341

As of the 9th of March, the banner shown to askers whose questions have attracted at least one duplicate close vote has a couple of new options:

That solved my problem! / No, my question is different. I will edit to explain how

If the author clicks the first button, they're shown a confirmation that clarifies the results of this action somewhat:

This will mark your question as a duplicate, directing future readers to the original question and preventing further answers from being posted here.

If they click "Ok", the question is instantly closed:

Community: This question's author approved a pending duplicate vote.

If, instead, they edit, they'll be offered this guidance:

Your question has been identified as a possible duplicate of another question. If the answers there do not address your problem, please edit to explain in detail the parts of your question that are unique.

...and once an edit is submitted the choice will go away until another duplicate close-vote is cast.

In just the past day, 80 questions network-wide have been closed in response to confirmation from the asker; edits by askers are up significantly as well.

We'll be monitoring the results to see if further adjustments are needed. A few open questions surrounding this change include:

I'm posting this separately from the original feature-request to allow room for discussion and work around a bug with featured posts.

  • 70
    Is there any reason why the banner says "That solved my problem!" instead of "Yes, my question is a duplicate." Wouldn't the latter be much more straightforward? – Rainbolt Mar 10 '15 at 20:08
  • 12
    Should the box show something different if the target question doesn't have an answer? The user might get their hopes dashed if someone flags their question as a duplicate of an old and unanswered question. – Joe W Mar 10 '15 at 20:11
  • 26
    @JoeW The target must have an upvoted or accepted answer in order for anyone to flag/close as a duplicate. – user259867 Mar 10 '15 at 20:12
  • 3
    @Woodface Then my concern still stands, just because it is meta does not mean the alert saying there is an answer when there is not and the target is an older question with minimal activity and views can be disappointing. – Joe W Mar 10 '15 at 20:18
  • 12
    If there are no answers, then I would think not clicking that button would be pretty obvious, @joe – Shog9 Mar 10 '15 at 20:34
  • 51
    If you don't know what "duplicate" means in this context, then that question is going to be hard for you to answer correctly, @rainbolt – Shog9 Mar 10 '15 at 20:36
  • 9
    It seems that 2/3rds of the time there is a close-as-duplicate vote, the person that made that vote is completely wrong. It's not clear to most folks that close votes are for duplicate questions, not questions with similar answers. – Brad Mar 10 '15 at 20:44
  • 9
    @Brad A question should be closed as a duplicate when the answers to the proposed duplicate answer the question at hand. The text (and general functionality) here is completely in line with that goal. Questions that are different, but to which the differences are not material to the answer, are duplicate questions, as far as the site is concerned. Of course, if the question is related, but an answer there fails to answer this question for some reason, then the questions are indeed not duplicates. That isn't at all the case in 2/3's of duplicate closures. – Servy Mar 10 '15 at 20:47
  • 8
    @Servy News to me. Every time this comes up on the StackOverflow Meta, it's been very clear that duplicate questions are for duplicate questions. I'll find some examples later tonight. – Brad Mar 10 '15 at 20:52
  • 13
    As for the first part of the open question ("Should askers be able to dismiss the banner without editing or confirming? Perhaps if the number of votes is below some threshold?"), why not check with User Experience; this seems like their specific forte. Also, why "dup-target"? I realise that's a placeholder, but ugh: can we use actual English? Please? "Dup" hurts. – David Thomas Mar 10 '15 at 20:56
  • 4
    @Servy I've seen a lot of cases where there are two very different questions (usually one of which because the person asking doesn't have a full understanding of the problem) that have the same answer. I don't think we should be voting to close these as duplicates. Leave a comment, sure, but a good answer could focus on different aspects of what's being asked. Suppose I had a question about a computer problem, and you had a question about a computer problem, and the solution to both our problems was just to reboot the computer. Not the best example, but this is what I'm seeing in the wild. – Brad Mar 10 '15 at 21:08
  • 3
    I ran into this a few hours ago on meta SO and thought it really helped save some time. – Travis J Mar 10 '15 at 21:20
  • 3
    I think the button should say "Sorry. I promise to search before asking my next question." (Only half-serious: I wish we could teach users that they'll get an answer sooner if they search rather than waiting for us to point out the duplicate.) – Jeffrey Bosboom Mar 11 '15 at 1:14
  • 16
    When I looked at that "That solves my problem" button, I didn't expect the proposed outcome to be "instant close"; I felt tricked. Why am I not being asked to confirm before closing the question? – oldmud0 Mar 11 '15 at 1:14
  • 8
    Conceptually, asker-dups are abusive, moderators are exception-handlers, and meta is An Abomination Unto The Lord. – Shog9 Mar 12 '15 at 0:20

12 Answers 12

158

The implementation of this idea is wonderful. However I have two suggestions.

  1. As OP approves the dupe vote, why attribute it to the Community User? It could be attributed to OP and his/her name should be placed in the blue rectangle, the same rectangle in which his/her name appears when (s)he comments. The tooltip should remain the same.

  2. With the implementation of this idea, it is possible for OP to close his/her question solely (not sure if it is a bug or by design), if he/she has 15 reps to flag his/her question for closure. I accounted this by voting to close my question and then approving my own vote with "That solved my problem".

