Wikipedia talk:Pending changes/Archive 2

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Autoconfirmed Users

"Can edit; a new edit is visible immediately if the previous version is already confirmed; otherwise not visible to readers by default until confirmed by a 'reviewer'" What does 'if the previous version is already confirmed' mean?

Also, I think that autoconfirmed users should maintain their right to edit semiprotected pages immediately. Quinxorin (talk) 05:15, 26 November 2009 (UTC)

Requests and usage

Crossposted to WT:PP, WT:RFPP

In light of rollout of the pending changes tool in 3 days now (June 14), I'm updating WP:RFPP and WP:PP to note some basics that will be needed by then. Otherwise it'll probably be a horrible mess of chaos on the day.

Usage and requests have no good (non-BURO) reason to be in a different place from semi-protection. The reviewing aspect is different but the scope, usage and requests (ie WP:RFPP aspects) are likely to be nearly identical to semi-protection and can usefully go on the same page.

Keeps it simple to have all forms of page protection and their requests in one place, even if they are in fact 2 tools for that purpose.

FT2 ^{(Talk | email)} 08:20, 11 June 2010 (UTC)

What happens to edits that are not accepted?

I'm not finding an answer to this in obvious places, so will stick it here, and send a link to RobLa to comment. It's likely that there will be a fair number of edits that are not accepted for various reasons. What happens to them? Do they remain in the article's history? Do they remain in the user's contribs? Are they specially marked in the article history/user's contribs so they can be identified as never having gone "live" for public view? I ask for a couple of reasons; first off, if they disappear into the ether, then it will be difficult to track problematic editors (or checkuser them, for that matter). Secondly, if they remain in the history, certain edits will remain eligible for revision deletion and/or suppression, and we need admins and reviewers to be aware of this, and to arrange for one or the other where appropriate. Risker (talk) 21:15, 12 June 2010 (UTC)

Hi Risker, in response to your first question, I think you may want to read my response in the What about edit conflicts? thread on Help_talk:Pending_changes. There's an example there that might help you understand better, but in a nutshell, the page revision history is just a linear series of snapshots. Without pending changes installed, the very latest revision is the one that people will see. With pending changes installed, the latest accepted revision will be shown. However, if someone goes to edit the page the very latest version is what they are editing. (BTW, I would love it if you or someone else edited Help:Pending changes to make this aspect of the feature clear; though just beware of creating a tl;dr situation).

In response to your second question, you're right that admins and reviewers might need to be made aware that non-accepted revisions will still hang around in the edit history unless/until someone does something about them (should that be necessary) -- RobLa (talk) 23:15, 12 June 2010 (UTC)

Well, I know what the page history *is*, what I wanted to know was whether unreviewed changes would appear in it prior to their review; it appears the answer is "yes". Therefore, by logical deduction, I can assume that unreviewed edits will also appear in a user's contribs. This is very important for checkuser purposes; if someone is trying to add the same unacceptable information repeatedly, we will actually have something to work with. Risker (talk) 23:55, 12 June 2010 (UTC)

Pending edits appear all at Special:Oldreviewedpages and are marked as such in recent changes and watchlists. An edit cannot be 'unaccepted' until it has already been accepted. So pending edits are either accepted, or reverted to the latest accepted revision (which automatically accepts), or overwritten then later accepted. Cenarium (talk) 23:39, 12 June 2010 (UTC)

I have to tell you that this is the most confusing part of the entire process. No matter what, bad edits are getting accepted and then something else happens to them; they can't be flushed straightaway. It is entirely counterintuitive; why would anyone do anything to accept vandalistic changes, even if only to revert them? This will be a key component of the learning curve. Frankly, I am quite concerned that *editors* won't realise this, and will wind up unintentionally adding bad edits. Risker (talk) 23:55, 12 June 2010 (UTC)

Of course they can be flushed straight away, by being reverted. Then there's never been any need to accept them, and the revert is automatically accepted, even if the reversion was made by an anon because when a new revision is no different than the latest accepted revision, it is auto-accepted. The editing process remains the same. Cenarium (talk) 01:15, 13 June 2010 (UTC)

To "reject" a pending revision, an editor will either need to rollback or undo (as appropriate). One feature that we decided not to delay the launch for is the "reject" button. We want to implement this as one of the first things we do after the launch. It's kind of a complicated feature because it's not always obvious what to do when the user hits "reject". Here's the specification for the "reject" button; we'd love your comments there. -- RobLa (talk) 01:22, 13 June 2010 (UTC)

To clarify, technically edits are in 'unaccepted' state until a reviewer 'accepts' them. If a new edit is made while the first edit has never been accepted, such as a reversion, then the first edit remains in 'unaccepted' state; then there's little purpose in accepting it, the new edit is accepted or not (and so on).

A 'reject' button could complicate things. If an edit is 'rejected', the likely actions are either to edit the page, undo or rollback the edit. All those actions are linked on the review page. It should be clarified on the documentation pages. Cenarium (talk) 01:32, 13 June 2010 (UTC)

Reviewer right

When will the reviewer right be granted? It is listed in Special:ListGroupRights and there is a backlog on Wikipedia:Requests for permissions/Reviewer. ‎ —Mike moral ♪♫ 23:07, 12 June 2010 (UTC)

Hi Mikemoral, the "Reviewer" group was just created earlier today. It doesn't grant any significant rights at the moment (basically, a couple that autoconfirmed users have already, iirc), but it'll become more significant when the feature goes live. -- RobLa (talk) 23:15, 12 June 2010 (UTC)

I was just thinking it'd be best to give it out before the trial begins so there isn't a scramble. Thanks for the response. —Mike moral ♪♫ 23:57, 12 June 2010 (UTC)

A noticeboard

We probably need a page where reviewers can report edits they're unsure about, or where specific issues with reviewing can be discussed, we could start doing this on the talk page Wikipedia talk:Pending changes/Queue, and if needed create a page such as Wikipedia:Pending changes/Noticeboard. Cenarium (talk) 04:05, 13 June 2010 (UTC)

Approval of edits by non-autoconfirmed editors

In the current version of the page there seems to be some confusion over who can accept edits by non-autoconfirmed editors. The lead says that (any) auto confirmed editor can. The table in the "Description" section says that only reviewers can.

(There is another sentence on this in the "Description" section and I notice that Delldot recently changed this to say reviewer, rather than unconfirmed.)

The idea that (any) autoconfirmed editor can confirm edits (presumably limited to pending changes level 1) was added to the lead on the 5th of June.

My first reaction is to think that this recent change must be based on a misunderstanding... but it actually seems to make sense. After all, on a level 1 page, an auto confirmed editor could make the changes directly should he/she want to.

Yaris678 (talk) 16:35, 15 June 2010 (UTC)

I've reworded the lead, but I'm not sure if this is very clear. Cenarium (talk) 17:32, 15 June 2010 (UTC)

Shortcut to the Queue

I have created WP:PCQ as a redirect to Wikipedia:Pending changes/Queue, to save having to type the full title -- PhantomSteve/talk|contribs\ 18:14, 15 June 2010 (UTC)

IRC Channel

Are there any plans to have an IRC channel to discuss the pending changes (something like #wikipedia-en-pending ^connect? Bug reports, help for new reviewers, etc? Or will people be directed to #wikipedia-en-help ^connect? -- PhantomSteve/talk|contribs\ 18:14, 15 June 2010 (UTC)

There is a wikipedia-en-pending channel created on freenode - if someone can make it an 'official' one, then that would be good! -- PhantomSteve/talk|contribs\ 18:26, 15 June 2010 (UTC)

I've registered #wikipedia-en-pending ^connect (after discussion on IRC) to prevent takeover by a malicious user, myself and Phantomsteve are ops for the moment. Quite happy to give op status to any other established editor who would like to help out with the channel (there will probably be a fair bit of people coming in). If nobody disagrees here then I'm going to add the channel to the policy page and the header about pending changes, so users can come into the channel for help and #wikipedia-en-help ^connect doesn't get flooded with requests. - EdoDodo ^talk 18:30, 15 June 2010 (UTC)

Also willing to transfer founder status to an established editor who is more knowledgeable about IRC than I am (I just registered the channel to prevent it being registered by a malicious user). - EdoDodo ^talk 18:35, 15 June 2010 (UTC)

To do when this is implemented

A few things:

Add a sitenotice for registered users inviting them to participate.
Update template:Pending changes trial, make visible Special:Oldreviewedpages in the navigation.
Make the first daily batch of pending changes protection listed at Wikipedia:Pending_changes/Queue#Day_1.
PC-protect testing pages at Wikipedia:Pending changes/Testing, most at level 1, so reviewers can test.
I added a warning at MediaWiki:Protect-text, will also need update.

(FYI, I won't be around for the scheduled time of implementation.) Cenarium (talk) 18:22, 15 June 2010 (UTC)

Major legal question about flagged revisions.

I have asked this on several occasions but I have not gotten an answer despite the extreme importance of the question. We all know that the real purpose of flagged revisions is to shield the Wikimedia Foundation from libel lawsuits on BLP articles. Two points. It is absolutely unclear to me that such a libel case would ever be successful against the Wikimedia Foundation for hosting the encyclopedia. That would be tantamount to blaming the owners of a building for somebody writing graffiti on it. What does seem clear to me is that a good lawyer could argue that those who have flagged revisions as acceptable have become liable for those edits. In other words, flagged revisions now place legal burden on editors that were not there before. Editors are now responsible for libel which might not even be obvious to the casual reader. This change, besides being unnecessary in my opinion, is fundamentally flawed. It creates a legal issue out of thin air and endangers the participants of the project. Jason Quinn (talk) 18:25, 15 June 2010 (UTC)

Jason, it's always been that way. As Mike Godwin, WMF's general counsel, said during the Office hours 11 June 2010 (Mike is mnemonic1):

[4:05pm] MC8: QUESTION: So, if someone does something illegal to do with WMF projects, is it the contributor that faces legal action, or the WMF? How much indemmnity doesthe average user have?
[4:05pm] mnemonic1: the contributor is normally the person legally responsible

If you are concerned about this, don't get the reviewer right, leave it to someone else. It's the same as if you revert a change which removed information, so that the information is back: if the information is incorrect, and you put it back, you would be liable - yet this happens every day.

As far as I am aware, no one has been sued for putting information into an article or changing it - are you aware of any instances? The reason why I ask is that a good legal counsel would say "my client may have approved it, but the actual contributor was xyz - so they are legally responsible for it, as they (presumably) knowingly added false information" - which would be exactly the sitation that currently exists: the person who actually contributed the information which is false is legally responsible. -- PhantomSteve/talk|contribs\ 21:28, 15 June 2010 (UTC)

It was implied in a previous time I raised this issue that there have been no cases. If that is true, then it has never been tested in court even if the Wikipedia Foundation is responsible for vandalism. This whole notion of protecting pages from pranks or "trying to open up" Wikipedia (these have been the angles used in the recent media coverage) is something of a smokescreen and it ignores the roots of this proposal. Plus since there have been no actual lawsuits in (almost) 10 years of Wikipedia's existence, it seems odd to make a change in response to perceived threats that may not exist. If a libel lawsuit was ever filed, I would have liked to know the outcome because (with non-expert opinion) I believe the Wikipedia Foundation may have been found innocent. It is users (if anybody) that should be responsible for vandalism. In that regard, you and I agree, Phantomsteve. As I have argued before, I have a strong concern that flagged revisions could be used by a lawyer to self-incriminate the Foundation since they seem to suggest the Foundation is vulnerable to these legal threats and action was needed to prevent them. What I'd really like is to hear from an actual practicing attorney specializing in this sort of thing: the exact type that the Foundation should have (did?) contacted shortly after the Robert Byrd incident. What were their collective opinions? That kind of information should be as openly discussed as the rest of Wikipedia. Jason Quinn (talk) 22:58, 15 June 2010 (UTC)

You know, we've got a pretty good lawyer on staff. You might just ask him :) —Preceding unsigned comment added by Philippe (WMF) (talk • contribs) 23:01, 15 June 2010 (UTC)

(Disclaimer - I'm not a lawyer, I've simply studied these issues). Welcome to the topic of "Section 230 immunity". This is indeed openly discussed in certain specialized law and policy areas, but Wikipedia talk pages are not good locations to seek out legal discussions. For example, the Wikimedia Foundation has been directly sued at least once, in the _Bauer_ case. It did win, though it could be claimed that the plaintiff's case was poorly argued. Note there's a big difference between having the law be a certain way in the abstract, and the practicalities of a situation. My favor quote on that matter comes from the writer of the Cyber Patrol break FAQ - "What I found out was that those organizations, through no fault of their own, were able to give me a lot of sympathy and not enough of anything else, particularly money, to bring my personal risk of tragic consequences down to an acceptable level, despite, incredibly, the fact that what I had done was legal. Ultimately, I couldn't rely on anybody to deal with my problems but myself. Some people learn that lesson a bit less impressively than I had to." -- Seth Finkelstein (talk) 12:31, 16 June 2010 (UTC)

Proposed policy?

I see no reason for this page to be policy. It's just the way the software works - that's not policy. Rich Farmbrough, 19:32, 15 June 2010 (UTC).

I suspect it's a policy in the same way that Wikipedia:Protection Policy is a policy - in fact, I think this is a sub-policy of that policy... (although personally, I prefer guidelines to policies). Mike Peel (talk) 22:39, 15 June 2010 (UTC)

The scope and reviewing sections have plenty of policy statements, which can't be reduced to what is (or can be) in the protection policy. Cenarium (talk)

Reviewer userbox?

Maybe this: [code redacted]

—fetch ·comms 23:17, 15 June 2010 (UTC)

See Template:User wikipedia/Reviewer. –xeno^talk 23:19, 15 June 2010 (UTC)

Aww :( —fetch ·comms 23:19, 15 June 2010 (UTC)

Oh my. How Orwellian. And people wonder why I don't have all those fancy icons and fezzes on my userpage. Risker (talk) 09:12, 16 June 2010 (UTC)

Administrator guidelines

It would be nice if along the way a simple guide can be provided for the administrators who don't delve into this area. I suspect that we will be getting a bunch of requests as this feature is ramped up. Having a short 'checklist' would help. Are we also going to replace 'forum shopping' with 'admin shopping' when one admin rejects a request and the user tries a second or a third? Will there be a way to see if a user has already been rejected? Vegaswikian (talk) 23:27, 15 June 2010 (UTC)

Obvious question

Sorry, this has no doubt been raised before, but something that immediately jumps out looking at this new arrangement is that, despite the hype that it's supposed to be increasing accessibility, auto-confirmed users are actually losing the ability to edit pages with immediate effect (if there are outstanding unreviewed edits on those pages). Is this intentional? Is there a way round it?--Kotniski (talk) 08:36, 16 June 2010 (UTC)

As I understand it, autoconfirmed users only lose the ability to edit pages with immediate effect when there are already pending changes from non-autoconfirmed users. This is presently rare as changes are reviewed within seconds I believe. And what's more, reviewer is being handed out extremely liberally. –xeno^talk 13:55, 16 June 2010 (UTC)

Hmm, if we could be confident of changes being reviewed within seconds, then we wouldn't need this tool or any other protection (we could just let people revert vandalism within seconds). But presumably as more and more pages come under the "pending changes" regime, and the initial burst of interest drops off, the time during which changes remain pending will increase quite significantly. --Kotniski (talk) 12:35, 17 June 2010 (UTC)

Two responses:

Almost any autoconfirmed user can become a reviewer simply upon request, which also helps the "unreviewed edit" problem.
Without PC, even if we reverted vandalism in less than a minute, it could be cached (by Google, for example) for hours. This is not merely theoretical, since Google is constantly reindexing to try to keep themselves up to date. One reason PC is useful is because it gets rid of this problem.

— Carl (CBM · talk)

"accepted by" → "reviewed by" ?

Thread retitled from ""accepted by" → "sighted by" ?".

I'm not particularly fond of the "accepted by" language [1], what about simply "sighted by" ? –xeno^talk 13:56, 16 June 2010 (UTC)

As in "sighting a UFO" or "sighting Bigfoot"? :-) -- RobLa (talk) 14:26, 16 June 2010 (UTC)

Agree, "sighted" is not a common verb used in that way in american english anyway. I'm not really arguing for "accepted by" though, just against "sighted" - cohesion 21:04, 16 June 2010 (UTC)

This could conflate with "oversight", which, while it is oversight, it's not the oversight used on Wikipedia. --Izno (talk) 21:33, 16 June 2010 (UTC)

And how about "reviewed by" or "checked by"? Salvio ( ^{Let's talk 'bout it!}) 01:30, 17 June 2010 (UTC)

Reviewed by is perfect... Makes sense, even. Why didn't I think of that! –xeno^talk 01:31, 17 June 2010 (UTC)

Questions about the reviewing policy

Several interesting questions about the "pending changes" system have come up at Wikipedia talk:Reviewing, and they would benefit from wider participation. They include:

Should a reviewer "approve" a revision of a page which she knows is erroneous? Discussion: Wikipedia_talk:Reviewing#Don't approve things that you know to be wrong
Is the "pending changes" system preferable to page protection for articles that are protected due to NPOV concerns? Is it suitable for pages where the only concern is massive juvenile vandalism? Discussion: Wikipedia_talk:Reviewing#Sea of Japan and Liancourt Rocks

I hope we can keep the discussion centralized at WT:Reviewing. — Carl (CBM · talk) 15:47, 16 June 2010 (UTC)

Pending changes and edit warring/content disputes

Here is a situation that came up not long ago on the World War I article:

Non-reviewer makes a (bold) edit.
Reviewer reverts non-reviewer's edit.
Non-reviewer restores said edit.

