@YuiToCheng If a question is poorly stated, we are encouraged to improve it. I see no reason not to improve the "initial question" to make it a better question. That would seem to benefit the site. However, the improvements to the question were removed and now it is closed and on the path to deletion.

The question, as amended to include an asymptotic expansion, is more interesting than the question whose answer is 0. Why revert it to the original question? That seems to concentrate on the OP, not the question itself.

Thinking about that question lead me to make some interesting discoveries, and I think allowed me to write a good answer.

@robjohn I have no problem with your amendment. However, I would rather support a newly asked question, that better models a good question, which you can copy and paste your earlier answer into. That will be a great example of self-answering!

@robjohn Please show me evidence that the asker gave in comments the integral obtained via the substitution. Otherwise it deserves to be deleted in my opinion (and so the closure reason is irrelevant to me). Thanks.

@user21820 so you're saying that if the improvement does not come from the OP, then the question deserves to be closed/deleted? There were many times when I claimed a question was interesting, but there was simply too little context, and I was told to add context (I believe it was by amWhy, but that is not important). That is what I've tried to do in several cases recently, all of which have met the same response: the OP didn't improve the question, so it doesn't count.

@robjohn You claimed "the OP discussed their attempts in comments". So I ask for evidence. If you can't provide evidence, then please don't make such a claim.

@user21820 They mentioned the substitution, but they didn't write out the integral. However, that is beside the point of my comment: if the improvement doesn't come from the OP, it doesn't count.

I am not trying to get that question undeleted necessarily, but to get a handle on how things are perceived here.

@robjohn I will tell you my viewpoint, but you'd have to ask the others for theirs, because some have stricter criteria and some have looser criteria. I firmly believe in not putting words into anyone's mouth. This excludes correcting typos/grammar or obvious careless errors in answers, but includes the kind of edit that you made.

In this specific example, how do we know that the asker didn't just copy a hint that said "use substitution t=ln(x)"? We don't. Likewise we have no evidence that the asker actually tried the substitution and did it correctly, since there is no working and not even the result of substitution.

However, I make no distinction between context in the question written by the asker and context in comments by the asker, though I would advise the asker to edit the question instead of just writing comments. Sometimes I edit for them to transfer their comments verbatim.

At the same time, this issue of putting-words-in-mouth is orthogonal to the issue of whether a question is bad. I want to see clear evidence of genuine effort by the asker or the motivation behind the question, and that for me is enough for a question not to be closed. In the absence of both, I think it is a bad question and should be closed.

Furthermore, if a bad question does not have any intrinsic mathematical interest or does not have an excellent answer like this, then I think it should be deleted.

@user21820 Thank you for your viewpoint. I think that a change to a question that improves the site, whether added by the OP or someone else, is valid. It's more about the quality of the site than a few points that might b gotten by the OP.

@robjohn My goal is not infinitesimal improvements to site quality on the scale of 1 post, but global improvement to the average site quality, and in my opinion that can only be achieved by setting a proper standard for questions and answers.

@user21820 So a good answer is grounds for keeping a questino?

@robjohn No. Only excellent answers on the level of the one I linked to, and the question must have intrinsic mathematical interest.

Bad question ∧ ( Uninteresting ∨ No excellent answer ) ⇒ Should be deleted.

@user21820 "excellent" vs "good" is a judgement call.

Yes of course, and so is "interesting". Everything ultimately is a matter of opinion.

depends on your interests, etc.

I think that this question is of intrinsic interest because it raises the question in my mind about the asymptotic expansion. The first term is simple, the further terms are difficult and I am very happy with my answer (whether it is "excellent" is a judgement call).

I could move my answer to the duplicate source on that one.

I had not seen the original before I answered

@robjohn That's exactly what I've always suggested; move answers worth keeping to good questions, and delete the bad questions. And of course, you can do that easily here since you're a moderator.

@user21820 being a moderator doesn't help. I still need to copy the answer by copy/paste. I can delete the question in one vote, but I don't like to do that except in extreme cases.

@robjohn Oh I didn't know moderators can't move answers, but anyway it's still merely a copy-paste, and the question itself can be deleted by 3 users.

However, the duplicate source is not really a better question.

just more answers.

The thing that got my interest was the asymptotic expansion.

and I am really happy with how that worked out.

@robjohn I agree that it's not a good question, but (in my opinion) it's better than the later one. As for interest in the asymptotic expansion, I'm not sure how it's relevant to this discussion. I also like asymptotic expansions, but my interests are a separate issue from site quality.

"Interesting" was a quality you were looking for above ;)

When asking about limits, the next question is if there is an expansion, that is how it is relevant.

How does it approach the limit

I know that, but to me the question at hand has no intrinsic interest. There are tons of limits I can solve via asymptotic expansion, and only a small handful of them are intrinsically interesting.

In that question, it was that getting the asymptotic expansion is not via one of the usual methods. A new method was needed.

That made it intrinsically interesting to me

I didn't see immediately how to get the expansion

And we seem to disagree on what is interesting in this case

@robjohn Not at all. It's a standard technique to cut it up into pieces and apply different bounds. It is obvious that the sum over k∈ω(1) is ο(2^n), so we just need to bound the sum over k<f(n) for some f such that f(n)∈ω(1) with f growing sufficiently slowly. That's easy.