    In the case if it is really by design, it is absolutely unnecessary to ask for confirmation by showing the window which allows me to choose to close my question or not. If I already voted, then I am sure I want to close it. So it can be closed immediately, once OP has voted to close the question as dupe, so as OP doesn't need to approve his/her own vote by choosing "That solved my problem".

(In case if it is bug, the ability to approve own vote should be possibly removed)

In the closed state, the banner would look like this:

Also there is a bug - this window maybe not really relevant when the question has an open bounty. The suggestion by Jon Ericson to disable this button if there is an open bounty is quite good.

  • 7
    No and no. 1 - users without close privilege can't really vote to close, so it would be misleading, confusing and wrong. The OP is not really voting to close, it's done by the new process. 2 - wrong, close vote cast by OP is not binding, it's just a single close vote. – Shadow Wizard Mar 10 '15 at 14:01
  • 2
    @Sha OK about the first one, but think about the second one. I can vote to close my question as dupe and then click "that solved my problem". E.g.: meta.apple.stackexchange.com/questions/2324/… – nicael Mar 10 '15 at 17:23
  • 8
    @ShadowWizard askers (having privilege to flag/vote close) have unilateral dupehammer on their own questions, I just tested how it works – gnat Mar 10 '15 at 18:00
  • 81
    @ShadowWizard Even users without the close vote privilege should be able to close their own question in a perfect world. – TylerH Mar 10 '15 at 20:26
  • 2
    Yeah, if you realise it is a duplicate why not? And flagging as dupe is better than deleting. – Tim Mar 10 '15 at 20:35
  • 8
    Although the effect is the same in this (one, narrowly-defined) scenario, question owners do not have binding close votes on their own post - I believe such a feature would likely introduce a great deal of confusion unless also coupled with binding reopen votes, which would of course introduce either an opportunity for abuse or a ton of extra rules for voting. The system is closing the question, not the user who confirmed the duplicate. – Shog9 Mar 10 '15 at 22:20
  • 13
    @ShadowWizard With your argument, please stop calling edits that a low-rep user approves on his own post as "approved", since the user obviously can't approve edits. – yo' Mar 10 '15 at 23:46
  • 20
    @Shog9 Well, they do have a completely assymetric binding close vote on their questions now. If you paint stripes on a horse, he doesn't become a zebra. He's just a horse with stripes. – yo' Mar 10 '15 at 23:48
  • 18
    Why do changes not go on ux.SE before going live? I feel like if you have a large group of people with experience in UX, many of whom are paid to do it, who are willing to give advice at no cost, you may as well run it by them. It's not like you hide pending changes anyway, even just putting it on the community bulletin would be a good idea. – Jon Mar 11 '15 at 1:28
  • 2
    This is on the bulletin, @Chipperyman. Right now. That's probably how you got here. I'm soliciting suggestions from the folks who actually use it, because... Their opinions are the ones I care about. – Shog9 Mar 11 '15 at 3:42
  • 9
    @Woodface better answers could still be posted on the older question. That's the best result: you don't want two identical questions, one with a good answer, one with a bad/dangerous answer, because then it's 50:50 which one someone finds when googling a problem. Have one with all the answers, and the other as a signpost to that one. – user568458 Mar 11 '15 at 9:04
  • 10
    @ShadowWizard New users don't have the delete privilege either, but we let them delete their own questions without putting the Community user's name on it. If that doesn't cause any confusion or problems, I don't see how it would a duplicate-close? – SevenSidedDie Mar 11 '15 at 21:36
  • 3
    @Shog9 In all honesty, I think you think I'm suggesting something I'm not, but I'm not sure. Here's the trail: The answer here proposes sticking the name of the "This solved my problem!" clicker in the "marked as duplicate by Alice, Bob, and Clicker" instead of "Alice, Bob, and Community". Shadow Wizard said, no, that would be super-confusing. All I'm saying is that, no, it would not be super-confusing (and would actually be informative). Are we on the same page now? Is that what you're saying no to for UX reasons? – SevenSidedDie Mar 12 '15 at 3:04
  • 2
    Yes, @SevenSidedDie - I don't want to spend the next few years answering "why was user able to vote on this question" questions with "weeeelll... He kinda didn't, but we put his name on it anyway... Uh..." and "take my name off this you bastards I'LL SUE" demands with "weeeelll... You sorta did..." – Shog9 Mar 12 '15 at 3:45
  • 7
    @Shog9 Get ready for these ones in the next few years: "How can Community close questions?" I got confused by this, and I am one of the people who had seen this announcement and who was aware of the new system. As someone said before: In ideal world, OPs would have more power on the closing of their question, like they have on suggested edits. – yo' Mar 19 '15 at 9:42
56

Could you add a link to a help center (or regular meta) page before/after the link for users to edit their post if they feel it isn't a duplicate, add banner at the top of the page while editing (when coming from that link), or something along those lines, to explain in a bit more depth how to edit a question in order to properly distinguish it from a proposed duplicate that is not actually a duplicate.

A lot of people in this situation tend to edit the post to say that the duplicate doesn't answer their question, but often not how or why. Good guidance is rather important at a time like that. Without some guidance on how the question should be edited I fear we'll just see a lot more people making edits that don't actually help readers understand the differences between the question and the proposed duplicate, creating a rather unenjoyable experience for everyone involved.