What should be the next step to take as a reviewer?

(another) Reviewer accepts or rejects non-reviewer's edits again. (which was what happened in the World War I article)
- This can be seen as an endorsement of one side in a content dispute or edit war.
- If a reviewer accepts or rejects again, then this allows the cycle to repeat, thus continuing the edit war.
- The reviewer should not be able to gain an upper hand simply because that editor is a reviewer.
Reviewer keeps said edit under review and tries to force discussion on the talk page.
- What if non-reviewer fails to participate in any discussion on the talk page?
- What if, after an agreement/consensus is reached on what editorial action(s) to take, said edits continue to happen?

Should reviewers take steps and prevent such edit wars on PC-protected pages from occurring or getting out of control, or should they take a more passive "see no evil, hear no evil" approach to situations like this and enforce edit warring sanctions as before when they occur? Discussion is strongly encouraged here. –MuZemike 17:43, 16 June 2010 (UTC)

I think that reviewers who edit articles should do so just as they would if the article was not protected. If someone is too assertive about an article, so that it looks like a WP:OWN issue, we have a system to handle that. If an IP adds the same material 5 times, we can warn and block as usual. If we need to protect the page to stop people from edit warring, we can still do that, too. But as long as the page is not protected (just under pending changes), editing should be as normal as possible. — Carl (CBM · talk) 17:59, 16 June 2010 (UTC)

In this case, personally, I would accept the non-reviewer's edit and start a discussion on the talk page, unless, of course, it's a BLP violation, blatant vandalism or a copyvio. Actually, if it's not vandalism or something like that, I would have accepted the change in the first place, even if I don't personally agree on or like it. Salvio ( ^{Let's talk 'bout it!}) 01:39, 17 June 2010 (UTC)

The reviewer function is instituted to protect against vandalism, not bold edits. In my opinion, proper etiquette would be for the reviewer to revert or accept the bold edit, then discuss it on the talk page and allow the normal cycle to begin. Their discussions should be performed in normal user capacity, not as a reviewer (I foresee the phrase "uninvolved reviewer" becoming common). I do believe review should be used to stop edit wars, but only once they've begun, not beforehand; also, Reviewer should most definitely not be used as an edit-warring tool in itself by not reviewing or unreviewing edits. -Quinxorin (talk) 02:10, 17 June 2010 (UTC)

I entirely agree with you and think that repeatedly unaccepting an edit to get one's preferred version should be treated as edit warring and lead to sanctions. Salvio ( ^{Let's talk 'bout it!}) 09:31, 17 June 2010 (UTC)

There's no reason an editor has to act "as a reviewer" at all. For most editors, the best way to handle a page with PC is to just edit it like normal and them mark your changes as reviewed when you are done. — Carl (CBM · talk) 13:52, 17 June 2010 (UTC)

Autoconfirmed users2

Just to clear this up, since I'm still a little confused - any edit made by an autoconfirmed user (like me, for example) would automatically be visible to all, right? This page states such edits will be "automatically approved" - but I'm not sure if that means the edit is visible immediately, or that the reviewer has to approve it regardless of any other factor. All Hallow's Wraith (talk) 19:10, 16 June 2010 (UTC)

Not with level 2, which could easily wind up being put to most high traffic articles, with the claim they're being harmed by sleeper socks. Gwen Gale (talk) 19:11, 16 June 2010 (UTC)

It depends on the level of pending changes. For level 1, which is currently what is used for most (all?) pages, edits by autoconfirmed users go live immediately and are not checked. For level 2, edits have to be reviewed by a reviewer/admin before they go live even if the user who made the edit is autoconfirmed. The exception to both of these is if you make an edit to an unchecked version of the page, in which case the page will still remain unchecked regardless of who made the edit. - EdoDodo ^talk 19:25, 16 June 2010 (UTC)

How would one make an edit to an unchecked version of the page? Or does that mean any edit made to a page that has any unchecked edits? (i.e. - does that mean that if a non-autoconfirmed edit goes unchecked, autoconfirmed editors can't make immediate changes to the page as a result?) All Hallow's Wraith (talk) 19:38, 16 June 2010 (UTC)

Edit it, but nothing will show up for readers who aren't logged on until the queued version, with its muddle of edits made after the latest live version, has been "accepted" by a reviewer. The outcome of level 2 will likely be a caste, an online ghetto of hundreds if not thousands of long-autoconfirmed editors whose edits won't go live outside the garden until someone clicks accept. Gwen Gale (talk) 19:46, 16 June 2010 (UTC)

So what you're saying is that, until a non-autoconfirmed user's edits are checked, every single autoconfirmed user is in the exact same position as a non-autoconfirmed user when it comes to being able to edit immediately. Well, doesn't that kinda suck? This fact should be highlighted more prominently on this page and elsewhere in discussing this trial. I don't think most autoconfirmed editors are aware that in many situations, their edits are going to have the exact same standing as anon edits. All Hallow's Wraith (talk) 19:52, 16 June 2010 (UTC)

Nothing for now, in this test, is PC level 2, but in about a year hundreds if not thousands of high traffic, core articles will be PC protected at level 2, if this PoV and source screening/gaming tool stays as it is now. Gwen Gale (talk) 20:04, 16 June 2010 (UTC)

PC level 2 is only going to be used where full protection would be used now, and it's a huge improvement over full protection for preventing autoconfirmed vandalism. PC-1 will actually make editing more available to editors where it is not now. The only case where an auto-confirmed user's edits won't show up is when a non-AC user has made an edit that has not yet be accepted. So far the time to accept an edit is less than 5 minutes, so the sky isn't falling just yet. I've actually put together a script that constantly refreshes the list of pending changes in the sidebar so that they will be accepted faster. I really don't think this is an issue, and if it is, well that's why we are running the trial. If you are a frequent editor, just ask for reviewer permissions, and you'll most likely get them if you have been around at all. --Joshua Scott (LiberalFascist) 04:14, 17 June 2010 (UTC)

The problem with marking edits by autoconfirmed users AFTER a non-autoconfirmed edit that hasn't been checked is that the autoconfirmed user could not see the vandalism by the anonymous user, and then it would stay in the verified version of the article for readers to see. - EdoDodo ^talk 05:33, 17 June 2010 (UTC)

Logged-in users see the latest pending changes by default, so they could see and remove problematic material in theory, but they are likely not looking for vandalism/BLP violations. --Joshua Scott (LiberalFascist) 04:53, 18 June 2010 (UTC)

The magnifying glass icon is already used for the search

Hello. Please have a look at this issue discussed at Wikipedia:Village pump (technical)#Vector & Pending Changes. Yours, Dodoïste (talk) 11:59, 18 June 2010 (UTC)

Two-tiered registered users vs accessibility

I, and it appears a large number of other editors, have been granter reviewer rights whether we deserve them or not. It is clear that this is in the design a two-tiered system of registered users, putting further power distance between IP users and the 'authorised' (read 'those with reviewer status'). Of course, for flagged revision to work, there would need to be a large number of ordinary editors reviewing, otherwise the system will grind to a halt. However, this business is now giving me the creeps because it smacks of removal by the back door of the long-standing principle of equality of access for all; the controls on such accessibility confer additional powers to admins (who I believe, for the record, are quite powerful enough) is diametrically opposite to what I understand WP is supposed to stand for. Ohconfucius ^¡digame! 16:30, 19 June 2010 (UTC)

Only (some of the) pages that were already semiprotected are in the pending changes system. So if an IP could edit a particular page without review before, she still can. But now she can also have a chance to edit semiprotected pages. I do not believe there are any plans to expand the system to include most articles, nor would there be agreement for such plans. — Carl (CBM · talk) 18:52, 19 June 2010 (UTC)

Unclear who gets to see the draft version of articles.

The documentation here and at Wikipedia:Reviewing isn't very clear about who sees the last approved version of the article and who sees the "draft" version with unapproved changes. This page says that unapproved changes will not be seen by "users who are not editors (unless they wish)." What is a user who is not an editor? Anyone can edit. How does the software define a "user who is not an editor"? If a "user who is not an editor" is the same as an IP user, this should be stated explicitly. Or does it include non-autoconfirmed editors or some other group of users? Users with accounts who have never edited a page? Whatever the definition of "user who is not an editor" is, it should be explained clearly.--Srleffler (talk) 20:12, 19 June 2010 (UTC)

Everyone can see draft versions of articles. Logged-in users see draft versions by default, whereas anonymous users only see them by accessing the "pending changes" tab. -- RobLa (talk) 21:58, 19 June 2010 (UTC)

Actually, the bigger part of the message is the use of the word "editor", which I believe is not being used in the same context as it is on this project. Here, EVERYONE who makes an edit is an editor, even before they are autoconfirmed. It seems there's a permission somewhere called Editor, though, which is something that is highly problematic if being linked to this project somehow. Risker (talk) 22:27, 19 June 2010 (UTC)

Thanks. If the distinction is logged-in users vs. anonymous ones, the project page should use those terms, especially in the summary at the top.--Srleffler (talk) 01:34, 20 June 2010 (UTC)

Can I check... if you are logged in... but you haven't been autoconfirmed yet, which version do you see? Yaris678 (talk) 18:03, 20 June 2010 (UTC)

A logged-in (yet still not autoconfirmed) user will see the latest pending revision by default. -- RobLa (talk) 18:07, 20 June 2010 (UTC)

Cool. I have edited the {{Pending changes trial}} template to match this fact. Yaris678 (talk) 23:13, 20 June 2010 (UTC)

That edit was reverted by User:Svick so I have asked him to explain things here. Maybe we will all learn something. Yaris678 (talk) 23:51, 20 June 2010 (UTC)

I made another pass at it. All logged-in users see the pending changes by default, not auto-confirmed only. --Joshua Scott (LiberalFascist) 00:22, 21 June 2010 (UTC)

Ah... I see I actually mis-read what RobLa wrote. Thanks for sorting that out. Yaris678 (talk) 11:51, 21 June 2010 (UTC)

Another approach?

So far, the approach has been to apply pending changes to articles that are known vandalism targets. There's nothing wrong with this approach, but I'd like to suggest another approach for consideration should pending changes be brought in following the conclusion of the trial.

This approach would be the ability to apply pending changes to IP ranges, and individual editors (I've no idea if this would be technically possible, but that's for the developers to consider). The way it would work would be that applying the restriction to an IP or editor (at either level, as appropriate depending on the scale of the problem) would not prevent editing, but mean that the edit would have to be approved before it went "live", in a similar way to how pending changes is working already. Such protection could be used to monitor edits from IP ranges where there is a known problem, and in the monitoring of individual editors whose edits may be problematic in some areas, but not made as intentionally vandalistic - in effect allowing some overseeing and guidance as an alternative to a block. Mjroots (talk) 16:36, 20 June 2010 (UTC)

I think this is a pretty good idea, and though I don't know enough about the MediaWiki interface to know how this might be implemented, I'm confident it could be implemented with no greater difficulty than the original pending changes scheme involved. This would be a good way to balance the benefits of an IP range block to prevent users of multiple IPs from vandalising wikipedia, with the potential downside to enforcing an IP range block; i.e. unrelated users may be adversely affected. This means that anonymous editors in this range who do not wish to register an account at the time would not be blocked, but would rather have the lesser inconvenience of having their contributions reviewed. Specific IPs in the range which are shown to consistently edit constructively could also be added as an exception to the action taken against the IP range. GiftigerWunsch [TALK] 21:17, 20 June 2010 (UTC)

This is a really interesting idea. I don't think this approach is practical in the current iteration of the software or as something that we'll likely be able to do in the near term, but this seems like a good idea to keep in mind as we develop FlaggedRevs and review systems generally. -- RobLa (talk) 21:32, 20 June 2010 (UTC)

I agree - this would be very helpful for handling certain types of disruption. — Carl (CBM · talk) 22:53, 20 June 2010 (UTC)

I think it's a wonderful idea. It'd need to be used with care as a prolific IP vandal might flood the pending changes system, but the principle is brilliantly simple. If this gets more support, can this be stuck on a list of ideas to be implemented later instead of vanishing into the archives? Fences&Windows 00:02, 21 June 2010 (UTC)

Obviously, there'll be a discussion towards the end of the trial, I've got a few ideas of how we should apply pending changes. Unlike other wikis where it is an "across the board" thing, I don't think we need to go down that road. Pending changes should be another tool for us to use, and implemented where it is shown that there is a need to do so. Mjroots (talk) 06:50, 21 June 2010 (UTC)

That was pretty much the plan all along. There's a very strong consensus against using it on all articles, as that would create humongous backlogs, and many people have a misguided notion that PC somehow makes "anyone can edit" false. ♫ Melodia Chaconne ♫ (talk) 14:14, 21 June 2010 (UTC)

FWIW, I'm not convinced that it would necessarily result in such a huge backlog. A few hugglers can already keep up with all non-whitelisted edits; They could be even more effective if they wouldn't redundantly review edits, and would I believe be able to review or revert the brunt of the incoming, trivial changes. Non-trivial changes of our core articles would be reviewed by the many watchers of those articles anyway, the long tail would be pending changes of low-key articles. Does anyone have statistics of mean review time from deWP? I think I heard of a statistics page, but I can't find it. Amalthea 14:41, 21 June 2010 (UTC)

On first glance it seems obvious that the editing load would be prohibitive if Pending Changes was active on all articles. On the other hand, I am seeing the potential for ip edits to be balanced with better oversight.

I think Pending Changes is going to act as proof of concept for Patrolled Revisions, and I can't yet see an argument against having this reviewing/accepting capacity (without the editing delay) on the majority of Wikipedia articles. It wouldn't be hard to operate the revisions oversight on different levels, so that articles requiring protection could make edits pend, while the rest of the articles would simply benefit from a loosely organized scheme of experienced editors checking in every once in a while.

At present, this is done through recent changes and watch-lists, but there is no coordinated way to, say, show every article on Wikipedia in order of the most commonly viewed and edited pages weighted for the length of their "review" backlog, and to accordingly patroll the entire encyclopedia in an efficient, thorough, and proportional way. IMO, while pending changes does work as an alternate or supplemental protection method, it will be most useful as a template for just such a broad and minimally onerous implementation of Patrolled Revisions. 69.142.154.10 (talk) 21:31, 21 June 2010 (UTC)

Accepting edits that are detrimental to the article

If an edit is pending and it is something like a worse expression in English am I supposed to accept it? As in..not vandalistic, accepted.. and then go back and edit, revert, addition was bad English? Off2riorob (talk) 21:32, 21 June 2010 (UTC)

Best advice seems to be to edit like normal, accepting edits you would normally approve of and rejecting those you wouldn't. It's one thing not to venture into a substantial pov dispute, but what would be the point of sending through sub-par grammar? Vandalism isn't the only problem, just the most obvious one. Edit like you would, understanding that if an edit is more substance/pov oriented, you probably wouldn't get involved in it (meaning for pov edits, you could either accept or not accept, depending on your interest in facilitating that debate). Just my best guess. 69.142.154.10 (talk) 21:46, 21 June 2010 (UTC)

Thanks, that is how I see it also, there is no reason to accept an edit that is detrimental to the article. Edit just like you would as you say. This is interesting (meaning for pov edits, you could either accept or not accept, depending on your interest in facilitating that debate) accepting a POV edit that you agree with is not going to occur very often (hopefully) regards . 21:56, 21 June 2010 (UTC)

If you're going to revert the edit, then there is no point accepting it first. If you think an edit probably isn't a good one (but isn't actually vandalism, etc.) but you want to let someone more experienced with that article decide, then accept it. --Tango (talk) 22:00, 21 June 2010 (UTC)

There was a related discussion on this page Wikipedia talk:Reviewing#No_expert_review which may inform this discussion. — e. ripley\^talk 00:30, 22 June 2010 (UTC)

Auto-accepting reviewers edits

I reverted vandalism using Twinkle on the World War I article (see the history, but my change was not automatically accepted (I am a reviewer). This doesn't seem to match what the table is saying. The same thing happened to User:RashersTierney who is also a reviewer, but the revision he reverted was actually already accepted. Why are we having to review the edits of a reviewer? --Joshua Scott (LiberalFascist) 14:50, 19 June 2010 (UTC)

See Wikipedia talk:Reviewing#Undoing vandalism with Twinkle. –xeno^talk 14:51, 19 June 2010 (UTC)

Fixed. You'll need to bypass your browser cache. Amalthea 10:04, 22 June 2010 (UTC)

How is success or failure to be judged?

