Wikipedia talk:Pending changes/Request for Comment 2012/Archive 1

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.

---

A couple of things before this goes live

• The discussion section is actually a transcluded subpage. Since this is almost certainly going to be a high-traffic discussion, this seems a good way to reduce edit conflicts. The original plan was actually to have each section for endorsements to also be a subpage, with the main page fully protected to prevent anyone but the coordinating admins from screwing with it while the RFC is underway. Up to you guys if you want to leave it as is or go with some form of that idea.
  • There is a limit, albeit quite large (not sure what it is exactly) defined at Wikipedia:Template limits#Post-expand include size which means that if the template (or subpage) that you are transcluding goes over this limit, then it won't be transcluded any more - a problem that you can see at Stewards/Confirm/2012/en. You might want to bear this in mind as I imagine there is going to be a lot of discussion on this RFC - you don't want to hit the limit and then realise the issue. Hope this helps! The Helpful One 19:33, 28 February 2012 (UTC)

• The draft policy is also a transcluded subpage located at Wikipedia:Pending changes/Provisional policy. It occurs to me it should probably be protected during the RFC, but as pages are usually not preemptively protected I'll leave it to you guys whether to IAR on that or not.
• Before going live, this should certainly be added to WP:CENT, I also think a full-on sitenotice may be warranted, but again, will leave that to you guys to decide.
• Part of the reason the other RFC was shut down was the promise that this would be open soon. Not immediately, but soon, so the quicker the team is all on the same page the better.
• The shortcut WP:PCRFC currently points at the giant RFC from last year. It may be a good idea to retarget it here or to come up with another shortcut.
• And with that, I'm done. I hand this baby over to the admin coordinators. Thank you so much for volunteering to handle this, I wish I had thought of doing it this way last time, Wikipedia always works better when we work together instead of trying to go it alone. Beeblebrox (talk) 17:18, 28 February 2012 (UTC)

Templates, subpages, and the like aren't an area of expertise for me, so I'll defer to bigger brains than mine on what's best. At a glance, Beeblebrox's proposed set of subpages, protections, etc sounds reasonable to me. A definite yes on WP:CENT, and I also think a site notice would be wise, to avoid the issue of interested users not seeing it (on an issue like this, where emotions seem to, for whatever reason, run high, it's best to cast a wide net so no one interested feels that we failed to notify them).

I think before we start, the triumvirate need to agree on some basic principles for what actions we expect to be taking while the RfC runs, so we're not tripping over each other's opinions. For instance, will we:

• Move tangential conversations from the RfC to the RfC talk?
• Strike double-votes, or ask double-voting users on their talks to remove one?
• "Police" the atmosphere of the RfC to keep personal attacks, emotional outbursts, etc to a minimum?
• Impose any restrictions on who may vote in the RfC (IPs, new editors, apparent SPAs...)?
• Other stuff?

These things are probably best hashed out now so we're all on the same page as far as our jobs. A fluffernutter is a sandwich! (talk)

I had some similar thoughts to you as I was reading the RfC - who is eligible to vote needs to be clearly defined. Discussions will be going on Wikipedia:Pending changes/Request for Comment/Discussion - this is what Beeblebrox meant when he said that it is transcluded - that subpage is transcluded. The Helpful One 19:58, 28 February 2012 (UTC)

I just read the AN notice and would be also happy to help as a co-ordinator or a clerk - whatever suits best. The Helpful One 23:28, 28 February 2012 (UTC)

I would have no objection to making it a "quadumvirate". The possibility of a deadlocked jury exists with the addition of a fourth person, but I don't expect or want the close to come down to a straight vote to begin with - four people ought to be able to reach a consensus as well as three would regarding what the RfC says. A fluffernutter is a sandwich! (talk) 00:02, 29 February 2012 (UTC)

I have no problem with Thehelpfulone living up to the username in any capacity with this RfC; whatever you want to do. Definitely a WP:CENT notification, and a watchlist notice can't possibly do any harm. I'd think that removing comments from obvious socks or comments that are blatant personal attacks would be well within our remit, as would removing obviously off-topic conversations (for the latter, provided at least two of us can agree that it's disrupting the process). As to handling double-votes, I think we could simply move the second vote up to where the original one was and leave it as an addendum; that should minimize teh dramahz. Anything else anyone thinks of would be worth mentioning here, and asking the community what they want us to do might help too. The Blade of the Northern Lights (話して下さい) 01:03, 29 February 2012 (UTC)

I also have no problem with THO helping out, if anything he will be my reminder of small technical details that I overlook by accident. As for the transclusions and avoiding (edit conflict), I think each endorsement should have it's own voting page, and that the discussion page should be linked, not transcended for WP:TLDR reasons. The main RfC page should be fully protected so we can coordinate things well and we don't have a random change that affects the RfC, and we should do the same with the proposed policy, but allow the proposed policy to be edited if there is a reasonable consensus during the RfC to do it. We should also encourage discussions to occur on the talkpage instead of the endorsements page, maybe by edit notice? If there are no objections tomorrow, I will implement such ideas. WP:CENT is a must, maybe a watchlist notice and a WP:AN or WP:ANI (i'm thinking AN, but ANI would stop people from saying 'we didn't know about it' because it's a very heavily stalked page). I or another RfC admin can change WP:PCRFC when it's time. Moving tangents or discussions to talkpage as I said above is the best idea. The Blade of the Northern Lights has a good idea to just move around and strike only when we need to. Excluding IPs although I can see the reason, would not really be fair. But qualifications to vote should be that the editor is not blocked or evading a block at the time of the post, otherwise strike it out. Double based IP/Account voting should be taken up at the SPI venue, and we'll go on whatever ruling is made there. As for "policing" the RfC, I would encourage the usage of {{hat}} and {{hab}} vs. {{cot}} and {{cob}} just so we don't have anyone trying to continue emotional outbursts or similar. Another note THO said on IRC was the names of the RfC admins should be in a box on the mainpage, and I tend to agree with him. Now that i've created a WP:TLDR post, and i'm probably forgetting a few things I wanted to say, I'm going to let you guys comment. -- DQ (ʞlɐʇ) 08:15, 29 February 2012 (UTC)

