9
$\begingroup$

I've seen many related questions including:

But I think few users are following any of the better ways to ask or answer proof verification questions, and it is still continually generating tons of unanswered questions that will never get answered. I also don't think it makes sense to expect people to write up community wiki answers to such questions, since the asker has already been completely satisfied with the answer given in the comments, and often there is little more to say. I'd submit that Math SE is bound to have this issue much more than other SE sites, and so perhaps it would be really helpful to have a closing reason dedicated to this:

This proof-verification question has been satisfactorily answered in the comments.

This would allow us to close a large number of such questions effectively and clear up the unanswered questions list significantly. Such has been suggested before but nothing concrete has come out of it yet. I think it is clear that time spent writing up answers would be better spent giving good answers to other questions than just to make such questions answered.

Edit

It appears that I may have given a slightly wrong impression. I do not advocate actively seeking out such questions to close them, but rather it is intended to be a passive measure whereby users who finish reading a question and find it answered satisfactorily in the comments can immediately close the question so that other users who are not interested in these localized questions will not waste their time. Even as a passive measure I am sure that it would be an efficient mechanism to settle questions that no longer need an answer.

$\endgroup$
  • 2
    $\begingroup$ There is some possibility of answering such questions well, as is pointed out in other threads. This does not entitle all such questions to the effort required to write such an answer, but at the same time, it should tell us that closure is not an acceptable option. $\endgroup$ – Milo Brandt Jul 20 '15 at 3:12
  • 1
    $\begingroup$ @Meelo: So what should we do when there is really nothing more to be said that isn't already said? I still it does not make sense to me for a Q&A platform to have so many such questions with a one-word answer in some comment and still leave them open for answering! $\endgroup$ – user21820 Jul 20 '15 at 11:15
  • $\begingroup$ Not sure what "passive measure" really means. Once a close vote is cast, the post goes into the review queue. So it will have to be read by the already-overworked reviewers, who will either vote to close or to leave open. In the latter case the close vote ages away and we are back to the starting point. $\endgroup$ – user147263 Jul 21 '15 at 3:04
  • $\begingroup$ @NormalHuman: In the first place I don't think people should overwork themselves on anything. Besides, reviewers who see that the post has been marked for "closure due to satisfactory answer in comment" can feel free to skip. So by passive I literally mean only those who read the question simply while browsing and not via the review queue. $\endgroup$ – user21820 Jul 21 '15 at 3:44
  • $\begingroup$ Okay. In that case the close vote still ages away in 14 days. The process of closing is optimized so that it either happens quickly, or not at all. $\endgroup$ – user147263 Jul 21 '15 at 3:45
10
$\begingroup$

You are bringing up "effectiveness" and "time", and then arguing that closing is preferable to posting a pro forma answer... I don't follow the logic.

  • At present, there are $1118$ users able to close questions (source), and you need five of them to close a question. (This is before we consider that many of these users don't close much.) Each of them would have to read the question and comments to figure out that a satisfactory resolution was reached.

  • There are $10142$ users with $200$ reputation or more (source); I picked this threshold to remove passers-by from the count. It takes just one user to post an answer. There is a bookmarklet for copying comments into answer, which automatically provides attribution to the comment authors. One click to invoke the bookmarklet, two clicks to mark the answer CW, one click to submit.

I think the latter scales a lot better than the former.

Although personally, I'm in favor of the third approach: do nothing. It doesn't really matter how many unanswered questions are there. In the next revision of SE site interface, the Unanswered tab will be removed. (Its replacement is a filter for "need answers", which is meant to highlight the questions that actually need an answer, rather than all those that do not have one).

$\endgroup$
  • 1
    $\begingroup$ I didn't know about the upcoming change to the SE site interface, which would also answer my question. As for my reasoning, it is based on the fact that many people including myself don't like the idea of having an answer that just says "Correct.". If you look at the links I provided, one typical suggestion would be to give a more illuminating answer where possible, such as possible points for improvement. That would surely take more time than just closing. $\endgroup$ – user21820 Jul 18 '15 at 5:23
  • $\begingroup$ Most of the work required is not writing as such, but evaluation of the proof. To type "Your proof is correct, but the second paragraph would be more clear if you fixed $\epsilon$ first, and chose $n$ later" isn't as much work as parsing the proof. And parsing is something that each of reviewers would have to do, because they are affirming that the question was satisfactorily answered... I suggest spending a few hours in the close review queue before deciding which option is easier. $\endgroup$ – user147263 Jul 20 '15 at 23:46
  • 1
    $\begingroup$ Actually I'm only referring to questions that users have finished reading through and then find that it has been satisfactorily answered in the comments. All it takes is five of those (or a moderator) to close the question under the reason I proposed, rather than to write up an answer. This would save the time of all other users who are not interested in such very localized questions. I guess my point is that we're strangely accepting of leaving localized questions open and unanswered, whereas on other SE sites it seems they are usually closed. But as I said, the new SE site interface is fine. $\endgroup$ – user21820 Jul 21 '15 at 2:15
2
$\begingroup$

Why should a question be deemed "closed" if somebody checked the proof? What if the check is superficial, contains a mistake, overlooked some subtle point?

If I asked for a check of a proof of mine, I'd be delighted if somebody, even long after "successfully resolved" ("accept answer" would be the right action here, IMHO) added a different proof.

$\endgroup$
  • $\begingroup$ Of course you should check the solution in the comments. It's exactly the same as with a normal answer. The problem is simply that it had been posted as a comment instead of an answer, so accepting it was impossible. Especially if the asker says "Okay I understand now." or something like that, we can assume that he would have accepted it were it an answer instead of a comment, so it should not stay as an unanswered question. $\endgroup$ – user21820 Jul 23 '15 at 8:57

You must log in to answer this question.

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