What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 150 Stack Exchange communities.

While typing up my answer to this question: Why did my suggested edit get rejected?, I came to realize that suggested edits can, in some cases be technically valid, but reduce the chance of a generating "a great answer".
We've probably all seen what looks like easy/simple/basic or even silly questions on the face of things, that have produced very interesting answers, or at the very least: hilarious ones.

I've had a quick look round for clear-cut rules on when to accept/reject an edit but didn't really find a conclusive set of guidelines. What I did find were a couple of questions or laments complaining about "robo-accepting" and general carelessness. here, for example. By providing a clear list of parameters, available to everybody you send a signal that reviewing edits should be taken seriously, as we all stand to gain from it
The closest I got was this answer, though it still falls short:
If the goal of SO is to be a resource of knowledge, neatly packaged in a Q&A format, surely the chance of a question fetching an interesting answer should be taken into consideration when reviewing edits, or should it not?

Using the list from the question I linked to, I'd like to see what you might want to add/alter

When to approve?

  • If there are some spelling mistakes in code/other text.
  • If grammar is improved (but doesn't change the meaning)
  • If broken links are fixed.

When to reject?

  • If it changes main code.
  • If it adds additional code. (Though it depends sometimes.)
  • If it adds another solution.
  • If edit summary fails to explain diffs I observe.

When to skip?

  • If you are not aware of the subject (i.e. tag)
  • If you can't judge what to do (obviously!)
share|improve this question
Adding additional code is an invalid edit. It changes the context for pre existing answers. There is no sometimes. – staticx Apr 5 '13 at 12:29
@0A0D: not sure if I agree completely. if a question has been answered, and additional code/info has been given in the comments, it might be a valid/useful edit to add that info to the question for future researchers – Elias Van Ootegem Apr 5 '13 at 12:33
Then post it as answer. That's why there is a reject button for invalid edits and the description gives this very reason. Or post a comment. If people want to read it, then they will read the comment or the additional answer. It will make everything look weird if you come by later on and edit the question that way. Certain questions are geared towards community wiki, but few with code work well as a community wiki. Which, by the way, doesn't go through the review process. – staticx Apr 5 '13 at 12:42
@0A0D: good point... just a side-note: this is a copy-pasted list from one of the linked questions, not my list... but it seemed like a good place to start – Elias Van Ootegem Apr 5 '13 at 12:45
Fair enough, thanks – staticx Apr 5 '13 at 12:46
I believe that you should post 1 reason to accept or reject per post. Otherwise I have to ask myself if I should upvote that answer or downvote because I (dis)agree with some parts. - OTOH this turns this question into a list question. – Johannes Kuhn Apr 5 '13 at 13:59
@JohannesKuhn: good point... that way, this "thread" can serve as some sort of election, where each argument can be up/down voted – Elias Van Ootegem Apr 5 '13 at 14:20
I really appreciate a post like this, but can we try and consolidate the efforts into another [faq-proposed] community-wiki post about review guidelines in general here: Review Guidelines. It was created a while back, and could use the help of more users in defining and explaining exact rules for each of the Review Queues – Rachel Apr 5 '13 at 14:28
I think the biggest problem currently is that there is some disagreement if a post should be accepted or rejected. In fact, you could suggest the same edit several times and the outcome will be different. Once a common guideline has been found in this thread, I'll suggest a change in the Review Guidelines. – Johannes Kuhn Apr 5 '13 at 14:34
@JohannesKuhn: That's exactly why I posted this question. As developers, we all know the value of standards, which often come from the community. I have very few privileges on the meta pages ATM (started using only today), and I have to go now, but I will take the time to clean this post up a little and tag it correctly – Elias Van Ootegem Apr 5 '13 at 14:41

1 Answer 1

up vote 5 down vote accepted

When to reject?

  • When there are (several) other issues in the post.
  • When the edit adds wrong information.
  • When the edit adds spam or profanity.
  • When the edit makes no sense.
  • When the edit changes the meaning of the post.
  • When the edit introduces another error.
  • When the edit adds salutations, signatures or thank you notes.
  • When the edit is a comment.
  • When the edit is another question.
share|improve this answer
I disagree (only) with the first point, so +1 from me. – Tshepang Dec 13 '13 at 6:15
@Tshepang: sure, if “there are (several) other issues in the post”, then the reviewer should either edit the post to fix it or pass by, leaving it to others. Rejecting an edit on the grounds that improvements are insufficient is a disruptive behaviour. – Incnis Mrsi Sep 5 at 9:53
@IncnisMrsi I disagree with that. If I, as a reviewer, don't feel confident I can edit that post into shape but the editor clearly only went for the low hang fruit, I reject happily. We don't need editors that only care for the easy bits. Those editors are disruptive. – rene Sep 5 at 10:32
@rene I don't see how going for the low-hanging fruit disruptive. It's not like there is a shortage of reviewers. – Tshepang Sep 7 at 11:51
@Tshepang there isn't a shortage of robo-reviewers, sure. But I'm in the camp that advocates to fix everything at once, without unnecessary bumping of the post multiple times. But you can also argue that any improvement is still an improvement. I can see a benefit in that but it is not what I personally feel is the most effective way to improve quality. And their might be enough reviewers on high-traffic sites but I'm unsure about the smaller beta;'s – rene Sep 7 at 11:58
@rene as suggested by Incnis, then don't reject it. Just leave it for someone else (who will hopefully improve other things before approving). – Tshepang Sep 8 at 10:25
What if the edit removes salutations? This doesn't really benefit the question at all, it provides no material value or benefit. Removing a "thank you" isn't any more helpful than including it it seems. – leigero Nov 13 at 21:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .