-3

TL;DR: Users shouldn't flood edits without a consensus. Read my post but skip the optional stuff... :)

Every now and then, we get an edit flood at our front page. This sometimes happens without any consensus and sometimes even at unacceptable moments / rates. Which is why I think we should get a better coordination on this when such edit floods want to take place.

I've frequently done took part of this myself, my main intent is that I want it to happen but that shouldn't cause one to act by himself. If we get this better organized we act more together as a community...

I don't to see users banned for a full week just because of a misunderstanding, that scares them away!

Main Idea: We should require users to get permission from a moderator.

Q: I want to perform these changes to the site, may I perform an edit flood?

Three options:

♦ agrees: As there has been a consensus on this, you are free to go.

♦ delays: Not at this point, it's [peak moment in day] right now. You can do it later.

♦ denies: We have not reached a consensus on this, you are not allowed.

Additionally, a moderator can give this permission when a meta question wraps around:

♦ makes a meta question OR adds a comment to a post explaining a consensus has been reached.

Optional: We could gather these edit floods to do edit flood X times a year instead.

Yes, it doesn't seem a bad idea for a moderator to collect the things that seem necessary to be done and then we try to get this through as fast as possible. Instead of doing these multiple days, we just do these X times a year such that the rest of the time there are no edit floods on the front page.

It's fairly easy to do this:

  1. Create a chat room where we organize this, assign each person a different topic to edit such that edits do not collide with one another. This makes an edit on one topic actually go faster.

  2. Get everyone to open as much tabs as possible on the topic he needs to edit and prepare the edits along, you could like for example set up X minutes to do this. Then afterwards people literally flood the site which requests while still trying to avoid the captcha / IIS request limit, we don't want to take Super User litterally down in the progress.

  3. The infrequent visitor sees a flood at that moment which resolves soon after, instead of our visitors seeing unfortunate edit floods at more frequent intervals.

Optional: The SE system could be foreseen from an edit rate limit, which shows a warning.

That warning could explain that one should raise the required editing on meta such that it can be added to the above organized edit floods, of course the rate should be disable by moderators such that we can have the above edit floods when we really need them. The rate limit idea is by nhinkle

The only thing left to do is, discussing these idea and filling in the X...

Main questions:

  • What are your thoughts about this?

  • What do you think is an acceptable limit on rate of edits, as in X edits per Y minutes?

  • Do you have any other ideas / contributions to the edit floods topic?

Optional questions:

  • Supposing that it is a good idea, how many times a year should we organize an edit flood?

  • What should the warning towards an edit flooding user say, if it gets implemneted?

Related:
Do we care about edit flooding? Should we be doing anything?
Why are there large number of high-view questions on the front page?
What time of the day is Super User getting the least traffic?
Recent edit floods, some of them which confuse / upset people or happened without consensus.

  • 2
    I like to go through the review page at random times and often become guilty of "flooding." But I don't schedule it, I just do it. This idea is good, but doesn't coordinate well – Simon Sheehan Feb 16 '12 at 2:47
  • I'm aiming at more severe cases @SimonSheehan, those where it is particularly very easy to do something repetitive where you don't need to take the additional step to read the post. – Tamara Wijsman Feb 16 '12 at 3:32
  • @SimonSheehan Did you ever "own" the entire front page after going through /review? – Daniel Beck Feb 16 '12 at 10:09
  • @Daniel maybe once, not usually though – Simon Sheehan Feb 16 '12 at 11:59
  • "I don't to see users banned for a full week just because of a misunderstanding, that scares them away!" That clearly wasn't the case here. Were the user interested in how things work, he'd have seen e.g. the Garrett topic(s) on Meta, or would have asked about it on Meta or in chat. // When users abuse their permissions, they are sent to the penalty box. The worse the offense, the longer the timeout. It's pretty straightforward, actually. And the user got a silver badge out of it. A pretty good deal for him, all things considered. – Daniel Beck Feb 16 '12 at 14:30
  • @DanielBeck: I haven't looked into his case at all, still there are users that don't go and check everything at meta when they want to do something. Simply seeing other users edit floods makes it look acceptable, I have never looked for edit flood meta questions before I did my edit floods so I completely see this as possible. I however have no idea about the history of that user though, as until 1 - 2 weeks ago I hadn't have much time to visit Super User. At best I've seen a few edits here and there about him. But haven't read about anything bad, I saw him mentioned in chat but didn't read it – Tamara Wijsman Feb 17 '12 at 16:00
  • @TomWijsman And that's the entire problem. If the current meta topics aren't enough to educate these users, what good will another topic saying "By the way, get permission from the moderators before you do something, or wait for this quarter's mass editing cycle" do? – Daniel Beck Feb 17 '12 at 16:39
  • @DanielBeck: I'm questioning if you have read this question, because I'm trying to attempt to have some way that the system at least warns the user and if possible disallows him from doing this, as well as getting something organized monthly rather than having random users do it at random times. Even something as simple as limiting the edit rate based on reputation (higher reputation can make more subsequent fast edits), could have an effect. If at best I've only left a meta question about this behind, they have another one to show up in their search results; which shows different thoughts... – Tamara Wijsman Feb 17 '12 at 17:09
  • @TomWijsman You mean the stuff you declared optional? :-) // I don't think this is useful, and it will still require a big button "Don't throttle editing" for moderators, and while that is active, users can do everything they want. It might even hurt, since, if this proves popular, nobody will be able to review all the changes that are bundled in a very short time frame. Additionally, you create a situation that just invites edit conflicts by having all edits of all users take place in a short period. – Daniel Beck Feb 17 '12 at 17:13
  • Other than Gareth and kino, I cannot remember a single major edit flood, and Gareth handled it very well after people talked about it. I don't think it's an actual problem on this site that requires this much effort to prevent. I think an optimistic approach would be more useful. – Daniel Beck Feb 17 '12 at 17:14
  • You're missing the point, it would actually not cause any edit floods at all because it would be organized every two months and during the event everyone gets assigned one thing to be editing in. This causes every user to do it's own thing without any overlaps. Although you're somewhat right: We'll see as time develops; I've seen some other users (including myself) do some smaller floods, probably at a time that you're not online. There's no hurry, we'll see what edit floods follow... I just wanted to share the thought on doing this organized rather than having anyone do an edit flood anytime. – Tamara Wijsman Feb 17 '12 at 17:24
  • Maybe a [feature-request] to have a special edit/retag that doesn't bump to the front page? – Nate Koppenhaver Jul 26 '12 at 3:34
5

