52
$\begingroup$

Suppose that you ask a question, and field some answers.

After reading them, you then unexpectedly realize that your question admits only trivial answers, due to some edge case you didn't realize you had to exclude when you wrote it.

It is useful that people have shown you this, but now you want to exclude the edge case and try again.

What's the right thing to do here? You've already gotten answers, and those people deserve upvotes and etc. Do you make a new question that's a variant of the old one, and try again? Do you edit your existing question?

For example, this just happened to me here. I asked about commutative operations distributing over both multiplication and addition. Since I didn't leave out the possibility that the operation be undefined at the multiplicative identity, the only such operations are trivial. But there may be highly interesting, non-trivial operations which blow up to infinity as one operand approaches 1, for example. Do I make a second question and ask this now?

$\endgroup$
  • 18
    $\begingroup$ When I see such questions, I try to elicit a clarification before posting the "trivial" answer. Sometimes the OP says okay, please go ahead and post that answer, but more often it provokes a reformulation that makes the problem interesting. Of course you are welcome to post a new question if there has been a rush to state the obvious in the Answer box. $\endgroup$ – hardmath Aug 6 '15 at 4:15
43
$\begingroup$

You should post a new question, providing a link to the old question (but making the new question stand alone as well). You could also post a link to the new question as a comment to the old one.

In general, one should take care that any edits to answered questions do not change the scope or meaning of the question - both to preserve the answers' value and in hope that someone with the same trivial question will be able to find your original question. Additionally, as you note, the reputation system and the Q&A style of the site are more conducive to having multiple questions than to having a single question change.

$\endgroup$
21
$\begingroup$

You could one of the following (and probably not more than just one):

  1. Edit the question with something like "User such and such showed in the answer below, that without adding the condition bla bla bla the answer is trivial. What happens when we also require/not require that bleebleeblu?"

  2. Post a new question, giving partial context and linking to the previous question, in which case you should probably accept one of the answers given on the old question.

  3. You could try and ask the user in the comments about this, since it might also lead to a quick and trivial answer.

$\endgroup$

You must log in to answer this question.

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