Wikipedia talk:Pending changes/Let's move

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

This is a proposal to re enable pending changes on English wikipedia.

Proposed trial

It's a long time since we have disabled pending changes on English wikipedia. I believe it's a very nice feature a lot of people worked hard on and it could be huge benefit for wikipedia to have it enabled. Unfortunately the previous trial wasn't really well managed and ended in a chaos which discouraged lot of users from having the tool. I propose to start another trial, strictly managed by community, and after that we should decide if we enable pending changes widely, or have another discussion related to it.

What are pending changes

Flagged Revisions allows for Editor and Reviewer users to rate revisions of articles and set those revisions as the default revision to show upon normal page view. These revisions will remain the same even if included templates are changed or images are overwritten.

Why do we need them

This feature can allow us to keep articles which are already well written and important pages to be constantly on best revisions.

How do we test it

This needs to be discussed now. I propose following:

Have a 6 month trial. After the trial, all protections will be immediately removed from all articles and no other protections will be set until community discuss if we want to have the feature enabled or if we need to discuss it more.

For a duration of the trial:

• Remove full protection from random 200 main space pages and replace it with semi protection + flagged revisions, where only reviewers can review the changes. All changes done by users other than sysops or reviewers needs to be checked.
• Remove full protection from random 1000 pages from all spaces and replace it with no protection + flagged revisions, where only reviewers can review the changes. All changes done by users other than reviewers needs to be checked.
• Remove semi-protection from random 20000 pages from all spaces and replace it with no protection + flagged revisions, where only reviewers can review the changes, but confirmed users can have their changes automaticaly checked. All changes done by users other than confirmed or reviewers needs to be checked.
• Protect random 1000 GA and FA which has no protection with flagged revisions, where only reviewers can review the changes, but confirmed users can have their changes automatically checked. All changes done by users other than confirmed or reviewers needs to be checked.

This will allow us to make it possible for contributors without account to participate on articles which are currently protected, while we keep the good revisions as head. After 6 months we can ask researchers from wikimedia foundation or community to analyze the changes done to all pages and to give us a result. If we find that pending changes are useful, we can decide to write a wide rules for their usage including instructions for sysops how to set it up, and use it. In case we find any issues, we can discuss it again, and try to improve the tool according to feedback.

Who will get the review rights

There are currently lot of people who already hold the flag. These people were once considered trusted so there is likely no reason to revoke it. All people who are interested in the tool, should apply for the flag just as for any existing permission, like rollback, and administrator or bureaucrat decide if they get it. In case they decide not to give it to user, they will explain their reason. After the trial, permissions will not be revoked, but become unusable as there will be no pages using the protection.

Discussion

Please give us your opinion on this trial bellow, remember this is not a vote, but a discussion, upon your feedback the proposal can be changed. Once it's clear how we want to test the tools, we can start a wide vote and based on result start with trial, or discuss it more. Thank you Petrb (talk) 12:43, 27 February 2012 (UTC)

Petr, did you know that Beeblebrox is about to launch an RfC on the same issue? :). Okeyes (WMF) (talk) 15:17, 27 February 2012 (UTC)

Why this is a jolly good idea, almost

Pending changes was a sensible idea implemented clumsily. The Foundation screwed up badly in the implementation of the previous trial. But pending changes is actually very sensible. There are two ways to see pending changes: you either see it as locking up the wiki further, another layer of bureaucratisation, a further insult to anyone can edit, or you see it as a way of handling the reality of where we are. The problem is that protection and semi-protection have already stopped 'anyone can edit'. Blocking people does that. Anyone can edit any page, so long as that person is an administrator: administrators can edit protected pages, and administrators are exempt from IP blocks.

Pending changes is simply a better implementation of {{edit protected}} and {{edit semi-protected}}. You want to edit a protected page? Currently, you can't unless you are an administrator. You can politely ask on the talk page, and then maybe an administrator will come and do it. Alternatively, you could edit, and it would be held in limbo, and an administrator could come along and hit 'Approve changes'. Easy.

