We all want to conquer that maddening advertisement, but haven't found a cure for the persistent habit of "humbly" answering questions in comments. Fine reasons for the answer-by-comment strategy are duly noted, but don't we all agree that some follow up Answer should, well, follow up?

Two ideal scenarios, with user NewB the asker and Mr. Wizard (no disrespect) the commenter, are:

  • NewB posts a self-answer after reading a useful comment, and nobody notices (ReputationGain -> 0).
  • Mr. Wizard posts a real answer later, and NewB doesn't follow protocol (ReputationGain -> 0) or does (\[Infinity] + ReputationGain \. ReputationGain -> 15).

An alternative scenario, involving a fledgling user Ian (just sayin'), offers real incentive:

  • Ian translates Mr. Wizard's comment into an Answer, and comments on the original question (@Mr. Wizard I fleshed out your helpful comment into an answer.). The benevolent Mr. Wizard is notified and upvotes Ian's answer (ReputationGain -> 10).

The alternative scenario incentivizes middling reputation users to search out and "answer" the easy questions, and they (we) learn stuff along the way! Without participation from the answer-by-commenters however (and you do feel a little guilty, don't you?), I doubt many will work for these lessons alone.

My demonstration/experiment is in progress! What do you think?

--Edit--

Sample size is now two. Should be enough for solid inference.

Keywords/phrases for searchability:

Turning another user's comment into an answer

  • In your demonstration you commented on your own answer, which means IIRC that acl will never be notified. So he might never see it. You have to comment on the question if you want acl to be notified. – C. E. Jun 5 '15 at 21:08
  • @Pickett Ah, excellent point. Will try to rearrange. – Ian Jun 5 '15 at 21:48
  • 2
    It may be that acl could not decide whether the question should be closed (as "easily found in the documentation"). In fact, perhaps close votes were cast, but not enough accumulated to close the question before the votes expired. In any case, I'm glad you answered the question. I think after a week or whatever, a commenter loses "dibs" on the idea in their comment, although one should acknowledge it in an answer, as you did. – Michael E2 Jun 5 '15 at 22:16
  • A week as a waiting period looks kosher to me. Long as you acknowledge the original commenter(s), you're okay. Even better: if you can add your own contribution, that is an even better reason to write an answer! – J. M. needs help. Jun 5 '15 at 23:25
  • @J. M. Crediting the commenter should be done with care, is there a way that is immune to user name changes? I've just learned that right clicking on the timestamp next to a comment gives you a direct link. Is that the best practice? – Ian Jun 6 '15 at 1:50
  • Linking to the comment will work, yes. I don't know a way that is immune to name changes, tho. – J. M. needs help. Jun 6 '15 at 2:08
  • I do not believe in a one week "waiting period" unless the comment says "I don't have time to write this up now, but I will write it up later" or something to that effect. Some people make a link out of a user's name protect against name changes, f.e. Ian allows the reader to check out which user it refers to even if that user has changed his name, because it's the user ID in the URL that matters. – C. E. Jun 6 '15 at 21:18
  • I think this issue has been well covered in the answers below but I want to give my support to the idea of expounding on comments as answers. Simply mention in the answer something like "as Michael commented you can use ..." If the comment is an answer but cannot reasonably be expanded -- perhaps Daniel Lichtblau comments "this is a bug; it's been fixed in version 10.2" -- but the question remains "unanswered" then copy the comment including the author's name and timestamp to a Community Wiki post. – Mr.Wizard Jun 18 '15 at 17:38
  • Sample size is now three, I suppose, although I have posted as CW per @Mr.Wizard's comment above. – march Jul 1 '15 at 18:52
  • @march Thanks for the additional sample. Now 2 of 3 have become accepted answers! I still would have reaped the meaningless internet points though. – Ian Jul 1 '15 at 19:00

I am not exactly sure how to answer this. Nonetheless, I post an answer, rather than a comment, because the latter would just be going too far, wouldn't it? (I feel that I am probably one of the more guilty ones, posting answers as comments.)