Colleagues, forgive me if it's set out somewhere and I've missed it: trials normally require at least some benchmarks by which their success or failure is to be measured. Yet I see none. If there's to be community comment during the trial, and especially at the end, subjective opinions will be important, but I am surprised that there are no solid expectations or criteria that we might consider, too. I was prompted to write this after seeing this post. Tony (talk) 08:50, 21 June 2010 (UTC)

Truth be told, this looks more to me like a rollout than a trial. If there are any benchmarks, I haven't stumbled across them. Am I missing something? Gwen Gale (talk) 08:55, 21 June 2010 (UTC)

There's a bit at Wikipedia:Pending changes/Metrics, but nothing concrete. My subjective impression of our current approach isn't particularly favorable, but that can have several causes. I'm not yet sure what conclusions to draw from it. Amalthea 09:52, 21 June 2010 (UTC)

My lay perspective permits me to see this trial for what it is, just a "see how it feels" kind of thing. There won't be hard numbers, just experiences and impressions, opinions and new perspective. Since nothing like this had been tried before, that's really the most important starting point. A rigorous numerical analysis would miss the point: before it "works", it has to be palatable to the editing community. This isn't satisfying to those who fear more endless straw polls and debate, but it's a necessary part of demystifying the concept and quelling discontent. The mere fact that Wikipedia hasn't yet exploded is itself an important data point. 69.142.154.10 (talk) 21:39, 21 June 2010 (UTC)

Just based on my experience with this after a brief while, something is starting to become apparent to me. I'm seeing a lot more need to revert vandalism on previously semi-ed pages (example: 4chan), which can easily make one conclude that semi-protection was better. But I have a feeling that this perception may be an artifact of testing the process on pages that had been semi-ed. It might well be a big improvement on pages that had some vandalism but were previously not protected at all. I point this out because the current structure of the trial may be masking this fact, by making it look like the process is worse than semi. --Tryptofish (talk) 21:28, 22 June 2010 (UTC)

Confused

I'm hoping someone can explain what the rationale of pending changes is, because I can only hope I've misunderstood it. As things stand, it appears we're adding level-one pending-changes protection to articles that would normally be semi-protected; indeed we're removing the semi-protection. Adding level-one PC protection means that any of our tens of thousands of autoconfirmed accounts will still see the BLP and other vandalism. So articles may still contain nonsense or libel, and this will be visible to tens of thousands of readers. Adding level-two PC protection instead reduces the numbers a little, but thousands of readers who are reviewers will still see the problematic material.

Can someone explain the point of this? SlimVirgin ^{talk|contribs} 05:55, 22 June 2010 (UTC)

It's the same reason we allow open editing on all the articles that aren't semi-protected in the first place. What's to prevent nonsense or libel from being inserted into any article? The difference is that some articles receive more vandalism than others, and thus some are semi-protected to alleviate this problem. Given your experience, I'd expect you to understand this concept. Pending changes offers an increase in openness on those pages which would otherwise be semi-protected—openness that is generally in line with the founding principle of allowing great freedoms in editing, while preventing the huge majority of readers from ever seeing nonsense or libel (in line with our principle of advancing quality and the rest). It's also unreasonable to say that "this will be visible to tens of thousands of readers" when what vandalism does take place on such pages is reverted very quickly: according to the current values on Special:ValidationStatistics, pages at the 95th percentile of review time wait only 13m44s; by the 75th percentile, this is only 3m45s. Few people, even considering autoconfirmed users and reviewers, will see vandalism in such a short window, except perhaps under exceptional circumstances. Should the volume of vandalism be too great despite pending changes, semi-protection remains an option. There is also the possibility, though it's a (wiki-politically) sensitive subject, that pending changes might be used more widely in some respects than semi-protection, should the community be able to review all such edits within a reasonable timeframe—which would overall reduce the amount of vandalism and libel seen by the general public. I find the prospect of wider use somewhat dubious but nevertheless intriguing. {{Nihiltres|talk|edits|⚡}} 06:51, 22 June 2010 (UTC)

Thanks, Nihiltres, but I'm still not seeing the benefit. We currently have a situation where problematic articles are semi-protected to prevent things like BLP violations, which can be damaging even if seen by just a small number of people, so the aim is to stop them from being made—to make the article less of an attractive nuisance to people of ill-intent. What's happening now is that the semi-protection is being lifted from some BLPs, and level-one pending-changes added instead, so that the vandalism can continue, seen by tens of thousands of readers. Where is the sense in that? SlimVirgin ^{talk|contribs} 07:23, 22 June 2010 (UTC)

The sense in pending changes is that semi-protection prevents all IPs from directly editing the article. The first level of pending changes allows those IPs to edit the article, with the requirement that their edit is checked before it is made "live". Vandalism can still be dealt with, but IPs making constructive edits are not penalised by the actions of IP vandals, which would be the case if the article was semi-protected. Mjroots (talk) 07:55, 22 June 2010 (UTC)

The vandalism may be visible to tens of thousands of logged-in users, but this is a tiny number compared to the MILLIONS of passive Wikipedia readers who don't have accounts. By using pending changes rather than semi-protection, we can discourage anonymous vandals by depriving them of the gratification of having their changes visible to this large group of people (including themselves), while only marginally inconveniencing good-faith anonymous edits, which I'm sure is how most good editors get started. – Smyth\^talk 10:15, 22 June 2010 (UTC)

You would not be happy if defamatory edits about you were read by tens of thousands of people because I'd downgraded the protection of your BLP. SlimVirgin ^{talk|contribs} 16:57, 22 June 2010 (UTC)

Potentially read != actually read. If the material is clearly defamatory and violates the BLP policy, then any logged-in user with their editor hat on will revert it as soon as they read it, and only a handful of people would ever have been exposed.

Having said that, I agree with what Off2riorob says below, that the choice between review-protection and semi-protection should be made based on evidence of what's actually going on in each article. – Smyth\^talk 17:24, 22 June 2010 (UTC)

I can see the point of adding level one for unprotected articles with problems that don't quite rise to the level of needing semi-protection. But removing semi-protection from a BLP that clearly needs it means the article becomes an attraction again for vandals. That "she sucks cock" is only visible for a few minutes to tens of thousands of people is not a good argument—if it is, we might as well get rid of protection entirely and rely on people reverting. And with certain kinds of BLP vandalism (involving real names, real addresses, serious allegations) the focus should be on prevention. SlimVirgin ^{talk|contribs} 19:10, 22 June 2010 (UTC)

One problem I see now is that we're not doing enough to discourage anonymous vandals. When this first started, there was a big template on top of the protected pages that at least sort of made it clear that their changes aren't going to appear live. What we have now is nothing before they go to edit, and just a little box at the top of the edit box that says "Edits to this page are subject to review (help)".

I would guess that most IP vandals don't pay much attention to what they're doing. They're not going to notice that box or anything about pending changes before they commit the vandalism. They will check to see that they're insertion is in the article. Okay, so their vandalism won't show, which is obviously good. But, it was already done, thus making work for reviewers.

We can do more to discourage the vandalism from occurring in the first place if we make it crystal clear that their changes won't be live until reviewed. Would you rather have to kill termites, or completely stop the termites from entering to begin with?--Omarcheeseboro (talk) 15:55, 22 June 2010 (UTC)

If an article is moved from long term semi protection and there is over a period of time (a few days a week), say 50 or 60 vandalistic type edits from unconfirmed accounts all of which had to be reverted by a reviewer then I would myself take the article back for return to semi protection, all the reviewing work is only worthwhile if there is actually a percentage of beneficial additions from uncomfirmed users. Some articles will not benefit at all from pending and it should not be considered as a replacement for semi protection but another tool that may be helpful in some situations. Really for it to be used you need a decent percentage of beneficial contributions from uncomfirmed users, if that is not happening then return the article to semi and the occasional constructive contribution from an uncomfirmed user will as they do now find their way to the talkpage to request the edit be made. Off2riorob (talk) 09:01, 22 June 2010 (UTC)

I agree with you, and my understanding is that this is how it will work. Even though vandalism wouldn't appear live, having such articles on level 1 PC may take up too much reviewer time and cause pointless entries in the log. — Martin (MSGJ · talk) 13:42, 22 June 2010 (UTC)

...at which point, we raise the level of protection from PC1 to semi. Mjroots (talk) 19:38, 22 June 2010 (UTC)

Key points that aren't being explicitly mentioned: BLP is always going to be a problem. The solution is not some ideal case of perfect security but of speeding up oversight and limiting consequences. Pending changes keeps a constant watch on BLP pages (those which use it). As for limiting the consequences, this is the crux of the matter: a BLP violation which a few thousand reviewers see is not ideal, but as long as an edit doesn't get picked up by rest of media and the web, sanctioned by public display to all users on the Wiki, it's damage is minimized. 69.142.154.10 (talk) 01:19, 23 June 2010 (UTC)

Using pending changes as a means of complete unprotection of articles

In a few of the articles under pending changes that I have watchlisted (for the sole reason of this trial), I note that vandalism by non-autoconfirmed users is rather far and few inbetween; examples of articles with little unconstructive edits over the past week include the following:

This is very valuable information that can be used to gauge whether or not articles can be completely unprotected – something that we would not have without pending changes. That is, with our current practice of unprotecting articles (without factoring in pending changes), it's similar to "rolling the dice" and seeing whether or not vandalism picks up again and hoping for the best. This gives us a good system in evaluating many of our longtime-protected articles and seeing if protection is still necessarily. –MuZemike 16:15, 22 June 2010 (UTC)

I think it cannot be stressed enough that during the trial the rate of vandalism from school IPs is normally vastly reduced from what we see the rest of the year. I would be wary of drawing too many conclusions without considering the seasonal trends. -- zzuuzz ^(talk) 19:46, 22 June 2010 (UTC)

Reviewer rollback edits need to be reviewed??

Since we're discussing the merits of WP:PC here I wanted to bring up something else that concerned and confused me. Why is it that when I went to check a pending revision at Avatar (2009 film) and I clicked on "rollback (vandal)" (diff here), the system required my reversion edit to be reviewed? I am a reviewer, so why is it that I needed to go back and accept my own rollback of the IP's vandalism?. — CIS ^(talk | stalk) 21:58, 22 June 2010 (UTC)

[Rollback (vandal)] is a Twinkle feature. What Twinkle actually did was creating a derivative revision based on the unreviewed revisions, which is not automatically accepted even if you are a reviewer.
I fixed that earlier today, see #Auto-accepting reviewers edits. You might need to bypass your browser cache to get the updated version.
Amalthea 22:04, 22 June 2010 (UTC)

Thanks, I've gone ahead and done that. — CIS ^(talk | stalk) 22:07, 22 June 2010 (UTC)

Unaccept on Ernest Hemingway

I rollbacked vandalism, then rollbacked my change by mistake and now it wants me to review my rollback and the unaccept bottom won't display. What do I do? Truthkeeper88 (talk) 00:20, 23 June 2010 (UTC)

Unaccept means remove the acceptance of an edit. If the edit hasn't been accepted, you can't unaccept it. If the current revision is good, then you need to accept it. --Tango (talk) 00:24, 23 June 2010 (UTC)

It was bad and vandalism got in, which then I had to change manually. How do you reject an edit? Truthkeeper88 (talk) 00:27, 23 June 2010 (UTC)

You just revert it. It doesn't matter whether old revisions are accepted or not, only whether the latest revision is (and, if it isn't, what the latest accepted revision is). --Tango (talk) 01:36, 23 June 2010 (UTC)

I Don't Like Where This Is Going...

This is totally against what I thought was the Wikipedia Ideal. It also appears as a "sneak in" in it's timing... It appears to me as blatant CENSORSHIP. It is time for the rival to be started: EUROPEDIA based in the EUROPEAN UNION (ie. Brussels) There is some weird neo-con evil force taking over Wikipedia.... I also really resent the recent changes in appearance etc. Why do things self destruct? It is all very dis-heartning. But as soon as Europedia starts or this doesn't stop I will be ouuta here! Also the "accepted version box" on top of the Aristotle article is like a fist in the face in an undisclosed location..... The Encyclopedia of the World should be out of the Hegemonic power of the world. And we all know that that power is no longer the United States of America. It is in the European Union. You can check that in the CIA World Factbook...--Oracleofottawa (talk) 01:24, 23 June 2010 (UTC)

What difference does the country the site is based in make? This is the internet, it's global. --Tango (talk) 01:38, 23 June 2010 (UTC)

A Better Example..

Think of the Founding Fathers as they wrote The Declaration of Independence all those years ago and in the top right corner their was a box that contained the words: Latest Accepted Version... --Oracleofottawa (talk) 01:58, 23 June 2010 (UTC)

I like this idea

I realize that I may be in the minority her, but I am so sick and tired of rolling back vandalism caused by IP editors. Since long term article protection is so difficult to get, this will help. Every time I have tried it end up being for a week at most, but the vandalism starts up right after expiration. I think this is great. It will prevent vandalism from ever getting threw, while still allowing the majority of good editors to make good edit. This stops the problem before it is a problem. However, I do realize it is a change, so there will be headaches for a time, but I think in the end it will be a good things.--ARTEST4ECHO ^talk 01:43, 23 June 2010 (UTC)

Automatic acceptance of IP edits

How come this IP edit [2] was automatically accepted? DrKiernan (talk) 07:14, 23 June 2010 (UTC)

It was a reversion to a previously accepted version. NW (Talk) 08:01, 23 June 2010 (UTC)

Compatibility with suppression / revision hiding

Sorry if this has already been brought up and it probably doesn't really matter that much, but there is a "feature" where if the username or IP has been suppressed, you can unaccept the revision, but you cannot accept it without unhiding the username. --B (talk) 19:43, 23 June 2010 (UTC)

Bots

Should some bots that deal with vandalism be given review rights? At the time of this comment, the Big Bang page had vandalism that was reverted by ClueBot (at 23:22 GMT June 23, 2010). The edit was still pending for review. If bots are given reviewing rights, they would be able to make the job of the reviewers easier. It might have to be given out selectively based on the bot's task, but for bot like ClueBot that are just reverting edits, they could automatically mark the edits as reviewed. I do think that if there are other edits pending, should be made to not mark as reviewed. Rabbitfang (talk) 23:50, 23 June 2010 (UTC)

The ClueBot edit was marked as "automatically reviewed" and so it does not need to be manually accepted. This happens whenever any autoconfirmed editor (bot or not) reverts to a previously-reviewed version. However, the bot will not actually be marking edits as "accepted" so it does not need to have review rights. — Carl (CBM · talk) 00:14, 24 June 2010 (UTC)

Comment. If Level 2 reviewer protection (see the article) does begin to be implemented, bots will require reviewer status; otherwise, reviewers will have to approve ClueBot's edits, among other bots. CycloneGU (talk) 06:07, 24 June 2010 (UTC)

Pending changes for talk pages?

I'd like to see if I'm just laughed at on this one: we generally don't semiprotect talk pages, but I've strongly considered it on some (like Talk:Lil Wayne) because of the tendency for anons to use them as discussion forums. Would putting them under pending changes control fall under the trial, or is it just too wild of a thing to test?—Kww(talk) 00:11, 24 June 2010 (UTC)

Question

I have a question about this. If pending changes is a success after the trial and is kept, will semi-protection and full-protection be deleted from Wikipedia? I am asking this because some pages need protection. The main page, for example, really needs full-protection; we can't have risks of anyone vandalizing it, and it is very possible for a reviewer to vandalize it, as a reviewer is not necessarily a veteran on Wikipedia and can suddenly start vandalizing right when they get reviewer status so that their edits will immediately be visible I know this is highly unlikely to happen to pages, but still, it can happen very easily, so I would like to know if semi-protection and full-protection will go down the drain or not. I know this seems like a weird question, but I am concerned about what will happen if semi-protection and full-protection do go down the drain. --Ha dg er 00:54, 24 June 2010 (UTC)

I don't think that will happen. There's still the possibility of vandals crapflooding an article's edit history, not to mention highly-visible and used templates which should not be edited. –MuZemike 00:59, 24 June 2010 (UTC)

Requests?

