47

I edited this post and noticed my reputation change from 1466 to 1470 (i.e +4). I went to check my reputation changes and found:

4 rep

Note: My reputation later went down from 1470 to 1468, so it appears the system realized it made a mistake and corrected my reputation, but hasn't gotten rid of the second revision or the history event.

My activity history still shows 1 suggested edit with 2 revisions.

19
+200

There is definitely something strange going on.

First, the time stamps on the edits are different:

  • Revision 2 was approved at 13:24:57Z
  • Revision 3 was approved at 13:24:58Z

Those timestamps match the +2 entries in your profile.

But when you look at the reviewer timestamps is when things get screwy:

  • talonmies approved the edit at 13:21:30Z
  • us2012 approved the edit at 13:24:57Z
  • Parado approved the edit at 13:24:58Z (which was 1 second after Revision 2 was approved).

If the timestamps can be trusted, it appears that something happened that allowed 2 approvals to approve the edit, but at then also approved the edit with the 3rd approval.

13
+200

Nope. It isn't (well, shouldn't be) possible to gain more than 2 points for making a single edit.

As Mat pointed out in the comments, the edit revisions history does look weird.

This could be a bug, but I'm not exactly sure what's causing this.

  • 4
    Both link to the same suggested edit though, something's wrong here. – Mat Sep 28 '13 at 13:44
  • 1
    @Mat Yes and also in my Activity It shows one edit suggestion and two Revisions – Walker Sep 28 '13 at 13:46
  • @Mat: You're right. Something's wrong. – Amal Murali Sep 28 '13 at 13:51
  • 2
    Is it possible that two people reviewed the edit at the same time, both improved it and marked that the suggestion was helpful? – jmac Sep 28 '13 at 15:24
  • @jmac: when an edit is "improved" it is visible in the post history: you get one entry for the suggested edit (as suggested) and a second one for the "improvement", associated with the reviewer. – Mat Sep 28 '13 at 15:32
  • 1
    @jmac Nice guess, but if you click through to the revision history you can see that nobody improved this suggestion. – Mark Amery Sep 28 '13 at 15:52
  • Then two people just accepted it as-is? Or clicked improve and changed nothing? – jmac Sep 28 '13 at 15:58
  • @jmac: No. I remember someone had suggested an edit on Phys.SE; two people approved, and I improved the edit, and no such problem had occurred. – Abhimanyu Pallavi Sudhir Sep 28 '13 at 17:00
  • 1
    @PreciousTijesunimi please don't mark this accepted. Bug report should be "closed" only when it's being either fixed or rejected by a dev. – ShaWiz Sep 28 '13 at 22:59
  • 1
    I agree with @ShaWizDowArd here. You can untick the checkmark and wait for a moderator/developer to respond to the question. – Amal Murali Sep 28 '13 at 23:01
8

After a ton of digging through the database trying to identify how many users hit this bug and the cause of it, we've determined that it's a race condition.

The fix will be going in this week, once we've identified all the places where we're hitting this same issue (Shog's got a list). The fix will include a redis lock (which we already use elsewhere), and it will prevent these very rare multiple entries.

As far as the rep from these aka the extra +2, there are very few cases in which this happened and it involves deleting from several tables to clean it up, so we're leaving it. It shouldn't happen again after this week, when the bug is fixed.

You must log in to answer this question.

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