  • 1
    I imagine if the asker is too lazy to write why they're different, they'd be too lazy to read linked content. A banner or similar would be good, though. – Veedrac Mar 10 '15 at 21:00
  • 27
    @Veedrac I imagine most users aren't too lazy to explain why the posts aren't different, but rather don't realize that that is what is expected of them. Many seem to think that their word that the post isn't a duplicate means that everyone else should just believe them and not close the question. When actually prompted to explain why the proposed duplicate doesn't answer their question, many people are more than willing to elaborate. Obviously some people are beyond help, but the point is that the information should be there for those trying to do the right thing. – Servy Mar 10 '15 at 21:02
  • 2
    Got any specific suggestions for advice that'd be useful here? – Shog9 Mar 10 '15 at 22:34
  • 2
    @Shog9 a banner when editing that says "Please explain what steps in [linked name of the duplicate post] you tried and why they didn't work." – nhinkle Mar 11 '15 at 0:45
  • We might even link to the faq on that topic :-) – Bergi Mar 11 '15 at 0:47
  • 5
    @nhinkle That seems specific to troubleshooting/debugging questions, and would be out of place on Christianity or Space Exploration. – user259867 Mar 11 '15 at 1:04
  • 2
    @Shog9 How about "please explain which parts of the question [linked name of the duplicate post] do not apply to your question, and why they do not apply"? That should work equally well on all sites, even Worldbuilding. – a CVn Mar 11 '15 at 11:54
  • 14
    Actually I'd flip that around, because the criterion for not being a duplicate is really that the new question has a part that the old question doesn't, as opposed to the old question having something that doesn't apply to the new question (as in @MichaelKjörling's suggestion). So perhaps "What do you want to ask that isn't covered in [linked name of duplicate]?" Or "Please explain which parts of your question are not addressed by [linked name of duplicate]." – David Z Mar 11 '15 at 12:16
  • 1
    I like @David's suggestion here; maybe... "Your question has been identified as a possible duplicate of [link] - if the answers there do not address your problem, please edit to explain in detail the parts of your question that are unique." – Shog9 Mar 11 '15 at 21:36
  • See edit - Anna added this guidance for editors, shown whenever there's a pending duplicate suggestion and the author opts to edit. If there are multiple duplicates identified, only the first is linked to from the edit-page banner. – Shog9 Mar 19 '15 at 21:34
29

I can imagine OP clicking the button thinking that it leads to the question listed above it. Or because people like clicking things.

Since the button has an immediate action that is not easily reversed, this action should be explicitly stated. I suggest expanding the text of the first option:

That solved my problem! | mark this question as duplicate

where mark this question as duplicate can be next to the button or under it. Tooltip text would not be enough.

  • 5
    If it solved their question, it should be marked as a duplicate. New users often seem to protest dupe / close comments, so explaining it fully may not help - they won't realise that closing as dupe isn't a bad thing. – Tim Mar 10 '15 at 20:12
  • 16
    Intentionally withholding an explanation of what a button does is not quite in line with above all, be honest. – user259867 Mar 10 '15 at 20:15
  • 10
    There is no point using fancy words like "mark this question as duplicate"... for a start, users often don't know it is a question - they think it is a thread. secondly, they won't want to perform a bad action, and thirdly, they are likely to not want to mark as a dupe - it makes it sound to them like they made a mistake. – Tim Mar 10 '15 at 20:35
  • 13
    +1, and probably I'd like to see the confirmation dialog. – nicael Mar 10 '15 at 20:35
  • It would be good to have the tooltip as well. – NoviceSEMetaGeek Mar 10 '15 at 20:57
  • 1
    @Tim Right on!! – NoviceSEMetaGeek Mar 10 '15 at 20:58
  • 10
    @Tim If they don't want to mark it as a dupe, that's too bad. But in this case, tricking them into doing it with UI design reminds me of Quora's "click here to read the answer ... except you won't". It's not the sort of tactic I associate with Stack Exchange. – user259867 Mar 10 '15 at 21:01
  • 5
    I'm the kind of person that clicks things to see what they do - I'd be right bummed if it got me question closed. – Veedrac Mar 10 '15 at 21:06
  • 4
    If it answered their question it should be closed as a dupe, no matter what a new user who is protective of their question seems to believe. – Tim Mar 10 '15 at 21:11
  • 21
    I've seen 2 cases already of the OP closing their question based solely on a wrong dupe vote; they could not possibly have thought the other post helped them. The confirmation appears to be necessary. – Martijn Pieters Mar 10 '15 at 21:16
  • 9
    The intention here is not to "trick" anyone; rather, the goal is to use terminology that prompts a response which lets us know which action would be in the best interest of the asker and then try to act appropriately as a result. A more explanatory confirmation may be useful here, but let's not get hung up on closing - the actual value here is in determining whether or not the identified duplicate is actually useful, and hopefully prompting clarification in cases where it isn't; closing just saves everyone else time and effort in cases where the identified original is sufficient. – Shog9 Mar 10 '15 at 22:24
  • 5
    I 100% agree with this and think it would be fixed with some sort of undo button. Perhaps something similar to an undelete. If they dup closed their own question, they can reopen it within a 5 min grace period or something – David Grinberg Mar 11 '15 at 3:54
  • @MartijnPieters A confirmation that asks them if they're sure that the other question solves their problem would seem sufficient, then, instead of throwing jargon at them that will make them uncertain. The issue is that honestly answering the question on the button shouldn't require knowledge of how SE works, and if they answer honestly, what should happen is obviously (to us) closing the now-unneeded question. – SevenSidedDie Mar 11 '15 at 18:48
  • 1
    Anna has added an informative confirmation pop-up to the button - see my edit for details. – Shog9 Mar 12 '15 at 1:14
21

This "that solved my problem!" text just further reinforces the misunderstanding that many newbies have that Stack Overflow (and its siblings, cousins) is a support service, or a personal helpdesk. It's not! I think for the button we should stick to something that the OP can recognise as agreeing with the suggestion of duplication, and leave it at that.

