I submitted my first suggested edit of an answer to a question I asked on Stack Overflow today.

It was rejected

After reviewing several other posts about rejected edits, I generally understand the reason for rejection but disagree with it and see room for improvement with the review process.

In this instance I was adding what the OP of the answer said himself in comments. These comments were made in response to an edit of my question which made it more precise, and these comments made the answer better.

Some say that the comments themselves are adequate. In my opinion, this is fragmented. As a user of SE sites, when I read the best answer to a question, I do not want to have to also parse what could be a multitude of comments in order to really know the answer.

Perhaps there should be some mechanism when making a suggested edit where I could assert that comments made by the OP of an answer support the edit, and the reviewer is provided with the comments as evidence?

share|improve this question
12  
looks like your rejecters were wrong. You said you were incorporating comments by the OP. I guess you could have said it more clearly. "Editing comments by answer's author into the answer itself" is more clear than "I have updated the answer to include relevant information made in comments regarding an edit made to the original question. " –  Kate Gregory Mar 29 '13 at 20:38
3  
I agree with Kate. Those edits get rejected too often because they often seem to be adding content. Also, some reviewers might not realize it's OK (even good) to do this kind of edit. –  Andrew's a Unitato Mar 29 '13 at 20:39
    
@Kate I agree, and now will use your well crafted phrase in the future. The only thing that comes to mind is it may need to be formalized in a feature. How many reviewers fail to even read the provided reason, and have it their head that no new content is to be added period. –  gravidThoughts Mar 29 '13 at 20:41
3  
"Some say that the comments themselves are adequate. In my opinion, this is fragmented." - You are absolutely right in your assertion. –  Aditya Mar 29 '13 at 20:49
add comment

2 Answers

up vote 11 down vote accepted

On a very related issue, I think It would be great if the editor's comment were made a little bit more conspicuous.

I personally never see it unless I'm specifically looking for it(which I often do when the purpose of the edit isn't clear).

And when I do read it, and it matters, It is made clear to me that there are a large amount of reviewers who don't read it at all. They might not even know that it's there.

It's pale yellow, and in a small font, It's the kind of thing that my eyes completely miss when I'm reviewing comments.

It just doesn't draw attention to itself very well.

For those of you who can't access the queue, here's a sample

enter image description here

share|improve this answer
    
I have never seen the reviewers UI because I don't have the rep! I agree improvements in that UI can help. Does the current UI only have the edit itself, or does the reviewer have access to all of the comments associated with the Question/Answer? –  gravidThoughts Mar 29 '13 at 20:47
2  
I completely agree with this; I've mis-rejected a number of questions because I've missed the comment. The combination of the background color, font color and size just camouflage the text –  LBT Mar 29 '13 at 20:51
1  
I'm not sure a UI change is needed here as much as a change in editor/reviewer attitude on comments. Improved Formatting is a comment that just makes me want to rip my hair out. –  psubsee2003 Mar 29 '13 at 20:52
    
Although your specific example is pretty good since that is accurate to what done. –  psubsee2003 Mar 29 '13 at 20:53
    
@psubsee2003 You hit on a very real issue. Free form text is problematic when trying to build computer systems that are meaningful semantically. Thats why I suggested the idea of an editor being able to make an "assertion" and providing "evidence". The reviewer would be required to explicitly reject the assertion. The point being is both would be required to do something explicitly in the system that is more meanigful than writing and reading free form text. –  gravidThoughts Mar 29 '13 at 20:58
1  
It's highlighted in bright yellow. This is a pretty standard convention for emphasizing things. How should we make it more visible? Add a <blink> tag? –  Cody Gray Mar 29 '13 at 21:21
    
Per my comment above, while I agree this would be an improvement over the current system, it is still free form text. The semantic meaning of the text in the mind of the editor cannot be reliably transferred to the mind of the reviewer. As Kate points out in a comment on the question, my reason for edit was simply unclear to the reviewers. –  gravidThoughts Mar 29 '13 at 21:24
    
@CodyGray maybe if it was actually highlighted in bright yellow instead of a pale wheat color it would stand out more. If the font was larger, and if it were more bold, it would also stand out more –  Sam I am Mar 29 '13 at 21:24
    
Selected this after letting question sit a while. An effective quick improvement Would still like to see the process more structured than free form text. –  gravidThoughts Apr 4 '13 at 13:48
add comment

Ok, so I just read the post on Etiquette for answering your own questions. I should have done this immediately sorry for that. This is my 5th post on a SE site.

I propose the following.

In addition to providing a free form reason for edit, the editor can alternatively select to make an assertion.

Some examples of assertions:

  1. The OP of the question/answer substantiate the information in the edit themselves in comments.

  2. The answer to the question is dependent on which version of a particular system is being used. The original answer is incorrect for newer versions.

If the editor makes an assertion, they must provide evidence.

The SE system can do this smartly based on which assertion is being made. For example, the reviewer could be provided with comments associated with the answer/question. A URL could be provided to an authoritative source for a breaking API change.

If the editor makes an assertion the reviewer must explicitly reject the assertion.

This type of system would also make it possible to review the quality of work done by the editors and reviewers themselves.

It strikes me that for as much work as it takes to become a fully empowered member of SE sites, members are not required to prove their reviewing skills.

Again sorry for the delayed answer, my original question was to evaluate this proposition. I now know better.

share|improve this answer
1  
an interesting assertion would be "This edit does not change the content or meaning, only formatting, spelling, grammar, and organization". Seeing this as something to assert might remind edit-suggesters that in general, they are not supposed to change meaning –  Kate Gregory Mar 29 '13 at 21:58
    
@Kate Nice example. And, again now your starting to produce meaningful data. You have a user that has used your proposed assertion 10 times, yet in reality they changed content every time and were rejected. The edit review data proves it, and now you are fairly certain that this user is an unreliable editor. –  gravidThoughts Mar 29 '13 at 22:04
add comment

You must log in to answer this question.

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