No objections? or did my TL;DR really become TL;DR? -- DQ (ʞlɐʇ) 00:46, 3 March 2012 (UTC)

I read it all the way through; nothing I can disagree with. The Blade of the Northern Lights (話して下さい) 01:04, 3 March 2012 (UTC)

All sounds reasonable to me too. I'm on board. A fluffernutter is a sandwich! (talk) 01:39, 3 March 2012 (UTC)

Ok. THO has commented with me off wiki, I might give him a poke tomorrow to comment here. But tomorrow, since there are no objections I'll start rolling the ball on implementations and make any extra modifications I feel are necessary, and we can bring them back here if there could be a possible issue (but that wouldn't happen because i'm perfect ;) just kidding). -- DQ (ʞlɐʇ) 09:52, 3 March 2012 (UTC)

I agree with the above, the comment that I made to Fluffernutter about the exclusion of IPs was that the outcome of this RFC will also affect them if they are trying to edit a page that has pending changes enabled, for example. Therefore, it would not be right to exclude them from expressing their views in an RFC that would potentially affect restrict their ability to (freely) edit a page. With regards to the names of the RfC admins in a box on the main page, we should possibly put that if users would like to contact us, to do so on a particular page (perhaps this one or a subpage) so that all of us will be able to see any requests easily - instead of searching around across multiple user talk pages. Yes, DeltaQuad, you are perfect. The Helpful One 15:22, 3 March 2012 (UTC)

Overlap in positions 2 and 3

I believe there is an overlap between positions 2 and 3, in that someone supporting position 2 would also support position 3, assuming insufficient consensus is achieved for the draft policy in position 2. Perhaps before the RFC is opened, a statement on how the results will be interpreted should be issued: for example, for purposes of keeping the pending changes feature available for future use, statements of support for both positions 2 and 3 will be considered in favour. Otherwise, the vote-splitting problem between those who favour the draft policy and those who do not (but still want pending changes to be used someday) may become an issue. (Another possibility is to try to establish a consensus "best draft policy" before going ahead with this RFC.) isaacl (talk) 23:43, 28 February 2012 (UTC)

Well, I designed it to be a three-way choice because the attempt to make it a black-and-white choice last time didn't work. Some people are satisfied with the status quo. They have every right to that opinion and therefore that view is represented. If I've failed to adequately express the distinction between the those two positions then some light rewording may be in order.

I don't believe there will be any way for the coordinators to state beforehand how they will perform the close. Despite the structure being employed this is still a discussion, not a vote. Beeblebrox (talk) 05:29, 29 February 2012 (UTC)

Position 2 is really a subset of position 3, though, so it isn't a simple three-way choice. I fully appreciate why three choices were given, as many people want to see a specific policy before stating support for pending changes, and so I don't see a good way to make the choices fully independent of each other. I think that the arguments for pending changes will get split between positions 2 and 3, which may complicate weighing the relative support for dropping pending changes versus keeping it in some form, either with the proposed draft policy, or some yet-to-be-drafted policy. That's why I am proposing that some guidance be given regarding how the arguments for positions 2 will be evaluated: that arguments for position 2 will also be considered in the context of position 3 as well. isaacl (talk) 05:59, 29 February 2012 (UTC)

Well, I'll leave this up to the coordinators, but I have to say I disagree with the entire premise of that line of reasoning. The way things are right now, pending changes is turned on, but it is not in use and will not be in use until a policy has already been approved. Saying we will not use pending changes and it should be turned all the way off indefinitely is one option. Using the policy we already have as a basis for resuming use of the tool with the expectation that it can be altered like like any other policy as needed in the future is another. Option three is really "do nothing." It represents no change whatsoever from the current situation. They are three distinct options, the first two have effects, and the third does not. It is fundamentally different from options one and two in that regard. Perhaps the proposals should have titles or names of their own, such as "option one: turn it off, option two:use the draft policy and modify as needed, option three: status quo. Having such distilled versions of the positions expressed as titles may help prevent further misunderstandings of this nature. Beeblebrox (talk) 07:04, 29 February 2012 (UTC)

From the previous discussion, I suspect that there will be some who will want a policy tuned to their liking before agreeing to support pending changes, and so will opt for position 3 if the draft policy doesn't meet their desires. Thus arguments for using pending changes will be split between positions 2 and 3. Everyone who supports the use pending changes can support position 3; a subset of these persons who are satisfied with the draft policy will support position 2. I believe this should be taken into account if it should become necessary to determine an appropriate compromise consensus view: Arguments made in favour of pending changes with a specific draft policy in mind should also be considered in the context of supporting pending changes with a policy yet-to-be-determined. For example, if someone puts forth an argument in support of position 2, saying that pending changes will enable the deployment of an improved workflow for evaluating edits to topics that are more sensitive to vandalism (such as biographies), this argument should also be considered for the case where the usage policy might not be the one specified in position 2, but will be determined in more detail later. isaacl (talk) 07:39, 29 February 2012 (UTC)

Technical issues

Some issues are not clear (at least to me) from the RFC page and presumably need to be discussed before the RFC goes live. Also, as I advocated elsewhere, a short summary of the previous RFC (what users generally see as positive and as negative) would help to minimize repeating the same arguments all over again. However, my points below are not about this.--Ymblanter (talk) 16:27, 29 February 2012 (UTC)