Otherwise, I think this is great. Good job!

  • 15
    Weird objection. Regardless of the theoretical implications, folks asking questions based on actual problems that they face are sorta the bread'n'butter of this whole juggernaut. – Shog9 Mar 10 '15 at 22:28
  • 16
    @Shog9: I'm looking deeper. Building a repository of Q&A (so that future readers can benefit from the actual problems that the OP faced) is the bread and butter of this whole juggernaut, and I feel daily like a virulent lack of comprehension therein is leading to a downwards spiral in quality. Thus any little hint that reinforces said misconception would be a bad thing. Or if there is a new policy that Stack Overflow is now for the OP alone, turning it into a personal helpdesk service, then there ought to be some announcement to that effect. – Lightness Races in Orbit Mar 10 '15 at 22:50
  • 9
    I think you're confusing the end and the means. The end is indeed re-usable questions and answers; the means is mostly enlightened self-interest, particularly when it comes to askers. The trick is to find ways of using what folks are already motivated to do in ways that benefit others. – Shog9 Mar 10 '15 at 22:57
  • 9
    @Shog9: I'm not "confusing" them. I'm presenting my opinion that the choice of means have an effect on how effectively we can achieve our end, in the long run. In this case, the text (the means) runs the risk of mistraining site newcomers, who will over time contribute yet more low quality content accordingly, leading, ultimately, to a more literal "end" than you wanted. :P It's a subtle and more complex issue than you are going for here, sure, but I don't think that makes it unimportant (nor worthy of being summarily written off as "confusion"). – Lightness Races in Orbit Mar 10 '15 at 23:03
  • 15
    Lemme try to put this a different way: I strongly doubt that a significant portion of the folks asking questions today or at any point in the past 6 years are doing so because they wish to build a repository of Q&A. As much as I'd love to believe thousands of people decide every day to make up questions they hope will be interesting to others, I've seen very little evidence of this. You're afraid we're going to inadvertently discourage a motivation that doesn't exist. OTOH, if you have suggestions for guidance that'll help folks write up their problems to be reusable, I'm all ears... – Shog9 Mar 10 '15 at 23:19
  • 2
    @Shog9: That's a reasonable counter-argument. I still propose that you run the risk of making the newcomer's eventual realisation of what this site is (and the corresponding leap towards high question quality, reusability and MCVEness in their contributions) come somewhat later than it otherwise might. – Lightness Races in Orbit Mar 10 '15 at 23:55
  • 4
    I would think that people come here to have their questions answered. 99 times out of 100 they don't care about how we run the site, they just want an answer to their question. It's the 1 in 100 that stick around and learn what it means for a question to be 'off-topic' or 'duplicate' - those are the users that would care. I would hazard a guess that everyone on this question is part of that 1 in 100. – Robotnik Mar 11 '15 at 6:16
  • 1
    Think of SE like a museum - we have people coming in and giving us artefacts, but we're the ones who work here, we're the ones deciding what's worth displaying vs. what isn't. If we already have 100 arrowheads, Do we really need to display the one the user just brought in? They may think their arrowhead is special or different, which may warrant further investigation, but most would just say "Oh, ok, nevermind then". – Robotnik Mar 11 '15 at 6:18
  • 1
    In other words, the curatorship can be handled by those that care, whilst the ones who don't can just ask the question they need answered, clicking "This solved my problem" if it, you know... solves their problem. – Robotnik Mar 11 '15 at 6:21
  • 2
    I agree, I think that the site's wording should be encouraging to those who are of the 1/100 ‘I care’ mentality and those who are at least tameable because these people are those of us who stick around and become decent contributors. Yes most of the questions that I ask are questions that I myself seek an answer to, but in doing so, I do try to write my questions (and other contributions) in a way that contributes to a repository of information, and occasionally I do use the ‘Answer your own question – share your knowledge, Q&A-style’ checkbox, which is clearly of the contributor mentality. – James Haigh Mar 11 '15 at 6:46
  • 1
    @LightnessRacesinOrbit - When did I say that? AAMOF I go out of my way to get people interested in how the site works. I dunno how many comments I've left to the effect of "Welcome to the site :)", or "If this answers your question, you should 'accept' it by clicking the check mark under the vote count, which awards the answerer (and you!) some extra rep points :)". It's people's choice whether they stick around or not, and I really don't think how we word a button is gonna change that. So why not word it for the 99% of the people that are actually gonna see it? – Robotnik Mar 11 '15 at 12:13
  • 2
    @Robotnik: You're missing my point. You're assuming that, if those 99% of people are the ones to see it, we must word the button according to their current misconceptions. Why not take word it to accurately reflect what this website is and what it stands for? Choosing to label a button to pander to misconceptions is harmful. Is what I'm saying. NB all your friendly comments are nice and great but irrelevant to this discussion! As is whether someone "sticks around or not". – Lightness Races in Orbit Mar 11 '15 at 12:18
  • 3
    @LightnessRacesinOrbit - No, you're missing mine. The users shouldn't have to care about our secondary objective of 'curating a repository' or the loftier 'building a better web'. Yes, they are secondary objectives - without the questions and answers that actually keep SE sites going, we would have nothing to curate, the 'misconception' is thinking that curating comes first. Further: you're the one that brought up "mould[ing] a 99/100-er into a 1/100-er.", and then called me 'harmful' for "Not pushing that mould" so don't tell me it's irrelevant after you've already made the point lol – Robotnik Mar 11 '15 at 12:37
  • 2
    @LightnessRacesinOrbit How in the world is "this solved my problem" reinforcing a misconception? I've read this whole comment stack and I still don't understand what distinction you're making. – SevenSidedDie Mar 11 '15 at 18:53
  • 3
    @LightnessRacesinOrbit I think you might have SO-focused blinkers on here. Yes, people treating SO as a helpdesk is a problem. That's not a new-user misconception that all of SE shares. Treating the whole network as if it's just like SO is not helpful. – SevenSidedDie Mar 11 '15 at 19:10
18

