2
$\begingroup$

If this has already been proposed, I apologize. I have noticed that many questions are answered through short discussions in the comments; it is obvious that the OP is satisfied, so no one bothers to post an official answer. You don't have to look too far to find this sort of question, but here is a recent example. I have found several Meta questions (e.g., this, this, and this) about this very issue; the prevailing solution that I have seen is to either ask the commenter to post their comment as an answer or to post it as an answer yourself. This seems like a perfectly fine solution to me, except for the fact that I have never seen it done. It may happen regularly without me seeing it, but it still remains that there are questions that are left in this condition, which seems like a problem to me. These questions will go down in the archives of the site as unanswered, and they will likely never receive an answer. This has a negative effect on the organization of the site and poses a problem for people browsing old questions.

This is why I think that there should be a review queue for old (one week, for example) questions that don't have an answer. The reviewer would be able to distinguish between good questions that no one has been able to answer yet, bad questions that slipped through earlier moderation efforts, and questions that were answered in the comments. The first variety would move on as intended by the site's design, the second would have another chance to be altered as needed, and the third would have a specific opportunity to have the comment turned into an answer (it seems like turning it into a community wiki would be the fairest option). Thus, potentially low quality questions would have a chance to be looked over before they are relegated to their future of likely inactivity in the archives. It seems to me that this process would be immensely beneficial to the quality of the site's archive of questions, which is good since Stack Exchange emphasizes that its questions are meant to be a permanent resource.

I haven't been active on this site for too long, so I would appreciate any response on why this is not a good fit for the site if it is not.


Edit (sort of): Having written my above question, I searched on Meta Stack Exchange to see if this had been asked there before posting; sure enough, I found this question that in essence is the same as mine, albeit without the emphasis on dealing with questions answered in comments. The answer states that review queues are best suited for resolving issues through community consensus, and that tasks like the one I am suggesting should be done on one's own. I would just like to say that I don't think that answer is satisfactory; sure, many review queues deal with gaining a community consensus on how to handle a question, but the First Post and Late Answer queues are similar to my suggestion and seem to work fine. Also, I am asking this question because a significant number of people have not taken it upon themselves to police questions that are about to go into the archives and because a review queue would make this process easier, more accessible, and more visible.


Edit: Now that I have had some responses, I would like to rephrase my idea (hopefully more clearly) and add some new information.

First, I am trying to express the fact that I think we need a way to filter out questions that are answered in comments rather than with "official answers"; the questions I linked earlier in this post seem to support this. The solution to this in the past has been to hope that users repost such answers as official answers, but this isn't entirely effective. Thus, it seems to me that the best way to filter out these answers is with a review queue that contains questions that are still unanswered after a week; thus, the reviewer can easily repost any such comment answers as real answers and make any other modifications that need to be made.

I would like to know whether:

  1. Others think that having answers in the form of comments is an issue
  2. The current method of hoping for users to notice this and modify it is sufficient
  3. A review queue is a viable option for fixing it (if necessary) or what a better method would be (if any)

I have gotten the sense that people do see it as an issue that is not completely resolved, but that the effort adding a review queue would entail is not worth it. I would like to know why.

Second, I used several searches to gather some information that may be pertinent:

  • Seven days ago, 162 questions were posted.
  • After seven days (today) 40 remain unanswered.
  • Out of those 40, 18 were put on hold, closed, migrated, or marked as duplicates; 2 needed modification to conform to site policy (not pertaining to being answered in the comments); 11 were fine; and 9 were answered in the comments.

This makes me think that we do in fact need to make some sort of change to address this sort of post; assuming these data are representative (which I think is fairly safe) and putting the first category aside, a user looking for old, unanswered questions has roughly the same chance of finding one that doesn't fit their criteria as they do of finding one that does, which would be very frustrating.

