0

A while ago I asked a question on the SFF Meta When I flag as dupe, can the system put the link to the answer and not the question?, but I have recently realised that it would be a better fit here. I will copy/paste the question here:


This might be a little bit confusing, please bear with me.

As per this discussion: http://meta.scifi.stackexchange.com/questions/3684/should-questions-tagged-as-duplicates-because-of-an-answer-be-treated-different?cb=1

What essentially happened is that I flagged the question as dupe, but when I did, I flagged with a link to an answer and not the question.
A very subtle difference, but it could help the OP and any mods / reviewers seeing the flag understand why I've made the flag, or where the duplication lies.

To clarify:
Could we please change the flagging process to accept links to answers and not just the question which contains the answer?

So when the users click on the link in

This question already has an answer here:
Link to Question x answers

they should, in these cases be taken to the answer, and not the question.

Notice how I'm linking to a specific answer (i.e. "https://scifi.stackexchange.com/a/40119/21267"). It wouldn't matter if the post has 100 answers, the user would be pointed directly to the answer they're looking for.


Here's how it would work:

I see a new question posted and open up the close-as-dupe dialog. Which will then prompt me for a link to a post. What I'm asking about is if I enter a link to an answer:

enter image description here

2
  • 4
    And what if there are multiple answers? The accepted answer? That can be changed. The highest scoring answer? That can change, too. Only do it if there's a single answer? What if someone adds an answer? Commented Jun 9, 2015 at 10:56
  • 1
    @FishBelowtheIce It doesn't matter how many answers the linked Q has, rather which question contains the duplicate info.
    – Möoz
    Commented Jun 9, 2015 at 22:31

1 Answer 1

11

There's only one problem with your feature request: answers don't make questions duplicates. Questions are duplicates only when the questions themselves are duplicates. In other words, you can't close a question as a duplicate of another question just because "you can find your answer over here."

The only exception to this are canonical questions; these are questions that are asked and answered specifically to address a category of questions that are always answered the same way.

For example, someone asks a question to the effect of "How do I fix this error," but the error turns out to be a null reference exception. Unless the error is, say, specific to a particular library where the answer might be "upgrade the library," such questions always have the same cause and fix, and are definitively answered by What is a NullReferenceException, and how do I fix it?

8
  • 1
    If the answer or correct information exists, then that's where a user should go. Questions marked as duplicate are not necessarily a bad thing, just another way of finding the same information, which is why they don't necessarily get removed. My point though isn't the validity of the close-votes when they're cast, rather once the question has been deemed dupe then point the users to the correct answer (based on the links provided in the VTC).
    – Möoz
    Commented Jun 9, 2015 at 5:46
  • 8
    But I always thought that the general SE policy would be to duplicate vote based on answers rather than questions (not that I would personally support such a policy, but it seemed to be accepted, especially on the site this meta question is primarily concerned about). Commented Jun 9, 2015 at 11:07
  • 2
    @ChristianRau there are quite a lot of people who seem to think so - I've never understood why since all the stuff in meta on this is consistent and the close reason even says "This question has been asked before and already has an answer" Commented Jul 13, 2016 at 10:59
  • This meta answer, with significantly more upvotes than this one, and multiple references to the Stack Overflow blog for background, seems to advocate the "close if answers are the same" approach.
    – tripleee
    Commented Apr 25, 2017 at 8:38
  • 1
    @tripleee: The meta answer you referenced consistently uses the phrasing "Duplicate Questions," and the suggestion that questions may be dupes if the answers are the same is just that... a suggestion.
    – user102937
    Commented Apr 25, 2017 at 19:48
  • Your answer here seems to state that this can never be the case. The interpretation that it is acceptable to close questions as duplicates if the answer would be identical is widespread, and further perpetrated by the meta post I linked to. It is linked from the FAQ and broadly cited. If it is incorrect, it needs to be updated. It would feel rather paradoxical to have to open up another meta question to clarify the situation but I guess that's what I have to do?
    – tripleee
    Commented Apr 26, 2017 at 4:08
  • @tripleee: You can if you want, but it's already been discussed exhaustively. The guiding principle remains: questions are duplicates if the questions are themselves duplicate. That the original question answers the duplicate is a nice bonus, but the question still has to be a duplicate, your interpretation of that meta post notwithstanding. Closing questions as "you can find your answer over here" doesn't always make sense; the questions can bear little resemblance to each other, and it is the question, not the answer, that is the defining characteristic of a Google Search.
    – user102937
    Commented Apr 26, 2017 at 22:26
  • Is there a canonical meta you can point me to where this is clearly established? Ironically, the question here is about a different topic...
    – tripleee
    Commented Apr 27, 2017 at 3:56

You must log in to answer this question.

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