TL;DR

I think this is an interesting idea, but causes more harm then good.

Long version

Just my two cents, but I think that there is really only one type of edit flood that you are trying to address, namely deliberate "mechanical" editing where the user knows that they will be processing a lot of edits in a short period (like out recent edit flood with the now aptly named user URGE TO RETAG.

As Simon says, and some of the posts that you linked to demonstrate amply, most edit flooding is good edits that occur when a user is going through \review, or something similar.

Another important thing to consider with this type of moderator approval required to edit, is that it is only effective if all the other ways of bumping a question to the front page are also disabled. To my knowledge, this means not just no editing, but no answering.

Yet another important fact to consider is that even if this raises a flag or something like that for mod attention, it might not get dealt with for an hour. While generally this is not a concern (and in many cases it is less) to a user that is in the middle of going through \review it just stops them cold. And they don't know when they can continue. This seems like it will have a chilling impact on the heavy use of \review.

Even if we coordinate edit flooding, and thus avoid it being a problem to the average user at the average moment, it still leaves the issue of the edits not really getting "reviewed." This is the reason that there is no minor edit flag. In theory, someone could edit a post during that time to be something offensive, and no-one would catch it for a long time, buried as it was under all the other edits.

It is my understanding that most things that will cause edit flooding that can be known in advance should be taken up on meta, and a mod/dev deals with it (like a mass retag, burning a tag, etc). These issues should never be done by a user on their own in any case, and so the proposed collaboration will not be helpful.

For your second question, I think that if something like you idea is implemented (and I am not for it to be honest) the rate should not be a fixed number / minutes, but rather having more than N on the front page at a time. Allows for faster reset on an incremental basis, and accounts for heavy traffic periods that way.

As to your third question: I don't think its really a problem when flooding occurs with "good edits" It is the same as it is the same as Community bumping old questions, except faster. It may cause some very temporary issues, but overall, not exactly what I would call harmful. It is only really a problem when someone goes ahead and makes loads of edits that are not useful.

Aside:

I think the optional part of your post has a lot of potential, but NOT as a remedy to limiting edit flooding. I think it would be great to clean up all the question in particular tags, all in one fell swoop, as it would make things like locating dupes much easier, would make things much easier to learn about, and would make the site much better in those areas. I think it would also encourage more people to edit. Seeing the editors for the past month shows a large divide between the 100+ edits this month, and the less than 100 edits this month, and I think that organizing edit drives, (if more people could get involved) would be great in reducing this divide, and getting more people involved in the community.