We already have substantively the same procedure as pending changes, we just have it implemented really clumsily. Anyone can edit... so long as they are familiar enough with wikiprocedure to know to put a template in a section on a talk page. The complexity increases the cost of changing things. It's too big of a burden to do {{edit protected}} to request an admin fix a few spelling mistakes or change a category with something like HotCat.

Plenty of other language versions of Wikipedia have pending changes... and it's fine. I edited Polish Wikipedia a while back to add an image I found while trawling an image archive to a page which had no English equivalent. I edited, and it went in to the pending changes queue. And it was no big deal. About 20 minutes later, a reviewer marked it as checked and it was on the page. Given that I know nothing about Polish Wikipedia policies, I'm quite glad that someone actually looked at it.

So, yes, pending changes good. We bloody well ought to have it as soon as possible and it's an utter travesty that the community has rejected it.

But, given the past pending changes trial, we need to specify what exactly would make the trial successful. That's not to say that the trial wouldn't end after six months, simply we should agree what are the issues PC ought to solve. If after the six months, people still have their positions on PC, that's fine. As they say, you are entitled to your own opinions but not your own facts. Agree up front what facts exactly we are hoping to find out from this trial. Having a sensible discussion about PC is dependent on actually having a mutually-shared set of facts rather than lots of bickering and bullet points with bold text.

There are a few parameters in Petrb's suggestion that probably ought to be "subject to administrative common sense". There are some fully-protected non-content-space pages that should remain fully-protected (MediaWiki JavaScript etc.) without any need for pending changes. They are exactly the sort of thing you need to explain to an administrator before doing. So I'd amend the "Remove full protection from random 1000 pages from all spaces" to "Unless common sense dictates otherwise, remove full protection from 1000 random pages from all namespaces".

I'd suggest that rather than all piling on and adding lots of bullet points with bold words in them, people who support having a new pending changes trial need to work out what exactly it is we want from pending changes, some sort of target almost. Opponents of pending changes should also consider what it is exactly that they object to, and set some reasonable challenges that PC needs to meet. It should be possible (perhaps with support from researchers at the Foundation) to see over an equivalent six month period, whether the randomly selected full and semi-protected pages have had edit requests, and then compare that data with the pending changes period and see whether pending changes enables non-confirmed and non-admins to edit semi and full protected pages more effectively. See whether it actually increases their ability to edit articles or not.

Also, track how long it takes on different classes of articles for pending changes to go through. If it is taking hours and hours, that's not good. During the last pending changes trial, I can't remember seeing PC edits in the backlog that were sitting there for more than about 10 minutes. Ideally, that queue should be kept empty most of the time, rather than creating another backlog that needs handling.

Finally, on the matter of approving reviewers, I'd say that the answer to how we hand out reviewer rights should be left deliberately vague and give admins scope to use their judgment. We can't predict how those rights will be used or misused, and it'd be better to formulate rules and guidelines on that as we go along. You can't codify cluefulness. —Tom Morris (talk) 15:20, 27 February 2012 (UTC)

I think the kernel of what Tom says here is very important: before any trial is started, metrics by which to judge it must be agreed upon beforehand. Otherwise, there will be no common ground for the pro- and anti- factions to use to judge whether the trial is successful. Does the community want to see a reduction of X% in visible vandalism on BLPs? Does the community want to see a maximum holding time of X minutes in the Review queue? We need to know what our goals are to meet before we can set about seeing if they're meetable. A fluffernutter is a sandwich! (talk) 15:28, 27 February 2012 (UTC)

