24
$\begingroup$

One of the changes to the closure system in the recent revamp was that to close something as a duplicate, the suggested duplicate needs to have an upvoted answer.

I suppose that's a good thing when you suggest a duplicate of a question from a month or two ago. However sometimes a user would post a question, and when that one goes unanswered they would go on to post another some hours later.

Is it possible perhaps to modify this behavior and allow closure as duplicate even if there are no answers, if the suggested duplicate was posted in the last 72 hours or so?


Is the countable sum of a set of measurable functions also measurable? and Is the countable sum of a set of measurable functions also measurable? is an excellent example.

The user posted both here and on MathOverflow. The latter was migrated here, but since the user is unregistered on both sites, we can't close as a duplicate.

This is exactly the reason that for new questions duplicates should be allowed without the limitation of upvoted answers.


Here is another example:

  1. Equivalence relation for set given as matrix
  2. https://math.stackexchange.com/questions/1003400/equivalnce-relation-for-sets-given-as-matrix

Are word for word identical, and the user gravatar indicates that the same user has asked both questions. But since the account has split, it is impossible to close as a duplicate without at least one of the questions have an upvoted or accepted answer.

Which again shows that relaxing a little bit the requirements from a duplicate suggestions might not be a bad thing.


Third example:

  1. https://math.stackexchange.com/questions/1023246/is-the-relation-ra-a-function
  2. Check if a relation on a set is a function
  3. Arbitrary Set A a Function??

Here is an example where first the same user asked the same question twice, the option to ignore the time/answer restriction in that case were used correctly. But then another user (possibly a classmate) comes and asks the same question within a two weeks span.

This is not an isolated incident, and probably most regulars can remember one time or another that their favorite tags were bombs by classmates asking the same question over a week or two.

This, of course, stretches the 72 hours I suggested above, but it still shows some merits to allowing a "cooling period" where duplicates can be suggested and closed at will, and not just under additional conditions.


Here is yet another striking example why this is a wonderful idea.

  1. Proving the existence of disjoint subsets
  2. https://math.stackexchange.com/questions/1108354/suppose-a1-a2-an-are-subsets-of-a-which-has-cardinality-n

Are posted within less than an hour of each other (35 minutes to be more accurate). It's the same question, and the users are completely different.

But alas, we cannot close one as a duplicate of the other, because neither has an upvoted answer yet. I still don't see the need for having two copies of the same question, posted almost at the same time, and not being able to close them as duplicates.

$\endgroup$
  • $\begingroup$ In the meantime, should we close such questions as "too broad" or something else? $\endgroup$ – Asaf Karagila Jul 15 '13 at 6:58
  • 1
    $\begingroup$ If the two questions are identical and by the same poster (as often happens with these), I would prefer if they were merged (or the new one simply deleted if it had not had time to attract comments on things apart from the fact that it was a duplicate). I do realize that this is somewhat harder to achieve than getting it closed. $\endgroup$ – Tobias Kildetoft Jul 15 '13 at 8:13
  • 8
    $\begingroup$ There is already an exception for questions from the same user, you can close those as duplicates even if they are unanswered. $\endgroup$ – Mad Scientist Jul 15 '13 at 8:39
  • $\begingroup$ @MadScientist: oh, I did not know about that. Is that announced somewhere? (I may have just overlooked it!) $\endgroup$ – Willie Wong Jul 15 '13 at 8:48
  • 3
    $\begingroup$ @MadScientist: Sadly it isn't rare I those cases that the unregistered user has split in th meantime. $\endgroup$ – Asaf Karagila Jul 15 '13 at 11:16
  • 4
    $\begingroup$ Another approach would be to flag for moderator attention. $\endgroup$ – Gerry Myerson Jul 15 '13 at 13:14
  • $\begingroup$ meta.SE: Duplicate question changes - can't mark duplicates by same user. Epimorphic provided this link in chat: How to handle a deliberate duplicate question. $\endgroup$ – Martin Sleziak Nov 3 '14 at 15:03
4
$\begingroup$

This is not a very common situation; I think I can still count the number of times I encountered it on my extremities. Therefore I would suggest flagging these cases for moderator attention -- it is what I usually do now.

I would be interested to get an indication from the mods on how often such flags pop up and if they think that it's happening too much (i.e., that the process should be community-sourced somehow).

Compare the discussion on Meta.SE when the restriction on dupe-closing was first implemented, in particular this answer, where Shog9 comments:

Merges have always been a mod-only function. As for dup-closing: unless there are a crapload of these scenarios that folks have been ignoring up to now, this is a drop in the bucket mod-load wise.

Edit: Shog9's answer to that question sums it up nicely as well.

$\endgroup$

You must log in to answer this question.

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