• Who can establish level of protection and protect the articles? Administrators only?--Ymblanter (talk) 16:27, 29 February 2012 (UTC)
• If a reviewer rejects changes, is it just a rollback, meaning there is no confirmation as in undo and the edit description is "reverted smth"? Or is it an undo, with the edit description which is supposed to explain what was wrong in the reverted edits? Can edits from different users be summarily reverted?--Ymblanter (talk) 16:27, 29 February 2012 (UTC)
• If it is a rollback, do we need actually two different flags for a rollbacker and a reviewer? May be just existing rollbackers become reviewers? If they do not like PRs, they may decide not to review anything.--Ymblanter (talk) 16:27, 29 February 2012 (UTC)
• Is there a record of changes accepted somewhere? For instance, an article is ordinary, and then because of some problems it gets moved to the protection level 1 (PL1). Is it then automatically clean at the moment when it is moved, and only later changes need revision? If it gets moved back and forth, what happens? This sounds like a minor technical issue, but actually it might be very much important is flagged revisions are introduced at some stage - then the version which was accepted can be marked automatically as flagged.--Ymblanter (talk) 16:27, 29 February 2012 (UTC)

I believe all your questions are already answered int eh "background" section. In order to make the RFC less cluttered that section is collapsed, you need to click the "show" button on the green bar in that section to see alll the background material, which explains how PC works and contains numerous links to past discussions and other relevant pages. Beeblebrox (talk) 16:42, 29 February 2012 (UTC)

Well, I obviously uncollapsed and read the green sections before posting this message. I checked some of the links and could not find answer to these questions. I do not think it is fair to say "hey, they are described somewhere, please check the previous RFC which is btw seven hundred thirty six screens long, and the answers should be there", and besides it is an efficient way to push people to vote against pending changes. The answers should be either on the RFC page (in collapsed sections), or on the dedicated FAQ page with a clearly visible link (may be the FAQ page exists but at least the link to it is not obvious), or it should be an invitation to request the details on the talk page, with someone volunteering to answer these questions. If this does not happen, people would start assuming the worst and to share their (mis)interpretations in the RFC.--Ymblanter (talk) 16:53, 29 February 2012 (UTC)

Now I realized my reply may sound harsh (and I hope it obvious English is not my mothertongue). I did not intend to sound harsh and as a matter of fact I am a supporter of pending changes, however, I am afraid that the RFC has a chance to be derailed again because of the issues which are actually not a part of it and are secondary to the main question. This is why I think it is important to prepare it in such a way that these issues have less chance to be raised.--Ymblanter (talk) 17:28, 29 February 2012 (UTC)

Sorry, your questions suggested you were entirely unfamiliar with the background, so I was just trying to point you to the relevant information to answer your questions. I really think your questions are answered either in that section, in the draft policy itself, or at Wikipedia:Pending changes, reading the previous RFC should not be needed to answer any of the queries you posted here, but here are some answers for you anyway:

• As with all other forms of protection, PC may only be added to a page by an administrator and is subject to the protection policy.
• The only flag needed at PC level one is autoconfirmed. PC level two requires the reviewer flag, although it entails a similar level of trust to rollback and users who have one are generally presumed competent to have the other. The differing protection levels, what they mean, and how they work are explained clearly in the large multicolored table in the background section.
• The pending changes log, linked in the box of PC related links at the top of the background section, logs all PC protection actions. Additionally, the links at the bottom of the background section lead to a page (blank at the moment as the tool is out of service) that would display all currently pending edits, and another page showing all articles with PC applied to them. Actual edits and their acceptance/rejection are displayed in page histories like any other edits. Adding PC does not automatically "clean" the article, it is another kind of protection, it prevents certain edits, it doesn't edit the article itself.
• I'm not sure what all the talk about "moving" is meant to be asking. Generally if a page is moved while protected, the protection settings move with it.

I hope this helps. Beeblebrox (talk) 17:39, 29 February 2012 (UTC)

Thanks for your answers, though my point was not so much that I need them now (for the record, I have three years experience with flagged revisions), but that other users would ask them at the RFC or assume smth which is not in the policy and will vote with this preconception. If I could not find the answers written down easily, I guess others also will not able to find them. Concerning moving a page, my question was not about moving a page from name A to name B, but from setting a protection and then lifting it again. But anyway, this is not so much important at this point, let us just see whether this issue would surface out at the RFC.--Ymblanter (talk) 18:02, 29 February 2012 (UTC)

It is possible that because of my intimate familiarity with these issues I have not made the links obvious/clear enough, I'm sure the coordinators can rectify that if needed. Beeblebrox (talk) 18:09, 29 February 2012 (UTC)

Option 4

After such a contentious debate we need to seek consensus, and that means trying to understand and accommodate each others concerns. It would be wrong to rule out compromise options at this stage, so I've proposed one. My own view is very much in favour of simply implementing flagged reviews as works on DE wiki and elsewhere, however I'm aware that there are others with deep concern about the idea of edits not going live. hence my proffered compromise. Obviously given past experience we can't expect the Foundation to change software to accommodate our concerns unless we first get consensus for that software. But that just means we need to agree the design before things are developed, it does not preclude software changes to implement things that we can agree on. ϢereSpielChequers 14:10, 3 March 2012 (UTC)

I would note for the record that I deliberately did not include such a position when designing this process because it has been rejected by the community in past discussions. The data from the last RFC shows that out of all the 100+ participants in phase two, only four users supported such a position. In fact it tied for fourth least supported position. Phase two is also a perfect example of how having too many options leads to a situation without usable results. If we are going to arrive at an actionable result, we need to keep this as simple as possible. If we start letting every alternate proposal that three or four users agree with get added to it, we will have failed before we have even begun. Beeblebrox (talk) 18:48, 3 March 2012 (UTC)

