I think this is a common scenario on every Stack Exchange:

  • User 1 asks question A
  • Question A gets answers X, Y, and Z
  • User 2 asks identical question B, but adds arbitrary condition excluding answers X, Y, and Z
  • User 2 asserts that question B is therefore not a duplicate, and may even reference question A.

A typical Stack Overflow example would be a question with two answers, one suggesting to use a specific framework, and the other providing a template of an equivalently complex hand-rolled solution. There is inevitably going to be a follow-up duplicate claiming that (a) the framework isn't an option and (b) the custom solution is too complex.

A recent example (today, in fact) on Seasoned Advice was a question about how to store something; the original question got plenty of answers (fridge, freezer, garden, vase, even one answer linking to a list of options). The new question is exactly the same thing, simply placing arbitrary restrictions ("don't want to use the fridge.")

So my questions to you are:

  • Is this type of question actually considered to be a form of duplication?
  • Should they be closed?
    • If so, is Too Localized a better reason than Exact Duplicate?
    • Should we show sympathy by leaving them for a little while, or close ASAP to avoid leaving "broken windows"?
  • How should we respond to the people who post the questions, especially if they object to closing?
    • ...as a normal user?
    • ...as a moderator?

Since I got criticized for not linking to examples (even though I think everybody should know what I'm talking about), here are a few:

And the one that inspired me to ask about this:

Basically any time I see the phrase "I saw this [question/thread] but...", it's obvious what's coming next 9 times out of 10.

  • What happens when the original question didn't get an answer that actually answered the full question, but marked a non-answer as accepted anyway (and the original question is at least a year old)?
    – Troyen
    Jun 21, 2011 at 0:21
  • 2
    @Troyen: That's an edge case I'm not really too concerned about at the moment. I'm talking about non-localized questions that received reasonable answers on the first take. Even under the conditions you mention, I'm not convinced that it's justified; the answers people gave were the answers people gave, you can either help by giving the question a (legitimately-obtained) bump or adding your own answer, or accept that you're unlikely to get the answers you're looking for.
    – Aarobot
    Jun 21, 2011 at 0:23

3 Answers 3


The burden of proof is on the asker - by default, you should assume that it is a duplicate, and he's either too lazy to read the answers, or isn't clear on how to implement the solutions they propose but can't figure out how to ask a question that would provide him with the information necessary to do so...

...Either way, asking a duplicate question with some arbitrary restriction tacked onto the end isn't the way to go: other readers won't be able to distinguish between these in search results, and will likely find themselves wading through an ocean of overly-specific questions with inapplicable answers when what they really need is a single, comprehensive answer:

Help us build a great library of canonical answers. If you keep seeing the same form of questions, whether it’s mod_rewrite rules on Server Fault, freezing computers on Super User, or how to use regular expressions to parse HTML, write a great, canonical answer, once and for all. Make it community wiki so that as many other people as possible can make it great. Work really hard on writing something that is clear, concise, and understandable by as wide an audience as possible.

By the time a duplicate question gets asked, it's too late: the stage is already set for a massive, time-wasting dance with its author wherein he struggles to clarify what he really wants, why the existing question doesn't provide it, and what needs to be asked to lead him to his eventual solution. Yes, you can and should close these as duplicates... but if this becomes a persistent problem, then it's worth considering that the original question's answers are not all they could be... That can be a good argument in favor of

So with that said,

Is this type of question actually considered to be a form of duplication?

Yes, unless the asker puts in the effort to effectively demonstrate that it isn't. Of course, if they do this up-front you don't even have to think about any of this, so...

Should they be closed?


If so, is Too Localized a better reason than Exact Duplicate?

Doubtful, but not unthinkable. I recall a scenario from a while back on SO, where users were prevented from using library routines and responded by posting questions asking for solutions that didn't involve referencing the library. Of course, the solutions tended toward posting library code, turning SO into a really bad mirror of the canonical source, all to work around someone's particular bureaucratic nonsense.

So yeah, possible. But when in doubt... Close as a dup.

Should we show sympathy by leaving them for a little while, or close ASAP to avoid leaving "broken windows"?

I don't ever see the point to leaving something open "just in case". Once you decide it should be closed, close it. You can always re-open if someone manages a convincing argument.

How should we respond to the people who post the questions, especially if they object to closing? ...as a normal user?

Vote to close and provide a link to the answer that answers their question. If they respond with a compelling argument, say so, and edit the question to make the distinction clear, and, if closed, vote to re-open.

If the author responds with confusion, consider editing the answers on the existing question (or providing a new answer) to clarify things.

If they respond with indignation, suggest a nice cup of tea.

...as a moderator?

Like normal user, but without the pesky "voting" bits.

  • I was hoping for something a little more... diplomatic. But I'll think about this and see what I can come up with.
    – Aarobot
    Jun 21, 2011 at 2:57
  • Amen and hallelujah, especially for your second paragraph!
    – jscs
    Jun 21, 2011 at 3:04
  • 2
    @Aarobot: you can be as diplomatic as you like about it, but... If the answer is there and they couldn't find / understand / implement it, the solution is a better answer not another question.
    – Shog9
    Jun 21, 2011 at 3:11

I think the important criteria would be whether or not they've explained why X,Y and Z don't work for them in a reasonable way that allows other to discern what could work for the asker. If they just dismiss the existing answers out of hand, then potential answerers are left stabbing in the dark. In this case it could be closed as a dupe, or as "not a real question" since you're unable to answer it reasonably when all the potential answers have been arbitrarily rejected. Closing as a dupe is probably preferable since it will point others to the question that has the good answers. It's not an exact dupe since they've excluded all the existing answers, but since no one else can meaningly differentiate the questions I think it's ok.

I might also look at whether the user has a history of posting these sorts of near-dupes. If not, I'd be inclined to give them some slack, but if they're doing this frequently then they're just a help vampire and it's not going anywhere.

  • Good point on the history, although very often these are new-ish users who have almost no history to speak of.
    – Aarobot
    Jun 21, 2011 at 0:54

I would vote to close these as duplicates of the referenced question unless the OP has provided some substantial addition to the question which allows for different answers. It is rare that they provide anything substantial, but it happens.

In the case that they are merely conjecturing I would close it as "Not Constructive" instead and direct the individual to chat referencing the old question.

You must log in to answer this question.

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