To get the limit is easy. I did that just looking at the question on the front page. The asymptotic expansion is not so easy.

So you're just supporting my point that the question is bad and you should simply have posted your question about the asymptotic behaviour of the sum with your answer, instead of answering the bad question.

It's your question after all, not that asker's.

There's a answer-my-own-question checkbox on the ask-question page, for precisely that purpose...

certainly that goes way beyond answering that the limit is 0, but what good answers just stop at answering the question asked?

@user21820 Is there? I have only asked one question and I didn't notice.

math.stackexchange.com/questions/ask

It may have been added after that, too.

> Answer your own question – share your knowledge, Q&A-style

I knew we promoted answering one's own question, but not about the checkbox.

Now you know. =)

Cool

What exactly does that checkbox do? Does it alter the question in any way?

@robjohn It makes both question and answer appear simultaneously with the same time stamp. Other than that, there is no difference from posting question first and answer later.

@robjohn Moderators can merge questions, can't they? (That also moves all answers to the target question.) Or did I misunderstood what @user21820 suggested there...?

Oh, I see. It's copying answer to a different question (which is not duplicate - so merging doesn't make sense in that specific case).

@MartinSleziak yes, we can merge questions, which essentially moves all answers. In this case, that is just mine, but in general it is cut and paste to move one answer (and that changes the author if the author is not doing the cut and paste).

@MartinSleziak It is marked a duplicate.

or am I missing what you're saying?

@robjohn @MartinSleziak: Just to be clear, I'm not at all against merging to preserve good answers, but I disagree with leaving the bad question locked. If a question is bad, then it should be deleted, and especially so after all the answers have been moved away. That's all I've to say about merging.

@robjohn I meant that the question "find the limit" is not a duplicate of the (hypothetical) self-answered question "find asymptotic expansion".

@user21820 is one of these questions locked?

@robjohn No, but merging will automatically lock. I never understood the reason for that.

@MartinSleziak Ah... It is no problem to cut and paste :-)

Oh, I did not know that merging automatically locks the question.

@MartinSleziak Another moderator told me that, can't remember who.

@user21820 I've merged a couple of questions recently. I'll check.

@robjohn You could try on the one we're talking about and unlock the question after merging, so that normal users can delete it.

Here is some list of merged questions, if it helps: data.stackexchange.com/math/query/966538/…

When I look at sources, they seem to be locked: math.stackexchange.com/posts/3163617/revisions math.stackexchange.com/posts/3162716/revisions math.stackexchange.com/posts/3238960/revisions

this question was merged into another. It was locked, but not the destination question.

> A merged question is a question where a moderator has chosen to move its answers to another question, because it is an exact duplicate of another and has good answers that would be useful on the target. Merged questions are also locked to prevent edits.

From this post (faq-proposed): What is a “merged” question?

@MartinSleziak Yes, I understand that lock. It prevents more answers being added. I thought it was being claimed that the destination was locked.

This might also be relevant for links from other sites or even from this site:

> If the merge stub gets deleted, it will redirect to the target post for users who don't have enough reputation to see deleted posts.

Anyway, if the "source" question was bad enough, it is more likely to end up with negative score and thus being cleaned-up by roomba.

@MartinSleziak merges are usually for duplicates. They may not be bad questions.

Sorry, what I wrote is incorrect.

> As merged questions are locked, they are exempt from the Roomba autodeletion criteria as those exclude locked questions (except those locked as migration stubs or rejected migrations). If you believe a merge stub is no longer useful, flag it for moderator attention and ask for it to be deleted.

It seems that there are a few things about merging I wasn't aware of.

@robjohn I am pretty sure I have seen some mergers done in order to "preserve" good answer. (In fact, in some cases the "source" question for the merge was already deleted; or at least went through some delete/undelete skirmish.)

In the case I mention above, the question was already deleted before I migrated the answers (resurrected, as it were).

@robjohn I said "I disagree with leaving the bad question locked" if it is merged with a good question to preserve good answers to a bad question. This circumvents the means by which normal users can clean up lousy questions, and is the only issue I have with merging. If mods always delete/unlock after merging a bad question, there is no issue.

Otherwise it's unilaterally preserving bad content.

@user21820 You won't see one unlocked because then an answer might be added and ruin the tidying that was just done. I think they can still be flagged.

@MartinSleziak yes, the purpose of merging is to preserve the answers. The good and the bad are saved, though the bad can still be deleted.

@robjohn Then why not delete? My point is that if a question ought to be deleted, then merging and locking would be a unilateral preservation of bad content. Flagging doesn't solve the problem, because a moderator still has to do the unilateral deletion after a post has been locked. Normal users cannot vote in any way on a locked post, nor even post a comment.

@user21820 when we merge a claimed duplicate, we are not judging the question. I guess the stub is retained for links to point to.

We can delete them if they are flagged for deletion.

the links will be forwarded, but the new destination may be confusing

For example, a link may refer to a poorly stated question, then the forwarded user will be sent to the merge destination, which is hopefully not so bad.

Ok. Sorry I got to go soon, so I'll just leave the usual cleanup lists.

C1, C2, C3.

C4, C5, C6.

C7, C8, C9.

D1, D2, D3.

D4, D5, D6.

D7, D8, D9.

D10, D11, D12.

D13, D14, D15.