Consider this suggested edit to this question. The question regards a particular error seen when attempting to embed youtube videos for iOS. At time of writing, there are no answers. The edit is a major rewrite, but the root question/issue is exactly the same.

Apparently the edit was sparked because the user who suggested the edit had the same problem, wanted it solved, but wasn't satisfied with the question as it stood. Editing was their way of asking their own question that would otherwise have been a duplicate. They also put a bounty on it.

On the one hand, this is good. The question's better than it was before and more likely to be answered well. On the other hand, there are aspects that don't fit well with the current SO model:

  • Should this edit even be approved?
    It would not be a stretch to call it a radical change. Outside of the actual problem, nothing would really remain of the original asker's question. The editing guidelines aren't totally clear. This is well beyond the minor changes suggested as editing reasons, but certainly falls squarely within the overarching goal of making the post substantively better.

  • If approved, it feels like the editor should own the question.
    The editor (evidently) put more time into writing the question, and has in this case committed a bounty to it as well. They have plenty invested, while the original asker just happened to note the problem first. Obviously their main goal is getting an answer to the problem, but it's unfortunate to see them get the short end of the stick reputation-wise because someone else happened to ask the question first.

  • If not approved, is there any alternative for the editor/bountier?
    If they ask a new question, it'll be (rightly) closed a dupe.
    If they put a bounty on the unedited original question they're likely to not get good answers because it's not as well written as it could be.

Fundamentally, the question is:

What should you do when your question has already been asked badly?

Looking through meta there are a number of similar discussions, but none that seemed to address the core problem:

(If there's a definitive discussion I missed, please point me towards it.)

  • 3
    It is possible to ask a new one and try to get the community to close the older one as a duplicate of it, if the older question is really that bad. – animuson Jun 29 '12 at 23:40
  • as long as I'm positive that root question/issue is exactly the same I'd approve the improving edit in a heartbeat, no matter how much words and letters have been changed – gnat Jun 29 '12 at 23:57
  • @animuson If the original question did obtain some useful answers despite its badness, a merge might be more appropriate – Tobias Kienzler Sep 28 '12 at 15:16

As long as the "rewrite" attempts and succeeds to respect the OP's focus and does not dismiss any data or sub-questions they asked, and does not invalidate existing answers on the question, the edit is completely in the interest of both the OP and the community. (This is however easier said than done as the intent may even be quite unclear or the OP even unable to recognize their own problem in the wider context.)

The editor may obtain fewer badges or reputation that way, but the reputation system is not intended to be complex and cover every situation directly.

The conflicting objectives of editing often cannot all be met with confidence. So there is also another clear option to enter a separate question and either:

  • reference the original question and explain why exactly that question does not frame the problem adequately or correctly as part of the new question, as a measure against being closed as a duplicate
  • or ask the community to close the worse question as a duplicate

If the older question was not just bad, but also obviously bad, that will work heavily towards the success of any better new question regardless of which replacement approach is chosen. Otherwise choose wisely.

  • 2
    "As additional insurance against premature closing, put a bounty on your question." This is not going to work on two accounts: 1) Using bounties just to prevent a question from being closed is heavily frowned upon; there are far better ways to make your case against closing 2) You won't be able to set a bounty until 2 days later anyway, by which time it'll probably already have been decided whether or not to close the question. – BoltClock's a Unicorn Jul 3 '12 at 2:15
  • 1
    @BoltClock'saUnicorn ...and 3) It would be quite an overkill. I agree with you. Paragraph replaced. – Jirka Hanika Jul 3 '12 at 5:44

In a case like this, where the question had been asked badly and was still unanswered, I would encourage posting a new question, and closing the old question as a duplicate of the newer, better-written one.

If the old question had worthwhile answers, I would find it legitimate to post a new, well-written question and close it as a duplicate of the old, very differently worded question. If the old question is really bad like in this case, editing the old question is ok too. If the old question is ok and merely a very different way of formulating the same question, both versions of the questions are useful and the one that has no (worthwhile) answer should be closed as a duplicate of the one that has answers.

  • 1
    Yet another reason why we should somehow modify the description so people don't whine about older questions being closed and pointed to newer ones "when it should be the other way around." – BoltClock's a Unicorn Jul 3 '12 at 2:17

I think that if you have to rewrite someone's post that much to get your ideas across that you should rather start a new post to not risk losing the original intent of the post. Since if it's that important, even to the point that one puts a bounty on it: then it obviously deserved to be put in a proper question.