$\endgroup$
  • $\begingroup$ I agree with the diagnosis, but not with the proposed remedy. $\endgroup$ – AccidentalFourierTransform Aug 2 '17 at 19:02
  • $\begingroup$ Perhaps also relevant physics.meta.stackexchange.com/questions/9846/… $\endgroup$ – JMac Aug 2 '17 at 19:07
  • $\begingroup$ @AccidentalFourierTransform What would be a good alternative? The idea of users just fixing it themselves when they see it seems to have been emphasized for a long time to no avail. A review queue may be seen as too extreme a step, but I don't know of any alternatives between the two. $\endgroup$ – user122423 Aug 2 '17 at 19:11
  • 2
    $\begingroup$ What do you mean when you discuss (v1) "questions that are about to go into the archives"? $\endgroup$ – rob Aug 3 '17 at 3:00
  • $\begingroup$ Something to consider: why haven't you posted answers to some of these old questions, especially the ones which have already been answered in the comments? It's free reputation, if nothing else... $\endgroup$ – David Z Aug 3 '17 at 3:41
  • $\begingroup$ @rob I just mean questions that leave the main page; I guess they aren't necessarily "archived", but most certainly don't receive much attention after that point. $\endgroup$ – user122423 Aug 3 '17 at 4:26
  • $\begingroup$ @DavidZ I haven't thought much about this issue until recently, which I think supports my point; if it is a task worth being completed, there should be some official structure for it rather than hoping a sufficient number of users come to this conclusion on their own, then regularly seek out and fix such questions. $\endgroup$ – user122423 Aug 3 '17 at 4:37
  • 2
    $\begingroup$ Well... okay, what is your point exactly? Are you saying that we could use some way to draw more attention to old questions which have not had answers posted (whether or not the information that constitutes an answer exists in the comments)? That I agree with. Or are you saying that we need additional structure to convert comments which contain "answering information" to actual answers? That's where I don't agree - I think we have all the structure needed, which is the ability to post an answer. $\endgroup$ – David Z Aug 3 '17 at 5:32
  • $\begingroup$ By the way, one reason you may never have seen comments converted to an answer is that the comments are often flagged and deleted once the answer is posted. $\endgroup$ – David Z Aug 3 '17 at 5:32
  • $\begingroup$ @DavidZ I'm saying that once a question reaches a certain age (I'm thinking ~1 week) without an answer, it should be placed in a queue to be reviewed; if it's fine, so be it. If it is low quality, it can be edited as needed. If it was answered in the comments, the reviewer can repost that information as an answer; no new structure other than an additional review queue that helps ensure that questions don't remain in a form that doesn't fit the site's intended format. $\endgroup$ – user122423 Aug 3 '17 at 6:10
  • $\begingroup$ OK, that would be option #1 from my comment. In that case, transferring answers which have been posted in the comments to actual answers is not really the main focus of what you're asking about (right?), and personally I think it would help clarify your post if you de-emphasized that case. $\endgroup$ – David Z Aug 3 '17 at 6:35
  • $\begingroup$ @DavidZ Transferring answers in the comments to actual answers is the main focus of my post; it just happens that the solution I'm proposing has wider implications. $\endgroup$ – user122423 Aug 3 '17 at 14:23
3
$\begingroup$

There already is such a review queue: the Unanswered Questions tab, easily accessible on the main navigation panel between the Badges and Ask Question links.

Moreover, you can also filter on a per-tag basis and even add any search terms that you want (click on any tag, or do a search adding [tag] [other-tag]) and then sort via the 'unanswered' criterion. There's also a wide array of search criteria so you can get some pretty fine control over dates as well; as an example, created:7d.. will pull posts created within the past week, lastactive:14d..7d will find posts whose last activity was the week before that, and so on.

Having a moderate archive of reasonable questions that can still be answered is not a bad thing ─ it is a resource that enables new site members to get up and running by finding old posts that can use some more work and putting in those improvements.

Now, I'm not saying your idea is terrible, either, but I'm not sure I see a problem with the way things are at the moment.

$\endgroup$
  • $\begingroup$ I'm afraid I disagree. I am not saying that it's bad to have unanswered questions. I am saying it's bad to have questions questions that are categorized as unanswered when they were actually answered in the comments. $\endgroup$ – user122423 Aug 3 '17 at 14:13
  • $\begingroup$ @Ziggurat You're entitled to disagree, but none of the arguments you've set forward show any evidence of how your proposal would be useful above and beyond the existing functionality in a manner proportional to the magnitude of the problem. Review queues take a lot of code and it is very hard to convince SE to implement them (and heck, we're already trying to push for new functionality on existing queues that would actually solve a problem), so you'd need to show a much stronger case that this is needed. $\endgroup$ – Emilio Pisanty Aug 3 '17 at 14:19
  • 2
    $\begingroup$ ... and I'll also disagree with that last sentence of yours: you were looking for something to answer and found something with answers in the comments? Good! now you can summarize those comments in an answer and help the site. I don't see at all how that is a problem. $\endgroup$ – Emilio Pisanty Aug 3 '17 at 14:20
  • $\begingroup$ The approach of asking people to take this responsibility upon themselves has been tried unsuccessfully for 4+ years from what I've seen in my research; I'm not necessarily looking for something that will make it easier for me to do it, I'm looking for something that will actually get it done. $\endgroup$ – user122423 Aug 3 '17 at 14:36
  • $\begingroup$ @Ziggurat Well, as I said, if you want SE to implement that much code for you, then you will need to show correspondingly strong evidence that the alleged current lack of "getting it done" is a big problem. (It's easy to underestimate how much code a review queue entails: as a starter, consider that it's a change to the database structure, and that it impacts the data explorer and the data dump, with extra code and changes to the schema to make sure that it accomodates them, and with a bunch of code references that need to be tested to make sure they don't break. None of that is easy coding.) $\endgroup$ – Emilio Pisanty Aug 3 '17 at 15:08
  • $\begingroup$ Thank you for your thoughts. $\endgroup$ – user122423 Aug 3 '17 at 15:13

You must log in to answer this question.