In addition to general description of implementation given here, askers can now unilaterally dupe-close their questions (provided they have sufficient privileges to cast close votes or flags).

I just tested this using one of my own questions here at MSE (see revision history).

It went as follows:

  1. I voted to close as duplicate
  2. Refreshed my question in browser to make new duplicate prompt appear
  3. Clicked the big blue button ("That solved my problem!"), the question is instantly closed

I like the way it works. Although, I am apparently biased:

It occurred once or twice to me and have to admit, it felt... great. Just think of it, you get your answer, what could be better?

In cases like this I flag / or vote to close my own question myself...

The reason for "self-closure" is simple: since I found an answer, I want to save self from trouble of further "maintenance" of my question - you know, from studying and replying to possible comments, answers, from stuff like that.

Since I found an answer, I just don't need my question anymore, and the sooner I get it closed, the less I will need to worry about it in the future...


Worth adding that as an active close vote reviewer I also find this feature very promising. Per my observations, reviewing duplicates is much more difficult than the rest...

duplicates review went slower (much slower) than the rest: I had to study both questions, and in some cases I also had to check the answers to both questions...

...so additional chance to get help from asker is really great.

If they "accept" the duplicate, this simply cuts all the cumbersome work that I would have to do as reviewer, and if they edit in a compelling explanation of the difference, this makes it so much easier to decide to leave open.

For the sake of completeness, I already saw some inexperienced askers adding rather senseless "explanations" in response to a new prompt. If this gets widespread, we may have a problem.

  • 1
    meta.stackexchange.com/a/250960 - reported, but thought it is by design :) and no, my previous comment was addressed to Shog :) – nicael Mar 10 '15 at 17:56
  • @nicael yeah I was thinking about this. But after testing it appeared more like a (relatively minor and natural) part of the implementation. I couldn't figure what to discuss / challenge but wanted only to clarify the way how things work now; a supplementary answer looked like a more appropriate media for that – gnat Mar 10 '15 at 18:45
  • 1
    HEY look where we are now :) – nicael Mar 10 '15 at 19:55
  • @nicael magic fly! – gnat Mar 10 '15 at 20:02
5

Is more guidance needed for the asker regarding what will happen when the asker confirms the duplicate?

Change the text on the first button from "That solved my problem!" to "Yes, my question is a duplicate."

The latter makes it way more obvious what will happen if you click on it.