I think that a lot of care must also be taken by the Editing Poster not to lose sight of original intent, especially when the proficiency in english language of the EP > OP.

In my opinion this is why you'll regularly see cleaned up posts with code tags and regular formatting that still keep the OP's 'broken english' (for lack of a better phrase) even though their EP has a very good or even native grasp on the english language. It's just too easy to move away from the OP's intended question.

e.g. In this specific edit information about the video format used gets lost (webM) WebM is not supported by my browser I'll be honest that I don't know how important that it is, but it might be and it just got tossed, hereby losing a chance it might trigger an "oh yeah that's how you solve that!" opportunity for a visitor that might know the answer. The general intent does seem kept.

The better written question will be the one people stumble on to when searching for their issue in search engines and can thus serve as the defacto link-to-as-duplicate question. Serving it's proper goal: to inform people of the answer.

Finally if you write your own question then at the very least you'll be sure that the original author won't think "what's this I didn't do this/mean this" and won't rewrite the question again. Protecting a great question is a duty as well, not for the reputation but for the site and people looking for your answer. (maybe people can't edit after someone with higher rep or a moderator has edited, I honestly don't know, in which case disregard this) It's better to have the user with the need for an answer and the best grasp on the english language to hold said question in my opinion.

  • How should an asker handle their better question being closed as a dupe (which seems reasonable in the cited example)? – blahdiblah Jun 30 '12 at 0:17
  • I think that's a job for the moderators to decide provided you're convinced your question is the better one. So best course of action would be to open a thread in the meta subsection of the site in question. I also doubt people will 'vote close to close' just because it's a duplicate. I think the better question will 'win'. – Harald Brinkhof Jun 30 '12 at 0:20
  • the main factor will probably be how much of such questions already exist: only 2/few or a gazillion like your typical closed PHP tagged question? – Harald Brinkhof Jun 30 '12 at 0:24

What should you do when your question has already been asked badly?

Fundamentally? Ask a new question.

Should I totally reword someone else's question?

This I regard as wrong. A question should always be the thing that is asked by a particular person, though sometimes it might take quite a lot of cleaning up to expose what they are asking. Nevertheless, they are asking it.

What can we do for new users who want to ask a question that's already been asked, but hasn't yet got an acceptable answer?

If it's not got an answer worth the name (and isn't likely to attract one) then asking a new question that tackles the problem from a different angle might be a good idea. Who knows, it might actually be a genuinely different problem.

Adding additional info to someone else's question?

Again, if it's more than a small amount it's more likely to be a significant alteration to the question. What is “small” and what is “significant” are by nature tricky to define. I suggest using discretion. (I wonder if anyone's suggested edit has ever been rejected for the combination of being too major and too minor…)

How to teach an OP to write questions? May I rewrite a badly asked question?

The bad question writer? Too often they're busy writing badly somewhere else in the world, rather than paying attention to SO. While yes, teaching them would be nice I really suspect it's not going to happen. OTOH, rewriting a bad question while preserving its author's voice is tricky too; it's a learnable skill for sure, but you really need to practice it to be good. It might be better to flag the question for moderator attention with the write-in text that you think it is poor and you're going to try to write a better question.

  • FWIW, the goal I tend to aim at in my editing is as if instead of writing the post themselves, they had a competent native-english-speaking secretary (that knew all of SO's formatting tricks and foibles) do it for them. I want the OP to think “this is how I wanted the question asked”. – Donal Fellows Jun 30 '12 at 0:04
  • How should someone distinguish their question from the previous one to avoid having it closed as a duplicate? – blahdiblah Jun 30 '12 at 0:14
  • @blahdiblah Reference it by URL and explain in words how it is different? Try the most obvious approach first… – Donal Fellows Jul 4 '12 at 14:53
  • That works fine when the question is at least somewhat different, but for one like this about an error message it seems like the extent of the difference might be "...but mine's better written." I'm not sure that would be enough for people not to close it. – blahdiblah Jul 4 '12 at 19:12

When you think the question is a duplicate of some other question, first of all you should comment on that question, with a link to the question which you think it is a duplicate of. Then you can close that question (at least you can do that).

You must log in to answer this question.

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