I have to agree with Beeblebrox here, we have to keep it short and tight nit otherwise it's just going to be all over the place again. Three options do make users to choose between options, but if were going to get anything going, we have to leave it at three options. -- DQ (ʞlɐʇ) 01:47, 4 March 2012 (UTC)

I also agree with both Beeblebrox and DeltaQuad, I believe the least supported position that Beeblebrox was referring to is that which is stated here on the RFC page of last year (which had far too many options to choose from - hence why there was no clear consensus). The Helpful One 01:55, 4 March 2012 (UTC)

Knew I forgot something before clicking save, it was that link :P -- DQ (ʞlɐʇ) 01:57, 4 March 2012 (UTC)

I would anticipate that this particular option would be very few people's first choice. But more importantly very few would actually object to it, as it doesn't contain the features of pending changes that were commonly objected to. The option thehelpfulone links to above is very different to option 4 and I'd agree that should not be revived. If we are going to reopen the subject of pending changes then to my mind we need to try and move the subject away from the features and options that large minorities object to, and instead see if we can find compromises that we can all live with. Option 4 might not be the best the community can come up with, but if we are to seek consensus one way to do so is to have people propose compromises that achieve part of what they want whilst resolving other people's concerns. Perhaps option 5 or 6 might be the answer that we can almost all accept. Options 1-3 just revive the old battle - to achieve consensus we need to move beyond them. ϢereSpielChequers 02:01, 4 March 2012 (UTC)

In actuality we haven't had this "battle" yet. No decision has ever been made as to whether PC should be permanently deployed here. This process was carefully designed to address 'that question only, although the option to keep the status quo is included. We can't let this thing descend into another mythical beast of a discussion, where every time one person doesn't like one particular option we just create a new position to be endorsed and then someone doesn't like that and they create another, and so forth until there are forty proposals and no possibility of a consensus. If we are going to start adding every possible permutation we might as well quit now. Ground rules were established before beginning this time because the lack of any controls is what made the last discussion such a disaster. If you want to propose something else that is another discussion for another RFC. Beeblebrox (talk) 02:55, 4 March 2012 (UTC)

I may raise this proposal as a separate RFC, but it does seem a shame to me if people are trying to constrain the debate. Consensus should not be about giving people an artificially constrained set of choices and ruling out the possibility of consensus solutions emerging from the discussion. ϢereSpielChequers 22:39, 4 March 2012 (UTC)

I don't recall if you were one of the participants in Wikipedia:Pending changes/Request for Comment February 2011, but failure to have controls in place before beginning is what caused it to fail to achieve its primary objective. Beeblebrox (talk) 23:48, 4 March 2012 (UTC)

It's more a case of postponing discussion on other options. As I understand it, Beeblebrox would like to establish a consensus first that moving ahead with pending changes is desired. Once this is done, the discussion can proceed with more elaborate possibilities, without the spectre of a lack of consensus for pending changes overshadowing the debate. isaacl (talk) 00:23, 5 March 2012 (UTC)

Well, said, that is exactly the intention. And of course there is the issue that the WMF has made it clear they will not expend any more resources on PC until we do have a consensus on its use. Beeblebrox (talk) 00:44, 5 March 2012 (UTC)

That's precisely why we need proposals such as my proposal for pending changes without branding edits as "not yet accepted". The Community won't agree to implement the version of ending changes that was trialled, the WMF won't make changes unless we have consensus for those changes, so if we want to revive the idea of pending changes we need to discuss options that achieve at least some of the benefits of pending changes without some of the perceived disadvantages. The flipside of the WMF being unwilling to make changes before we get consensus to implement them is that we can reasonably assume that if we do get consensus for a compromise form of pending changes the WMF would then do the development needed for that compromise to happen. ϢereSpielChequers 18:13, 5 March 2012 (UTC)

Not to be a pain, but....

Are we waiting for something in particular? There's been no action here in over a week. Beeblebrox (talk) 15:21, 14 March 2012 (UTC)

Thank you for poking. I've been meaning to poke the rest of the RfC admins, I got one or two on IRC today, but afaik were about ready to start...unless i'm forgetting something which is very possible at this hour and in my state. Anyone else have any objections before we start moving this thing? or last minute ideas? -- DQ (ʞlɐʇ) 06:24, 15 March 2012 (UTC)

My apologies for not responding earlier; I've been busy whacking people in Indian caste articles and cleaning up after the kana vandal made yet another couple of raids. I'm set and ready to go; I think we have things nailed down about as well as we can for now. The Blade of the Northern Lights (話して下さい) 05:14, 16 March 2012 (UTC)

I've also left a note for Fluffernutter; hopefully we'll hear from her soon. The Blade of the Northern Lights (話して下さい) 22:02, 19 March 2012 (UTC)

Sorry, I saw these posts a few days ago and it just sort of fluttered out of my mind since then that responding would be, you know, a useful thing to do! I plead rl stress :/ Anyway, I'm prepared for the RfC to open if all us admins think it's ready, but I don't know that the actual responsibility of opening it should fall to us rather than to the initiator of the RfC once we're agreed that it's ready to go. So basically I'm ready to see it launch, but I'd rather Beeblebrox or someone else be the driving force behind actually hitting the big red "GO" button. A fluffernutter is a sandwich! (talk) 22:49, 19 March 2012 (UTC)

I can do that if needed, but then I really am done, except to add my endorsements in the rfc itself. I've just asked for a sitenotice [1] which hopefully wil get the green light soon and we can begin. Beeblebrox (talk) 00:42, 20 March 2012 (UTC)

I forgot that this issue is not deemed important enough for a sitenotice, now requested as a watchlist notice [2]. Beeblebrox (talk) 02:59, 20 March 2012 (UTC)

• Watchlist notice is on, RFC is open! Here we go! Thanks again to the volunteer coordinators! Beeblebrox (talk) 17:40, 23 March 2012 (UTC)