Is there a way to request articles be PC enabled? Or will that have to wait until the testing period is over? Glenn Danzig is a pretty high-profile BLP that could really benefit from PC1. <>Multi‑Xfer<> (talk) 05:15, 24 June 2010 (UTC)

You could add it to the queue yourself and one of the admins administrating the trial will apply the PCP to it. 山本一郎 (会話) 05:47, 24 June 2010 (UTC)

Ah, now I see the "Queue" link up at the top. Thanks. <>Multi‑Xfer<> (talk) 06:02, 24 June 2010 (UTC)

it sucks

No one wants this. 152.31.193.40 (talk) 18:26, 22 June 2010 (UTC)

^{[citation needed]} –MuZemike 18:56, 22 June 2010 (UTC)

Except for a large majority of those that responded to the poll, you mean? --Tango (talk) 19:03, 22 June 2010 (UTC)

As I recall the people responding to the poll thought it would provide extra protection, especially for BLPs. What we're seeing here is BLP protection being removed by downgrading from semi-protection to level one PC. SlimVirgin ^{talk|contribs} 19:13, 22 June 2010 (UTC)

It seems to be pretty effective to me, for the reasons explained above. I didn't feel that this thread was particularly constructive so I collapsed it in good faith, but I'll leave it expanded since SlimVirgin reverted. GiftigerWunsch [TALK] 19:19, 22 June 2010 (UTC)

It seemed like a good idea, but after experiencing being a reviewer, I must agree with the anon, the implementation sucks. Any time a "pending change" shows up for me to review, the page takes about 10 times longer to load than normal. This isn't efficient, and seems to be putting a burden on the server. ~Amatulić (talk) 20:27, 22 June 2010 (UTC)

I was just about to come here to post this about the excruciating loading time wait. It feels like I'm on dial up when I'm trying to review any pending changes. If something cannot be done to make this more speedy, it seems to me that it's unnecessarily wasting the valuable editing time of reviewers. — CIS ^(talk | stalk) 20:30, 22 June 2010 (UTC)

Yeah, it does take a LOT longer to load than it normally would. I didn't know if it was just me, or if it was taking longer for others too, but now i know that it's not just me. At first it sounded like a good idea, but now it's not very efficient. as Amatulic said, it can't be doing much good for the servers if it takes this long to load. Maybe if you can make it faster without spending a load of money, it would be a good idea but until then, it's just not gonna work Just Sayin' Pilif12p : Yo 20:33, 22 June 2010 (UTC)

FWIW the next WP:TFA, Ernest Hemingway, will be our dry run at PC on the main page. Should be a barrel of laughs, since the article is a constant vandal magnet. Truthkeeper88 (talk) 20:36, 22 June 2010 (UTC)

Heh, and Michael Jackson two days after that, which will be even worse I think :) Amalthea 20:50, 22 June 2010 (UTC)

Concerning slowness, I think that this is solely due to selection bias: Many featured articles are now downgraded to PCP1, which are all very template-heavy. I was about to file a bug report about the slow load times as well, but after comparing it with other FAs of similar length and with similar number of citations, I couldn't see a difference in load time any more.
If changes could be accepted from the page history, after checking a diff with popups, the process would be much faster.
Amalthea 20:39, 22 June 2010 (UTC)

Hi Amalthea, you wrote: "Many featured articles are now downgraded to PCP1, which are all very template-heavy." Do you mean the FAs are template-heavy, or that there's something about PC that makes interacting with templates slower than it normally is? The reason I ask is that it always very slow, which is one of the reasons several of us argue to avoid using citation templates. So I'm wondering: is it the citation templates causing the problem, or is there something over and above that with the PC protection? SlimVirgin ^{talk|contribs} 20:43, 22 June 2010 (UTC)

I think that it is independent of PCP, and solely due to the templates. Looking at this diff from the unprotected TFA takes 18 seconds to render as well. The problem with PCP is simply that patrollers now have to open these diffs much more often than before. Amalthea 20:50, 22 June 2010 (UTC)

Well, in that case, if PCP is going to show the community how citation templates make articles slow to load, then I'm all for it. :) SlimVirgin ^{talk|contribs} 21:38, 22 June 2010 (UTC)

No, I don't believe it's selection bias causing slowness. Articles that I used to edit regularly with no problem are now excruciatingly slow to load when reviewing a change. It's gotten to the point where I am now ignoring "pending changes". I'm donating valuable time to this project, and I don't need it wasted. If I see a pending change, I move on to other things. I daresay others may feel the same way, and the last thing we want is for a new "feature" to chase off constructive editors.

Also, have a look at Talk:Avatar (2009 film)#"Pending changes". This new PC thing has resulted in the destabilization of a pretty good article, with reviewers allowing in vandal edits. ~Amatulić (talk) 20:51, 22 June 2010 (UTC)

Do you have an example? It's possible that it generally got slower altogether, I don't want to rule that out. But as far as I can tell there is no difference between editing an unprotected page and a PC-protected page if they have the same content. Amalthea 20:53, 22 June 2010 (UTC)

I gave an example. Avatar (2009 film) was never so slow in displaying a diff. ~Amatulić (talk) 20:56, 22 June 2010 (UTC)

I turned PCP off for that page since it only got noise. Diff rendering doesn't get fast though, still takes 20 seconds to render this diff, as it did before. Amalthea 21:08, 22 June 2010 (UTC)

It was on for less than 24 hours. How is that long enough to determine if it is useful? We all knew this would increase vandalism, the question is whether there are enough good edits to be worth it. You can't tell that in less than a day. --Tango (talk) 22:28, 22 June 2010 (UTC)

If it disrupts the article to the point where the regular editors must complain, then there's no reason to keep PCP active. It probably works better on some articles than others. ~Amatulić (talk) 23:05, 22 June 2010 (UTC)

Sure, but you need to give it a chance. --Tango (talk) 23:06, 22 June 2010 (UTC)

It takes edits to judge that, not necessarily time. With ten out of ten anon edits reverted it's not worth it, in my eyes. Amalthea 10:08, 24 June 2010 (UTC)

Just compared current rendering times with old ones. Reference taken from WT:Citing sources/Archive 27#Making pages faster to load, I used this revision and this PCP copy. Ten tries, render times in seconds:

Late 2009	26	23	16	16	22	16	28	28	16	16
June 2010, unprotected	25	17	26	25	26	17	18	25	17	17
June 2010, pending	17	26	25	17	17	17	17	17	26	29

Based on that limited test the render times appear practically unchanged, in both protection scenarios. The two classes of rendering times depend, I assume, on two classes of server setups that render the page. Amalthea 21:46, 22 June 2010 (UTC)

That is also one of my concerns with pending changes is the increased loading times as mentioned above. I daresay it's about as bad as the edit filter, knowing how long it takes to load a filtered edit. –MuZemike 22:03, 22 June 2010 (UTC)

Actually, I think I just showed that there is no indication that load times of diffs got significantly slower? Amalthea 22:09, 22 June 2010 (UTC)

Hmm, yes. I wonder why the Avatar article seems slower now than a week or so ago? By the way, thanks for removing PCP from that article. ~Amatulić (talk) 23:05, 22 June 2010 (UTC)

Amalthea, are those figures while there are pending changes in the queue? –xeno^talk 23:15, 22 June 2010 (UTC)

I also seem to be getting slower load times, more notable when there is a pending change in the queue. History, dif, 'review' all seem slower— Very unscientific observation by First Light (talk) 04:04, 23 June 2010 (UTC)

Anecdotally, I've seen this issue at the Barack Obama article - definitely one of our most template-heavy ones, to be sure, but I've seen a perceptible difference in loading times depending on whether an edit is pending approval. The kicker for me is that loading appears to be faster if someone approves a revision before I've loaded it. I think this is a real issue, not only an issue of perception. — Gavia immer (talk) 04:28, 23 June 2010 (UTC)

Makes no difference whether the diff I look at is accepted, not accepted, or not PC-protected at all.

Diff, PCP, unaccepted:	17	17	17	17	17
Diff, PCP, accepted:	25	17	17	17	17
Diff, no PCP:	17	26	17	17	17

That's a bit weird though, actually. When I look at a diff to the current revision, it shouldn't have to render anything since the latest revision is almost always cached somewhere. Was that always that way? The whole perceived slowness may be a general regression. Amalthea 10:05, 24 June 2010 (UTC)

Crap idea

This accentuates the elitist, panel-based editing practices of Wikipedia. I don't like it. 80.225.185.63 (talk) 03:45, 23 June 2010 (UTC)

Nobody's forcing you to edit, are they? Situation here is exactly the same as when I'm making a small contribution to Wikinews - I make the edit, another editor reviews it and all can then see the result. If you're not making problematic edits there's nothing to worry about. Mjroots (talk) 19:25, 23 June 2010 (UTC)

And pending changes protection is only being applied to pages that are already semi-protected. So the IP editor would not have been able to make that change at all without posting it on the talk page and having someone else put it in the article. So I consider pending changes protection an improvement over that. Reach Out to the Truth 21:38, 24 June 2010 (UTC)

Pending changes

Not sure this is the correct place to ask this question, but with regard pending changes, if you review an edit and it is not accepted is it counted as a revert, in regard to 1RR articles would it be counted as a breach if you made another revert? Mo ainm~Talk 16:10, 23 June 2010 (UTC)

Yes, it should be. Reverting an edit that is under pending changes is no different than reverting an edit that is not under pending changes. The whole editing process is identical; the only difference is that some revisions are marked as "reviewed". — Carl (CBM · talk) 18:28, 23 June 2010 (UTC)

Thanks for the reply, I think this should be made very clear on article that are under review and also 1RR, because I can see some drama about blocks resulting from editors saying they didn't realise it counted. Mo ainm~Talk 18:41, 23 June 2010 (UTC)

Yes. I'd suggest putting it in the edit notice of the page. — Carl (CBM · talk) 19:23, 23 June 2010 (UTC)

There is no option for rejecting an edit, you just don't accept it and revert it using the usual methods. I can't see why anyone would think that didn't count. --Tango (talk) 22:16, 23 June 2010 (UTC)

Well Tango I assumed it counted but wasn't sure and as I said I can see drama when blocks come and peple claim they didn't think a review counted so what harm in having a notice on

pages to avert any claims of not knowing. Mo ainm~Talk 07:56, 24 June 2010 (UTC)

Question about the scope of this trial run

Ok, let me admit right here and now that I stopped paying attention to the debate on this a long time ago as it seemed it would drag on forever. I had the impression that this was going to be trial run on BLP articles, but it seems it is being used elsewhere now as well. Did I miss something or was the scope expanded to cover other articles? Beeblebrox (talk) 21:44, 23 June 2010 (UTC)

It appears to cover other articles too. I know because I generally don't look at or edit BLP articles at all. I imagine there are other reviewers like me out there. Restricting the trial run to BLP articles would basically shut out reviewers like me who have no interest in BLP articles.

I thought it was a good idea, but in practice, I find it more irritating than useful. I have enough busy-work to do without the addition of pending changes. ~Amatulić (talk) 21:48, 24 June 2010 (UTC)

Comment Isn't Pending Changes meant to be handled by the new group of reviewers? While admins. have access to it, they are not required to use it themselves if they don't wish to, am I correct? After all, that's the only thing I presently do myself so I would handle them, and report to the requests page if I find vandalism excessive in any one article, BLP articles included. CycloneGU (talk) 22:22, 24 June 2010 (UTC)

Official poll area?

Is there a specific place to vote on passing this new policy yet? A p3rson ‽ 02:58, 24 June 2010 (UTC)

As far as I know there will be a poll for this proposal after the trial finishes. 山本一郎 (会話) 05:49, 24 June 2010 (UTC)

My understanding is that the trial will finish, there will be a short break whilst some statisticians analyze the trial, and then there will be a poll. I did see this written down somewhere, but I'm afraid that I can't find it again... Mike Peel (talk) 07:46, 24 June 2010 (UTC)

I know that there will be a poll after, but what about during the trial, to see if anyone even wants it? A p3rson ‽ 02:44, 25 June 2010 (UTC)

Suggestion: 'Recent changes' list specifically for pending changes

Unless there's a link I haven't been shown, pending changes essentially get dealt with as they show up on the recent changes page. Now granted, that's not a huge problem now, but sometimes edits can go through the list pretty quickly, and it would be convenient to just be able to look over all the pending changes there are instead of only concentrating on a page at a time. HalfShadow 19:50, 24 June 2010 (UTC)

Special:OldReviewedPages will show you pages that currently have pending changes. Reach Out to the Truth 21:49, 24 June 2010 (UTC)

High-profile pages

I'd like to request that we stop adding high-profile vandalism targets to the queue, such as Adolf Hitler, Enimem, and John Lennon. Invariably the vandalism continues and dealing with it creates extra work, which seems pointless. Would the PC test not make more sense on articles where the effect was harder to predict? SlimVirgin ^{talk|contribs} 19:09, 24 June 2010 (UTC)

It makes sense to test PC on all types of articles and maintain semi-protection on articles that are deemed to still require it. I thought Eminem was reprotected or requested such already, surprised if it isn't. Same with Hitler. I know even Barack Obama was unprotected, and put on PC, but I think that one is still being debated to be kept at a minimum for testing (if I'm wrong, I'll find out, no need to edit a reply).

In short, I don't mind putting even possible vandalism targets in the testing queue for PC. We'll reprotect those that still need it. It's good to test now and then whether a page under protection still needs it. CycloneGU (talk) 21:58, 24 June 2010 (UTC)

It's extra work for no gain. SlimVirgin ^{talk|contribs} 04:33, 25 June 2010 (UTC)

Depends on your view. I saw an admin. somewhere within the last couple of days (I forget where) saying that articles are routinely removed from semi-protection - not sure how frequently - as a test to see whether the status is still needed. Sometimes, that is a fail. Is that what you also call "extra work for no gain"? This is no different; it's another "test" of an article's semi-protection necessity. CycloneGU (talk) 05:29, 25 June 2010 (UTC)

We don't normally test articles where we know what the result will be. SlimVirgin ^{talk|contribs} 14:21, 25 June 2010 (UTC)

The problem here is that too many were 'tested' at the same time. I'll cite what I said at WP:PCF: "There are plenty of pages where one could expect substantial vandalism would result, but actually didn't: Ronald Reagan, Richard Nixon, Rome, etc. The vast majority of pages downgraded from SP to PC didn't get substantial vandalism. Though I think we should have spaced more the downgrades from SP to PC on articles where substantial vandalism would be likely, and should do so from now on. I'll try to come up with a policy statement.". Cenarium (talk) 22:12, 25 June 2010 (UTC)

One of the problems is that this is being discussed on RfPP, protection policy, this page, the PC feedback page, AN, AN/I, and individual article talk pages. It would be good to have one page for discussion so that consistent themes are identified. SlimVirgin ^{talk|contribs} 22:27, 25 June 2010 (UTC)

Agree with SlimVirgin - it's hard to keep track of all the discussion pages. Just added Wikipedia:Pending changes/Noticeboard to watchlist, and haven't really tallied how many pages and threads are taking places in various places, but the conversations are quite scattered. Truthkeeper88 (talk) 22:46, 25 June 2010 (UTC)

Yes, discussions happen a bit of everywhere. Here should be the place for substantial discussion of policy, the noticeboard for specific cases, the feedback page is just for feedback, not thorough discussion. I've proposed the policy statement here. Cenarium (talk) 23:11, 25 June 2010 (UTC)

Query

I'm trying to copy-edit the page but find I can't understand parts of it. Does anyone know what the following means (confusing parts in bold)?

1. Logged-in users, and anonymous users who click the "edit this page" or "pending changes" tabs, will see the latest changes as usual. Any pending changes on such pages are visible to all users via an additional tab.

Is the additional tab in the second sentence different from the "pending changes" tab in the first one?

2. Any pending changes on such pages are openly visible to all users via an additional tab, and the presence of new edits in the queue is explicitly clear to help reviewers.

What does it mean to say the presence of new edits is "explicitly clear to help reviewers"?

SlimVirgin ^{talk|contribs} 14:17, 25 June 2010 (UTC)

Now this is a new phrasing for me, I thought anons. didn't see changes until approved. Also, I think the "additional tab" is the pending changes one, but I have not seen the tab yet.

They appear in an orange background at the top of the edit summary. This is the only way I know of, and it's certainly clear. Go to OldReviewedPages and click any article with a pending edit, and visit the edit history without reviewing the edit. Once reviewed, it behaves as normal. CycloneGU (talk) 14:25, 25 June 2010 (UTC)

Yaris has cleared up the first point (thank you, Yaris!), so it now says: "Both logged-in users and anonymous users who click the "edit this page" tab will edit the latest version as usual. Any pending changes on such pages are visible to all users via an additional "pending changes" tab."

I've removed the second point about "explicitly clear to help reviewers." SlimVirgin ^{talk|contribs} 16:46, 25 June 2010 (UTC)

Articles where considerable vandalism is probable if downgraded from SP to PC

I propose we add that "Attempts to downgrade from semi-protection to pending changes protection on articles where it is probable to result in considerable vandalism should be spaced on the duration of the trial.". Because too many at the same time can add too much burden on recent changes patrollers, reviewers and editors, as it did to some extent since the beginning of this trial. On the other hand, several articles where one could expect substantial vandalism have received little or not at all. So it shouldn't be abandoned completely, but spaced in time. Cenarium (talk) 23:09, 25 June 2010 (UTC)

Protection tactics

Depending on the cases, it can be more appropriate to use (1) PCP indef or temp, (2) SP indef or temp, or (3) PCP temp or indef and SP for a short time. (1) best for articles meeting the requirements for protection but with constructive edits by unconfirmed users and no massive vandalism (2) best for articles (recently) subject to massive vandalism with no constructive edits (3) best for articles with (recently) massive vandalism but also constructive edits by unconfirmed users. To explain the later, in such cases PCP may not be able to 'break the vandalism cycle' so a short use of SP would be needed, but the article may still be 'vulnerable' for some time, so PCP could be applied for a while, or indef if persistent. Cenarium (talk) 02:47, 26 June 2010 (UTC)

I don't get it

Anything marked "pending changes" "automatically accepted," or "accepted by" seems to be incredibly slow to load, and when you click on the latter two you are taken (very slowly) to the current revision of the article. The rollback feature doesn't seem to work on things marked "pending changes," yet the "unaccept" option (which is a very peculiar word; what does it mean?) isn't selectable.

