It's not the first time I see that my edit was rolled back right after it was accepted. I think this is maybe because my edit contained some bad choice and the user who accepted it realized (after acceptance) that it's not really good for some reason.

But it's not so good because I may not realize this and

  1. don't learn that it wasn't good enough
  2. cannot edit it again, correctly

My suggestion is that I should get a notification of the rollback (and maybe I should lose those points as BoltClock's a Unicorn♦ says in a related question).

What do you think?

  • 1
    Wasn't an approver who rolled back. Just another 2K user who saw your edit, didn't think it was as appropriate, and rolled it back. Jun 13 '13 at 14:09
  • 2
    But what is your question here? Are you trying to understand why it was bad? Or do you want to discuss a new feature to notify you when this happens? Jun 13 '13 at 14:10
  • 1
    OK, thanks for clearing it to me. I will add an update to make my intention more clear.
    – qben
    Jun 13 '13 at 14:17
  • 2
    Three people approved your edit; one disagreed after the fact. I don't think you should lose points for that. The edit introduced grammatical errors in this case, which is why it was rolled back. I would not have approved it (or at the least, would have edited it to fix the grammatical issues) but that doesn't change the fact that several people did approve it. Jun 13 '13 at 14:18
  • @GeorgeCummins Thanks for the link.
    – qben
    Jun 13 '13 at 14:19
  • 3
    @GeorgeCummins as you say it is more wise to edit it again than rolling it back. I feel my work is just thrown away and still don't know what the problem was.
    – qben
    Jun 13 '13 at 14:23
  • 9
    @qben: when a reviewer tries to reject an edit but is beaten by robo-approvers who approve it without paying attention just to get badges, the person trying to reject is told by the system that they can rollback the edit if they feel it shouldn't have been approved. Sure, editing the question themselves may be more appropriate, but it's hardly surprising if they take the system's advice.
    – Wooble
    Jun 13 '13 at 14:32
  • @Wooble maybe an other choice is needed - "edit it again" or something similar.
    – qben
    Jun 13 '13 at 14:36

I rolled your edit back, for exactly the reasons George Cummins and Wooble said in the question's comments: it introduced grammatical errors to the post and was approved while I was reviewing it, so my reject vote failed. I didn't have the patience to go through and edit the question myself.

Regarding getting a notification when an edit is rolled back, I'm at best ambivalent about it. It would be useful in the case of helpful users, such as yourself, who want to sharpen their editing skills. However, there are a few problems with this idea.

Getting notifications about rollbacks will greatly increase the likelihood of edit wars. This, I think, is probably reason enough not to issue notifications of rollbacks.

Another problem is that for the most common workflow (edit post > click rollback on a previous version), the edit comment is "rollback to revision X". This doesn't tell you why the rollback happened, so the notification wouldn't provide any information beyond "someone with 2000+ rep thought your edit was inappropriate". I suppose there's the possibility of requiring comments with rollbacks, but I predict a lot of users' being annoyed by the extra requirement and writing nothing other than "rollback" in the text field. Certainly not all of them would do that, though.

  • There is not risk of edit war from the user who suggested the edit, since that user needs to suggest edits. If 3 users approved the suggested edits, and other 3 users approve the second suggested edit, then there are two chances: The users who approve the suggested edits don't pay too much attention, or 6 users really think the suggested edit was correct.
    – apaderno
    Jun 13 '13 at 15:43
  • @kiamlaluno That's almost entirely true (the <2k rep user could suggest and get it approved repeatedly by auto-accept reviewers, but this would be a lot of work and take some luck), but they aren't the only ones who can get their edits rolled back. I've had 2k+ users make improper edits on my posts, which I promptly rolled back. If they got a notification about it, that would likely encourage them to return and rollback my rollback. Jun 13 '13 at 15:45
  • 2
    Example: this edit is too minor: stackoverflow.com/posts/15769429/revisions Jun 13 '13 at 15:48
  • 1
    @EsotericScreenName thank you for your answer. I feel now that the best way for me now is to simply ask the user who rolled back my edit what he found wrong. I see that getting notifications about rollbacks is problematic.
    – qben
    Jun 14 '13 at 8:53

I've done this several times, when I reject a suggested edit but four other people with accept-reject rates like 500-0 decide to accept it beyond belief. It is accepted by people with <2000 rep that just mash the accept button. That is frustrating, so I go ahead and roll it back myself.

