I am seeing a great deal of suggested edits that extend non-CW answers in a significant way. I believe they are good faith attempts to "improve" an answer by filling in details, but the resulting answer may depart significantly from the original answerer's intent (especially if the question appears to be homework and the answerer intentionally left some details to the reader). In any case, the sudden switch in narrative voice is usually apparent, and the original author surely would not appreciate such an alteration of his/her answer.

My question is two-fold:

  1. Should such edits generally be rejected? My guess is that they should, but I would like to hear others' opinions.
  2. If so, could some kind of brief guideline be provided on the edit screen regarding acceptable edits? I think this will reduce the number of such radical suggested edits. As I said, it is primarily coming from new users that (I think) simply misunderstand the modus operandi of the site.
share
10  
I vote "yes" on 1. –  Gerry Myerson Dec 9 '12 at 23:00
2  
I am happy that you raised the question. I have cravenly passed on rejecting/accepting, because of uncertainty about policy. Would be most comfortable with a policy of mostly reject. –  André Nicolas Dec 9 '12 at 23:55
1  
I recall feeling uncomfortable once when someone make such an edit to one of my posts, then feeling relieved when I discovered the edit was reviewed and rejected. –  Hurkyl Dec 10 '12 at 0:08
    
This happened to me once... math.stackexchange.com/posts/231501/revisions –  The Chaz 2.0 Dec 10 '12 at 6:19
2  
I routinely reject them, with a note that substantive changes should be made only by the OP or at least with the OP’s explicit approval, and that the appropriate action is to leave a comment. Added: And now I see that I’m essentially repeating Asaf’s answer. –  Brian M. Scott Dec 11 '12 at 8:13
2  
Here's a question I have; in the case where such edits are unwelcome, is it better to post your own answer in parallel? My concern with doing that for some questions was that I was repeating content. –  Joe Z. Dec 18 '12 at 2:38
    
Unless the edits are of a typographical or stylistic nature they should be generally be rejected (with short explanation). –  copper.hat Dec 21 '12 at 23:22
add comment

2 Answers

up vote 10 down vote accepted

I think we should reject edits which change too much of the intention of the original author using the option radical change. There is already a description on the reject screen describing what is considered as such. As with all things in life, this is subject to individual interpretation.

It seems that many of these radical edits come from unregistered users who probably don't know about how the site functions. It is a pity that their contributions are lost in this way, but then again this problem doesn't happen on too large a scale.

There is already a guideline provided on how to edit, saying that one should always respect the intention of the original author. However, this appears on the right side of the screen not directly above the post (you can see for yourself if you try edit something while not logged in), so I would like to suggest to our SE masters that this section gets moved to a spot directly above the post to make it more visible.

share
add comment

If you are uncomfortable with "too radical" you can use the custom message and give a short explanation that such changes should be made by the OP, and best pointed out as a comment or another answer.

You may also post a comment yourself in case the editing user cannot comment themselves.

However I find that edits that change any content of the post are to be made either by the OP or by someone which pointed out a mistake and that was confirmed by the OP.

share
    
I see my question was a little misleading. It really should say simply "Should such edits be rejected?". You address my intended question in your last paragraph. –  Austin Mohr Dec 9 '12 at 23:12
add comment

You must log in to answer this question.

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