I am looking for guidelines on suggested edits. Specifically how drastic can they be?

I have read this http://blog.stackoverflow.com/2011/02/suggested-edits-and-edit-review/ and numerous posts on suggested edits that are against small/cosmetic edits, but haven't found any guidelines on how major they can be.

I made my first edit today https://stackoverflow.com/suggested-edits/216254 and it was rejected with "This edit changes too much in the original post; the original meaning or intent of the post would be lost."

Certainly the edit changes a lot of the text in the post. However, in my opinion, it distilled the actual question from a lot of other guff (look where the first question mark appears in the original). In terms of helping other users in the future I believe the edit is an improvement. I also wouldn't mind seeing the post merged or closed as it is hardly an original question.

Is it because of the added suggestion in the accepted answer? This was essentially a code review, and unrelated to the question asked.

I think there are a lot of questions like this on SO. I am happy to spend some time improving some of them, but first I would like to read some guidelines on what I can do (no doubt they already exist, and I have failed at search).

note: since my edit there has been a bit more back and forth in the comments. It seems the original asker wasn't really clear what he wanted. Is it not still better to edit the question into a concrete one and allow him to add new questions, or expand himself?

  • You might want to correct the second link – Bart Mar 7 '12 at 19:22
  • thanks, I changed the link to the question directly. The other link apparently was only visible to me? – Sean Mar 7 '12 at 19:34
  • Just a thought: I probably would have approved this edit. – Cody Gray Mar 7 '12 at 20:46

You removed code which materially changes the meaning and intent of the question.

Please, please, please do not do that.

A suggested edit which removes--or sometimes just substantially reformats--code is suspect at best, and most likely invalid.

I like to think of it as the user's code is the chain of evidence, as this is where their problem most likely exists. The text accompanying the code is all hearsay. If you destroy the chain of evidence, you may substantially decrease the value of the question!

| improve this answer | |
  • thanks. Could you link me to where I can read about this please? I read a lot about the power of community editing in the "official" documentation, but most users seem to share your view. Strikes me as a disconnect. – Sean Mar 7 '12 at 19:36
  • Also what was the original meaning and intent and what did I change it to? (Yes, this is hard, because the question was badly presented... maybe it needs an edit :)) – Sean Mar 7 '12 at 19:37
  • For reference see meta.stackexchange.com/questions/113149/… and meta.stackexchange.com/questions/78845/…. If you read these you'll find the prevailing sentiment is code changes in a question is bad, while some code changes in an answer are Ok. This is consistent with the official documentation. – user7116 Mar 7 '12 at 19:40
  • 2
    @Sean: There's also meta.stackexchange.com/questions/101583/…, and the question it's closed as a duplicate of (which is more thorough): meta.stackexchange.com/questions/88627/… – jscs Mar 7 '12 at 19:42
  • Additionally, you don't have a high reputation on SO. Right or wrong, reviewers may take that into consideration when you make a radical change to code in a question or answer. – user7116 Mar 7 '12 at 19:42
  • 5
    There is no official place where this is officially presented. Just think of it this way: you changed "I started with sugar, flour and four oranges. Why doesn't my apple pie taste right?" to "I started with sugar, flour and two apples. Why doesn't my apple pie taste right?" The problem was that the asker used the wrong fruit, and by "fixing" that in your edit, the question was destroyed. – Pops Mar 7 '12 at 19:42
  • @all thanks for the links. What if the code posted is superfluous? In this case the actual question doesn't require the code at all, or only one line of it (which he repeats inline). What should be changed in the original post? Nothing? My (subjective) opinion is all of it. But I seem to be swimming against the tide. – Sean Mar 7 '12 at 19:56
  • 2
    @Sean: usually I leave a comment to let them know that they need to clean up their code to only include the relevant parts. I understand in your situation you lack the ability to do so. I would still err on the side of "don't touch code in a question", no matter how egregiously wrong or superfluous. Supply an answer or comment instead and let the OP handle it, so that the chain of evidence is preserved. – user7116 Mar 7 '12 at 20:09
  • This sounds reasonable, but then why does the edit function exist at all? – Sean Mar 7 '12 at 20:29
  • 2
    I agree with this in 99.9% of cases. The only exception I would add is when a question is hopelessly, completely unsalvageable without your edits and going to inevitably be closed/deleted, you should feel free to improve the question however you see fit, even if you change the author's meaning along the way. I see it as a special case of optimizing for pearls, not sand. – Cody Gray Mar 7 '12 at 20:44

You must log in to answer this question.

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