21

I just got two down votes on a question I answered in August. After I provided a short answer to the original question, the poster revised his question the next day. Now, my answer makes little sense.

I understand the dilemma of finding a bad answer to a question, and not wanting to give the impression that it is useful for future readers, but it's also weird to see you've been 'admonished' for an answer to a question that no longer exists.

Isn't there a better way to handle a situation like this?

7
  • 9
    Does your answer fit the revised question? If not, you should delete it.
    – nobody
    Commented Sep 19, 2014 at 14:20
  • 1
    I did delete it.
    – joel goldstick
    Commented Sep 19, 2014 at 14:33
  • 4
    It's a known issue, with no real solution IMHO then educate the user that keeps changing his question. If it's a moving target I usually vote to close as 'unclear'.
    – Patrick Hofman
    Commented Sep 19, 2014 at 14:44
  • 40
    don't worry too much about internet points
    – Sam I am says Reinstate Monica
    Commented Sep 19, 2014 at 16:03
  • 10
    This situation seems confusing because you're looking at it as 'admonishment' or 'punishment'. It's easier to use a viewpoint where the goal is forming a useful question/answer combo. If you deleted it, then it sounds like you've pursued that goal.
    – DCShannon
    Commented Sep 19, 2014 at 22:12
  • 5
    Is the person who provides the answer notified when the question is changed? I believe this little feature of the site engine would be enough for this kind of issues.
    – Wiktor Zychla
    Commented Sep 21, 2014 at 18:51
  • 3
    This is why really it might be good for the notification system to notify people of edits to questions they have answered.
    – Sammaye
    Commented Sep 21, 2014 at 19:07

1 Answer 1

Reset to default
36

Generally this comes down to two different cases:

  1. You mis-understood what the question was asking, the edit clarified it, and now it's clear that you mis-understood the question.

  2. You answered the actual question the author meant to ask, but they edited the question to change what they wanted to know (possibly because they now have an answer to the first question, and have a follow up) and the new question is quite separate.

For case #1 the answer needs to be fixed/removed; it was incorrect all along and it's simply clearer why it wasn't a valid answer now. For case #2 the edit to the question should be rolled back (or flagged if it seems a rollback war has started) and optionally asked as a new question.

15
  • @Unihedron He can suggest the edit.
    – Servy
    Commented Sep 19, 2014 at 14:53
  • 3
    His suggested edits has to go through the suggested edits queue, and reviewers may reject it for radical changes.
    – Unihedron
    Commented Sep 19, 2014 at 14:55
  • 5
    And for case 2, I disagree that you should always roll back. You should considering the other answers as well. What if someone answered the revised question. Rolling back would invalidate their answers.
    – psubsee2003
    Commented Sep 19, 2014 at 14:55
  • @Unihedron I assumed you meant the OP of the actual question, not the meta question; I edited my comment accordingly.
    – Servy
    Commented Sep 19, 2014 at 14:57
  • I stand my point that it may be rejected. It might be better, if a rollback was necessary, for him to bring it up on meta or ask a privileged user to roll it back instead.
    – Unihedron
    Commented Sep 19, 2014 at 14:58
  • 4
    @psubsee2003 I don't see that as a reason not to roll back the question, although the problem should ideally be avoided whenever possible by addressing such cases as soon as the question is changed, rather than a while later.
    – Servy
    Commented Sep 19, 2014 at 15:00
  • @Unihedron It certainly might be, although it shouldn't be if there is a good revision comment. He can flag the post if he needs to though, yes.
    – Servy
    Commented Sep 19, 2014 at 15:01
  • 1
    @Servy I agree that if it is found immediately, then rolling back is the correct option. But when discovered weeks or months later, and rolling back would invalidate other answers, then that is unfair to the users who posted answers to the changed question. This is a question without good answer unfortunately. I don't have a solution in this situation, though.
    – psubsee2003
    Commented Sep 19, 2014 at 15:15
  • 7
    @psubsee2003 So why is it okay to invalidate the answer to the original question then? If the new question and its answers really are that valuable, a new question can be asked and those answers provided to it. Unless there was some particularly compelling reason to do otherwise, I'd default to keeping the original question, not the altered version.
    – Servy
    Commented Sep 19, 2014 at 15:18
  • Thank you all for the enlightenment. I didn't know about rolling back, and I am sure I don't have that authority. In my case, I think Item 2 applies, but with the revised question came a good answer(not mine), so that seems like a good outcome. I deleted my answer because it was of no value to the revised question
    – joel goldstick
    Commented Sep 19, 2014 at 16:22
  • @joelgoldstick You don't need any authority to roll back an inappropriate edit. Anyone in the world is entitled to do so.
    – Servy
    Commented Sep 19, 2014 at 16:25
  • 6
    second case is known as chameleon question
    – gnat
    Commented Sep 19, 2014 at 16:26
  • 2
    @psubsee2003: So your recommendation is to not roll it back, and therefore punish the poster who actually answered the original question asked in the form it was asked? That hardly seems like the proper solution to me. The edit to a new question after there was an answer posted was invalid, IMO, because at that point it became a chameleon question. Once a question receives an answer, it's inappropriate to then change the entire meaning of the question. IMO, the rollback to the first answered version is the appropriate action, and the user should be informed why it was rolled back.
    – Ken White
    Commented Sep 19, 2014 at 21:38
  • 3
    @KenWhite No one's being punished. The idea is to get the most useful question/answer combination for posterity. If rolling it back does that, then good. If that would invalidate good new answers to the good new question, then the best combination might be achieved by deleting the earlier answer. It's case-by-case.
    – DCShannon
    Commented Sep 19, 2014 at 22:09
  • 2
    @DCShannon It's also about being consistent: in the rules, we state that if you have follow-up questions, or a different question, then you should ask a new question, and not edit the current one. Does the fact that there are new answers suddenly make this OK? Also, as an answerer you can see that it's an old question, with an old answer that does not seem to address the new edit. If you then answer, it's at your own risk.
    – MicroVirus
    Commented Sep 21, 2014 at 11:52

You must log in to answer this question.

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