I haven't had an edit rejected before. Today i made a change to fix an answer on which people were commenting that it did not work for them. I tried out the answer, it was indeed broken, then fixed the answer, and it works.

In my edit I made a code change, to fix an answer.

What's stange is that the change was rejected by users who seem to have little experience with python or pip which is what the edit is about. I hope i'm not going too far with this, but is it really the case that questions are being Accepted and Rejected based on other's Rejects, or at random?

  • 2
    It looks as a fix in code which in general people stay away from and let it to the user to the poster of the answer to fix... – rene Jan 16 '14 at 12:49
  • 4
    You fundamentally changed code. While there are different views on this, plenty of users consider that a nono. That will be the reason to reject. – Bart Jan 16 '14 at 12:49
  • 2
    It its code, and they have no experience on it, can they not skip it? I have never reviewed myself, are they forced to Accept or Reject? Should there be a warning for editors if code-only edits are considered bad? – Philipp Gayret Jan 16 '14 at 12:51
  • there are more question and discussion on this but this give you an idea: meta.stackexchange.com/questions/211601/… – rene Jan 16 '14 at 12:52
  • Why not post your own competing answer? – Richard Tingle Jan 16 '14 at 12:57
  • 1
    It is a one line fix to an existing answer, the question already has a ton of answers. It is not different in any way and just fixes the answer. – Philipp Gayret Jan 16 '14 at 12:58
  • 1
    @AlienArrays that is a question on code style and best practices – Philipp Gayret Jan 16 '14 at 13:01
  • @Philipp just an FYI - all reviewers see is the page you linked for your edit, nothing more, so any comments under the post are not shown. Reviewers can click through to see the original post and all of the comments, but most don't and the ones that do usually do not end up completing the review for other reviewers – psubsee2003 Jan 16 '14 at 13:15
  • What's with the driveby downvotes..? I understand the edit comment could have been prettier, an answer here is already in place. Please leave a comment ( or edit the question ;-) ), just -1ing won't make improvements – Philipp Gayret Jan 16 '14 at 13:19
  • 1
    @Philipp voting on meta is different in some respects, so it can be very different than Stack Overflow (but it can also be the same). Most likely, the downvotes are because people disagree with your edit or it could also be due to lack of research as this is a heavily discussed topic. – psubsee2003 Jan 16 '14 at 13:25

Editing code in answers1, except to fix obvious typos, is discouraged by a portion of the community. This premise is based on the idea that you are changing someone's work to something that it isn't, and that person, not you, will feel the impact (in terms of downvotes) if your edit is not correct. Because of this, many reviewers will reject almost all edits that change the code of an answer regardless of the reason, whether they are not familiar with the topic or not.

The flip side is there is a sizable minority that thinks this practice is acceptable and proper.

So you can say that it is, at best, a disputed practice.

The best practice in most of these cases unless you want to deal with the likely potential of rejected edits is to leave a comment and let the original author make the change, or adding your own answer. If the code isn't working, then feel free to downvote because a non-working answer is probably not helpful.

1 - Editing code in questions is always discouraged, except to fix spacing in non-whitespace sensitive languages.

But if you want to continue fixing code in answers because you think it is better, then read on...

For this specific edit, you could have done yourself a favor - you could have made the edit description more clear:

Added easy.install as dep, this worked for me

Ok, so what, it worked for you, but that doesn't say why you did it.

You should have expanded it slightly to say something like:

Per the comments, this solution is not working for many users. Edited so that it is correct.

When a reviewer sees this, they will understand why you did it and why you thought it was necessary to make the edit.

It still may result in the edit getting rejected because of the disputed practice of editing code in answers, but it stands a much better chance of getting accepted.

DISCLAIMER: I have no idea if your edit is correct or not, I'm not familiar, so I am assuming your edit is correct

| improve this answer | |
  • 1
    I've followed your advice and the edit was accepted. :-) Thanks, I do think a warning should be added for code-only edits, as there seem to be a lot of questions on meta about this issue – Philipp Gayret Jan 16 '14 at 13:20
  • 1
    @Philipp I see your point, but it is difficult for an algorithm to differatiate a minor typo (like missing a ; at the end of a line) to a more major change – Richard Tingle Jan 16 '14 at 13:22
  • @Philipp I feel your frustration with that, but a warning would only reinforce that idea that it is unacceptable, and as I mentioned, it is a dispusted practice. There are very high profile people who think it is ok, including one of the site's founders. – psubsee2003 Jan 16 '14 at 13:24

Let me give you an analogy. Imagine you run a cleaning service. You come to a house and tidy, sweep, vacuum, mop, etc. And then the neighbors can stop by and rate your work. One day, you move the couch into another room. Or perhaps you replace it with a brand new couch.

When the neighbors give you a poor rating for doing that, you ask "what do you know about interior design? And several people said it was not working for them. It is clearly better this way." And they say "You are not supposed to move the furniture or add new furniture. Just to clean up." That's what's happening here. The reason we don't need tag expertise is that suggested edits (which need review) are supposed to be no more than tidying up.

If you can improve the answer with a line of code, add an answer of your own that does so. Or comment on the flawed one. But don't suggest an edit that improves the content or code of an answer beyond tidying up.

| improve this answer | |
  • 1
    I'll add a 19th answer instead next time. – Philipp Gayret Jan 16 '14 at 13:38

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