Note about the use of PC elsewhere

Pending Changes is extensively used on the MediaWiki wiki as well, not just enwiki.Jasper Deng (talk) 00:52, 20 March 2012 (UTC)

A note about watchlisting

I understand the reason this was set up this way, but people who are used to watchlisting pages like ANI are probably going to get cranky. It might be too much information to include on the page, but is it worth noting somewhere on the page that people can't watchlist Wikipedia:Pending changes/Request for Comment 2012, they'll need to watchlist all 3 subpages? Or do we assume people will figure that out for themselves eventually? --Floquenbeam (talk) 18:26, 23 March 2012 (UTC)

I just ran into this same frustration - and I'm one of the quadrumvirate who agreed to use this setup in the first place! I'll add some information to the top of the RfC about watchlisting. A fluffernutter is a sandwich! (talk) 18:29, 23 March 2012 (UTC)

 Done with something I quickly drafted up. A fluffernutter is a sandwich! (talk) 18:43, 23 March 2012 (UTC)

To add your endorsement, you must edit the section it is in....

Please change {{imbox|text='''To add your endorsement, you must edit the section it is in, editing of the main RFC page has been disabled'''}} to {{imbox|text='''To add your endorsement, you must edit the section it is in; editing of the main RFC page has been disabled.'''}} →Στc. 03:06, 24 March 2012 (UTC)

 Done, although the grammar of our notices is probably the least important thing here to fret about... A fluffernutter is a sandwich! (talk) 04:42, 24 March 2012 (UTC)

Just keep asking

Moved to the discussion page. -- DQ (ʞlɐʇ) 06:03, 26 March 2012 (UTC)

Where was everybody when semi-protection was proposed?

Moved to the discussion page. -- DQ (ʞlɐʇ) 06:03, 26 March 2012 (UTC)

Discussion section not transcluded?

Hi. Why is the discussion section of the RfC not transcluded? Thanks! Allens (talk | contribs) 11:10, 24 March 2012 (UTC)

I can't answer that question, because I haven't seen it. FormerIP (talk) 01:23, 25 March 2012 (UTC)

It's linked to at the bottom of the RfC, just as the edit links are done above there for the other three sections that are transcluded. Wikipedia:Pending_changes/Request_for_Comment/Discussion. Allens (talk | contribs) 04:13, 25 March 2012 (UTC)

Why don't we just merge that page to this talk page, as having discussion in the "Wikipedia talk" namespace makes more sense? --MuZemike 17:10, 25 March 2012 (UTC)

Well, I would think that page would be for things like questions about the proposed policy (or possibly for opposition to one of them without favoring either of the other two), while this page should be more for discussion of the process, not the proposed policy. Allens (talk | contribs) 17:16, 25 March 2012 (UTC)

Our main reason for not transcluding the discussion page was so that extended discussions could happen away from the core RfC. This is because there are transclusion limits, (see Wikipedia:Template limits), and so if the inclusion limits are exceeded, then we wouldn't be able to transclude any of the subpages. However, it appears that this seems to be causing some transparency issues and we do not wish to discourage discussion. Therefore, we are going to transclude the discussion page, but bearing in mind the template limits, if we start to face problems due to extended discussions, then we may need to remove the transclusion of the discussion page.

For the RfC admin co-ordinators, The Helpful One 18:20, 25 March 2012 (UTC)

Thanks! Allens (talk | contribs) 20:04, 25 March 2012 (UTC)

Which of the two discussion pages do you mean? Why are there two discussion pages in fact?--Victor Yus (talk) 18:24, 25 March 2012 (UTC)

The page that's just been transcluded to the main RfC page is for discussion of Pending Changes, people's votes, etc. This discussion page - the one that you get to when you click the "discussion" tab attached to the RfC - is intended for discussions about the running of the RfC. So if you want to comment on, say, someone's !vote rationale, use the transcluded discussion page. If you want to tell the RfC admin coordinators that we've got a typo, or can we please transclude something, or something else about the RfC at a meta level, you use this talk page. A fluffernutter is a sandwich! (talk) 18:43, 25 March 2012 (UTC)

Thanks, that even sounds logical. But you need to make the instructions for doing that clearer than they are now. Also, how about transcluding the discussion ABOVE the vote sections, to make it a bit more likely that people will read the discussion before they vote (which is what we ought to want, no?)--Victor Yus (talk) 19:17, 25 March 2012 (UTC)

That's what we do for RfAs, so I wouldn't have a problem with that. The one issue is that the discussion is largely going to consist of comments on different votes, so that would mean people have to scroll down to the vote being mentioned and then scroll back up. I'd like to hear from the other coordinating admins; I'm pretty much fine with whatever they want. The Blade of the Northern Lights (話して下さい) 02:58, 26 March 2012 (UTC)

I'm still a little iffy about transcluding it for TL;DR reasons, but i'm open to a change. But yes, we should effectively move it up above the !voting section. We may have screwed this up by putting the RfC Process comments here, and the other one at /comments, instead of the reverse, but i'm not inclined to flip it aorund right now considering the pages are already being used quite heavily. -- DQ (ʞlɐʇ) 06:09, 26 March 2012 (UTC)

Draft policy

Moved to the discussion page. -- DQ (ʞlɐʇ) 06:03, 26 March 2012 (UTC)

Excuse me, but why is discussion being shuffled off to some further distant page? Why is it not here on the talk page? This very much gives the impression of gerrymandering. Risker (talk) 03:19, 29 March 2012 (UTC)

It's been moved to the discussion page that's transcluded onto the main RfC body, Risker - it's more visible there than it was here. A fluffernutter is a sandwich! (talk) 04:26, 29 March 2012 (UTC)

