I know that this is a common problem, one very much discussed in meta. But what I want to know is if I made a good call in this particular instance.

I posted this question a few minutes ago on MSE. However, I made a blunder. I forgot fundamental assumptions for the question (essentially, the hypothesis of Mayer-Vietoris).

Previous to my update, an answer to the question as posed was submitted. It was (then) correct, and I have upvoted it. However, it was not correct to deal with the question I wanted.

The problem is: the context I gave (explicitly talking about Mayer-Vietoris) implicitly states that I am working on the assumptions that it makes. This made me think that the question was not intrinsically badly posed and misleading, so I decided not to delete it and simply state things explicitly.

But of course, it was my error not to put everything explicitly. Therefore, I left the previous answer upvoted and accepted another one.

Was this the best path in this situation, or was it better to follow another course of action? (As of now, the previous answer explains very well what was the situation, so I think the outcome was not bad. But I would like to know people's opinions, since I don't like the situation of "changing the question while invalidating answers").

  • 9
    $\begingroup$ Why not ask a second question (and think more about the question you really want to ask, next time)? $\endgroup$ – Did Mar 13 '17 at 9:32
  • $\begingroup$ @Did Because of the fourth paragraph I wrote. If I had not mentioned any context, I would ask a second question. But given that I did, I think it would be "artificial" (for lack of a better word) to post another question. $\endgroup$ – Aloizio Macedo Mar 13 '17 at 11:00
  • 9
    $\begingroup$ This paragraph does not change the fact that the first version of your question was (rightfully, according to you) interpreted differently. If the first answer posted was indeed addressing the question as it was written at the time, I see no reason to invalidate this answer. And a convenient way (to which I see zero defect, to be frank) to avoid this, is to post another question. $\endgroup$ – Did Mar 13 '17 at 11:06

Changes to the question within a few minutes are OK in my opinion. But do not change the question a week after posting and getting an answer.

We have often seen a question asked, then an immediate answer (or comment) saying "a constant function satisfies this". Feel free to add the hypothesis (such as "nonconstant") immediately. Do not make a new question with this trivial change.

  • 1
    $\begingroup$ But the concept of trivial solution really depends on the level of the user. Since math.SE is supposed to be a site for asking math questions for any levels, so it's better to have a more objective rule that is easier to execute. quid on math.meta SE "As a rule a question should not be a moving target, and to alter it in such a way as to render existing answers wrong or even just incomplete is discouraged. It is alright to rollback such edits and to ask OP to ask a new question. " $\endgroup$ – GNUSupporter 8964民主女神 地下教會 Mar 19 '17 at 10:03
  • 1
    $\begingroup$ @GNUSupporter in that post, the sentence directly following the one you quote is: "However, one should also strive to answer questions in the spirit in which they are asked, asking for clarifications first if needed." In my mind, and I think this is the spirit of the answer, with which I agree, there should be some room for individual judgment. I am personally not keen on preserving misstated questions with (non-interesting) literal answers. That said, in cases of conflict, it is likely best just to ask a new question. $\endgroup$ – quid Mar 19 '17 at 12:43

You must log in to answer this question.

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