We’ve talked before (inconclusively) about reactions and public votes. Allowing anybody to click a “this worked for me” button on an answer – separate from voting – would be similar and would seem to address the use cases here. Thoughts?
(If you click “this worked” and didn’t upvote you might be prompted, up to some level of experience TBD.)
I think this misses the use case of allowing someone to indicate there’s no answer that worked for them (other than the asker, who can do this by omission).
It’s an alternative allowing an expert to give a “stamp of approval” and serves to validate an individual answer, but doesn’t tell you whether anything is missing from the existing answers.
What about some sort of bounty system, to temporarily “mark” a question as missing something, possibly with an explanation of what that is (but probably not give reputation as on SE)? Maybe this would need to go through a review process to prevent abuse. Maybe this should be treated as a distinct Q&A (which exists in parallel with the old Q&A until it’s resolved), where you can see the original and the original answers, but the focus is on the explanation about what’s missing and any new answers posted in response.
Having the asker do this wouldn’t be the ideal use case (in my mind), but it can also work for that.
“Bounty…not give reputation”
So no “cost”/“payment”. Which effectively means two things:
Putting those two together, that simply means (to me):
Comments (on SE):
Of course this could be improved with a good implementation, but I don’t think it could be improved that much. And, if we do manage to improve / change it far enough, you’d probably end up with something quite similar to what I’m proposing.
I think the idea of “bumping” a question is flawed in way too many ways and that’s not what I’m proposing here.
What I’m proposing is much closer to a bounties, in the sense that:
Reputation is only a small part of what makes bounties bounties.
(I haven’t read every single post here, so apologies if I’m repeating something.)
I often would face a conundrum at Stack Exchange when accepting answers: what if there are two equally good answers? In this case, I don’t feel right accepting either answer. (And accepting one because it merely came chronologically first feels “meta” [i.e., not based on content] and therefore improper to me.)
Accepting an answer seems almost equivalent to an asker-endorsed upvote (except you can endorse multiple answers; and answers can have multiple endorsements), so maybe that’ll work.
(Bear in mind, if it’s not part of the MVP, there might be a whole bunch of old posts without accepted answers down the line.)
One thing to explicitly avoid is the Microsoft Community method whereby an “expert” can mark an answer as the answer which forever floats at the top. I’ve seen far too many answers there where the accepted answer is posted by someone who appears to be from Microsoft, yet the discussion has gone on long after that one was checked with many comments indicating that it doesn’t work at all.
Additionally, there are times when there is an excellent solution proposed, however, for “reasons” the OP cannot use that solution (for example, don’t use Excel as a database, but OP doesn’t have and cannot install any other tools). Allowing someone other than the OP to select the “correct” answer would probably lead to that being the selected one, even though it didn’t help the OP at all.
Voting on this answer should show that it’s a very good one. Hopefully, the OP would comment that it’s good but that he can’t use it, and she can then accept the answer that he was actually able to take advantage of.
On several forums I participate in, there is a request to mark topics “[Solved]” once they are. Unfortunately, this usually means editing the OP and adding text to the title. This leads to A) people forgetting to do it, and B) all sorts of different markings “[Solved]”, “(Solved)”, text at the beginning of the title, text at the end, etc. How about adding a “Solved” check box to the Question and an “Accepted” check to the answers?
Yes, if we have “acceptance”, it will not confer special status to the answer. Answers will be sorted by score.
It will confer special status - that is the nature of “accepted”. What it won’t do is affect sorting. For example, there may be queries (whether options on user search of the site or as API calls/data dump selections, etc.) for “accepted answer”.
I think there should be a sorting option which shows accepted answers first, at least. Because if I’m looking for a solution, that’s generally what I want. Another answer might be voted higher because it came in first, or on some topics because it sounds right and looks shiny.
This might be too complicated, but: what about a way for high-reputation users to collectively identify or change which answer or answer is “correct”?
All true. But also very often:
For me, if the Accepted Answer is clearly & prominently marked then it will be obvious enough for anyone who wants to zoom in on it. Unless it is 10th answer down - in which case there is probably a very good reason why it isn’t near the top.
The community should express its views through voting. We shouldn’t be in the business of deciding who is an “expert” who is allowed to change things like this.
“Acceptance” should be informative, a way for the asker to say “this worked for me”. Possibly the asker could mark multiple answers (if each contributed a piece); don’t know yet.
Public votes (if we allow that option, and it would be an option not required) would be another way for experts to annotate answers.
Whether it’s the asker or some other user making a declaration about the utility or correctness of an answer, the information is only as good as the evidence presented in the answer itself, the repute (real-world or otherwise) of the person saying so, and the reader’s own evaluation of that data.
One thing I’m struggling with is the one-dimensionality of upvotes. Consider two answers, one long and detailed with a huge amount of background and many references, and another which gets right to the chase — also accurately, and with sufficient but succinct reference.
I’d definitely vote up both — and in fact, the longer answer might get more votes in reflection of the effort put in — but I think I’d like for the succinct one to come up first.
This is why I suggested the “experts” thing that @cellio doesn’t like. 
I think in these cases there should be some way to at least remove the checkmark.
I encourage you to embrace the power of “accepted but score -20”.
LOL. Yeah, that’s a pattern I’d love for the new system to not copy. I understand what it means, but it’s objectively weird and confusing.
Probably in that clear case, the system should just remove the accepted checkbox.
This is part of why I don’t even want to have an acceptance mark. It’s one person’s opinion; so what? Voting is what really tells you which answers are better.
I understand the desire to be able to mark a question as “resolved as far as the author is concerned”. Maybe we should do that – mark the question as “I’m satisfied” but don’t tie it to any specific answers at all.
I think if we don’t have some kind of acceptance mark, we’re going to spend a lot of time reverting edits that add “[SOLVED]” to titles.
When the mark is right (which is most of the time!), it’s very useful. When I’m looking for an existing solution to a problem I’m having right now, I want to be able to quickly drill down from search results to a solution. I don’t want to open up ten tabs of threads and check each one in hopes there’s a resolution somewhere — and that includes filtering out highly-voted “well, this doesn’t really help you, but here’s some related information that could be useful”.
Right, but when you’re looking for a solution, do you want the opinion of the person who asked the question, or the opinion of the community who voted on answers?
The answer might be different for domains where answers are verifiable (did the code work?) versus not (many subjective topics).