I've seen numerous revisions (e.g., this) where the editor is attempting to apply a fix to the code. Oftentimes, I don't know whether the fix is appropriate or not. In these cases, my potential behavior is:

  1. Accept, and assume that if it's wrong someone else will correct it back
  2. Reject, as "invalid attempt to comment"... essentially assume the ensuing discussion will result in the OP fixing it
  3. Ignore and leave for another reviewer, because I don't know enough about the topic

What's the appropriate action?


If the answer was wrong, and the edit makes it (more) right, then approve the edit. This is exactly why suggested edits were introduce: to improve good-but-not-perfect answers.

If the answer was completely wrong and correcting it is impossible without completely rewriting it, don't edit, but instead downvote (and, preferably, post a comment to explain the error in the answer if there isn't one already) and post a competing answer.

If the edit is a matter of style or recommendation rather than a matter of correctness, then reject it. If you think the issue is important, comment on it. Generally speaking, don't correct an answer that's not actually wrong.

Don't edit code in questions unless you're absolutely sure that the error in the code is not part of the confusion. That's pretty much never. The exception is whitespace (especially indentation).

In this particular case, I don't know the subject matter, so I can't tell whether the original answer was wrong (in which case the edit should be accepted) or merely suboptimal (in which case the edit should be rejected).

  • I deleted my answer, sorry; I was confusing how moderators should review flags for moderator attention with how edit reviewers should review suggested edits and mistakenly assumed that the standards of review are the same in this respect.
    – Peter O.
    Nov 3 '11 at 21:26
  • I'm accepting this one, but both posted answers here are correct... yours deals with the question as it pertains to answers, and the answer from @meager as it pertains to questions. I didn't think to separate the two questions; my fault. Thanks to both of you!
    – eykanal
    Nov 4 '11 at 4:54

In the case of an edit to a question, reject it. The entire format of SO is geared towards fixed code being posted as an answer, not edited into the question. Fixing the code in the question renders the entire question/answer and voting process pretty moot.

In the rare case where you can say for certain that the typo was introduced while pasting the code into the SO editor, then maybe a fix should be accepted, but for the vast majority of cases you shouldn't make that assumption.

  • 1
    Your answer doesn't make sense. That was an edit to an answer, not to a question. Nov 3 '11 at 21:15
  • 1
    But it would make sense if the edit was to a question; it's a good answer, but not to the question that was asked. (Hmm, should I edit it?) Nov 3 '11 at 21:50
  • 3
    So it is, I took "original post" to mean the question itself.
    – meagar
    Nov 3 '11 at 23:11

You must log in to answer this question.

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