What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 130 Stack Exchange communities.

One edit I did in a Stack Overflow question was approved by two reviewers and rejected by a third, and this actually ended up in my change being rolled back.

I don't think my edits should be automatically accepted, and I understand that my point of view may be different than that of other people (and I've also read other Metas and blog posts on this matter), but I can't understand the reason of the rejection/rollback.

The reviewer that rejected the change motivated the action with the following reason:

This edit changes too much in the original post; the original meaning or intent of the post would be lost.

Which generally speaking is a fine reason, but in the specific case I feel it's wrong:

  • The original answerer said that, as far as he knew, there was no way to access the Control inside the DataGridViewCell besides inside the DataGridView.EditingControlShowing event handler: in my edit I deleted this statement, and added sample code that proves that it is possible (and how) to always access the Control.
  • The answer still answers the original question, and it doesn't change the original meaning and scope of the answer
  • The code sample that I added also has the benefit of explaining how to exploit what the first sentence of the answer said, which was missing in the original answer.

That being said, it is true that my edit doubles the length of the answer (which may be seen as changing much of it), but it also true that the edit improves, completes and makes the original answer easier to understand and apply in actual code.

What should I do this time and in the future in a similar situation?

Disclosure/note: I also happen to have another questionable rollback being discussed in another Meta Question of mine, but it covers a different scenario, that's why I posted this question.

In the other case I also ended up making a long discussion in the answer's comment section, which is not really useful and this time I ended up posting directly my concern here on Meta Stack Overflow.

share|improve this question
2  
Note that the rollback was done by one of the reviewers who approved your edit originally. Apart from that, your edit did change the meaning of the answer, so I see how one would reject it as a "radical change". –  Daniel Fischer Jun 13 '12 at 17:30
3  
Why didn't you just post your own answer, instead of trying to make major changes to someone else's? –  Robert Harvey Jun 13 '12 at 17:33
    
@DanielFischer: I noticed it was rejected by one of those that originally approved it, and I thought he did it because he thought that the reject reason was actually better than the one he had to accept. I don't find anything strange abouth this! –  Fulvio Jun 13 '12 at 17:47
    
@RobertHarvey: this is the point, I didn't post a new answer because it would be made up of the 90% of the accepted answer + my improvements..and this seems a good case were editing is better than adding a new answer to me. –  Fulvio Jun 13 '12 at 17:48
1  
No. Your entire edit could have stood alone as a complete answer, without copying any of the accepted answer's text. –  Robert Harvey Jun 13 '12 at 17:50
    
@RobertHarvey: ok I have got your point, you say "you could simply post the additional information" while my approach was "I could make the original answer more complete" (which I get is subjective and not so objective as I thought). –  Fulvio Jun 13 '12 at 17:58
1  
While questions and answers definitely belong to the community, we still defer to the original author, to a certain extent. –  Robert Harvey Jun 13 '12 at 18:06
    
@RobertHarvey: ok, I can understand this. I think I'll post another answer in the end! ;) –  Fulvio Jun 13 '12 at 18:14

1 Answer 1

up vote 13 down vote accepted

The threshold between editing someone's answer to improve it and posting your own answer is somewhat subjective, so it isn't surprising that different reviewers have different opinions. Here's my analysis of your edit:

  • The original answer is well-written, with both English and a short code sample.
  • The original answer is accepted, so it has been useful to someone.
  • There are no comments under the answer pointing to a flaw that the suggested edit might be correcting.
  • The edit would approximately double the length of the answer.
  • The edit has correct English and formatting.
  • The edit concerns a special case: when a view or cell is read-only.
  • The edit is substantial enough that it could be posted as an answer on its own (addressing the read-only case).

I don't know the subject matter. In principle, this should mean that I should pass on the suggestion. Unfortunately, the suggested edit review queue is short-staffed, and the list cannot be sorted by tag, so the few of us who hang out there sometimes have to take a decision without subject knowledge. If I had answers to the following, it would influence my decision:

  • Is the special case that the suggestion addresses common?
  • Are the additional code and accompanying explanation correct?

What clinches it for me is the edit is substantial enough that it could be posted as an answer on its own. So I would reject it, as radical change or invalid edit or with a custom message “please post this as a separate answer”.

If I knew the answers to the two questions above to be yes, I might approve the suggestion. Might, note: it would be borderline even then.


As a more general note, please keep in mind that reviewers might not be fully familiar with the subject of the post. They often wouldn't know enough to answer the question, and sometimes they won't even know the programming language that the code is written in. It helps if your edit comment explains what you did to an untrained audience. If you do more than fix spelling, formatting or typos, you should leave a comment for the original author, too.

Don't be shy of correcting factual errors (as long as you aren't changing so much of the post that you should instead be downvoting it and posting a separate correct answer), but do take care to make it clear why your changes are correct and necessary (in the edit comment or in a comment on the post: never add things like “edit by X: …” to the post).

share|improve this answer
1  
What clinches it for me is the edit is substantial enough that it could be posted as an answer on its own. So I would reject it, as radical change or invalid edit or with a custom message “please post this as a separate answer”. Agreed. –  Matt Jun 13 '12 at 17:35
4  
Worth noting: In general, moderator activities (including suggested editing) should not require substantial technical knowledge of the subject matter. This is why edits that change the meaning of a post are frowned upon; such edits make the assumption that the editor knows better than the poster what the intent is. –  Robert Harvey Jun 13 '12 at 17:39
2  
@RobertHarvey Suggested edits do require some subject knowledge. For example fixing a syntax error or the name of a library function in someone else's code is a good thing, and as a reviewer it requires that I know the language (or look it up). –  Gilles Jun 13 '12 at 17:43
    
@Gilles: Hence the word "substantial." Perhaps a better way to say it is "Seldom should suggested edits require expert technical knowledge in the subject matter, since the approver would also have to have the same level of knowledge." –  Robert Harvey Jun 13 '12 at 17:46
    
@Gilles: thank you for the complete overview, it is now clearer there are many more reasons for a rejection than I could think of, mainly because I dont know the workflow behind the approval/rejection. There is still something that seems to be out of place anyway: in the edit comment I said that the added text and code corrected a wrong part of the answer which said that something was not possible, when it actually is. It may be difficult to judge if this is right if you don't know the matter, but I didn't forget to state that the answer contained an inaccuracy and the edit fixed it. –  Fulvio Jun 13 '12 at 18:01
    
It may not be a good choice to have approvers wielding the approvalstick without understanding the answers- not the first time I've seen that and it discourages. And the edit here should have been approved, because it is the correct answer to the question. covering special cases is VITAL in giving an answer, otherwise you can imagine the programmer trying to find the problem elsewhere for hours because he believes it can't be there. –  Morg. May 24 '13 at 9:21

You must log in to answer this question.

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