyone is a good teacher. Yet teaching programming is part of the job of a programmer, and very much has its place on a questions and answers site about programming.


  1. This question is perfectly on-topic.
  2. This kind of beginner questions is welcome here. There is no justification to close them. And I should add: comments to the contrary tend to come out as rude or non-constructive. Flag away (or in your case, Robert, delete away).
  3. This question is not “very low quality”. The English was sub-par but comprehensible; that's a reason to edit, not to delete (which is what “very low quality” is supposed to entail).
  4. Compelling people to show code doesn't strictly restrict the site to a “coding support site”, but it would be an arbitrary, harmful restriction on a site about programming. (If you want every question to have code, try Code Review.)