Well, I want to downvote the acceptation of an answer. How I can do that?
ps. I don't want to downvote the answer, I want to downvote the acceptation. For example in this question I think that the question owner should not mark this answer as accepted. I want to downvote the acceptation.
EDIT: Thanks to the owner of the question for kindly marking-back his question as unaccepted.
This is a very poor feature-request. You have not even explained how such a feature would benefit anyone. No justifications whatsoever...
First of all, it's always and only up to the OP which answer he decides to accept. You can't force or demand him to accepts yours or someone else's answer. It's always has been this way AFAIK and this is very reasonable. The voting system is designed to allow the community to vote on their most favorite(to some extent - see why voting is important and voting down and up) questions and answers.
To contradict your idea I can't resist to introduce you to a Populist Badge. It's been awared nearly 4,5K times which (at least to me) means that it makes sense and proves the idea that an accepted answer is not always the best answer provided.
Golden rule to remember: ACCEPTED ANSWER ≠ BEST ANSWER
If you personally disagree with an accepted answer, you can either leave a comment, or downvote the answer. You can have a go an provide your own answer - however, this approach may or may not change the way OP thinks.
I honestly do not see any benefit in an additional feature which allows people another way to express their disagreement against the accepted answer. Like I've already said, it's always down to the OP which answer has helped him the most and no one here on SO/META can change that behaviour.
I think mehow's answer explain the general situation very well, so I'm not going to revisit anything there. But given your most recent question, I think I've finally figured out what the whole purpose of this is feature request is.
You believe that the presence of an accept answer somehow changes whether the feature would get implemented or not, so if a feature request has an accepted answer that somehow says that the feature isn't necessary then it means it won't get done. That is absolutely not the case, especially when the individual is not speaking for Stack Exchange, Inc. (the employees usually have diamonds just like moderators, but their profile will indicate that they are employees). In fact there is an answer from an employee and he says nothing that suggests it won't get done (or will get done - so it is still up in the air).
An appropriate status tag (usually status-completed, status-declined, status-review, but there are others) usually indicate the intentions of SE and until then the feature is still open and subject to consideration. And even then, there have been cases where features were previously declined and have still been implemented. If you want to attract attention to a feature because you think it would be beneficial, then you have 3 options:
It is unnecessary to "downvote the acceptance" or, in most cases, to ask a new question if you just want to revisit and draw attention to an old feature question.
The reason asking a new question is usually not a good idea is you end up hurting your own cause. There are 12,000+ open feature requests, so new questions only cause posts to get lost in the mass of feature requests. Plus duplicates only serve to split the discussion and make it more difficult to assess community support, which will only make it harder to decide if it is worth pursuing a specific feature.
However, it is easy to understand why this might have been confusing, and also why some users try to discourage accepting answers to true feature requests unless there is some indication by the SE staff that the feature will or won't get done for that reason.