  • 18
    The user doesn't need to know or care what will happen, they need to know which choice to choose. All that we need to know from them is, "Did this solve your problem, if no, why not?" The user doesn't need to know what the consequences of their answer is in order to answer the question correctly. In fact, they're probably less likely to answer truthfully (or to understand the question) if the question is phrased about site mechanics that they are very likely to be unfamiliar with. – Servy Mar 10 '15 at 21:05
  • @servy I find it unlikely that knowing the consequences of the choice will cause more users to answer untruthfully. – Rainbolt Mar 10 '15 at 21:07
  • 7
    I do. So many users protest close votes, and object to it... I think that adding words like close as Duplicate is gonna put them off, either they don't want to admit they already made the mistake of asking, or they don't want to close it if it isn't really a dupe - and make a mistake now. – Tim Mar 10 '15 at 21:10
  • @Tim I can see that some might be put off by the thought of acknowledging that their own question is a duplicate. But wouldn't you also be put off if you clicked that button and then realized that you had accidentally closed your own question? – Rainbolt Mar 10 '15 at 21:21
  • 4
    Hmm perhaps. but if it answered them, surely it should be closed? – Tim Mar 10 '15 at 21:22
  • @Tim Yes, exactly. Here's the expanded logic: If Foo is solved by Bar, then Foo is a duplicate. If Foo is a duplicate, then Foo should be closed. – Rainbolt Mar 10 '15 at 21:28
  • 3
    Right. so why would we make the user not want to do that? – Tim Mar 10 '15 at 21:40
  • @Tim I feel like I'm being asked a loaded question. What exactly are we making the user not want to do and how does it relate to my answer? – Rainbolt Mar 10 '15 at 21:53
  • 4
    We don't want the user to be scared off from clicking that button. So why use long confusing jargon when we can ask them a simple question that they understand? – Tim Mar 10 '15 at 21:54
  • @Tim Sorry, but "Yes, my question is a duplicate." is not "long confusing jargon". It's actually just a simple six word statement written in words that everyone who knows even basic English can understand. – Rainbolt Mar 10 '15 at 21:56
  • 11
    New user. What does close mean? What does duplicate mean? These phrases are specific to SE, and mean different things elsewhere - especially on forums. Think back to being a new user, you probably understood very little. – Tim Mar 10 '15 at 21:57
  • 2
    @Tim What, do you think we were cavemen before we got here? What does duplicate mean? What does solved mean? What does problem mean? These questions are simply idiotic. – Rainbolt Mar 10 '15 at 22:09
  • 18
    Note that the banner doesn't use the word "duplicate" anywhere else right now; to use that terminology on an action button we would have to define it somewhere, which might just open up a bigger can of worms. The intent here is to motivate the asker to either confirm the suggestion or clarify his question - in other words, even if the question is a duplicate, he might well be better off editing to make it not a duplicate if by doing so he's better able to eventually find a solution to his actual problem. – Shog9 Mar 10 '15 at 22:15
  • 7
    Even experienced users get caught up in "is this a duplicate?" a lot, for example worrying about whether the questions are exactly the same even if all the answers that apply to one would apply to the other. Making it dead simple for new users avoids all of that. It doesn't mean we're acting like they can't understand English, we're just asking them the core question. – Cascabel Mar 11 '15 at 4:02
  • 1
    @Rainbolt And that definition is apparent given the current wording. In this case, "different" means, "That didn't solve my problem", which is made clear from the first option. If the first option is changed to "this is a duplicate" then that information is lost. One does not just need to know that "different' questions are "questions to which the answer to one doesn't answer the other" because the first option is already explaining that. – Servy Mar 11 '15 at 14:43
5

Another case for "this question is not a duplicate, and does not need to be edited":

Tracking & Synchronization of Multiple Sheets with a Master Sheet

This question was flagged as a duplicate although there was already verbiage within the question that clearly demonstrated the difference. The verbiage pre-existed my awareness of the proposed duplicate, and therefore does not explicitly reference it, but it still applies just as well and I don't think it needs any modification.

I have added a comment to the person who proposed the duplicate, pointing out the requirements that make my question distinct. However, with the current implementation, there is no way for me to appropriately dismiss the prompt to confirm/reject the proposed duplicate. That prompt will be there for the life of the close-vote (or, at least, I assume it will disappear if the question is not actually closed after a certain time or the close vote is rescinded by the voter) even though the question is neither duplicate nor in need of modification.

It is not at all uncommon for someone to inappropriately propose a duplicate despite a question already being well-defined enough to clearly be distinct. And I'm sure it's far from unheard of for someone to propose linking to a completely unrelated question. There really should be some way for the OP to dismiss the banner in these cases, without forcing unnecessary modifications to the question.

I suppose I could always just add a string of &nbsp; or <!-- HTML Comments --> and be done with it.

  • 3
    Or you could just ignore the banner. – Servy Mar 11 '15 at 15:35
  • 2
    @Servy I'm no expert on the subject, but I don't think that's good UX. – Iszi Mar 11 '15 at 15:36
  • Additionally, even in the information distinguishing the questions existed before, if readers are voting to close it as a duplicate it may not have been clear. Expanding on that point, emphasizing it more, perhaps mentioning it earlier in the quesiton, etc. is frequently going to be worth doing. – Servy Mar 11 '15 at 15:36
  • 3
    While it may be the case that the post isn't a duplicate and you don't know how you can better improve it, such situations are likely to be rare. Any option to just "ignore" is almost certainly going to be used far more by people who shouldn't actually be ignoring it than it will by people who really have nothing to do. – Servy Mar 11 '15 at 15:38
  • @Servy HTML comments it is, I suppose. – Iszi Mar 11 '15 at 16:00
  • If you're that opposed to actually improving your question, I guess there isn't anything anyone can do to force you to improve it, and if you're willing to make an abusive edit just to not see a banner, that's your decision. – Servy Mar 11 '15 at 16:02
  • 1
    @Servy I don't know about anyone else, but I personally spend a good deal of time drafting, reviewing, editing, and proofreading my questions before I post them - especially one as lengthy and detailed as that. So I hope you'll forgive me if I'm not inclined to go and re-write, or post an addendum with equal care and nearly as much time, just because someone has tagged it as being duplicate with something that the question already distinguishes itself from in its current form. – Iszi Mar 11 '15 at 16:10
  • 3
    Then just do nothing, rather than making an abusive edit. – Servy Mar 11 '15 at 16:12
  • 3
    The review interface, for users who have access to it, already has a "Leave Open" option that logs the reviewer's opinion and removes the review item from their queue. In fact, it's a completely separate option from "Skip" - the latter being analogous to your "ignore it" suggestion. Why not at least permit similar functionality in this proposed duplicate banner, for those with enough reputation to work the same review queue? – Iszi Mar 11 '15 at 16:12
  • 1
    I already explained why. You'd find that just about everyone would press "ignore" if you give it to them as an option, in this context, rather than either indicating that they have their solution or taking the time to actually improve the question. It would be appropriate in so very few situations, and yet would be used in so many, that it would almost certainly be a net harm to add it, more or less defeating the purpose of having the feature in the first place. – Servy Mar 11 '15 at 16:14
  • @Iszi The review queue UI is designed for uses who have proven they already know what they're doing, so it's naturally more trusting that users will use it according to SE principles. That is in fact the whole point of the review queue. – SevenSidedDie Mar 11 '15 at 19:06
  • 2
    In which case that's the threshold you want: allow users with close-vote review queue privileges to dismiss the banner. If a user would be able to vote to Leave Open as not a duplicate, they understand the site well enough to know be able to let stand if they think their question is already suitable, without extra annoying UX elements which won't go away. – E.P. Mar 11 '15 at 20:09
  • (I'd put the threshold much lower than that as the consequences are not terrible and the 3k+ population is not that big on many sites, but that's another discussion.) – E.P. Mar 11 '15 at 20:10
  • @E.P. Sounds like that would make a good feedback answer to post on its own, yes? Edit: Ah, it has been already! :) – SevenSidedDie Mar 11 '15 at 21:43
  • I don't see why it is so critical to dismiss the banner. The banner is only shown to you, so it's only a mild annoyance, but it won't affect what answers you get. I suspect your specific scenario is very rare, so putting extra developer effort in to accommodate that scenario doesn't seem like it makes sense. In addition the benefit of adding the extra functionality you suggest is minor (and the number of people benefitted is likely small), whereas the potential harm is significant (as Servy explains), so the feature you request might do more harm than good. – D.W. Mar 11 '15 at 22:05
4

