2
$\begingroup$

I'm curious: how are we supposed to handle questions that have been changed such that older answers no longer reflect the questions being asked? In the context and spirit of this meta question what do we do for (or to) existing questions who've been edited such that answers:

  1. are invalidated
  2. are out out of context
  3. are answering questions no longer present.

Here's an example of the 3rd case (which I answered yesterday). The highest rated answer at the moment, Motosubatsu's answer, answers a third question which has since been edited out. Personally I like his answer to the 3rd question, but was confused as I had not seen it when writing my own answer (because it was edited out).

In the example above, should the question that has been edited out be edited back in to prevent confusion to future readers?

In general case, what is the recommended procedure?

$\endgroup$
4
  • $\begingroup$ The question should be rolled back I think as it is unfair to those who put time into answering. $\endgroup$ Commented Nov 16, 2017 at 21:22
  • $\begingroup$ @Bellerophon The question was edited so as to be not put on hold. I don't think the answer has been invalidated, completely, since it still answers the premise of the question in it's current state. $\endgroup$
    – sphennings
    Commented Nov 16, 2017 at 21:25
  • $\begingroup$ @sphennings Sorry, I was answering in a general case. I haven't looked at the specific question. $\endgroup$ Commented Nov 16, 2017 at 21:29
  • 4
    $\begingroup$ meta.stackexchange.com/q/302373 This is the closest I could find on this topic. The answers don't fully address your question but it might help. $\endgroup$ Commented Nov 16, 2017 at 21:33

2 Answers 2

3
$\begingroup$

The post was edited because it was at risk of being put on hold. Each post should ask a specific answerable question. The second question of the post was edited out, because in addition to being a second question, it was about creating plot not building a world. By removing the second question the post was made specific, answerable, and about building a fictional world.

It's policy to not make edits to questions that invalidate existing answers. It's also policy to edit questions so that they aren't placed on hold. I don't think we've had a discussion about what to do when these two policies come into conflict.

In this case since the answer still answers the core premise of the edited question, I don't think the answer has been invalidated.

$\endgroup$
1
  • $\begingroup$ This might be the closest to what you're looking for, although it only sees from the author's edit right perspective. $\endgroup$
    – Vylix
    Commented Nov 17, 2017 at 4:40
1
$\begingroup$

General issue was already discussed network-wide meta:

Exit strategies for "chameleon questions"

Generally if answer was invalidated, then there are two options:

  1. Question was looking OK when answer was posted. It is OK to roll it back and comment - unless rollback would invalidate other answers. Either way, it is not right to downvote such answer.

  2. Question was looking close-y when answer was posted. Especially if there was close votes and person who answered had view close votes privilege. In that case the answerer is at fault, and should vote / flag to close, or work with OP to clear up the question, instead of answering. In such case, treat bad answer as bad answer: delete if it's yours. Request edit and/or downvote if it is not.

    Bottom line - we should edit too broad and opinion based questions, or or put them on hold faster, to prevent such answers from even being posted before question is in good shape.

$\endgroup$
2
  • 1
    $\begingroup$ I disagree with 2. Although the person has close vote privilege, for him, the question might be answerable, and thus it's not his fault for answering the question. If he VTC and answer, then it's his fault. $\endgroup$
    – Vylix
    Commented Nov 17, 2017 at 4:33
  • 1
    $\begingroup$ Overall, this is a good answer. But personally, I would not edit a too-broad or opinion based question without OP's consent, just to make it a good question. $\endgroup$
    – Vylix
    Commented Nov 17, 2017 at 4:35

You must log in to answer this question.

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