37

There has been a user active for the past 3 days destroying posts. I'm seeing the edit approval prompts and have consistently rejected them. The typical edit is deleting content and replacing it with a dumb remark. Or adding random characters.

I cannot find out who it is, the page doesn't provide a way to get to this profile. Only "anonymous user" is visible. It could well be more than one. There are two basic patterns, erase content and adding random keystrokes. The total number of edits I've personally rejected is around 20 or so.

No real harm done but this is getting old. I have to say that I see no reason at all why users that don't log in should have edit permissions. Any of the users that make legit edits have been named users. Please consider making these kind of users at least traceable.

6
  • 13
    I agree. While it's community-run, I don't see a need for someone to be able to edit anonymously. Commented Feb 11, 2011 at 0:09
  • Do you have permalinks to the destructive suggested edits? The SO team probably has IP info and can ban.
    – waiwai933
    Commented Feb 11, 2011 at 0:12
  • @wai - I don't, these edits disappear quickly. Commented Feb 11, 2011 at 0:13
  • 3
    @Hans On the suggested edits, on the left, there is a permalink which will be preserved even if the edit is rejected. If you see this happening again and could grab the permalink, I think that would help the Team.
    – waiwai933
    Commented Feb 11, 2011 at 0:14
  • @Hans .. if the edit is rejected it disappears from the page. If we dont remember the question how can we get the perm link?
    – Shoban
    Commented Feb 11, 2011 at 0:23
  • Here is one good example: stackoverflow.com/suggested-edits/64485 - judge for yourselves. (Yes the "anonymous user" is back at full strength probably took him this long to bypass the obstacles set by Jeff) Commented Jun 22, 2011 at 11:29

2 Answers 2

5

There are some additional restrictions on anonymous user edits that should prevent these sorts of anonymous edits from entering the queue at all. These restrictions are based on the history of the known anonymous bad faith edits.

I can't describe them in detail because that would be a recipe for bypassing them.

6
  • 11
    Security through obscurity? Commented Apr 8, 2011 at 17:08
  • It failed. the anonymous user is back and destroying posts again for at least 5 hours straight now. Already used all my votes, rejected most of the suggestions, though few were good enough to approve. Naturally, he's also clogging the queue, preventing registered members from suggesting edits. Commented Jun 22, 2011 at 11:26
  • @shadow there was a bug which made all edits show up as anonymous for a bit; see meta.stackexchange.com/questions/96000/… Commented Jun 22, 2011 at 22:07
  • Thanks, it just looked like the same person making most of those suggestions, many times adding meaningless contents to existing posts or just doing useless changes in old answers. Commented Jun 23, 2011 at 7:52
  • I agree with the sentance. Message me if you're curious. Commented Mar 24, 2014 at 11:19
  • @MichaelMackus what sentence you agree with? There is no private messages in Stack Exchange and nothing in your profile page so how you want anyone to "message you" exactly? Commented Mar 24, 2014 at 12:04
10

Some stats about anonymous:

From inception we had 91 rejected edits and 82 approved edits. The vast majority of edits come in from registered users (I am counting about 4200 now from registered users)

Of these anonymous users the top number of submissions we got from a single ip address is 4.

The 3 anonymous users that have more than 2 submissions all submitted good content.


So it appears that DISASTA PASTA, is coming our way not from a single anonymous source but many distinct IP Addresses .

Even at this low hit rate, many of these edits anonymous is posting are good.

At the moment if a single anonymous account is rejected N times per day the IP is auto banned for a week. So we already have some protection.


To add to that we could add a second param that auto reject edits where anon is clearly vandalizing a post. Perhaps if more than 80% of the post is changed by anonymous it is silently auto-rejected.


Here is a full list of all the rejected anonymous edits from the start, let me know if you can think of simple heuristic to cut this down:

11
  • I like the silent auto-reject. Perhaps apply the same heuristics as we use with detecting crappy answers?
    – Ivo Flipse
    Commented Feb 11, 2011 at 0:48
  • low entropy comment seems like a good one as well ...
    – waffles
    Commented Feb 11, 2011 at 0:58
  • How about introducing a "reject as spam/trash" vote that needs to be confirmed by one other user, with much higher limtis? Not sure whether this can ever be reliably filtered out automatically...
    – Pekka
    Commented Feb 11, 2011 at 1:06
  • 7
    I also find it fascinating that people actually mistake the edit function for the comment field (or simply don't give a damn). See "Very nice answer. i love ur solution" and "can you give some example sir ???? that will be helpful for us.."
    – Pekka
    Commented Feb 11, 2011 at 1:08
  • @Pekka: my thoughts exactly! Grabs his jaw from ground
    – user138231
    Commented Feb 11, 2011 at 1:15
  • I'm pretty active on reviewing edits, and I've not yet seen one of these, so well done everyone that's rapidly rejecting them Commented Feb 11, 2011 at 1:21
  • I wonder if there's more of this: stackoverflow.com/edit-suggestions/834 (reject somebody's edit) → stackoverflow.com/posts/2108300/revisions (then make it yourself). It could be that the user clicked reject accidentally then applied the edit anyway, or it could be an attempt to game the system.
    – badp
    Commented Feb 11, 2011 at 1:26
  • @badp ... most likely an accident ... it is too much effort to do this yourself imho
    – waffles
    Commented Feb 11, 2011 at 1:36
  • 2
    @Pekka: Anonymous users don't have the ability to comment. Perhaps, seeing the lack of comment field, they think that's how to comment.
    – Borror0
    Commented Feb 11, 2011 at 1:59
  • @badp: Yes there is: stackoverflow.com/edit-suggestions/179stackoverflow.com/posts/4779835/revisions ;) - (And yes, I realize you didn't include the code markup, though I'm not really sure why)
    – sth
    Commented Feb 11, 2011 at 2:09
  • @sth [ I submitted that edit anonymously.](chat.meta.stackoverflow.com/transcript/message/468687#468687) I did it to make a point about why diffs should default to show markdown diff instead of rendered output.
    – badp
    Commented Feb 11, 2011 at 3:16

You must log in to answer this question.

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