Example: https://stackoverflow.com/review/suggested-edits/2303945

  • 2
    Your rollback on this was wrong. OP approved the edit themselves.
    – hjpotter92
    Jun 13 '13 at 15:49
  • What I saw: A user (Cloud) who did not participate in any other part of the question or answer, did a drive-by edit without commenting or answering themselves. Drive-by edits that change the answer are rejected. Jun 13 '13 at 15:51
  • Also: How am I know that the OP approved it, when I am not given that information with it's in my Suggested Edit queue? Jun 13 '13 at 15:52
  • 1
    Yes, but when OP thinks that the edit was helpful, you shouldn't have rolled it back. Your judgement on the review-queue was correct but rolling back the accepted edit was not.
    – hjpotter92
    Jun 13 '13 at 15:52
  • How am I to know that OP thinks that the edit was helpful? It's not apparent. Jun 13 '13 at 15:53
  • 1
    Username is given a different background color: rgb(224, 234, 241);
    – hjpotter92
    Jun 13 '13 at 15:54
  • Where is that ? Jun 13 '13 at 15:54
  • 1
    – hjpotter92
    Jun 13 '13 at 15:58
  • That is not apparent. Jun 13 '13 at 15:58
  • Rollbacks are as dangerous as too minor edits are. In both the cases, you are making the question bump in the front page without making it any better.
    – apaderno
    Jun 13 '13 at 16:08
  • Did you really think your edit was legitimate? The code in my answer is a quote from the question. And everyone here agrees that under different circumstances, the edit under discussion was invalid. Jun 13 '13 at 17:20

Assuming that the suggested edit is rolled back because it is really bad, and nobody of the users who approved it noticed it, I don't see why the user who suggested it would need to be notified about the suggested edit being rolled back.
If a user wants to learn how to correctly suggests edits, he can start looking at his past suggested edits, and see which one have been rejected (the URL is http://stackoverflow.com/users/<user id>/?tab=activity&sort=suggestions). I don't think a user for which a suggested edit has been rolled back should do anything about that. Suggesting edits is not a "try the first time; if it is rolled back, try again a second time" process. If the user didn't did a good edit the first time, I don't think the same user can get it right at the second tentative. And if the user gets it right the second time, that would makes me wonder why he didn't get it right the first time.

Notifying a user about something that is negative is not something for which notifications are used. Rolling back a suggested edit is potentially negative, as it means the user suggested something that was not salvageable.

Removing the points gained from the suggested edit it is not what I would suggest, but since the reputation gained from suggested edits are removed when a post is deleted, probably that could be done. I can still imagine the situation where the rollback was wrong. What happens in that case? I think it is preferable not to complicate the code, and simply let the user who suggested the edit keep the reputation gained.

  • I see your last two points. But about the first I don't think you're quite right. At least in my case, I always try to edit again correctly if my edit gets rejected first time.
    – qben
    Jun 14 '13 at 8:59

After reading the answers and especially Woobo's comment which I quote here:

"When a reviewer tries to reject an edit but is beaten by robo-approvers who approve it without paying attention just to get badges, the person trying to reject is told by the system that they can rollback the edit if they feel it shouldn't have been approved. Sure, editing the question themselves may be more appropriate, but it's hardly surprising if they take the system's advice."

Let the system advise one more option: "edit again". I know someone could do that without offering this option of course, but this would draw its attention to this possibility.

My point is not to lose valuable effort to make a question/answer better. This would save those cases when lots of good changes were made but the edit contained also one or two mistakes which could quickly be corrected.

As for the notifications I'm convinced now that they could be dangerous. But the "reviewers' side" could be better with this option.

You must log in to answer this question.

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