Making a significant improvement to an answer already accepted or upvoted, where to put update? [duplicate]

Question

This would be a person making an edit to that's person's own prior answer that makes a significant improvement. Say it's an algorithm improvement that reduces time complexity (such as from O(log(n)) to O(1)), but that the answer was already accepted and/or upvoted.

I'm thinking that for significant improvement, for the benefit of others reading the asnwer, the improvement should be inserted at the front of the answer, noting the change and using a separator to retain the original answer after the inserted improvement.

---

I'm updating this question with a specific example, an answer about a Visual Studio 2015 change to std::list::sort. The improvement in this case was a switch from directly accessing nodes (std::list::splice does this), to using iterators. Rather than replace the prior answer, the updated answer has the improved version at the start of the answer, with the prior answer at the end. I left the prior answer there because of comments. However, the prior answer could be considered obsolete, and I'm wondering if I should just remove it, and leave a brief description of the prior answer so that the comments would still make sense.

`std::list<>::sort()` - why the sudden switch to top-down strategy?

Answer 1

In order to make "a significant improvement to an answer [which has] already [been] accepted or upvoted" I think the most direct thing to do is to edit, or suggest an edit, to the answer.

For all edits to posts, even if that edit is to your own post, I think they should as much as possible blend into the prior content rather than being kept apart from it.

Stack Exchange users vary greatly in the degree of ownership they assert on their questions. I do not rollback even the most trivial of edits to my own questions if I think it makes an improvement to it, or even if I only think that it does not degrade the question.

However, be prepared for the asker either not seeing your edit as an improvement, or them being someone who routinely rolls back all or most edits to their questions. If they roll your edit back then it is best to just accept that you may be wrong, or that "you may be able to lead a horse to water, but you cannot make it drink".

In that case, I think your indirect options for trying to improve the question include:

• Making a comment to describe the improvement and hope the asker does that themselves. This can also be done before trying to edit directly. I think whether a direct edit will be allowed to stand by the asker will often depend on your relative reputations, even though it should only be the content quality that influences that.
• Adding a new answer that links to the one you are considering editing and saying you endorse that content but want to add something extra of your own to improve/qualify it.