Should you accept an answer that doesn't help you?

Question

Hi,

In a couple of occasions I've accepted an answer to my questions that didn't actually help me. This is because they were correct answers and would answer the question a large section of the people who had a similar question. Also the questions were open for a period of time first.

example: SVNServ deny write access to a directory via wildcard match

The answer was correct but wouldn't work in my particular environment.

My question here is, is this an acceptable way to accept answers or does this preclude getting new answers that may be more useful to the community? I.e. is a correct incomplete answer marked as a correct more useful or less useful than the same question marked unresolved?

Related Question:Should Discussion Questions have an Accepted Answer?

Actually about if I should accept an answer in this thread.

Answer 1

The answer was correct but wouldn't work in my particular environment.

If you clearly mentioned that particular environment, then don't accept, but point out that it doesn't work for you and wait for further answers.

If you didn't mention it, the fair thing is to accept and ask a new question.

Answer 2

The mark of acceptance represents what the author believes to have been most helpful in concluding the question. So if the answer doesn't solve your issue, but you think will be more useful in a large scale, then an upvote is probably preferable. People will also read good voted answers, so if the answer is truly helpful to more people then this will be established by an eventual high vote count.

Usually, a question with an accepted answer will get most of its traffic from users who are interested in seeing the answers, not the question. So the chances of getting a new answer are lower if you already have an accepted answer. It's not absent, but it is lower, so you may consider that if you are, in fact, still looking for a solution.

Answer 3

You should always seek to make your questions as specific as possible.

The best resolution would have been if you had added the extra context immediately on discovering that it didn't work. Then you could have added a comment on the failed answer as to why it did not work.

You say that the answer would help others but it didn't help you. If you can't or haven't verified this, then an upvote is the most you should do. Your supposition about what works on a different environment is as good as anyone else's. The acceptance checkmark is for what solved the issue experience by the OP. If the answer didn't solve it, then it isn't "the" answer.

Answer 4

I would maybe accept that answer and start a new question that references the old one as not working and adds more detail. No?

Answer 5

Several times I have asked a specific question and received a technically correct answer to it (here's an example). However, I then discovered that what I was trying to do was probably not the best idea. The answer, however, was still correct. In these cases I've marked the correct answer as accepted.

If I were to have another question about the different idea I wound up using, I would post it as a new question.

It's not anyone's job to read our minds and point us to the best solution to a problem we inadequately described. The best we can do is post specific questions, and the best anyone else can do is provide specific, accurate answers.

_{(In other words, if you ask a stupid question...)}

Most importantly, in my example question above, a user may visit the site later with the same question. That user would see that I marked a technically correct answer as "accepted", and would try that solution. For all I know, that's exactly what this other user needed.