I liked this suggestion so much — and think it makes a poor 20k ability (sorry Andy) — that I'm suggesting it directly:

Temporarily lock questions from being edited, to prevent ridiculous edit wars like this one.

This could be handled by automatically locking any post with E edits in T minutes. It could be limited to just a day or two, then automatically unlocked, if moderators finding time to review the post is a concern. However, it could require a new locking type, as I believe posting comments on these questions might still be useful.

For E=10, T=60, I don't remember any false positives (barring quickly-wiki'd, discussed below), but it would cover the mentioned case, the "bomb effing USA" troll (okay, he used several posts, but it would've slowed him down and reduced frustration for important users), and others.

For the quickly-wiki'd problem, I have a solution as well, which would eliminate them as false positives for this feature.

I expect E and T need tweaked, with an attempt at gauging effectiveness.

  • 1
    Even cooling it off for 1 hour would help. – yhw42 Nov 13 '10 at 6:03
  • I concur with @yhw42. Overall, though, an excellent idea. – IAbstract Nov 14 '10 at 6:41
  • I think giving the OP the final say and then transfering control to moderatore would be a good idea. Meaning OP can lock his own post, so that no one except a moderator or OP himself/herself can unlock the post for editing – One Face Mar 3 '15 at 14:46

It's a good idea -- we had this slated during The Great Edit Wars but demand for it seemed to wane over time.

It's probably safe to auto-lock (for a set duration) based on a heuristic of a very large number of edits by a small # of users in a small amount of time.

  • Would a 1 hour cool-off period (for example) be triggered if a post is edited 5 times within 2 minutes? It is rare, but in the event that I continue to find grammatical errors or ways to clarify a question, would I trigger a cool-off? – IAbstract Nov 14 '10 at 6:40
  • @dboarman: That sounds like accidental conflicts (when a question is new) rather than intentional edits/reverts, and the latter is the focus of this. Incidentally, the linked suggestion fixes those (as I'm sure other suggestions would as well), which should eliminate them from being false positives on the "edit war" trigger proposed here. – Gnome Nov 14 '10 at 8:10

Edit wars are so extremely rare that we don't need an automated system to deal with them.

Just wait until a moderator can jump in, usually it doesn't take very long.

  • 2
    I also thought about this, then realized I'd seen it happen three times with plenty of frustration and aggravation; even counting for only finding one through meta, how many times have I not seen it? My usual 10:1 ratio to estimate that is still a drop in SO's bucket, but I'm throwing this out there because I think it could be worthwhile. – Gnome Nov 13 '10 at 7:57

I'd like to suggest a better solution : the system should automatically cast a moderator attention flag when an edit war is taking place, but without locking the post.

Your current solution has the major downside of having a 50% luck of locking the post while it's on the wrong state, and then it'll stay in that wrong state until the lock times out. It may not be a big deal in that particular question, but when truly wrong and potentially dangerous information is at stake it becomes a real issue.

  • 2
    The system already does this when it detects a rollback war. The problem is that moderators aren't always online, and aren't always looking at the flags when they are online, so it may be a long time before they actually see the flag. – BoltClock's a Unicorn Mar 3 '15 at 10:21
  • @BoltClock'saUnicorn the only disruption caused by an edit war is just bumping the post up on the homepage, so we can disable that in case of such a war, but other than that I don't see what harm is caused by waiting for a moderator to solve the issue. Offensive/spam edits will be taken care of anyway (by flagging the post) even if a moderator isn't present. – user230564 Mar 3 '15 at 10:26

You must log in to answer this question.