I disagree with that, actually, since only people who watchlist the discussion page will see actual ongoing discussion. More importantly, why is the discussion positioned below the voting? The discussion should come BEFORE the voting. For that matter, why is there absolutely not a single word in this RFC providing analysis of the trial? Risker (talk) 04:18, 30 March 2012 (UTC)

You'll also notice that this page was not even transclused to the RFC page until significantly after the RFC went live. That combined with the preloading of something like 50 support votes for position 2 and the "Just keep asking" issue that both myself and WFC have brought up and... "gerrymandering" seems somewhat generous, actually. Our concerns are nothing more than personal attacks though, so... *shrug*
— V = IR (Talk • Contribs) 03:41, 1 April 2012 (UTC)

We initially thought that it wasn't a good idea to transclude the discussion page because it's almost certainly going to hit the transclusion limit; enough people asked for it that we decided to do it. And rest assured, none of the four of us have a position on PC; that's why we're the ones coordinating it. The Blade of the Northern Lights (話して下さい) 03:58, 1 April 2012 (UTC)

Concern about the structure of this RfC

I agree wholeheartedly with Risker. I'll try as hard as possible not to reiterate my arguments from the RfC, but am concerned about the way this is structured.

The whole point of turning PC off was to evaulate what was right and wrong with the trial, and then come up with a policy which mitigates the technical shortcomings of PC. We already knew with near certainty that once this process had happened, there would be consensus to try again. Yet, for no obvious reason other than the passage of time, we are now effectively (if unintentionally) being asked if there is in principle consensus for PC, and are denied the option to express a "support, but" position, instead forced into either "support" or "Hang on a minute". A cursory look at position two shows just how many people feel that there are improvements we can and should make.

Maybe my previous paragraph is unduly harsh on the RfC drafters (who I commend for getting this discussion back on track, whatever I think of the precise manner). It is fair to say that formal, recent consensus for PC needed to be re-established, however sure we were that it was already there. But we should not simply pretend that time and a rewrite of the same policy will make PC any better than it was last time. To close this RfC and immediately restore PC based on this RfC would be to do precisely this.

My alternative? A second phase to this RfC. It should make clear at the start that PC will be introduced at the end of that process, absolutely regardless of the outcome – if there is no consensus for any amendments to the policy, then we go with the policy from position two. Once we give supporters of PC the piece of mind that there is zero chance of a critical comment being interpreted as being in some way against PC, we can then be confident that they will give their full and frank opinions on what lessons we feel need to be learnt from last time. —WFC— 02:51, 31 March 2012 (UTC)

But the policy from position two is no policy at all - look at all the unanswered questions it's generating on the discussion pages. It seems from the current voting that most people want PC to be available as an option alongside the other options that protection policy provides - but that's it. (I don't believe all the voters have considered the wording of the draft policy, or even necessarily looked at it, since it's hidden from view by default.) Now we need some reasoned discussion to reach conclusions about the unanswered questions and apparent logical contradictions, so as to reach a situation where we have a full and transparent policy, and without some default pseudo-policy hanging over the process (which would allow people who feel comfortable with such vagueness to hold the rest of us to ransom, to use a rather extreme metaphor).--Victor Yus (talk) 08:49, 31 March 2012 (UTC)

A quite likely errant assumption - for example, I most absolutely read the "draft policy." And I suspect most people do read material even if "hidden". Collect (talk) 12:09, 31 March 2012 (UTC)

A quite likely errant assumption. A large number of people have supported without comment, which is often albeit not always an indication of spending little time considering the matter. Others openly express reservations about the previous trial, completely ignoring the fact that there has been no material change in policy since then. Yet more say that they support PC but want policy to be improved – these people are either cavalier to the point of wanting PC but not caring if it fails, or haven't read the RfC (if they had they would have gone to option 3).

It is beyond question that the community wants PC and must get PC. It is equally beyond question that the structure of this RfC makes it systemically difficult to discuss the policy and nigh-on impossible to meaningfully change it. Whether by accident or design, a wishy washy carbon copy of previous policy has been rubber stamped before even being discussed. —WFC— 05:59, 1 April 2012 (UTC)

• As the principal author of the RFC I am rather alarmed at the comment that the option to say "hold on a minute" isn't available. That would be position 3, and if that is what you feel you should endorse that position. I put it there specifically because I knew there would be users who liked PC but didn't want to use the policy we used during the trial, which is essentially what the draft policy is. I am honestly quite surprised at how few users have endorsed position 3, but it seems like a majority want to turn it back on and work on the policy as we use the tool instead of trying to perfect the policy first. Beeblebrox (talk) 18:31, 31 March 2012 (UTC)

No, that is not Option 3. Option 3 is "yes, but we have to work on the policy", and not "hold on while we actually look at the facts here" - which is the step that is missing entirely from this process. Since you're the primary author, Beeblebrox, could you please explain why you've provided no factual information about the trial in this proposal? Risker (talk) 01:33, 1 April 2012 (UTC)

What data is available is at Wikipedia:Pending changes/Metrics which is linked in the "background information" section up near the top of the RFC, along with links to every previous discussion, poll, and RFC . I wasn't really involved in this issue until after the trial and despite asking about it a few dozen times in the last RFC I have yet to hear a satisfactory answer to the question of who was supposed to be gathering and managing the relevant information, and apparently nobody did any follow up in the intervening year that has passed since that discussion. So, the short answer would be that it appears that nobody was ultimately responsible for that and nobody was interested in doing the work and therefore the data already linked to this RFC is all we have. In my opinion the only real positive to come out of the badly mismanaged trial is that users got some hands-on experience with PC and how it works and it seems like the vast majority feel able to make their decision, be it for or against PC, based on that and the information presented here. The goal here is not to present a perfectly finished policy but rather to finally eliminate the ambiguity regarding where consensus is as regards this tool. We simply can't move forward until that question is answered.