Also, I don't really get the point. As the current discussion over the protection level on Eminem shows, this isn't actually preventing vandalism at all. As I understand it, the idea is that casual users won't see the vandalism, only registered users. But Wikipedia currently has 47,323,110 named accounts, and all of those users can potentially see any vandalism/libel/deliberate errors. So pending changes is simultaneously downgrading the protection level, increasing the amount of work required to deal with vandalism, and increasing the amount of vandalism that will be seen on a given article. What is the point of preventing "vandalism or inappropriate changes from ever being displayed publicly outside the editorial community" when the "editorial community" is so vast? This really would only be a sensible idea if the number of editors was very small -- in the hundreds, perhaps.

In fact, it seems that the only type of user this is helpful for at all is a user who is editing while not logged in. Such an editor who wished to make a useful edit would previously have had to request changes on the talk page of a protected article, where the changes might have actually had to be discussed, while an anonymous vandal would have been prevented from vandalising the article. But under this system, vandalism isn't prevented at all: any vandal who isn't logged in can simply edit the article, and their edits will still be seen by potentially millions of users, while the task of preventing their vandalism is now passed on to admins and the new "reviewers." Really. What is the point? Exploding Boy (talk) 19:31, 23 June 2010 (UTC)

Very nice summary. I don't get it either. Truthkeeper88 (talk) 22:09, 23 June 2010 (UTC)

Me neither. It all seems like make-work. SlimVirgin ^{talk|contribs} 03:17, 24 June 2010 (UTC)

Eh. I think the "12 million accounts" argument is total hogwash. Do you have any idea how many of those accounts have been used in the past... 90 days? So many studies have shown that less than 1% of those accounts are even active and an even smaller number represent the core of Wikipedia editors. Yes, in some cases semi-protection should still be used, but this a wonderful solution to, as you said, protect our casual (not logged in) readers. JBsupreme (talk) ✄ ✄ ✄ 21:18, 26 June 2010 (UTC)

If a registered user does not want to see the vandalism, they can specify in their preferences that they want to see the latest accepted version instead. I'm not sure if this is documented in the help pages. If not, it should be. Reach Out to the Truth 22:24, 23 June 2010 (UTC)

I wasn't aware of that, but it's not really the point. For one thing, if everybody did that, then no edit by an anon editor would ever get through. For another, Pending Changes doesn't actually do anything to prevent vandalism, it only creates extra work in getting rid of it. Exploding Boy (talk) 22:28, 23 June 2010 (UTC)

(ec) It's worth remembering that the vast number of people that use Wikipedia aren't logged in. Also, the hope is that people who would make useful edits but would otherwise be stopped by semi/full protection can go ahead and make those edits without being put off / having to figure out the discussion pages as well as the edit functionality right from the start. Try putting yourself in a newbie's shoes, and thinking how they would react to protected pages the first time they encountered them and tried to edit them.

The slowness of the system is irritating, but I would hope that it is an issue that will be resolved over time (has anyone reported it to the developers?). Mike Peel (talk) 22:30, 23 June 2010 (UTC)

We often hear that argument, but how true is it really nowadays? Maybe when Wikipedia was just getting going people who encountered a protected page might think the "anyone can edit" thing wasn't true (even the notion that you could edit a webpage that wasn't your own was new), but Wikipedia is much better known now (often the very first search result when you look for something), and people are much more internet savvy than in the earliest days of WP. Also, we know that anon editors are more likely to vandalise than logged in ones. It would be very simple to make a better notice on protected pages telling users that they cannot edit this page at the moment, but they can edit nearly every other one -- we could even incorporate a "random page" link. I doubt that many people with useful edits to make would be confused or turned off to the point where they'd never try to edit WP again (but people intending to vandalise might be). Exploding Boy (talk) 01:05, 24 June 2010 (UTC)

Rollback can be used on articles with PC, to revert the latest editor as usual (and it automatically accepts the new revision if the revision to which it was reverted was accepted, at least native rollback, Twinkle rollback has issues). There are many registered accounts, but probably only a small minority used to read articles and not to edit. I don't expect that even vandalism on high profile articles would be seen by more than a hundred of logged-in users. And the editing community is relatively accustomed to vandalism happening on articles. So I don't see an issue in displaying by default the latest revision to logged-in users, setting which can be overridden.
While statistically there are more vandalism by unregistered users than by registered ones, it does not imply that unregistered users are more likely to vandalize than registered users, at least not to a proportional level. There's research work been done by the usability team which shows that many people are not fully aware that they can edit articles in a (relatively) straightforward manner. And since we made MediaWiki:Protectedpagetext more accessible to users, there has been many more edit requests; so there are definitely people out there trying to edit, most in good faith. And I've seen many good edits which would otherwise not have been made.
If the level of vandalism is too high, articles can be put back under semi. Citing my reply at AN: "Let's say PC adds a more flexible level of protection, which can be used in cases where SP is justified in terms of policy, but where using SP would be unnecessarily restrictive. There are several needs to balance: to allow constructive editing, to protect the encyclopedia from harm, to avoid wasting resources unnecessarily, etc; they should be considered on a case by case basis.". Cenarium (talk) 03:14, 24 June 2010 (UTC)

By the way, what happens when an unregistered user makes an edit to a ... Pending Changes ... protected? ...page? Are they able to see the change right away? Exploding Boy (talk) 14:38, 24 June 2010 (UTC)

Yes, they're directed to the 'pending changes' tab. Cenarium (talk) 14:47, 24 June 2010 (UTC)

So the system logs edits by IP, and if you've edited the page without logging in you're able to see... what? Only your own edits? Every unchecked edit? Exploding Boy (talk) 16:56, 24 June 2010 (UTC)

I'll take this one. You're able to see only the most recently "accepted" version. If another IP edits, and you, not logged in, go to view the page, you will not be able to see it; meanwhile, if you're logged in, you WILL see it unless you set not to; once it's approved by a reviewer, EVERYONE sees it no matter what. Now, if YOU are not logged in and edit...I presume here that you still only see the most recently approved version. CycloneGU (talk) 17:08, 24 June 2010 (UTC)

So then if an unlogged in editor makes an edit, s/he will not see that edit in the article until it's been accepted? Exploding Boy (talk) 19:01, 24 June 2010 (UTC)

I think Cenarium means that if an IP user (not logged in) edits, they will be informed that it has to be reviewed. They can continue to edit a second or third time, though that can also cause some of the mess we've seen with undos, etc., and likely would merely be a spammer NEway (though occasionally IPs do edit in good faith). However, an IP user would not be able to see their contributions in the article itself.

A good way to test this is to go to the PC testing page and make an edit while logged out, then see what happens on your screen. That gives you an idea then what any IP editor would see, then you can come back and verify with us. Heck, even in the edit say "Do not accept until 19:20, testing IP edits" or "Exploding Boy is testing something in regards to IP edits, please do not act on this edit" or something like that. CycloneGU (talk) 21:52, 24 June 2010 (UTC)

I was trying to avoid having to log out. Anyway, my point, if I've understood what happens correctly, is that it's only slightly less potentially frustrating for an unlogged-in user to try to edit a PC page than a protected one. Exploding Boy (talk) 22:11, 24 June 2010 (UTC)

Actually, anons. cannot edit ANY protected pages, full or semi, at all. They can edit talk pages for them, however, in proposing changes. CycloneGU (talk) 02:11, 25 June 2010 (UTC)

You don't have to log out, just register a new account for testing (so you don't have reviewer rights) and set your preferences to view by default accepted versions. On pages with pending edits, there are two tabs: 'article' (showing the latest accepted version) and 'pending edits' (showing the latest version), logged in users see the pending edits tab by default, and logged out users see the article tab by default, except if they just made an edit to the article, in that case they're directed to the pending edits tab, with a note that their edit is not visible until accepted. Cenarium (talk) 14:01, 25 June 2010 (UTC)

I know that, CycloneGU. My point is that in one case the user can edit the page but not see the results, while in the other they can't edit at all; for them, very little difference. For logged-in users, more work in reverting vandalism on protected pages. Exploding Boy (talk) 21:14, 26 June 2010 (UTC)

Why bother?

This seems entirely too complicated. It will make Wikipedia harder to explain to new users. (I'm not all that new, and it's hard for me to figure out how this would work.) Since it's so easy to register, why not simply "semi-protect" all articles, and let anonymous IP users suggest changes on the talk pages (or maybe a "suggested changes" page connected to the article)? An anonymous user who doesn't get a response to a suggestion could simply register and do it oneself. Ruckabumpkus (talk) 00:09, 25 June 2010 (UTC)

Yes, I find it curious that an additional burden is being placed on registered users here just so anonymous users can experiment on actual articles, when the burden should be on the anons to register if they want to edit badly enough, and learn to edit responsibly. ~Amatulić (talk) 00:12, 25 June 2010 (UTC)

The system also serves for new users as well, I think, who are new to editing. I've seen a few actual user edits that required approval. CycloneGU (talk) 00:30, 25 June 2010 (UTC)

Why is it any more complicated for an anonymous user? They attempt an edit, it takes a few seconds to show up? That's probably less work or confusion than making an account, right? --j⚛e decker ^talk 01:11, 25 June 2010 (UTC)

Indeed. Even though there are small annoyances going on with the system at the moment, its greatest benefit will be a decrease in the amount of vandalism gets out there to the public (i.e. the media), which in turn will hopefully also benefit Wikipedia's reputation in the long run. My only beef with the system at this point is that the default for registered users is to see the pending changes at the article before they are accepted. I think the default should be the opposite (except for admins and users with "reviewer" status), because I'm sure that quite a few people in the media have accounts on here, and it would probably most benefit Wikipedia on the whole for the default to be not seeing the pending changes. — CIS ^(talk | stalk) 02:53, 25 June 2010 (UTC)

Don't forget that Stephen Colbert also apparently had an account on here once, too. Sure, it's satirical comedy, but it's television (media) and he refers to Wikipedia all the time. Not sure how often he's on here after his show, but who knows? CycloneGU (talk) 03:24, 25 June 2010 (UTC)

Why bother? To decrease the easy access of the ever present vandals that denigrate WP's reputation as a "reliable source." After a long running and ongoing battle with a banned editor who using a cell phone, changes ISPs as often as diapers, she has overtly demonstrated that some restrictions need to be placed. After a community ban and much discussion, the banned editor's response is a blatant "Catch me if you can, and you can't!" DocOfSoc (talk) 22:02, 25 June 2010 (UTC)

I still don't get why it wouldn't be sufficient to simply restrict all editing of articles to registered users. Given the ease of registering, and indeed the anonymity of a username, why is it important to allow anonymous IPs to edit articles? Ruckabumpkus (talk) 00:28, 26 June 2010 (UTC)

I agree with Amatulić, and since, as I type this, we have going on 13 million registered users, it seems utterly pointless to restrict the ability to view vandalism only to registered users, while simultaneously increasing their burden in preventing it. I would much prefer to limit the number of edits an unlogged in user can make, which would both automatically reduce vandalism and make tracking, reverting, and blocking vandals easier. Exploding Boy (talk) 21:11, 26 June 2010 (UTC)

Once you start restricting editing to only registered users, Wikipedia ceases to the "the online encyclopedia anyone can edit". Isn't that catch phrase supposed to be part of the draw to Wikipedia, or is Wikipedia to now become just another online social club where those who want to take part must first pass muster and be considered worthy? If that's what it becomes, then the Wikimedia Foundation can start to kiss its donations goodbye. 70.197.197.195 (talk) 01:41, 27 June 2010 (UTC)

I can only imagine what the reaction was when "semi-protection" was implemented some 5 years ago. –MuZemike 01:45, 27 June 2010 (UTC)

It's never been strictly true that "anyone" can edit Wikipedia. First, you have to have internet access. Second, you have to be using an internet service that doesn't filter out Wikipedia (as many public schools do in their computer labs, and I expect some autocratic regimes do, too). Is there in fact anyone who meets those two criteria who can't also register a username? If not, then allowing only registered users to edit articles would change nothing about the truth value of the assertion that "anyone" can edit Wikipedia. Ruckabumpkus (talk) 02:10, 27 June 2010 (UTC)

In any case, requiring people to go through a quick, simple, and free registration process doesn't restrict them from editing, they only have to do it once, and there are benefits to doing it, not least of which is increased privacy. Exploding Boy (talk) 02:22, 27 June 2010 (UTC)

Exploding Boy has a valid idea, actually. I like having unregistered IP editors only allowed to edit a little more periodically, say once every minute, maybe two. That way they are still able to contribute, and they can be alerted even on the editing page of this restriction. By the time a spammer finishes one edit, within two minutes it's likely reverted and admins. have time to start typing a warning to the user before they can contribute the next spamdit (my term). CycloneGU (talk) 03:49, 27 June 2010 (UTC)

Or, say, five times in any 24 hour period. The actual limit could be tweaked up or down until a good balance is found. Ruckabumpkus (talk) 03:54, 27 June 2010 (UTC)

That wouldn't be a good balance. Sometimes, an editor may want to edit several articles, make a typo that they waste an edit correcting, realize they could have said this thing better...and by then, they've used three or four of their five edits. I don't like setting a 24 hour limit, but a limit based on a few minutes at a time sounds like a good idea. And while Jimbo and anyone in that hierarchy would generally want to keep IP editors here, I'd hope this is something that might be feasible as it's a very minor restriction I'm proposing here, and any non-spammer editor will need a minute or more to prepare their edits NEway. CycloneGU (talk) 04:03, 27 June 2010 (UTC)

Vandals I've dealt with wouldn't be deterred by having to wait mere minutes to strike again. In one case, the vandal was interested only in inserting one particular POV statement in a lead paragraph (see Special:Contributions/173.53.251.130). I don't think it's too onerous to ask frequent editors to register. Ruckabumpkus (talk) 11:39, 27 June 2010 (UTC)

Regular editors frequently forget the following:

You likely started out as an i.p. too

Registering raises concerns about email confirmation, spam, new passwords, divulging personal information, etc. It's on Wikipedia to convince users how easy it is to register. We haven't.

Having a username requires logging in EVERY time you get to the site and want to edit it. Many non-advanced computer users do not employ form-savers or automatic logins, and thus want to be free of the inconvenience of multiple or repetitive logins.

Solutions to vandalism which compromise the ease of usability for newcomers are likely to undermine its growth. Tools like Huggle and Twinkle are ideal technological solutions to complex social dynamics. Better to let the technology handle it if possible: people are much messier.

As mentioned above, pending changes keeps an ongoing log of edits to controversial pages. Though the number of reverts might increase dramatically, the difficulty of getting rid of them does not.

i.p editors are not just vandals 69.142.154.10 (talk) 17:29, 27 June 2010 (UTC)

Most people start as ips, but you only have to register once. The procedure can't be that onerous or I wouldn't have done it. I don't recall having to provide any personal information, and you don't have to activate email. In all my years as a registered editor I haven't received one spam email through Wikipedia.

You only have to log in again if you are using a different computer, otherwise there is a clear and easy option to remain logged in for 30 days.

I don't see how Pending Changes is really making it any easier for newcomers to use Wikipedia. It seems to me it would be more confusing. But limiting the way that ip editors can use the site seems entirely reasonable, especially when we know such users are more prone to vandalism. Exploding Boy (talk) 17:48, 27 June 2010 (UTC)

The IP is saying that people will be worried about "email confirmation, spam, new passwords, divulging personal information", simply by virtue of the fact that most sites DO have that issue. It's not that WP doesn't, it's that they don't know it doesn't. And the IP has a point about needing to log in. Some of us may have an auto-log in, but many people edit from libraries and schools, or may have a shared computer and not want/be able to use a saved form. ♫ Melodia Chaconne ♫ (talk) 18:10, 27 June 2010 (UTC)

I'm going to go way out on a limb here and say that unregistered editing has outlived its usefulness on Wikipedia. Our goal here is to produce the best encyclopedia we can. The value of unregistered editing is that it lowers the barrier to entry. That was important in the early days when not many people knew what WP was. Today, WP is one of the most well-known sites in the world. I'm sure there are some people who get the idea in their head that they want to try editing, but wil be dissuaded by the need to create an account. Some of those people have the potential to make valuable contributions, and the fact that they don't sign up will be our loss. But, on balance, we'll be better off. I don't expect this opinion will be very popular, but there it is. -- RoySmith (talk) 18:33, 27 June 2010 (UTC)

:"Our goal here is to produce the best encyclopedia we can". The problem with that statement is that while it may be the best known editable encyclopedia online, it is not now, nor will it ever be, the best online encyclopedia. As long as anyone can edit and it is a continually evolving entity, Wikipedia will never be a true encyclopedia that is used as a reliable source and reference. If college and university (or even public school) students are not allowed to use Wikipedia as a reference for school work, it's not a real encyclopedia.

Further, as long as the cliques, clubs, rogue editors, and those with a certain stamp of approval are allowed to go unchecked as they bully and intimidate new and experienced editors alike, Wikipedia will remain the joke it truly is among the educated (and those seeking to be truly educated). And let's not forget the behind the scenes politics where deals are made as to who can stay, who is expendable, and who gets blocked or banned (if you think there are no such agreements, you're wrong). Wikipedia has become a social networking site that only looks on the surface like an encyclopedia. In reality, it's become no different than Usenet, Facebook, or MySpace (with the exception that Wikipedia appeals to shut-ins, geeks, nerds, and obsessive, neurotic control freaks). Set alongside the social messaging and networking sites, Wikipedia is the same brand - it just has different packaging.

