What is the etiquette for providing an answer already in the comments?

Question

Most of the times when I see a question that I (think! I) can answer, I leave a comment first checking if that solution is really helpful to the OP, or not. This is to prevent providing answers that are misguided/ just plain wrong, because of a misunderstanding of the question etc. I assume others are practising the same. That is, only providing answers if they are absolutely sure that it answers the OP's original question.

Suppose someone else leaves a comment that I believe solves the problem as well. What's the etiquette to follow here? I feel like if I leave an answer to the question, then it's like "stealing" from the original commenter/answerer.

An example: how to center equations in parentheses (disclaimer: I am in no way targeting any one in that question, but it does exemplify what I'm trying to say here. This is not about rep as well.)

There are a few things an incoming I-know-the-answer-but-someone-else-has-provided-it-in-the-comments person can do:

• Leave a comment for the commenter to add the comment as an answer
• Just answer the question because it doesn't matter who makes it, as long as the answer answers the question

With the second option, how long should we wait before actually providing an answer (going back to the point about "stealing")? Is this just something really minor and that I am overthinking (or maybe over-feeling)?

Ps. The title is a little clumsy. If anyone has a more concise way of phrasing it, feel free to edit it.

Answer 1

I will express my opinion as an answer instead of a comment do to its length, but please mind that mine is not the “correct” etiquette, simply my opinion.

I often find myself commenting instead of answering for many reasons: an answer usually has to be written completely, compiled, tested… maybe I don't have the time to compile the code, make sure it works, pretty print it, etc… maybe I don't have the answer but some pointers or suggestions, which nevertheless may prove decisive to solve the original problem. I usually write my comments as answers only when requested.

I usually think that stealing an answer from a comment is kind of wrong, unless one takes the care of expanding the comment (after making sure that the original commentator has no interest/time of doing that himself). After all it's not my idea, and maybe the original comment-poster has some code going already (so I would probably waste time).

Nevertheless, one should always strive, in my opinion, to form a complete and thorough answer whenever possible, not only for the benefit of the person who asked the question, but for making the question/answer pair more easily searchable in the future. A comment is not that visible.

On another note, sometimes it can also happen that, while creating a thorough (or at least tested) answer (which can take some minutes: creating the code, explaining it, compiling, snipping the pics to be uploaded) someone else answers it better or summarizes all you want to say in a poignant comment. Well, it happens! (And it happened in the question you linked)

Answer 2

Further opinion:

A fairly new user (or new to answering) should be encouraged to write the answer themself. They could use the rep (of course they've got enough to comment if we're asking this, but it could be a self-answer).

An old hand can be assumed to know what they're doing. By providing a quick fix in the comments they sort out the OP's problem and save the effort of writing the rather thorough answers that are common here. I do this myself, especially if I'm on the train (patchy mobile signal). I can only speak for myself at this point, but if someone comes along and fleshes out my comment into an answer -- great, the system is working. A mention is nice of course. Rep is nice, but I doubt it's why many of us are here -- in fact some are still active on comp.text.tex (I might be if I had decent usenet access) where rep isn't measured.

Best case: a fairly new user asks a question that turns out to have a simple answer (but of course not a duplicate), an answer appears in the comments, and the OP writes their own answer.

If I throw down a quick comment ("try \command from package") and I intend to write an answer immediately, I'll normally say so in the comment: something like "I think xxx should work, I'll have a play". On the other hand I've also been known to comment things like "xxx might do the trick. I'll write an answer if I get the chance later, but someone else might get there first" as a hint that I'm not bothered if I'm beaten to it (and will quite likely forget to come back in any reasonable time).

I tend only to answer when someone else has commented if (i) there's still a bit of work to do (this could be a very snmall but IME important step), (ii) some time has elapsed, and (iii) I'm interested in learning something from the attempt. As I'm not the quickest at writing an answer, especially if it's going to take a bit of debugging, I'm unlikely to get there before anyone else who starts ahead of me.

I think one factor behind the thorough answers I mentioned is that we're populated by people who not only are helpful, but come from a background involving academic or technical writing and think about their writing, leading to a reluctance to commit to the permanence of an answer if it's not done properly.