Should askers be able to dismiss the banner without editing or confirming? Perhaps if the number of votes is below some threshold?

Yes! But only for people with a certain amount of rep, like 100 or more. The reason for me thinking that is, that it is not always the case that a supposed duplicate actually is a duplicate. I have seen a number of cases where a close vote on "duplicate of..." not is a duplicate at all, but merely seem to have been a hasty judgement based on similarity in title, use of certain expressions etc, or simply a misunderstanding of what the question is all about.

You could expect, or there is a chance for, that a to an extent experienced user with 100 rep or more already have seen a suggested "duplicate", therefore this user should be able to dispute / reject / walk through a suggested self-close by clicking on a "Disagree"-button or similar.

However - new users, or users with very little rep, seems to be more likely not to have done any research before asking, so they should of course be encouraged to realize that they have asked a question already answered before.

  • 4
    Instead of 100 rep (that's trivially easy just by associating accounts and doesn't demonstrate understanding of SE), attach the dismiss ability to having the close-vote privilege. Having that privilege demonstrates (one hopes!) that they know how duplicates work, and can confidently dismiss the banner. – SevenSidedDie Mar 11 '15 at 21:47
  • 1
    +1 to this and @SevenSidedDie's comment. I suggested a similar threshold (access to close-as-duplicate review queue) in a comment to my own answer. – Iszi Mar 11 '15 at 22:29
  • 2
    Not sure this is necessary, but if we do it it'll probably be controlled by the number of close votes rather than the rep-level of the asker - it's not like new-user questions are somehow immune to bad dup-suggestions. – Shog9 Mar 11 '15 at 23:30
  • @Shog9, you are right of course - new users also faces questionable dup-suggestions. Number of close votes a good idea. – davidkonrad Mar 12 '15 at 23:49
  • 1
    @SevenSidedDie, when a user receives a starting 100 rep he or she is considered "experienced Stack Exchange network user with 200 or more", so you would expect such a person know what a duplicate is. I agree with you to a certain point, just think the close-vote privilege is rather high (3000) - it is hard to get there. And compared with the point of "nagging" users who have posted a supposed duplicated question, it seems to me to be an exaggerated high demand. – davidkonrad Mar 12 '15 at 23:56
  • It's true that the association bonus is fluffed as "because we trust you elsewhere", but it's still a trivial amount of rep to get. The trust is also at the level of "you know how to tie your shoes elsewhere, so we're pretty sure you can here too"; while trusting that a user understands how closing should be done isn't assumed until much higher levels of rep. It's moot if it's not going to key off rep, but if it was, it should be somewhere higher than 100, though possibly not as high as the 3000 needed for closing privileges, I admit. – SevenSidedDie Mar 13 '15 at 0:00
  • 1
    @SevenSidedDie Funny analogy :) Yes, 100 is perhaps too low. I think Shog9 has a point. If I understand correctly, you could say that if there is two or more close votes pointing to the same supposed duplicate, then the user should not be able to click on a dismiss / reject button. This would make sense. – davidkonrad Mar 13 '15 at 0:09
3

I recently ran into this and I liked the notification / call to action.

However, do you really want people adding justification for Why This Is Not A Duplicate into the question? IMO, this is promoting Bad Question Writing by creating a Second Class Question: it has to justify it's own existence within the text of the question itself. This creates questions that are wordy, defensive, and conversational. I fell into this trap here. That's administrivia, not content!

  1. I don't like writing Second Class Questions. I want my question to be concise and stand on its own, just like all other Questions.
  2. I don't want to read questions that defend their existence.

    If it's not a duplicate, then why should I have to read the defense? Shouldn't that be irrelevant? I just want to read the question and the answer.