I agree with the IP editor above: shut down the possibility for IP editors to edit, and donations will likely suffer. 97.50.112.120 (talk) 19:25, 27 June 2010 (UTC)

I'm going to throw my support behind Exploding Boy here for his point on registered users viewing vandalism while increasing their burden in preventing it. I also want to underline 69.142.154.10's points that IPs are not just vandals, and we all started as IPs too.

I am concerned that this policy will alienate new IP editors, and deter them from registering. Does anyone remember that we encourage people to be bold in updating pages? I think this is a barrier to that boldness. I think it is also a barrier to the learning curve of Wikipedia, because I am also concerned that this function will lead to some reviewers and admins rejecting IP edits based on 'poor grammar' and unsourced edits. Has anyone thought about notifying these IPs of why their edit was rejected, and the burden that will come with that? Who can guarantee that we are going to be able to do that when, in my early editing days, I saw alot of IPs who didn't even get warning messages on their talk pages when they committed blatant vandalism?

While vandalism is a chronic problem and I'm not about to deny that, I think this proposal would be a greater evil when it is Wikipedia's goal to be bold, and teach editors of how to constructively edit here with neutral and cited material. And the only way we can continue creating a reliable, free encyclopedia is making the editing experience more welcoming so as to encourage more users and more editing ,and addressing these faults in quality directly at the source, not erecting new barriers around our encyclopedia to prevent some vandalism and trying to filter the best edits.

13 million registered users is quite an accomplishment, but its not even half of the population of my country, and this is open to the world! I can see the growth of that userbase start to decline if you implement this, and that will mean less editors, more burden and burearacy for our existing users, less content contributed, and therefore a worse encyclopedia that anyone^*††‡ can edit. --Natural R_X 19:14, 27 June 2010 (UTC)

Not liking this

This was implemented on Burger King and withing an hour the IP vandalism started. What good is this? --Jeremy (blah blah • I did it!) 18:35, 27 June 2010 (UTC)

A little too early to complain, isn't it. I mean, if you look at the page history, there's only been three edits since pending changes was added to the page. Sure, one of them was IP vandalism, and another was the reversion, but that's hardly evidence to say, as a whole, that the "IP vandalism" has started. WTF? (talk) 22:02, 27 June 2010 (UTC)

A list?

Is there anywhere where I can get a list of articles that currently have pending changes. I'm not talking about a list of articles that are using the feature, I want a list of articles that currently need reviewing. Why? Because I want to experience reviewing before making up my mind about the trial. And I haven't done not one review edit, because every time I go to an article that uses the feature, the edits have already been accepted by others. Feedback (talk) 09:05, 26 June 2010 (UTC)

Use the 'Special pages' link in the toolbox, and near the bottom you'll find the link to Special:OldReviewedPages. -- zzuuzz ^(talk) 09:21, 26 June 2010 (UTC)

Excellent, thank you. That link should be made more available to the reviewers. Now that I've "reviewed" an edit, I don't seem to understand something. Isn't this the same thing as semi-protection except that vandals get their edits in the history? Feedback (talk) 09:59, 26 June 2010 (UTC)

Errm, if you want to put it that way, pretty much. Another way to think about it is that good faith editors get their edits in the article where they couldn't previously. Although the most vandalised articles have been used so far in the trial, you should also expect number of articles under some sort of protection to be extended beyond what semi-protection is currently used for. So you could also say that more vandal edits will stay in the history as opposed to staying in the articles. -- zzuuzz ^(talk) 10:40, 26 June 2010 (UTC)

Well I don't see the good in that. Semi-protection only allows established editors to edit. I think it works better than this flagged protection because this flagged one allows everyone to fill up the history section without even making a single edit. And when IPs and non-experienced editors want to make good faith edits, all they have to do is post it on the talk page. I don't see what flagged revisions is adding to our project. Do these non-edits end up joining one's contributions? Feedback (talk) 16:59, 26 June 2010 (UTC)

You seriously think it's just as easy to post on a talk page as it is to edit normally? I'm 100% positive that many good edits are regularly not done because of (semi-)protection, because either people don't want to bother, or may not even know that's what they should do. With PC they can simply edit as normal. ♫ Melodia Chaconne ♫ (talk) 18:07, 26 June 2010 (UTC)

Not only is it just as easy, but only people with real/good edits are going to post on the talk page. Vandals aren't going to waste their time with that. And people who take their time referencing the material they put in the articles probably have no problem with the extra time they use posting on the talk page. Feedback (talk) 18:38, 26 June 2010 (UTC)

That may be your view - but some people are going to look at "Wikipedia - The encyclopedia anyone can edit" and the fact that can't edit and say "Screw this, I'm leaving". We've thus lost an edit. CycloneGU (talk) 18:49, 26 June 2010 (UTC)

I agree that a small amount of people will think like that. But the people who really want to be constructive will go through the appropriate channels. If they don't and are hysteric as you said, then their loss isn't significant. Maturer writers are welcome on Wikipedia, and if they read closely, they just need to know that all they need is a little more time and experience to be able to edit semi-protected articles. If they don't want the time and experience, its their loss, not Wikipedia's. Feedback (talk) 19:08, 26 June 2010 (UTC)

Actually, it is Wikipedia's loss. The attitude you display, that one is hysterical to be concerned about the accessibility and edit-ability of this encyclopedia, is pretty off-putting. Where's the maturity as an editor in understanding how trivial vandalism is when it's easily tracked and quickly corrected? Why should every anonymous editor be limited because of the deeds of vandals they have never met nor intended to emulate? Editing this encyclopedia is not an "honor" that people must train and wait for: it's the very foundation on which the project rests. —Preceding unsigned comment added by 69.142.154.10 (talk) 17:43, 27 June 2010 (UTC)

No, for every edit that needs reviewing, someone needs to review it. Basically, it takes 2 people to edit now when before, it would take 1. People's time will be sliced in half and only 1 person will get the contribution for it. Personally, I won't spend my time reviewing edits "that might be constructive" when I could be making constructive edits of my own. Feedback (talk) 17:55, 27 June 2010 (UTC)

Obviously you don't have to do anything on Wikipedia you don't want to. That's what's great about it. Other editors will see value in keeping heavily viewed pages open to unregistered or not auto-confirmed users, and they will do that vandal-catching work like they critically and dutifully do. You can say that your time is too valuable, but the more important question is whether your encyclopedia is too incomplete. Which will help improve it faster: more editors or fewer. My hunch is that the positive benefits of the unregistered far outweigh the minor burdens of each additional vandal--particularly when much of that vandalism is not even being "accepted".

There is also the important consequence that as pages come under Pending Changes, that vandals will lose interest. Though not currently designed to give editors any bad ideas, it will (or could) quickly become apparrent that vandalism on PC pages is invisible. Seems worth waiting to see how that pans out--or to consider how to maximize that effect--before applying a mathematical accounting to the project, which I suspect and hope is exaggerated in the negative. 69.142.154.10 (talk) 10:11, 28 June 2010 (UTC)

Our anonymous friend is correct and validates the point I've been trying (albeit unsuccessfully) to make. Many users might not have the sources for information, or know where to find information on certain subjects to add, but see value in confirming that a new addition and the given source is good for adding to Wikipedia. And yes, I referred to Stephen Colbert earlier; he referred to the ability to make an edit to Wikipedia and actually did it on his show with that RedWolf ID or something (I'd have to research the edit). He was discussing how we can make these changes and they are immediately there for everyone to see. Now...they're not. The same piece today would be greeted with, "What? It's not there? I demand Wikipedia that my work be submitted for immediate approval!" or something like that (he plays an ego-centric character on the show, I'm sure in person he's a nice guy). While this strays from my point, I'll get back to that: pending changes keeps spammers (including our dear nitetime friend) from being able to glorify themselves as editors because it wouldn't be shown as part of the encyclopedia, but as "awaiting approval". There is great value in keeping spammers out, possibly moreso than editing the encyclopedia yourself. It depends on your values. CycloneGU (talk) 21:06, 28 June 2010 (UTC)

Pending changes with semi-protection

I'm not sure but would having an article semi-protected with pending changes be appropriate? It would seem to allow all the autoconfirmed users to edit but only the reviewers to approve, which seems like a more restrictive version of semi-protection and just bizarre. I'm asking because that's what LeBron James is now under and frankly, I want to make sure before I get into any more with that admin. -- Ricky81682 (talk) 08:43, 28 June 2010 (UTC)

For most instances, PC is only used at level 1, which means any autoconfirmed edit is autoreviewed. Level 2 with semi would be as you describe, but I would think the usecases for that would be fairly few and far between. No comment on the particular case you're in. --Izno (talk) 08:53, 28 June 2010 (UTC)

Lots of complexity, little benefit

I think that my title says it all. People who don't know the intricacies of this won't know what is going on. Also, you are removing the all-important element of user interface design (feedback...see what you did, see what has been done). And the benefit is small. Basically reduction of (vs. fast correction of) vandalism. North8000 (talk) 13:08, 28 June 2010 (UTC)

This is a technical problem and possible easily fixed. That's what the trial is designed to elicit: info about how pending changes works--both so that it can be improved and so that it can be evaluated. It doesn't make much sense to determine that before seeing the best version of pending changes at work. Since this is the first trial, it's likely that further improvements to the application and functioning of pending changes are in the pipeline. So, maybe the reduction of fast corrections will just be a blip in the system, easily tweaked during or after the trial.69.142.154.10 (talk) 05:21, 29 June 2010 (UTC)

Much too complicated

I'm looking at the table in Wikipedia:Pending_changes#Description. There's 4 columns (which really describe 5 kinds of editors) and 5 different levels of protection. That's 25 possible combinations. Nobody can keep all that straight in their head. Well, maybe some people can, but I can't.

As an admin, I don't see how I'm going to effectively administer this scheme. Right now, when I see an article that's getting vandalized a lot, I've got to pick which of two levels of protection I want (semi or full) and the duration. I pretty much just always use semi-protection, and take a WAG at the time, so that's pretty simple. Now I've got to pick which of FOUR levels I want? I just don't see any rational way I'm going to be able to evaluate the level of damage, compare it to the multitude of possible protection levels, and come up with the right solution. The more choices you give somebody, the greater the chance that they'll pick the wrong choice.

I appreciate the effort that went into designing this, but I think it's got to go back to the drawing board to come up with something simpler. Even if the level 1 / level 2 stuff gets merged, I still think this is going to be too complicated. What/s the elevator pitch here? -- RoySmith (talk) 13:33, 26 June 2010 (UTC)

Absolutely. All flagged revisions have done is destroy the simple conceptual model of a wiki. Complication is a huge barrier than will result in a loss of new editors (and potentially old editors, as I have started to question my involvement with Wikipedia over the way big changes occur). Jason Quinn (talk) 20:39, 26 June 2010 (UTC)

I agree absolutely. Exploding Boy (talk) 21:07, 26 June 2010 (UTC)

I agree too. So many big changes like Vector and Flagged Revisions. I personally don't use Vector. Why try to fix something that isn't broken? Feedback (talk) 21:10, 26 June 2010 (UTC)

I couldn't agree more. SlimVirgin ^{talk|contribs} 23:59, 26 June 2010 (UTC)

I'll disagree. Those changes are designed to make Wikipedia better. For more users--including new users. I don't want to seem like I'm devaluing your perspective as a presumably tech-savvy, long-time editor, but many people are not. They benefit from a simple, clear, clean user interface which looks like other webpages they visit and functions predictably without off-putting surprises or overbroad prohibitions on their easy involvement. Also consider that you had no problem switching back to the monobook css and accordingly adjusting the javascript on your userpage. If you know what that sentence means, you're probably in the half of 1% of Wikipedia;s users. Who does it make more sense to optimize the site for: those who are less technologically flexible, or those who can quickly and effortless change their preferences and incorporate new scripts to their liking?

I am also a presumably tech-savvy (semi) long-time editor, and I think Vector is a huge improvement in design, usability, feel, and function. I look at the Citron prototypes and wonder what the heck took them so long to incorporate the kind of high-functionality features which have been standard on most websites for years. It's critical to keep a dedicated base of editors (like you) but without that core group's existence being an artifact of poor design or policy which deters the participation of others. 69.142.154.10 (talk) 22:26, 29 June 2010 (UTC)

This isn't relevant, but I wanted to point out that Wikipedia seems to be one of those few online communities where people can discuss major changes before they are permanent. There are a great many websites out there where changes take place regardless of whether or not anyone likes them. At least Wikimedia asks their users. I admire them for that. Shannon! talk 21:47, 26 June 2010 (UTC)

Well without us, the editors, they'd be nothing. They wouldn't have half the articles they do, and those articles wouldn't be nearly as detailed. It's certainly in their best interest to listen to what we have to say, unlike for example Facebook, which has no reason to care if we like a redesign or not. — CIS ^(talk | stalk) 21:53, 26 June 2010 (UTC)

Try not to put yourself too high up on a pedestal there. ;-) Any website without users (in some form) is nothing and will quickly disappear. JBsupreme (talk) ✄ ✄ ✄ 21:55, 26 June 2010 (UTC)

"Seems to be". I agree with you in more ways than you intended. Jason Quinn (talk) 22:43, 26 June 2010 (UTC)

I will note that there is this Wikipedia "old guard" that perpetually sees the website in a permanent state of 2004 or before and opposes any change to anything. Things have certainly changed since that time, and at the least I appreciate that we're trying to make efforts to move forward as opposed to arguing back and forth and doing nothing. –MuZemike 23:08, 26 June 2010 (UTC)

You mean the "old guard" that helped turn Wikipedia into the best site on the Internet? What a bunch of backwards know-nothing farts! Jason Quinn (talk) 23:13, 27 June 2010 (UTC)

Is that seriously the best you can do by launching attacks at others? Give me a break. –MuZemike 23:16, 27 June 2010 (UTC)

If you had offered anything worth responding to, I would have. The answer to your quip is simple: if the changes are an improvement and have been backed up with good statistics, then the "old guard" is all for them. What they are not for are changes spearheaded by the Wikipedia Foundation, Jimbo, and back-alley straw polls and petitions and eventually sold on the basis of lies like "it's opening up Wikipedia". The "old guard" are also against changes that fundamentally alter the Wikipedia mission and call the "free encyclopedia" motto even more into question without urgent need. This trial is nothing more than a charade. It will be implemented permanently. Oh, sure, there will be a grace period for "tweaks" when the trial runs out but this has nothing to do with making Wikipedia better and all to do with protecting against paranoid legal worries. This change has a good chance of driving away editors and slowing down the growth Wikipedia by undue complexity. The whole tangent about needing to encourage newcomers is a unrelated issue that somehow got thrown into the mix as selling-point spin. Wikipedia's slowdown in growth are due to its maturity as a project. It's success is responsible. This is so obvious that it's painful. Yet many editors seem to think that policy needs to be changed to combat it. This is a mistake and one that has a chance to permanently cause damage to the editorship. The "old" in "old guard" just refers to the wisdom of the editors that watched and experienced the growth of Wikipedia. You come to understand the deeper meaning behind the issues with time. This is something that the newer editors, by definition, cannot understand. Newcomers always — in all software projects — want to come in and change things (because "change" itself somehow automatically equates to "better" in their minds). This change is completely unneeded. Plain and simple. Jason Quinn (talk) 00:45, 28 June 2010 (UTC)

Well I'm just glad VECTOR isn't mandatory and I can go back to Monobook because the lack of compatibility with all the scripts I have is very frustrating. And I hope this Flagged Revisions thing doesn't pass. This is something that can damage the whole aspect of "Anyone can edit". Feedback (talk) 23:47, 26 June 2010 (UTC)

Except it allows MORE editing, not less. ♫ Melodia Chaconne ♫ (talk) 00:07, 27 June 2010 (UTC)

As Melodia said, flagged revisions will allow more editing, as they will be used instead of (semi-)protection, allowing users who would not normally be able to edit the page to edit it. - EdoDodo ^talk 06:12, 27 June 2010 (UTC)

Yeah and it also has good editors whose time is good on actually editing articles wasting that time reviewing other people's edits. Sure, thats great. Feedback (talk) 17:47, 27 June 2010 (UTC)

But people are working on RC patrol anyway, this is no different. --Tango (talk) 18:25, 27 June 2010 (UTC)

Review = Contribution?

Does reviewing an article count as the reviewer's edit or as the the guy who actually wrote it? Feedback (talk) 23:47, 26 June 2010 (UTC)

The edit is still the originator's. I could accept a review by 90.10.128.32 (random possible non-existent IP), and the edit in the history says, basically (removing most Wikilinks and such)...

