I sometimes run into a situation where I realize my own question is a dupe, or the problem goes away, and I need to vote to close as too localized, or with another close reason. It it proper to closevote my question, or should I flag, wait, or do something else?
There's certainly no question about propriety. If your question should be closed, it probably should be closed. But there are a few options:
Delete. If you completely messed up and your question was a problem caused by you, and it's never going to be useful to anybody else, you could nuke it.
Answer it yourself. I've actually had pebcak issues that I've solved, posted the answer and have since turned out to be very common issues. Posting the question and answer helped other people.
Accept what would have been the most useful answer. Sometimes I'll do something very wrong, float a question out. I'll get a few valuable (but incorrect) answers back before I notice that I accidentally moved
An answer of
sudo mv /bananahammock /etcis not going to help anybody else because few people are as stupid as me. So then I look at the answers that were there. People suggesting good diagnostic issues that could help the next person in a similar situation.
It's not just a case of clicking accept. If you pick a useless answer you might lead future visitors up the wrong path completely.
Or vote to close it, as you suggest. This is preferable to deletion in cases where it's still valuable to have the content around. One such way is where it's a duplicate. By close-duping your question, you make the master question easier to find for people in the future.
You just have to think what's most useful to other people. What option delivers most value?
It's up to you, but I would recommend only close-voting it if you would do the same had the question been asked by anyone else.
I realize my own question is a dupe
Mark is as a dupe like you would anyone else's question who you realise is a dupe.
the problem goes away
These questions should NOT be closed. It's unfair to others who may experience the problem.