9

Pretty straightforward question. When should one edit someone else's answer vs. Adding your own? Editing as in adding additional information to the answer, not correct something you think is wrong, not grammatical errors, etc.

Improve this question
3
  • Related: How aggressively should we maintain and improve very popular questions? Shog9's view there was adding information is fine when it improves the answer and is in keeping with the spirit of it (including adding info from other lower answers to make it easier to find) .
    – Martin Smith
    Commented Feb 17, 2014 at 23:27
  • 4
    @PeterMortensen The discussion in that question is four and a half years old - and only 2-3 months shy of being as old as MSO itself! Probably better to have it rebooted here and close that one as a duplicate of this.
    – doppelgreener
    Commented Feb 18, 2014 at 0:12
  • 1
    Marked as a duplicate yet it is old and the rules have actually seemingly changed since then. If the rules have changed then it isn't really a duplicate anymore is it?
    – Richard Le Mesurier
    Commented Mar 2, 2014 at 13:36

4 Answers 4

Reset to default
5

Personally, I'll almost always reject edits as invalid that substantially change the original post, including those that add information. I've seen an exception or two, along the lines of adding links to updated documentation or documentation for new versions of whatever product is in question, but other than that, I can't think of any.

I'm of the belief that it's better to either add a comment, if the information fits in a comment, and the poster you're responding to is still active, or otherwise, add your own answer.

"To expand on what [user] said in [linked answer], ..."

Improve this answer
3
  • 1
    The problem with this is that a highly popular post that fails to mention critical information will stay popular, if I merely add that critical information in an answer of mine. I prefer to edit the popular post, which will keep getting upvotes, instead of adding yet another obscure answer. Example - Datejs not being updated since 2007, Example - years-old obsolete information about how to install npm
    – Dan Dascalescu
    Commented Feb 19, 2014 at 7:10
  • @DanDascalescu This is true, though I'm not sure how applicable it is outside of StackOverflow. I'm having trouble, for example, thinking of anything on ServerFault that falls into this category. Perhaps I lack imagination, but generally, for the SE sites I haunt, it seems that obsolete information isn't much of an issue, because as new versions of [foo] come out, you get new questions and new, appropriate answers about [new version of foo], while the questions and answers about [old version of foo] are still perfectly valid, and remain applicable to that version.
    – HopelessN00b
    Commented Feb 19, 2014 at 20:53
  • What if [foo] turns out to cause certain errors (e.g. data loss), but nobody has mentioned them yet? You could leave a comment to that effect at the end of the long list of comments, and have it buried under "Show more", or you could edit the post and add the warning.
    – Dan Dascalescu
    Commented Feb 19, 2014 at 21:45
5

If you haven't already done so, I suggest reviewing Editing Help in the Help Center, as it covers some of this.

The cardinal rule for me is that you should not change the meaning of an existing answer.

For instance, you should not remove the word "not" from this answer, because it would change the meaning.

Things like updating broken links and fixing spelling and grammar are OK.

Adding new information is a serious gray area, though. This does run the risk of changing the meaning of the answer. I find it usually best to leave a comment with the additional information, and leave it up to the person who answered as to whether they want to incorporate it.

One example where I would find this acceptable is: an answer recommends using some command with a specific set of options, and your additional information explains what the options do.

If your own additions can stand alone as an answer, then by all means post a new answer.

Finally, don't be too afraid to edit; if it turns out to be really bad, either the original author, or 2K+ community members, can roll it back.

Improve this answer
1
2

My view is that as long as

  • The content you are adding is highly relevant to the answer.
  • It makes it obviously better
  • You are 100% certain the information added is accurate (and tested to work if code).

Such that you reasonably expect the post author will feel the edit has made the answer better, not worse, then edit away.

Occasionally I have made quite substantive additions to answers that are fundamentally correct but lacking in details and never had any complaints.

The original author can always roll back if they really object.

If your edits are subject to review you should note that this isn't a community consensus opinion however and probably most such edits are automatically rejected as "radical changes" rather than being evaluated per the points above by people with domain knowledge.

Improve this answer
1
1

If you feel you have information that someone else has not yet provided and would clearly benefit the poster and community, post it as your own answer! If you have advice or knowledge as to the improvement of another's answer, try leaving a comment for them and giving them the opportunity to change it themselves, and learn from what you have to share as well. I would stay away from changing an answer which completely changes what the original person attempted to post, this is even worse if what you are doing destroys the original answer and reduces the amount of benefit the original answer provided.

Improve this answer

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