20:00, 31 June 2010 90.10.128.32 (talk) (58,246 bytes) (→Plot summary) (undo) [accepted by CycloneGU ]

Something like that. =) CycloneGU (talk) 04:19, 27 June 2010 (UTC)

So basically all the good editors out there waste their time reviewing other people's edits and it doesn't even become part of their contributions? Thats excellent; good editor's time is going to be spent reviewing the edits of IP's and newbies while they could be using that time to improve the encyclopedia by editing themselves. Feedback (talk) 17:49, 27 June 2010 (UTC)

Reviewers have the ability to determine how much time they want to spend on reviewing articles each day. Let's say I choose to keep watch over the database page for an hour a day, and I spend three hours on Wikipedia. I still have two hours that I spend editing my own work, and an hour where I help out with pending changes. Keep in mind that anyone accepting the reviewer userright...it's like being nominated to admin. and helping out with semi-protected articles, watching them for occasional user spam. It's exactly the same thing, a responsibility that the user willingly takes on for the overall improvement of Wikipedia. That's how I see it, and accepting reviews IS a contribution in itself. CycloneGU (talk) 18:19, 27 June 2010 (UTC)

Exactly. Honestly, I like to think of my contributions to Wikipedia as whatever difference I've made here, not in terms of edit count. --j⚛e decker ^talk 18:39, 27 June 2010 (UTC)

Its not about the edit count, its about wasting valuable time. You say spending 2 hours improving the encyclopedia and 1 hour editing like if its a good thing, but its not. Reviewing is not "making a difference" or "improving the encyclopedia". Its just a new needless obstacle that prevents a lot of users from editing the encyclopedia. In no way is this a "contribution", its just a waste of time. Feedback (talk) 22:12, 27 June 2010 (UTC)

Its just a new needless obstacle that prevents a lot of users from editing the encyclopedia. → Same with semi-protection. –MuZemike 22:21, 27 June 2010 (UTC)

At least semi-protection doesn't allow vandals in the article's history, doesn't make editors waste their time on reviewing other's edits and doesn't pretend that everyone can edit the article whereas in this flagged protection, vandals edit like normally, editors must take their time reviewing other's edits and newbies might think they're edits are live when in reality they're just waiting for other editors to accept or decline their changes. Feedback (talk) 23:01, 27 June 2010 (UTC)

A mature editor would be able to see that their changes are submitted to the pending queue and will be available in a few minutes. They later refresh the page to see their changes added. What's wrong with that? CycloneGU (talk) 00:55, 28 June 2010 (UTC)

No, not a maturer editor, a more experienced one. Newer editors or one-time IP editors will have no idea whats going on. Wikipedia is not only for the users, but also for the every day Wikipedia reader who decides to edit one day. Those people won't be reading the fine print; all they know is that their edits are supposed to be immediately viewable, something that flagged revisions prohibits. Of course, all of those people would agree this is a bad idea if they had a say in this, but they don't because they aren't users. Its sad that you and most of the others want to make decisions without considering what they might think. Feedback (talk) 02:18, 28 June 2010 (UTC)

(This and above are two different posts, BTW. Anything regarding the above paragraph please reply above this.) MuZemike mentions semi-protection. Semi-protection is used to keep IP editors from editing, this is correct, and as much as it seems it hinders anonymous editing, I understand its use to protect articles that are all subject to possible endless amounts of spam. An alternative is provided when a user tries to edit, explaining go to this page and suggest what you want added...and don't harass multiple times to add it if it's deemed unsourced fancruft. =D

Pending changes, however, does NOT hinder editing. ANYONE is able to edit with pending changes protection, which is closer to what Jimbo wants, I think. However, some pages DO still need semi-protection, this is very true. We also want to encourage honest users to be able to contribute to articles they might otherwise not be able to. I like the idea another editor brought up (and I supported) of restricting edits within a timeframe (one edit every 1, 2, 3 minutes for instance) and this WOULD work with semi-protected articles in some sense in that a spammer cannot edit multiple spams into one article at a time, or even into different articles; it slows them down, and legitimate editors would see that they have to "wait another 32 seconds" to post. The slowdown allows for more constructive editing from legitimate IP editors, and gives admins time to block spammers before they cause heavy damage. If the idea is combined with semi-protection, I think we might have a winning combination here and would like to see some discussion on that; maybe as a separate topic. CycloneGU (talk) 00:44, 28 June 2010 (UTC)

This proposal seems very interesting to me. When you say "one edit every minute", do you mean as a semi-protection feature or on every Wikipedia article? Also, is that only for IPs or for users too? Feedback (talk) 02:18, 28 June 2010 (UTC)

I am referring to a sitewide policy. Forums tend to do this to protect themselves from people posting a new spam thread somewhere every five seconds before they could get him banned (thus leaving a huge mess to clean up); by limiting the number of posts made per minute or two to "one", the number of posts that has to be cleaned up is greatly reduced if the user is banned in the same amount of time. A similar idea can be used here to stop spammers from going to different articles multiple times before an admin. has a chance to act. CycloneGU (talk) 20:47, 28 June 2010 (UTC)

I forgot to add. It's only for anonymous editors. After all, registered users can be identified to a nickname instead of a rotating IP, and thus eliminated quickly if they cause trouble. CycloneGU (talk) 21:09, 28 June 2010 (UTC)

Every idea I hear is always to limit the editing of IPs. Why not just obligate people to register? Surely, it is still "the encyclopedia everyone can edit", all you have to do is use a username. Most websites obligate you to register before gaining the perks and posting (MySpace, Facebook, GMail, YouTube, Bugzilla, etc.) Showing your IP is a dangerous security risk anyway. Feedback ^☎ 02:26, 29 June 2010 (UTC)

The reason is the Wikimedia Foundation wants everyone to, in general, be able to edit. This is one of very few sites where you are not required to register to participate. This trial leads further towards the policy that anyone can edit the encyclopedia; in fact, look below (well below) for the thread about the World of Warcraft article, where the trial has worked spectacularly despite initial worries. My idea is to allow anyone to edit anything...but not be able to cause massive vandalism to all sorts of articles all at once, leaving a mess of a cleanup for admins. to deal with. It seems a good combination to me; less spam to revert if someone is caught quickly, and almost everyone (barring the banned) can edit. CycloneGU (talk) 01:54, 30 June 2010 (UTC)

Some positive feedback

I can't imagine even at this stage of the trial that WP:PEND will be anything but an unqualified success and ultimately will be expanded well beyond the limited test group of articles. The reason is quite simply that it's a sensible solution to a common issue on the wiki--that of admins' choices being limited to just semi-pp when anon vandals, disrupters and POV pushers are in the mix, thereby denying the good-faith IP user the opportunity to contribute productively without necessarily becoming part of the community.
..... Seems to me the occasional need for admins to revoke reviewer rights will likely be vastly less common than the current (or at least previous) demand for admins to answer requests to implement edits for protected pages. Levels 1 and 2, as has been noted by others already, greatly expand the capability for productive editing by anon IPs in semi-pp situations, in part because the anon is more likely to submit an edit on a semi-protected page--the offer to edit the page is already there in front and the IP doesn't need first suffer a rejection and then decide to take the extra step of requesting an edit on their behalf.
..... Using the current number of rollbackers to make a bald speculation of how many reviewers we'll have, it seems plausible to me that Pending-changes procedure will at least triple the number of trusted users willing to attend to these situations. (That's taking the existing admin pool plus roughly double the number of rollbackers as there are admins, to make a very speculative guess about numbers--seems to me there could well be many more than that in the end.) Whatever the overall number of reviewers turns out to be, I can only imagine there will be a significant net reduction of the total admin workload and a net increase in the productivity provided by anon IPs. In any event it looks like the trial period will at least go though to its completion, so we've got roughly another six weeks of experimenting to go, and much more direct feedback to get from a lot of users and admins. So, my personal response to the trial thus far is unequivocally positive. ... Kenosis (talk) 15:26, 28 June 2010 (UTC)

This is one of the best trials I have seen is years, the system works wonderfuly, and I cannot see how this could not be approved for full-scale application. Ronk01 (talk) 17:37, 28 June 2010 (UTC)

Don't be fooled, the good faith IP user will be pissed that his edits have to be reviewed by someone else. Distrust is something people take personally as they should, IMO. Feedback (talk) 17:34, 28 June 2010 (UTC)

(edit conflict) How do you know that? You seem to be speculating. Besides, wouldn't they be pissed if the article was semi-protected and they couldn't edit as an IP or new user? –MuZemike 17:39, 28 June 2010 (UTC)

Its not purely speculation that there is a message that says "When you click Save, your changes will immediately become visible to everyone." Officially, Wikipedia is lying to them with that statement. I'd be pissed too. Feedback (talk) 18:22, 28 June 2010 (UTC)

As an aside, I thought this RFC was a bit interesting but nonetheless got shot down. Anyways, perhaps there is a way that page can change to something different when "pending changes" is enabled – the same way different MediaWiki messages and miscellany pop up, depending on the status of pages. –MuZemike 18:32, 28 June 2010 (UTC)

Wikipedia still has:

When you click Save, your changes will immediately become visible to everyone.

plastered on each edit page. No template or tab is going to change the fact that as of right now, with Pending Changes, we are lying to those who can't immediately edit them. Feedback (talk) 18:41, 28 June 2010 (UTC)

Admittedly there will be some unforeseen programming changes needed. If one of them is the misleading note on the edit page, I'm sure that'll get covered by the programmers in due course. Has anyone informed Rich Farmbrough of this little glitch yet? ... Kenosis (talk) 18:56, 28 June 2010 (UTC) ... OK, I left a note about this on Farmbrough's talk page. I imagine he'll be able to get it into the proper information channel(s). ... Kenosis (talk) 19:43, 28 June 2010 (UTC)

Did you read a single word I wrote above? It is clear you haven't. –MuZemike 18:44, 28 June 2010 (UTC)

I read it, now you read what I wrote. "AS OF RIGHT NOW" means that we are currently lying to them. Regardless of any future plans, right now we are lying to them. That was my point. Feedback (talk) 18:52, 28 June 2010 (UTC)

Not lying, simply wrong. Of course the changes are visible to everyone, just not on the article page. I have raised a query at MediaWiki_talk:Edittools#Pending changes. Rich Farmbrough, 20:08, 28 June 2010 (UTC).

Oh, its not lying, its just saying something that isn't true? Right, thank you for the correction. Its totally okay now. Feedback (talk) 20:13, 28 June 2010 (UTC)

Rich, my suggested text until the trial is over (and then restored if, later, this is fully implemented) would be something like this:

When you click Save, your changes will immediately become visible to everyone.

Some pages will require edit approval before being visible in the article.

If you wish to run a test, please use the Sandbox instead.

That way the expectation is there that the edit may not be there immediately. Optionally, we can state that it applies generally to anonymous users. CycloneGU (talk) 21:30, 28 June 2010 (UTC)

There is an explanation at the top of the page, when you edit a PC page, about the PC process. I am reluctant to tinker with the wording, even, or especially, temporarily. Rich Farmbrough, 18:30, 29 June 2010 (UTC).

I think it would be informative to know, what percentage of unconfirmed edits have had to be reverted? Will this detail be possible to collect? Off2riorob (talk) 18:46, 28 June 2010 (UTC)

I am sure the number crunchers will have that at the end of the trial period. Right now, it might be hard to say as it's constantly updating. CycloneGU (talk) 21:30, 28 June 2010 (UTC)

It might also be worth finding those articles that have had no edits and considering removing PC, and using patrolling instead. It is tricky, obscure articles are ones where vandalism is rare per page but can stay a long time. Rich Farmbrough, 18:35, 29 June 2010 (UTC).

World of Warcraft - Pending changes working better than I expected

I was very very dubious when WoW was picked for pending changes. It used to *drown* in juvenile vandalism (edit to add) and I expected that to return. With pending revisions, some anon edits worth keeping have been made, and some of the usual vandalism has popped up and been addressed. But it seems that if the vandals can't see their graffiti at once, some of them wander off, or even better just "leave the paint in the can."- Sinneed 20:55, 29 June 2010 (UTC)

That's pretty cool, actually. Not only is pending changes making edits that are bad unable to be seen, but it's making some people who vandalize Wikipedia stop vandalizing. I guess pending changes does more than just make bad edits unable to be seen. --Ha dg er 21:47, 29 June 2010 (UTC)

Global semi-protection would accomplish the same thing: prevent bad edits from being seen (actually prevent them from even happening) and make some people who vandalize stop vandalizing. I am not convinced that pending changes has value. In order to preserve a tiny minority of constructive anonymous IP edits, we add a burden of reviewing to established editors. I'll point out that the creed "anyone can edit" is still preserved by requiring account registration. ~Amatulić (talk) 00:53, 30 June 2010 (UTC)

It would make them create an account, edit a bit, and wait a few days. No. It does not stop vandalism. It does, however, entirely prevent anonymous editors from making valuable edits.- Sinneed 01:19, 30 June 2010 (UTC)

Glad to hear such great news! It's great to hear that PC is providing a means for vandals to lose interest, this is a great side effect, if you will, of the protection. CycloneGU (talk) 01:14, 30 June 2010 (UTC)

If some vandals are not vandalizing because they can't see their edits immediately, isn't it obvious that some productive anons are doing the same thing? Leaving because their edits aren't immediately viewable? Feedback ^☎ 02:24, 30 June 2010 (UTC)

Maybe they are, but what is happening is that the vandals want everyone to see the edits they make, so some of them may find it pointless to vandalize Wikipedia when they know that many people will not see their edits. However, non-vandals also want people to see their edits, so they just wait for it to be reviewed. Their is a banner saying that edits to an article with pending changes must be reviewed when editing an article with pending changes. What's happening is that vandals probably know that their edits will be declined, while good-faith editors probably know that their edits will be accepted, so they decide to make the edits. However, it is true that maybe some good-faith editors are leaving for the same reason, but it is unlikely that it is happening. --Ha dg er 03:42, 30 June 2010 (UTC)

Maybe I'm reading it too deeply, Feedback, but I really don't see your logic in your last objection. Spammers choosing not to edit will never see their editrs in the article. Helpful anons. who edit will be rewarded with visibility of their addition within the article. Maybe "rewarded" is the wrong word, mind. CycloneGU (talk) 05:07, 30 June 2010 (UTC)

No, this wasn't an objection, it was a "clarification". They are celebrating vandals stopping to edit because they can't see their edits, ignoring the fact that non-vandals are doing the same thing. Even if the non-vandals are helpful, they probably have no idea their edits will become viewable in a matter of minutes. And if their good faith edits are reverted, they'll never get to see them and they probably will abandon editing all together (just like vandals). Feedback ^☎ 19:31, 30 June 2010 (UTC)

" ignoring the fact that non-vandals are doing the same thing" - An anon editor has told you this? I see anon edits that could not have happened before... happening. It is, and was in the pro and con discussion, a concern. You might read the extensive dispute about whether to even try the experiment. It was long.- Sinneed 01:08, 1 July 2010 (UTC)

Jimmy Wales

A frequent vandalism target (including once by myself back on April Fool's day 07:) it has been semi-proted since that particular padlock was created. It is a controversial BLP and, in short, a perfect candidate for this interesting little experiment.--R.D.H. (Ghost In The Machine) (talk) 11:59, 30 June 2010 (UTC)

Seohyun

I don't see how Seohyun meets the criteria for pending changes. It would not have met the criteria for semiprotection, and p.c. only applies to pages that could otherwise be semiprotected, right? — Carl (CBM · talk) 12:24, 30 June 2010 (UTC)

Edit notice?

I think that the edit notice for protected pages is not large enough. Maybe saying "Vandalism, or edits by unconfirmed users will not be added, unless a reviewer accepts the edit. Therefore, any non-Wikipedia appropriate statements will not appear" Maybe?, A p3rson ‽ 01:46, 1 July 2010 (UTC)

See here. It's been brought up. CycloneGU (talk) 03:07, 1 July 2010 (UTC)

The lack of "Review Conflict"

As I continue to experiment with this trial, there seems to be a lack of "Review conflict" a la "edit conflicts". When I'm reviewing an edit, and someone is reviewing it at the same time, there is no "edit conflict" notice. When I thought my review was done, in reality, another editor had reviewed them and Wikipedia did not inform me. This happened a total of 3 times. Feedback (talk) 22:21, 27 June 2010 (UTC)

Yes, I have had that happen several times also. --Funandtrvl (talk) 00:57, 28 June 2010 (UTC)

It happens every time. The most annoying is if the edit conflict creates a situation that the previous rollback is rollbacked again, thereby reintroducing the bad edit to the article. The only way I can find to determine that hasn't happened is to wait, then revisit the article and check the history to determine that the correct revision was accepted. In my view it's cumbersome. Truthkeeper88 (talk) 01:02, 28 June 2010 (UTC)

I think I might have experienced it, too. It is problematic, especially given the slow reaction time with the server (which goes back to my one main complaint in performance issues). –MuZemike 01:48, 28 June 2010 (UTC)

Yes, I've definitely experienced this several times - I load a page with pending changes, I rollback the top edit, which is supposed to be the recommended way to deal with bad top edits, and nothing happens. Then I try Twinkle rollback (which luckily has sanity checks built in) and it tells me I already made an edit. Not good. — Gavia immer (talk) 01:55, 28 June 2010 (UTC)

When I look at Special:OldReviewedPages, some of the pages will say "(under review)". Presumably, this means that a reviewer is looking at them already. Two questions:

How is the "(under review)" statement generated? Does it only work if the reviewer has accessed the page in questions from Special:OldReviewedPages? Does it only work if the reviewer is looking at the page that compares the current version with the last reviewed version?

Is this problem mostly caused when reviewer B clicks on Special:OldReviewedPages after reviewer A clicks on Special:OldReviewedPages, but before reviewer A has clicked on the page that reviewer B also (eventually) clicks on?