I think if we want to have smth sensible out of it, a good start would be to summarize the most important arguments from both sides which surfaced from the previous discussion. And only then we could start thinking about designing the trial, since the trial is supposed to confirm or disprove these arguments, not really to provide a bunch of numbers nobody would know what to do with. (I must add that I have a strong opinion concerning pending changes/flagged revisions since I was involved a lot with them in Russian Wikipedia for about three years, and when I quit a year ago I was ranked second in terms of the number of manual flagging actions over there, so that I have plenty of first-hand experience, but I do not think my opinion and arguments belong to the discussion right now - otherwise we will get into the old argument again).--Ymblanter (talk) 15:37, 27 February 2012 (UTC)

why this is not a good idea

I have a more general suggestion qhy this should not be considered: that only proposals for changing procedures at Wikipedia be accepted that will decrease, not increase, the layers of bureaucracy and complication I gather this proposal will apply protected changes to several tens of thousands of pages or all degrees of current quality, though, since the proposal at the top of the page also refers to flagged revisions, it shows no attempt has been made to even harmonize the various versions. I know exactly how I will react to it: I will stop editing all such pages. If it works at all like it did before. I will not know exactly which changes I am accepting and on what basis--whether I am certifying accuracy and NPOV, or just the freedom from vandalism. for example, if I fix a grammar point, since I am an admin, that would seem to mean I must examine everything else in the article to make sure its OK, because my edit will have certified them,. If I were not an admin, I would do just what I do on deWP, which is stop editing altogether, because I would not be sure what parts of what I am doing would last in the article. The motive for a beginner to edit is to see their work in public. We have hundreds of interlocking layers of bureaucracy, but for most part, unless one is deliberately taking part in a bureaucratic process, one can ignore them. This way, everyone would be faced with them immediately. The deWP, often used as an example, thinks the amount of current content is enough, and it is only necessary to improve it. That's a far different principle of building an encyclopedia than ours. If we wanted an essentially static encyclopedia, this change would be worth considering. It might well prevent some kinds of deterioration, but it will also hinder improvement. I know the bureaucratic procedures I use, and with this, I'll find it easier to stick to them, and leave off content editing. At least it will increase the amount of time I spend on WP space, where I am also needed--I used to spend a lot more time at AfD, actually, and maybe that was more useful. DGG ( talk ) 19:23, 27 February 2012 (UTC)

Mabdul's view

One of the reasons many opposed to expand the last trial (or voted so that PC gets removed again) was that the software wasn't good enough. I got the review right a few months ago on the German Wikipedia and I really don't like it. So just to clarify: Are there any improvements on the software since the last year? mabdul 20:08, 27 February 2012 (UTC)

Pending changes is not the same thing as flagged revisions, which is what they use on the German Wikipedia. And no, it has not been improved recently because we aren't using it. Beeblebrox (talk) 20:43, 27 February 2012 (UTC)

please don't do this this way.

I'm quite puzzled by the timing of this, You posted a link to this at AN just a few sections below where I am attempting to get volunteers to administrate an RFC that is currently not live, that was carefully developed to avoid the pitfalls of the last discussion, and that will be run from the get-go by at least three uninvolved admins who will keep the discussion on track and do a close as a team, helping to insure that consensus is respected. This looks like it was thrown together in the space of a few minutes and has no such controls. Please, PLEASE, let the other carefully designed and more informative procedure go forward instead and close this RFC. Beeblebrox (talk) 20:42, 27 February 2012 (UTC)

Agree, this proposal was not well thought out and will not attract support. Its goals will be best accomplished by withdrawing it and letting another similar proposal go forward, contributing to that. The numbers of articles under pending changes here seem kind of random and don't stem from any clear goals, nor is an evaluation process given. Dcoetzee 22:16, 27 February 2012 (UTC)

It appears from the statements above that the user initiating this is not really even aware of the very substantive differences between "pending changes" and "flagged revisions". The other rfc that is not quite live yet is very thorough, highly structured, and specifically engineered to force a usable result. I am asking everyone to please hold off just a few days and not comment here, a more practical, carefully planned rfc is on the launch pad, and will be run by a trio of uninvolved admins who will keep it from becoming a circus like last time, when I foolishly rushed into an rfc about this without any real plans. Let's all learn from my mistakes and not do that again, please. Beeblebrox (talk) 22:48, 27 February 2012 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.