WP:FPPR, the original proposal that eventually led to the development of PC, went live thirty six months ago, the official trial period ended about a year and a half ago, and the last RFC ran for 101 days last year and is one of the longest discussions I've ever participated in, and I didn't even stick around till the end. My personal preference was always for the "questionnaire phase" that I designed last year that would have allowed each participant to submit their own ideas and comments along with replies to ten questions about their position on PC, but that idea was blocked and not allowed to proceed. I kniow this isn't perfect, I did the best I could with what was available. I'm not mathematically or technically inclined so gathering statistics and data is not something I would have been very effective at and it has become clear that nobody else was willing to do it despite the fact that it has been asked for again and again. Beeblebrox (talk) 02:52, 1 April 2012 (UTC)

It is not my intention to misquote or misrepresent you here, hence me going out of the way to be clear that the following sentence is my interpretation. My reading of what you have just said is that this RfC launched despite the principal author knowing full well that the single most important thing that needed to be done wasn't and will not be done.

Could you point us in the direction of where the idea of a questionnaire was blocked? I find that claim incredible. —WFC— 04:51, 1 April 2012 (UTC)

It could take some digging to find the exact section due to the extreme length of those conversations but the general gist of what happened there is that any third phase of the discussion that was not an attempt to discuss and resolve the short-term issue of the trial period never having ended was blocked. We had a general conversation and a rather long list of position statements for endorsement, but this discussion we are having now was put on hold. The actual questions are still lying around at Wikipedia:Pending changes/Request for Comment February 2011/Review Recommend phase. We did get two of them back before the third process was halted.

As to data/statistics/etc, if those who wish to know more could be explicit in what it is they are asking for that is not in the data we already have (which as I mentioned before is collected at Wikipedia:Pending changes/Metrics) perhaps a method of gathering this missing data could be devised and your questions could be answered. Beeblebrox (talk) 07:25, 1 April 2012 (UTC)

Thanks for that Beeblebrox.

My recollection is that the turning off of PC was considered a prerequisite to constructive discussion on its future, but in turn that further discussion on how we should use PC was a prerequisite to it being turned back on. I can therefore see why the questionnaire was rejected while the trial was still going, but not why it shouldn't have been used afterwards.

I think that there is an accepted consensus on the answer to some of those questions. But discussions on questions two and five in particular need to happen, now that the decision that PC is definitely going to come back has (barring the mother of all reversals) been taken. The best way to critically evaluate something is for opponents who recognise this isn't a black and white issue highlight elements of PC or possible applications which might have a degree of merit, and for supporters who recognise that what they are supporting isn't perfect elabourate their concerns. The way you achieve that in this instance is to take the decision on activation first, make clear that the decision has been taken and will not change, and finally discuss the precise mechanics. From where this RfC is today, we can still follow that path. But bearing in mind that dramatically changing a live Wikipedia policy is literally impossible, we ABSOLUTELY MUST ensure that the policy has been shaped properly before activation. If we don't, we are condemning ourselves to further divisive, unproductive discussions, and quite possibly condemning PC to long term failure. —WFC— 07:59, 1 April 2012 (UTC)

See Wikipedia talk:Pending changes/Request for Comment February 2011/Archive 2#Phase three draft and subsequent sections for a discussion on a proposed questionnaire, which turned to disagreement on the questions and the continuing trial. Wikipedia talk:Pending changes/Request for Comment February 2011/Archive 3 follows up with more discussion on different possibilities in soliciting ideas, and the desire to halt the continuing trial before proceeding with anything else. isaacl (talk) 07:00, 2 April 2012 (UTC)

There was also an ill-fated attempt by me to get mediation to help us decide how to proceed. I was at the end of my rope trying to move the process forward and how the mediation unfolded (or didn't) was what caused me to quit the last proceeding, but let's not get into all that. My point is that shortly before withdrawing the request I assembled a list of all the ideas that were on the table at that time, you can see them here. As you can see they range from simple and sensible to complicated and a bit wacky. Everyone was a bit sick of the discussion by then and none of these ideas was pursued until now. The one defined on that list as "go for broke" was an earlier version of what you see here. Beeblebrox (talk) 19:15, 2 April 2012 (UTC)

I just thought it worth mentioning that I don't have a concern about the structure of this RfC. While I agree that implementing PC without hammering out every detail will cause issues, this has become a matter of shit or get off the pot. Editors have been trying to debate the details and it has gone nowhere because they can't even agree that they want PC. This is a wiki and when a problem arises it can be undone easily. There has never been one policy decision or software change that has crippled the wiki. I don't know exactly how this will be implemented, and I don't care. I firmly believe it will be implemented poorly, that we will have countless discussions, and in a few weeks/months it will reach a pleasant stability. If you want safe and stagnant, I'm sure there is a firesale on the complete Brittanica. Progress is progress, go team, etc. Thanks Beeblebrox for kicking this into gear. ▫ JohnnyMrNinja 23:16, 4 April 2012 (UTC)

To summarise Johnny's point: "it will definitely be shit when it starts, but will be nigh on perfect a few weeks or months down the line". There are more flaws in that logic than I care to go into, but the biggest is the assumption that Wikipedia is capable of sorting out noticeable problems with current policy in a timely manner. Lest we forget, the problems if we launch in current form today are not new: they are precisely the same problems that PC encountered two years ago and was ultimately shut down over.