  • Please note that when I mean edit flood that I'm not targetting users that are using the /review page, thus the rate should at least lie such high that it doesn't trigger on people that are just rapidly reviewing. My edit flood I thus mean the "mechanical editing" as you've described in your first paragraph. On questions I believe there is already a limit, I'm not sure about answers but you wouldn't flood these (and if so you are probably suspendend), and comments do not even bump a question. – Tamara Wijsman Feb 16 '12 at 3:36
  • For your "yet another important" paragraph, raising flags is indeed a bad idea which is why it should be forced only on the people attempting to do an edit flood, the system can by the way easily identify whether an edit was part of /review or not as that's just the matter of adding an extra hidden field / action parameter to the edit review form. Your issue about "reviewed" is irrelevant, it doesn't matter whether the edit floods happens coordinated or not as in both cases its not "reviewed" because of the rapid flood. One could possible go through some activity the participating editors. – Tamara Wijsman Feb 16 '12 at 3:40
  • @TomWijsman, From a practical standpoint, the \reviewing and mechanical editing differ only in quality not quantity. For the raising flags, I just mean as a way to get a mod's attention (something that the system would do, not a user). With regard to the reviewed issue, it would be much larger if multiple people were flooding at the same time, as opposed to just one person. – soandos Feb 16 '12 at 3:43
  • Your "it is my understanding" is partly true, partly false. I want these things taken to the meta / moderators and the user should never act alone, which is what we agree on. However, flood edits are required when the moderator / developers don't have the direct tools to act. It's more easier/correct for people to go through some posts than to write a specific, than to write a script to do this and have half of it incorrect. My motive with this question is that I just don't want edit floods without permission from the community / moderator whether or not enforced. A warning would help. – Tamara Wijsman Feb 16 '12 at 3:43
  • Example of " flood edits are required when the moderator / developers don't have the direct tools to act." That should not be implemented by a mod (even manually)? – soandos Feb 16 '12 at 3:44
  • For my second question, having more than N on the front page seems like a good idea. I've though about scaling this by reputation, but yours is actually better in some way; although one has to filter out the edits coming from /review which is somewhat trivial. On the other hand, this is a problem at less active moments; so I'd rather have a rate limit... – Tamara Wijsman Feb 16 '12 at 3:46
  • 1
    For your last question: After seeing others edit flooding I kind of feel the bad thing about it; Me (myself) causing title dits on the frontpage for prolonged times, Gareth (no bad intention) causing image edits on the front page for mulitple days, todays bad tag edits; I would love to see something permission based with a warning for peronal edit flooding and organizing these together to lower the impact. It saddens me that today's flooder got banned... – Tamara Wijsman Feb 16 '12 at 3:49
  • [All my previous comments were a full response to your question, it might not be the best format for them to be in; feel free to suggest an alternative. My next comments will be in response to your replies.] @soandos: Mods can't do everything automated, but only certain things. Developers can do a bit more but still have things they would need to write a custom script for, that script can fail when human logic is required. A random example is an ambiguous tag that has to be split in multiple tags, volume was one and was split in logical-drive and audio / sound... – Tamara Wijsman Feb 16 '12 at 3:52
  • I don't think raising a flag is necessary as a mod would see the whole front page flooded. It might be that the system already auto-generates a flag though, there are quite some in place. Posting the exact same answer twice is a random flag that exists... – Tamara Wijsman Feb 16 '12 at 3:54
  • I like your aside part, we could tempt to swap each month between either an edit drive (get together things that have really to be done) or an audit question (discuss what we will do about stubborn popular question). – Tamara Wijsman Feb 16 '12 at 3:57
2

I believe we don't need additional measures.

We've had the occasional edit floods before (like with the re-uploading of images), but ever since then, we've been very cautious, and I guess the Meta community has developed a bit more sensible approach to mass-editing.

Whenever I call for editing of posts (like here, or here), I make sure to include a disclaimer. It'll inform users about not flooding the front page, but doing it in a coordinated fashion. This is what we agreed on earlier:

  • Doing 10 posts or so at a time.
  • Doing everything at once in low traffic hours (which would be 03:00 to 07:00 UTC).

… and it worked out pretty well so far.


Recent events.

The only editing floods we've had since then were caused by a particular user who would literally take one comment like "I'd retag this manually", or an answer at a score of 2/0 – and suddenly go ahead and do everything.

I don't [want] to see users banned for a full week just because of a misunderstanding, that scares them away!

This wasn't just a misunderstanding. The user has been repeatedly told about taking their edit privileges more seriously, but was still performing pointless edits, even to a point where it's not funny anymore, and mass retags without community consensus.

Daniel more or less anticipated what would happen a few days before. It's not like this was a misunderstanding anymore.


What we can do better:

To summarize, the way it currently is is totally fine. If it's obvious editing, like with the "Need Help" questions, then a Meta Question alone will suffice. Include a disclaimer, let people do their jobs.

I think we should just be more verbose when it comes to s. Maybe the Question/Answer approach is flawed here when the solution to retagging is not yet clear.

We should probably use the questions here for stating a problem, and then let the answers decide on how to best go about resolving the issue. Something like this:

Q: The tags X and Y are ambiguous, how do we change this?

A1: I'd go for creating Z and merging X into Z, keeping Y.

A2: It's fine to keep both, but retag only some.

Once we have a winner, and there is a lot to retag, then everyone is fine to retag, under the conditions we defined before (e.g. 10 at a time, only during low-traffic hours). We'll survive this.

  • 2
    Additionally, I commented on one of his edits early that evening regarding the missing community consensus, but the user continued anyway without responding. Editors can be @-addressed. – Daniel Beck Feb 16 '12 at 10:37
  • By the way, the [annoyances] edit was extremely funny, probably the one edit of the user that actually was. – Daniel Beck Feb 16 '12 at 10:38

You must log in to answer this question.

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