As ever, I agree with Leonid, in the linked thread. I think that a major reason why people post answers as comments is that they feel their contribution (whether for lack of time, or feeling unqualified, or whatever other reason) would not do justice to the question, yet still feel moved to try to help in some small way. But, it can certainly be imagined that one person's trivial contribution might be another's perfectly decent answer, and I can definitely say for myself that, if I post a comment, I am not trying to gain any points (or real reputation) from it. (In fact, I think on this particular site, people are much less motivated by the points than they are on some other sites--although it has to be said that, as the SE community has grown, many sites seem to be the same way. Perhaps rep-farming as an occupation was something peculiar to SO?)

So, at least for my part: if you see a comment that could be an answer, please go ahead and post it as one. Claim the points for yourself by all means--you are doing the community a favor by reducing the number of unanswered questions. And if nobody else agrees that the comment was useful as an answer, you will soon know about it, although there is no harm done anyway.

I remember turning the comments of such luminaries as rm -rf into answers like two years ago when the community was in a "let's answer some of those unanswered questions" mood. The idea back then (this lasted a few days, and several users were involved in the effort) was to drive down the number of unanswered questions which, as you say, is good for the statistics if nothing else.

For older posts (and it seems that over a week satisfies the notion of old), I doubt a simple answer upgraded from a comment earning more than one upvote, so even if it feels like stealing, it's not stealing much. While CW posting is an option, I guess I would discourage it to dispel any sentiment that pinging the commenter for that upvote is greedy.

Plain and simple: turning a comment into an answer is not "stealing" in any sense of the word. I don't care who writes the answer, I just care about the answer itself. If someone chose to post the answer as a comment instead of as an answer that's his problem, and he has no right to reputation that might be awarded if the comment is turned into an answer. The license on this site is very clear in this respect; as long as you give credit where it's due you have the right to reproduce content from the Stackexchange network.

Ian translates Mr. Wizard's comment into an Answer, and comments on the original question (@Mr. Wizard I fleshed out your helpful comment into an answer.). The benevolent Mr. Wizard is notified and upvotes Ian's answer (ReputationGain -> 10).

The author of the comment might choose to vote for the answer, or he might not. Maybe he doesn't think the comment completely answers the question and that's why he wrote it as a comment. So don't expect a specific user to vote for your answer, just hope that someone else who also thinks that the answer is good happens to see it when the question gets bumped to the top page. It will probably happen more often than you think, after all an answer that is written to a two day old question is no more likely to be read than an answer that is written to a fifty day old question.

I sometimes copy the comment into a Community Wiki (CW) post. Sometimes, if I have felt I have added something significant, I might not make it CW.

I don't like posting CW answers simply because I do not get notified when they are upvoted. The importance of the upvote is that someone else has agreed with me that what I had to offer in the way of an answer was useful, and it's annoying to have to remember to keep polling the answer to see its status. (At this point, rep matters little to me, unless and until rep can be converted to bitcoin or some such thing.) I think it's important that we have useful answers on the site. And I think answers should accumulate upvotes primarily based on their usefulness. And personally I like to know which of my answers are well received and which aren't.

The point about the CW posting is that since SE has a personal reward system, even though the reward is worthless, I (probably everyone) want to avoid poaching someone else's rep. Making the answer CW means I don't accrue any rep from upvotes, and other users probably won't withhold their votes on the grounds that I'm stealing.

I have nothing against the OP's approach. It's a service to the site and its users, and deserve some sort reward or appreciation on those grounds alone. I'm just sharing what my approach has been. I have to say, implicit in my hedging above, that I'm not entirely consistent. When wanting to be notified of upvotes gets the better of me, I don't make an answer CW. Sometimes I go back and forth to the point I say, just forget it!, and don't even answer.

  • 1
    For older posts (and it seems that over a week satisfies the notion of old), I doubt a simple answer upgraded from a comment earning more than one upvote, so even if it feels like stealing, it's not stealing much. While CW posting is an option, I guess I would discourage it to dispel any sentiment that pinging the commenter for that upvote is greedy. – Ian Jun 6 '15 at 1:47
  • How does one post an answer as CW? – chris Jun 14 '15 at 13:38
  • @chris, Under the edit window on the right, there is a "community wiki" checkbox. – Michael E2 Jun 14 '15 at 14:47

You must log in to answer this question.

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