22

In recent weeks a lot of questions get pushed in the close review queue with the custom close reason

I’m voting to close this question because it was solved by a comment.

I propose to retire this close reason.

Case 1: The problem and solution can be useful for future users with the same problem.

In this case, closing a question without a proper answer is detrimental to the library of knowledge we are building here. If the question does not have enough votes, it might be automatically deleted after closing and the useful information will be lost, see https://tex.stackexchange.com/help/auto-deleted-questions for more information about the automatic deletion.

Instead of closing the post, it would be more helpful to

  • ask the user who solved the problem to convert their comment into an answer
  • if there is no feedback from the user in a reasonable time frame, their comment could be posted as a community wiki answer quoting their comment.

This will ensure that future users with the same problem will find the information more easily than searching through potentially long comment threats.

If the post gets closed for other reasons, e.g. because it is too broad, needs details etc., the appropriate close reason should be chosen instead of a generic "solved in comments". This will give the OP and other users actionable feedback how they can improve the post. The presence of an answer will save the closed post from being deleted in this case.

Case 2: The solution is unlikely to help others.

If other users are unlikely to have the same problem, e.g. because the problem was a typo or a similar unique situation, then the post should be closed for this reason. This allows the OP and others to understand why the post has been closed. A generic "solved in comments" close reason fails to communicate this.

We have a dedicated close reasons for this:

  • This question was caused by a problem that cannot be reproduced or is a typographical error. While similar questions may be on-topic here, this one was resolved in a manner unlikely to help future readers. This can often be avoided by creating a minimal working example (MWE) necessary to reproduce the problem before posting.
16
  • 5
    A common use case for answering in comments is because the question is a duplicate or near-duplicate, but the commenter does not have time or interest to actually find the duplicate. In this case closing as "solved in comments" can be ok imo, there is nothing lost if the question is deleted (because the duplicate will still be there to be found if needed), but it is not fit for "cannot be reproduced or a typo". While it would be better to actually find the duplicate and close against it, many times this is not worth the trouble because the new question does not provide a good signpost. Commented Jan 12 at 13:31
  • 2
    @Marijn If that's the use case, I'm sure there could be a better formulation which would convey this information to the OP and other users. Commented Jan 12 at 13:48
  • 1
    It is often in the comment itself, something along the lines of "I'm quite sure this is a duplicate but you need to put \label after \caption to get the correct reference". Commented Jan 12 at 14:01
  • 3
    @Marijn The comments from the past few weeks just read I’m voting to close this question because it was solved by a comment. (or variations thereof) without any additional information for the OP or other users. I'm afraid that this might make some users think that just the fact that there is an answer in the comments would be a reason to close the post. Commented Jan 12 at 14:25
  • I didn't mean the closing comment, I meant the comment that is the reason for closing, which actually contains the answer. tex.stackexchange.com/questions/558641/… is an example, where the answer with the duplicate link is given in the comment by hesham and the closing comment is given by Mensch. The comment by hesham is the one that provides additional information for the OP and other users, in particular that the issue in the question can be solved by following instructions in the other question. Commented Jan 12 at 17:47
  • I have to admit though that this use case is not as common as I thought, when I was searching for an example I mostly found questions where the answer was given in comments without mentioning that it could be a duplicate. Commented Jan 12 at 17:49
  • @Marijn Thanks for the example! Personally I think the "not reproducible" close reason would have worked there too. The problem was no longer reproducible after a reinstall. Commented Jan 12 at 17:56
  • I agree that "not reproducible" would have been ok for that question as well. However, this does not solve the issue you mention in this meta post, i.e., that this question could be useful for other people but risks being deleted when closed, future visitors needing to look through comments, and no actionable feedback is given on how to reopen the post. Commented Jan 12 at 18:05
  • 1
    In some cases an alternative possible action is to write a community wiki answer capturing the information from the relevant comments. Commented Jan 14 at 9:57
  • @AndrewSwann Indeed! That's a good way to avoid loosing knowledge! Commented Jan 14 at 10:01
  • 1
    Some questions that people initially think is just a trivial useless typo turns out to get quite a lot of votes over time in StackOverflow. So we can't really tell that in advance. ■ Besides, if everyone goes "surely there must have been a duplicate" then the problem is it is entirely possible that all the duplicates doesn't have a proper answer ("closed because answered in a comment"), which would make it difficult for someone to search. Ideally we can find a good question with clean small MWE and well-written answer. Commented Jan 23 at 5:40
  • Here is an other example I'd vote to close, as it is solved by comments: tex.stackexchange.com/questions/736218/… . // Once the close-dialogbox is modified as discussed by me, there is a chance to create and preserve spelled-out answers more frequently. Commented Jan 31 at 17:14
  • @MS-SPO I don't see a reason to close this. The question looks clear and other users might also have this problem. Commented Jan 31 at 17:29
  • I come from the ticket world: answered, problem solved, ticket closed, no further action needed, alternative solution almost unlikely. That’s why I once again stress to implement reasonable process steps, provided, this information would get lost otherwise. Commented Jan 31 at 17:37
  • 1
    @MS-SPO Stackexchange sites are a lasting library of knowledge. Closing should for for posts not suitable for the site. Commented Jan 31 at 17:39