Yaris678 (talk) 12:33, 28 June 2010 (UTC)

The second question is kinda garbled, so I'll answer the first and give an explanation along the lines for the second.

The "under review" statement with yellow background is generated automatically when one of the 5,000 or so reviewers begins to review the edit. Due to my own edit conflict issues, I now stay away from those, typically. I recommend this to allow the reviewer starting to finish, and then maybe view it shortly after if it doesn't go away. Speaking of going away, I do not know how long the "under review" flag stays visible without another reviewer clicking.

When multiple editors try to make the same edit/rollback/both reverting an edit, an edit conflict results for the second (and third, and fourth, and fifth, and...). Based on my experience, blindly going ahead with your action puts the spam BACK into the article. This happened to me, in fact. So now, if I find an edit conflict or rollback notice, I flash back into the history and check before I try to do any additional work. CycloneGU (talk) 21:18, 28 June 2010 (UTC)

Fair enough about the second question. I couldn't think how to phrase it any more intelligibly. But you haven't answered the first question either! :-) When you say "begins to review the edit", can the editor be looking at any aspect of the page (such as the page itself, or the page history) or does the editor have to be looking at a special "reviewing the edit" page? Yaris678 (talk) 13:00, 29 June 2010 (UTC)

My sincerest regrets. I'll clarify.

When a most illustrious editor clicks on the "review" button (text basically) next to the edit requiring such, it brings up on his or her screen a page similar to that accessed from the history, when you compare the differences between two edits. The difference is that the edits are pre-selected (last accepted and pending), and one has the option of either accepting or rejecting the edit. If accepting, a reason is typed and submitted, and the history says, for instance, "accepted by Random123" next to the edit, and the edit exits the review list. If rejecting, they then either use rollback (if given the access to it) or merely undo the edit (needing a second step, with the typical edit box and a review of exactly what you're removing/changing). When the latter is submitted, that editor's change is automatically accepted and then the review taken from the review list.

If you require any further clarification, please do not hesitate to post further. CycloneGU (talk) 18:41, 30 June 2010 (UTC)

So I think the answer to my question is "yes". I was just thinking that people might not review the changes using the "review" button. They may look at the page history, at diffs etc.. (this may or may not be related to the possibility of a reviewer finding an unreviewed page "by mistake" rather than via Special:OldReviewedPages). Therefore, it may be a good idea to mark a page as "under review" if a reviewer has looked at it recently, rather than necessarily if they are using the "review" feature. Yaris678 (talk) 06:01, 2 July 2010 (UTC)

Its obvious what's going on...

Wikipedia is trying to take away privileges from anonymous editors. Why not just obligate everyone to make a user before editing, while we're at it? At least that makes more sense than "flagged revisions". Feedback (talk) 21:35, 28 June 2010 (UTC)

While there are some who advocate that, you won't help yourself by making obviously exaggerated statements like that. Not everyone who supports this supports that. This reminds me a lot about the semi-protection debate though. -- Ricky81682 (talk) 00:35, 29 June 2010 (UTC)

Feedback, that is not what Wikipedia is trying to do. If anything, this makes it easier for anonymous editing to take place. Think about it for a second. Semi-protected articles don't allow anonymous editing whatsoever. By replacing semi-protections with pending changes protections, anonymous users will be able to edit. In addition, we can effectively keep the amount of vandalism low by hiring reviewers to filter out the bad stuff. Lawl talk 02:26, 29 June 2010 (UTC)

But, to the point above, it seems that it doesn't work well against heavily vandalized pages, that were already semi protected. It does work well against relatively unwatched BLPs such as Seohyun, where previous protection was not applied. NativeForeigner ^Talk/_Contribs 04:30, 29 June 2010 (UTC)

My suspicion is that if the trial goes favourably and it gets implementation, almost every encyclopedia article will have this automatically applied (anything in User: and other template categories being exempt). The exception would be semi-protection, which if applied at the same time as pending changes (i.e. both present TOGETHER) would result in really no difference to such pages (and let's not get into fully protected ones, which have NO changes and really don't need PC). The only difference for SP: the odd fairly new registered user who makes an edit and it still has to be approved; unless the person creates an account for the prime reason to spam a specific article that they normally couldn't as an anon., most registered users will make legitimate good faith edits. Some may still have to be reverted, and sure, admins. watching can always reverse anything "automatically accepted", but they don't have to act otherwise as we can review these edits before action is required. I'd like to think SP makes an administrators job easier if used properly, and I think combining with SP is good because we can also report registered users spamming in those pages for swift action. CycloneGU (talk) 05:39, 29 June 2010 (UTC)

Is that what's done on German wikipedia? Or do you think English is somehow special in how it's going implement it? I think, like semi-protection when it was first introduced, it will (and still does) be abused in some level, some of which will be caught, some of which will not be seen with some people complaining that having it at all is a horror until the entire encyclopedia while others who have never wanted anonymous editors finding it reasonable. The more interesting question will be what happens as recent changes loses its value since you're just reviewing articles that nobody is really seeing (which in theory should have more eyes, not less). -- Ricky81682 (talk) 06:31, 29 June 2010 (UTC)

Frankly, I don't use the German Wikipedia, so I am not too concerned with whether it's implemented there. Also, other Wikis may be so much less busy and attracting of spam that they may not need it. With English being THE universal language, this Wiki seems to benefit the most. CycloneGU (talk) 06:47, 29 June 2010 (UTC)

I meant, seeing as how the German encyclopedia already has this, I think that's a better guess of our possible future than mere speculation. -- Ricky81682 (talk) 01:11, 2 July 2010 (UTC)

Rollback problem

I was reviewing a page, and there were 7 edits by an IP that were accepted by an editor. Great. Well, the IP did 3 more edits which I wanted to reject. So I should rollback, right? "Wrong. The rollback would remove the 7 accepted edits. Then that means I should undo, right? Wrong. The undo will only remove the last edit and keeps the first two. Basically, the first 2 edits the IP did don't even have to be reviewed to get into the article, all an IP has to do is make a nonsense 3rd edit, it gets undone and boom, his first 2 edits get in an article. Do you guys not see the problem there? Feedback ^☎ 20:35, 29 June 2010 (UTC)

View the revision you want to revert to, click edit, accept the edit, click save (explained in more detail at Help:Reverting#Manual reverting). –xeno^talk 20:38, 29 June 2010 (UTC)

Xeno, I know how to manually revert. Hell, I could even go to a past revision and copy/paste into an edit. But why should I be wasting my time doing that? The flagged revisions are forcing me to do something that I normally would do with the push of a button. Feedback ^☎ 20:41, 29 June 2010 (UTC)

So what would you have done prior to pending changes? –xeno^talk 20:43, 29 June 2010 (UTC)

Well, actually, Feedback, even if the page had no protection, you would still have to manually revert those three edits to keep all the good edits on the page. Pending changes doesn't really make a difference on what you must do to revert an edit. If I misunderstood what you were saying, then please tell me what you meant if you want to. --Ha dg er 20:47, 29 June 2010 (UTC)

What is it you guys aren't getting? Its pretty simple. Due to pending changes, the IPs are making "almost edits". In semi-protection, if someone wanted something added, they would have had to post it on the talk page and I would have added it manually with a click. Instead, they get to post whatever they want and I have to review each of their edits one by one. Its obviously more taxing to review the edits one by one. And even if they aren't one by one, have to manually edit what you want to go in or not. If you still don't understand, I'll explain graphically:

current sentence: "The quick brown fox jumped over the dog."

IP makes constructive edit: "The quick brown fox jumped over the lazy dog."
IP makes bad edit: "The quick brown fox jumped over the lazy cat."
IP makes 2nd constructive edit: "The quick brown fox jumps over the lazy cat."

With Pending Changes, I'd have to analyze each edit, diagnose which edits were productive and manually edit from scratch to create "The quick brown fox jumps over the lazy dog", a combination of both constructive edits which were intervened by a bad edit.

With Semi-Protection, the IP would have posted what he wanted on the talk page, and there would be no time wasted on figuring out the correct way the article should be (and no rollbacks, reverts or any space in the history whatsoever). Feedback ^☎ 00:16, 30 June 2010 (UTC)

Semi-pp still exists, if the acting admin chooses to go that route. Here, the page (whatever it is--it would be nice to see the actual diffs) is chosen to be subject to Pending changes. In this instance, because of the mix which includes some arguably constructive edits, it has the same issues as it would if no protection were in place. Thus it appears to be between the two reviewers and/or between other autoconfirmed users to hash it out just as they would normally need to do. Pending-changes protection obviously wasn't designed to solve everything; it solves what it can, and the rest is pretty much business as usual. ... Kenosis (talk) 01:22, 30 June 2010 (UTC)

"Pending changes was designed to solve what it can"... What exactly is it solving? Feedback ^☎ 01:25, 30 June 2010 (UTC)

I can now see this discussion is going exactly nowhere that's productive. You apparently disagree that there's any merit to the things with which it's argued to be helpful, which you've already been well exposed to and can readily find elsewhere. G'day. ... Kenosis (talk) 01:29, 30 June 2010 (UTC)

No, there's just no good answer to that question. Pending Changes solves no problem. Its just a different way to deal with something that was already being dealt with (semi-protection). This is basically another form of semi-protection plus the defects (slowness, edit conflict errors, rollback errors, etc.) And you're right, I do not see anything thats productive with this change; this just follows Vector as another unnecessary drastic change for the sake of change. Feedback ^☎ 01:41, 30 June 2010 (UTC)

Comment. Why not just click to the version of the article you want to roll back to, open it, and click Edit and Save Changes on that one? It reverts the article to that version. Then, if there are in-between changes you WANT to keep, go back and manually add those with a summary of something like "Readding good faith changes deleted in rollback" or something. CycloneGU (talk) 01:05, 30 June 2010 (UTC)

Thats exactly what I just said, and I stated its a hassle. Possible sure, but its still a hassle. Imagine if Wikipedia suddenly didn't let you Copy/Paste. You'd have to manually type in character-by-character all the codes in templates and boxes. Its possible, sure, but its a hassle. We should avoid hassles as much as we can and this is one of those situations where its just an unnecessary hassle. Feedback ^☎ 01:15, 30 June 2010 (UTC)

It takes one, maybe two additional clicks. It's not that huge a hassle, in particular if every edit was spam or unconstructive. I agree finding that one good edit can be frustrating, but it's a quick spam reversion and if you think nothing of value was added between the good one and the latest one, just rollback that way and forget about it. CycloneGU (talk) 01:44, 30 June 2010 (UTC)

Wait a second. I just saw this:

With Pending Changes, I'd have to analyze each edit, diagnose which edits were productive and manually edit from scratch to create "The quick brown fox jumps over the lazy dog", a combination of both constructive edits which were intervened by a bad edit.

No offense, and no insult here...but you have GOT to be kidding me. It's better to put the requests all on the talk page so that a user who is watching the page can go and personally edit the article? Reviewers are spending time verifying edits THAT ARE ALREADY MADE. Talk page viewers and administrators are spending their time making edits THAT SOMEONE ELSE COULD HAVE MADE. I see a similarity between those two groups; why is one group different from the other? BOTH GROUPS ARE SPENDING TIME IMPROVING WIKIPEDIA IN IDENTICAL WAYS. ~~The only difference is who gets credit for the constructive edit!~~ So if two editors edit the talk page and we end up adding the information to the article ourselves, all of a sudden semi-protection is wasted because we have to make these edits, forcing us to waste our time by not making our own constructive contributions to the encyclopedia...do you see what I'm saying here? Saying reviewers are being forced to waste their time is a completely void argument because WE CHOOSE to do it! Just like you choose to add someone else's edit to a semi-protected page!

Please clarify if I am just not understanding, but I see you contradicting yourself here in a backwards way. CycloneGU (talk) 05:18, 30 June 2010 (UTC)

Indeed you are just not understanding. Not only that, but it seems you are determined to misunderstand. "Credit" is not an issue, because if the reviewer ends up manually editing, then he does get credit for his contribution. Regardless, that just means there were 2 edits that provided exactly the same to the article and had absolutely no difference in content. The real different was that one of them was made by an anon and the other by a reviewer. If you don't see how "2 edits to make 1" is a "waste of time", then, as I already said, you are obviously determined to be blind in the situation. Feedback ^☎ 19:26, 30 June 2010 (UTC)

The reviewer's work is NOT AN EDIT. It is approval of another's edit for inclusion in Wikipedia. Without this step, spam can get into the article unchecked until someone comes along and notices it, and removes it. I do not consider this a waste of time, and have a lot of people who will back me on that.

Let's just agree to disagree; so far everything I've considered good with this system is something you see to be a flaw, and while I agree with some minor concerns, I disagree that the system is a complete "waste of time". CycloneGU (talk) 02:56, 1 July 2010 (UTC)

It seems what you take issue with here are problems with the MediaWiki software itself. Merging problems are not and should not be fixed by FlaggedRevs. It should be fixed in core. And as a user, you can fix this the same way you would in the past: manual editing. Nothing changes in this regard. Reach Out to the Truth 03:38, 30 June 2010 (UTC)

It seems your position is that you just don't like pending changes at all and think the anons should continue having to post their proposed changes at the talk page. You haven't identified a problem, you've given your opinion. –xeno^talk 19:32, 30 June 2010 (UTC)

Well what some may find "an issue", other won't. Its subjective. Personally, I hate Vector for many reasons which to me are "issues" like the lack of supporting all the scripts I use in MonoBook. Those who don't use the scripts I use or don't use any at all will continue to use Vector as if there is no problem. Now, I'm not going to rant about Vector, because I have the ability to turn it off and just stay in MonoBook and never do I ever have to experience it again. Flagged Revisions however, post-trial, will be implemented in a lot of articles and if I ever want to edit one of those articles I am going to come across this silly system that I don't find beneficial. Yes, it is my opinion, but it isn't a hard one to understand. Consensus is built upon the merging and understanding of opinions and my argument isn't one that is unsound. I have found many flaws with this system and the answers I've gotten are basically "there are flaws elsewhere in Wikipedia too". That is true, but why would we want to knowingly instill something that is flawed when we gain little out of it. No one in any way has been able to tell me what is the significant gain out of Pending Changes; and I am beginning to believe its because there is none. Feedback ^☎ 20:35, 30 June 2010 (UTC)

You can safely ignore pending change and simply edit like you always have, leaving others to review it. That's what I am (presently) doing. The significant change is making the byline "The free encyclopedia that anyone can edit" true more often. –xeno^talk 20:59, 30 June 2010 (UTC)

False, you're loopholing around the term. Are these edits really edits if they have to be re-edited? Feedback ^☎ 21:05, 30 June 2010 (UTC)

Uh...he is speaking truthfully. Pending changes lets ANYONE edit the article. Semi-protection doesn't unless through the talk page and someone ELSE adding. PC is closer to what the Wikimedia Foundation sees Wikipedia as. How is this not obvious? CycloneGU (talk) 02:23, 1 July 2010 (UTC)

Well, although it may seem tiring to have to view edits and review or edit them, it's better than having to make edits for anonymous users on semi-protected articles. Also, I am not trying to insult anonymous users at all, because they cannot control an article's protection, and it's good for them to want to improve the encyclopedia. I'm just saying that it's better for anons to be able to edit articles instead of users having to make edits for them. However, although you don't mention this, the lag when trying to view an edit that is not reviewed yet probably seems annoying too, and I hope that is fixed very soon. Anyway, like Cyclone said, PC is closer to how the Wikimedia Foundation sees Wikipedia: the online encyclopedia that anyone can edit. --Ha dg er 00:57, 2 July 2010 (UTC)

Autoconfirmed Users

More Questions: What about edit conflicts?

Requests and usage

What happens to edits that are not accepted?

Reviewer right

A noticeboard

Approval of edits by non-autoconfirmed editors

Shortcut to the Queue

IRC Channel

To do when this is implemented

Major legal question about flagged revisions.

Proposed policy?

Reviewer userbox?

Administrator guidelines

Obvious question

"accepted by" → "reviewed by" ?

Questions about the reviewing policy

Pending changes and edit warring/content disputes

Autoconfirmed users2

The magnifying glass icon is already used for the search

Two-tiered registered users vs accessibility

Unclear who gets to see the draft version of articles.

Another approach?

Accepting edits that are detrimental to the article

Auto-accepting reviewers edits

How is success or failure to be judged?

Confused

Using pending changes as a means of complete unprotection of articles

Reviewer rollback edits need to be reviewed??

Unaccept on Ernest Hemingway

I Don't Like Where This Is Going...

A Better Example..

I like this idea

Automatic acceptance of IP edits

Compatibility with suppression / revision hiding

Bots

Pending changes for talk pages?

Question

Requests?

it sucks

Crap idea

Pending changes

Question about the scope of this trial run

Official poll area?

Suggestion: 'Recent changes' list specifically for pending changes

High-profile pages

Query

Articles where considerable vandalism is probable if downgraded from SP to PC

Protection tactics

I don't get it

Why bother?

Not liking this

A list?

Pending changes with semi-protection

Lots of complexity, little benefit

Much too complicated

Review = Contribution?

Some positive feedback

World of Warcraft - Pending changes working better than I expected

Jimmy Wales

Seohyun

Edit notice?

The lack of "Review Conflict"

Its obvious what's going on...

Rollback problem