Should questions with accepted answers be closed?

[VolkerBarth]

JAMQ (Just another meta question):

On the former SQLA site, there have been only a few questions that were closed.

As Justin has obviously cleaned-up his bunch of older questions by closing them (as "answered and accepted") - is this something that should be done generally when questions have an accepted answer?

Or is this irritating as the questions might still be of interest for other readers?

What do you think?

---

Note, I'm not refering to other reasons to close questions (offensive, off-topic and the like). I do not know whether there have ever been such cases but then closing would be appropriate ...

[justin_willey]

My theory was (and this was really just an experiment) to close the questions to show that they were complete (as question and answers) and to encourage people who had further related questions to start their own, rather than as sometimes happens on the old groups, to tack on a new question to something that maybe several years out of date. The questions still remain searchable etc, for people looking for information.

Very interested to know what people think the policy should be.

[VolkerBarth]

That seem to be senseful reasons (which is not unexpected...) and an intersting experiment. - Apparently the NNTP groups do miss a "closed" action much more:)

For the "aging" of questions, I hope the version-related tags should help to filter out topics for older versions. Though I guess it will take a long time until that aging happens - compared to the much older NNTP groups.

What I do not like about the closing is that it put all these questions to the top of the questions list. That is somewhat confusing - as is when lots of questions are re-tagged.

[justin_willey]

I think this business of what should promote a question back to the top of the list is interesting. I suggest that administrative actions - eg closing, re-tagging etc should not.

[graeme_perrow · Accepted solution]

In my opinion (note that this is only my opinion and not necessarily that of Sybase), most questions should be left open, even after an answer has been accepted. In some cases, the answer could change in future versions and we want to make it easy to modify the answer.

At least one of the questions Justin closed is answered by "This is a bug and it's fixed in version XXX", and that answer will not change. If someone else sees the same problem at some point in the future (and has the bug fix applied), which would we rather them do:

1. add a comment to the existing question saying "I'm seeing this behaviour but I have the bug fix" or
2. ask a brand new question?

I'm not sure of the answer, but we're not likely to be getting StackOverflow numbers here, so it's not really going to come up all that often. Also, once we fix a bug, it never comes back so it's all moot anyway. 🙂

Questions that are sqla-specific can be closed.

[Breck_Carter · Accepted solution]

I agree with Graeme... I don't care if questions stay open, and the very act of closing a question draws fresh (and unwarranted) attention to the question IMO.

[justin_willey · Accepted solution]

Might that though be an argument for admin activities not promoting questions rather than not closing questions? Not sure.

[VolkerBarth · Accepted solution]

I agree that one should be able to influence whether an edit operation puts the question on top of the "active" list.

As long as such a feature is not available, I think closing questions seems not too necessary. And generally, I guess there's a much greater need to mark questions as accepted... (Not for questions asked by the members of this discussion, methinks.)