1 Answer 1

Reset to default
1

A few thoughts on your opening.

Process view

Instead of closing the post, it would be more helpful to

ask the user who solved the problem to convert their comment into an answer

if there is no feedback from the user in a reasonable time frame, their comment could be posted as a community wiki answer quoting their

comment.

How to avoid this process-wise? Extend the close-dialogbox, as was already done for other choices. Your 2 cases can probably be put directly into the dialogbox.

When do I use a close with "solved by comments"?

I just did it for Latex PDF adds line for areas with low opacity. Let's use it to study my behavior.

Observations:

  • Question is 3.5 years old
  • it's about an artifact (in hindsight)
  • Comment 1 says so
  • in Comment 2 the OP confirms "you are right! it was just the preview, thanks!" (and this was quick, i.e. not buried in the 20th comment)
  • sounds like a (ticket-) close to me
  • OP was here last time "more than 3 years ago"

Hm. When looking for unanswered questions, why do I see these types at all? It's no longer a problem. OP didn't care anymore.

What do we lose in information/solutions? In this case probably little to none. Visual artifacts are posted as questions every now and then. It may be different for other questions.

My criteria

I tried a randomized review on my "close because solved by comments" actions. However, via my profile/Votes/Closure this is ... not well supported: my randomized hits hit everything else, but not "solved by comments". So, I used it obviously quite sparingly ;-)

So recalling from memory these are more or less my criteria to vote this way:

  • Most important: did the OP comment "it's solved"?
  • how old is it? (i.e. might it be outdated?)
  • what other activities have there been or might be expected on this question?

Hope this helps

5
  • 2
    "What do we lose in information/solutions? In this case probably little to none." There are indeed such cases and I think it would be more helpful for the site if the close reason would explain exactly this and not say "solved by comments". The fact that the question is solved in comments isn't the reason for not keeping the question around, the reasons for not keeping a question around are the questions are off topic, lack information, are unlikely to help other etc. Commented Jan 28 at 8:52
  • Sure. In my example it was an artifact, which will probably not happen this way today. What's the harm, not knowing this exact one years later? // However, "known artifacts" could be a useful topic for a summary-thread as a kind of "did you think about this and tried ... ?" // Do deleted question really get lost? If so, how to identify and preserve the vital few? Commented Jan 28 at 9:11
  • Deleted question are viewable to high rep user if they have saved the link to the question. Recently deleted posts can be found from the tools page, but after that they are basically unfindable, so lost to the library of knowledge. Commented Jan 28 at 9:16
  • 1
    "What's the harm, not knowing this exact one years later?" The harm is that other users might not understand under which circumstances questions should be closed and then go about and close anything which has a comment as "solved in comments". Better not make them believe that being solved in comments would be a reason to close the post. Commented Jan 28 at 9:19
  • 1
    Ok. First: not all harm is the same. Second: The answers from Quality Management are Guidelines and Process Design. The answer from Quality Engineering is making it fool-proof, insensitive to unwanted variations. So let's implement that. Commented Jan 28 at 9:53

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.