Like I said, I really did like the notification / call to action. But I'll put any defense about "why my question should exist" in a comment. Many others won't, and over time, we'll have more and more questions containing Why This Is Not A Duplicate.


FYI: I think this Answer should exist.

  • 9
    Don't really want to have to read comments to figure out what you think is unique about your question. And CAN'T search comments to find your question if whatever makes it special is needed later. – Shog9 Mar 11 '15 at 8:11
  • 2
    You're quite right that a post stating that it should exist isn't helpful to anyone; it's just noise for the reader. What we really need people to do is to edit the question to explain why the proposed duplicate doesn't answer their question. They should explain, clarify, and/or emphasize how their question is different from the proposed duplicate and why the answers to the proposed duplicate don't answer their question. When done well it's not a defensive argument about the state of the question, but rather making what the original question was more clear to readers. – Servy Mar 11 '15 at 14:02
  • 3
    Of course, getting people to do that instead of just do what you did and assert, without explanation, that the post should exist, is hard. We could adjust the wording to help people with that but better instructions can only do so much. – Servy Mar 11 '15 at 14:02
  • @Shog9 Perhaps consider auto-appending a starter for the edit, along the lines of "I've looked at answers on [link to proposed duplicate], but they do not help me because...". This is something I do, and have seen others do, even before a duplicate is proposed if I've found some close matches that don't quite cover what I need. – Iszi Mar 11 '15 at 15:31
  • 1
    But… we do want questions to be written in a way that justifies their existence, don't we? That's a first-class requirement of a question, here, to the point where lacking apparent purpose is on the downvote tooltip. Doing it badly is still better than a question that doesn't explain why it's being asked. – SevenSidedDie Mar 11 '15 at 18:59
1

Should users who press this button get the same rep as for accepting an answer? That's effectively what they've done. You shouldn't be able to accept an answer on your question as well (and net a, cool, extra +2 rep).

It incentivises accepting the question a dupe, and also stays consistent with accepting an answers when there wasn't a duplicate question.

  • 2
    So... There are a few other proposals that would allow for this by actually showing duplicate suggestions as answers. I'm a huge fan of this idea, but unfortunately we're not there yet. – Shog9 Mar 18 '15 at 18:43
  • One day @Shog9, one day. – Pureferret Mar 19 '15 at 0:19
  • 1
    What a user who gets +2 for accepting an answer did: (1) posted a question that wasn't covered on the site yet; (2) thus prompted creation of new answers; (3) selected one of those answers as the preferred one. What a user who clicks the button did: (1) posted a question that was already covered on the site; (2) clicked "OK" button. Not quite the same. – user259867 Mar 19 '15 at 0:24
  • @Woodface good points. But consider some duplicates are from poor searching before hand. Others create genuine new routes to the existing answer. – Pureferret Mar 19 '15 at 0:29
0

I already pointed, that I can flag my question for closure, if I don't have 250 reps yet but already have 15 reps, and then click "That solved my problem".

As you may know your closure flag is approved when someone after your flag votes to close.

But with introducing the new feature to approve dupe votes, quite a serious problem appeared. If I flag it and then click that button to solely close it, my flag automatically becomes helpful, because Community actually votes to close.

  • Is this really a problem? The number of helpful flags isn't really used for much of anything, nor is it even visible to anyone else, and it's quite apparent if someone is doing this, making any possible abuses easy to resolve. – Servy Mar 11 '15 at 14:27
  • 2
    @Servy Helpful flags actually increase priority of your custom flags and number of flags per day. But it is not fair in this case, I think. – nicael Mar 11 '15 at 14:29
  • 2
    It doesn't affect the priority of the flags. It does increase the max flags that you have, but that would require doing this quite a lot, not just one or two times. If you start seeing dozens of self-closed duplicates with no close votes from other users, then one could look into it. It just isn't prone to abuse, it's easy to find and address if/when it happens, and the incentives to abuse it are just so low (very few people actually use all of their flags). – Servy Mar 11 '15 at 14:34
  • 4
    Hmm... Got more flags, lost ability to ask questions. Such a tradeoff... – Shog9 Mar 11 '15 at 14:54
  • @Shog Lol :D but then there should be away to buy an ability to unban yourself with the power of flags, no inconsistency please! :P – nicael Mar 11 '15 at 15:04
  • 4
    Farming helpful flags by closing own posts is ridiculous enough to not be worth guarding against. Easier and less noticeable opportunities for flag farming exist, which do not put a user on moderators' "had many questions closed" radar. – user259867 Mar 11 '15 at 19:45
-3

I hate that there are only two options. What about, "No, this question is not a duplicate, and does not need to be edited"? or "I think the other question should be closed as a duplicate of mine, and not the other way around"?

I just had one of my two-year-old questions flagged a duplicate of a brand new question. Why should I have to either edit my question or mark it as a duplicate just because someone came along and asked the same question two years later?

That's an unpleasant experience. It's one more of the "we are out to get you" sort of feelings that so often occur here.

  • 11
    There aren't actually just two options. There's also the third option of "press neither of the buttons" which is of course always an option. Note that your question being older doesn't mean it should be the canonical question. The best/most useful question (taking their answers into consideration as well) should be the one left open, not the oldest one. If your question is the one that should be left open than that's typically an unusual enough case that an extra option doesn't need to be added just for that. – Servy Mar 11 '15 at 14:05

You must log in to answer this question.

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