The question of whether PC should return has been settled beyond dispute: over 70% have actively supported it in some form. That gives us a window of opportunity to define how it should be deployed before it goes live. Doing so is the best and quite possibly only way we have of preventing PC from being shut down again down the line. —WFC— 05:34, 15 April 2012 (UTC)

Is it possible for people to say explicitly what they think the problems were that PC encountered in the previous trials?--Victor Yus (talk) 10:53, 15 April 2012 (UTC)

@WFC: For the record, PC was not shut down due to problems with the tool itself. It was shut down because it was supposedly turned on on a trial basis, but then when the trial period was over..... nothing happened, it was just on without a clear consensus on whether or not it should be used. That is what both this RFC and the previous one aim to resolve. The second the RFC is closed and the we are allowed to use the tool again anyone who wants to can open discussion on what they perceive to be flaws in the policy. That discussion probably won't require the sort of highly structured environment employed in thos RFC as it will probably not involve near as many users as this one.Beeblebrox (talk) 18:43, 15 April 2012 (UTC)

I'm acutely aware that to those who have followed this in its entirety, I am beginning to sound like a broken record. But I really want PC to work, and consider getting this right important enough to continue until I have some confidence that we are either going to make a conscious effort to define PC's strengths and limitations, or until the closers go on record as saying that in their opinion the community actively wants PC to start off in the same manner as before.

Beeblebrox: that's true, but ultimately a clear majority favour the principle of PC, and yet there was clear consensus to shut it down. Yes, there were those who called for it to be shut down because we said all along that we would do so after a certain period of time; I was one of them. But notwithstanding that, there wouldn't have been consensus to shut PC down if we felt that it was a net positive that we were not prepared to lose for any period of time.

Victor: the primary problem was that PC was deployed in situations for which the interface was unsuitable. Articles, generally on very high profile subjects and/or subjects in the news, in which there were large quantities of both good faith and bad faith edits, making it difficult for admins and reviewers to keep up and increasing the likelihood of bad edits getting through/good edits getting rejected. Statistically users are particularly likely to see these articles, so opinions of PC can be disproportionately affected by a small number of "bad" implementations. And that's the key. Unless we have some sort of definition of what a "bad" implementation is, PC's future is largely dependent on whether or not every admin prepared to use it understands its strengths and limitations. —WFC— 20:31, 15 April 2012 (UTC)

So perhaps the way to proceed is for someone to draft a new version of the WP: Protection policy (which is not even a great piece of documentation as it stands) so as to cover PC and to include guidance as to which kinds of situations are suited to PC and which kinds are not. I suppose we would also need a WP:Reviewers policy, analogous to WP:Administrators. Victor Yus (talk) 07:12, 16 April 2012 (UTC)

Closure of this RfC

[I will leave the honour of deciding upon a suitable caption to the RfC closers.]

I'm not really familiar with RfC procedure. How long does an RfC remain open until it gets closed?—cyberpower ^Chat_Online 20:16, 15 April 2012 (UTC)

The short answer is that it generally closes either after 30 days, or sooner if the closers feel the outcome is not going to change, which is clearly the case here. I think the only reason to delay at this point is that it's not entirely clear what should happen as a result of this RfC's closure. In my opinion there are three plausible scenarios:

1. PC is switched on immediately using the draft policy in the RfC.
2. The closers specify a date (presumably in a small number of weeks' time) that PC will be return come hell or high water, and encourage the community to discuss the policy in the meantime.
3. The closers acknowledge that despite the human tendency to vote yes or no on a polarising issue, there is an underlying consensus that we need time to develop policy further, and therefore that while there is consensus that PC will return, it is too soon to set a date.

My preference would be for three, but in my opinion two would be a more reasonable interpretation of the consensus. An argument can be made that there is consensus for option one, but in my opinion to do that would be more cavalier than bold. —WFC— 20:50, 15 April 2012 (UTC)

I see.—cyberpower ^Chat_Online 21:14, 15 April 2012 (UTC)

Yep, that's more or less the plan. Although we'll probably need a couple days to hammer out a statement. The Blade of the Northern Lights (話して下さい) 00:30, 16 April 2012 (UTC)

Being an admin coordinator, if this were to be closed now, what would be the result?—cyberpower ^Chat_Offline 00:33, 16 April 2012 (UTC)

For now, I'll have to go with 無. The Blade of the Northern Lights (話して下さい) 01:19, 16 April 2012 (UTC)

@WFC: You speak of the "outcome" not going to change, but there is no outcome yet. Correct me if I'm wrong, but it has been my understanding throughout my years of editing here that the outcome of an RfC is established when it is closed, and anything before that is only prediction or speculation. He/she/they who close it need to assess the relative strength of the arguments to determine (1) whether consensus exists and (2) where consensus exists. While you're welcome to suggest that the return of PC is a foregone conclusion, and you may well be correct, the process hasn't reached that point at this time. Rivertorch (talk) 01:06, 16 April 2012 (UTC)

A valid point, although I'm personally of the "let's not pretend that this is anything less than certain" school of thought. —WFC— 02:12, 16 April 2012 (UTC)

My single-person perspective, from within the admin-coordinators camp, is that 30 days will be the minimum and that we should be open to leaving the discussion open longer than that if we're still getting a steady flow of input, as we seem to be currently. I haven't had a chance to discuss that with the other admin coordinators, however, so please don't take as coordinator-gospel anything said so far regarding the close schedule. A fluffernutter is a sandwich! (talk) 02:16, 16 April 2012 (UTC)

But is it certain? How do you know? Surely not because of the relative numbers. As of this writing, 78 Wikipedians—most of them old-timers, including quite a few longtime sysops and at least two sitting arbitrators—have presented an array of cogent arguments for Position #1 in an RfC the outcome of which has momentous implications for the future of the Project. Unless you know something I don't—and I wouldn't care to think what that might be—I assume that those arguments will be considered with great care and thoroughness. While I sympathize with the reasons for the streamlined configuration this RfC has taken, let's not pretend that its format lends itself to easily amalgamated positions. This isn't about another trial; it would be the real deal, and very hard to undo even if it proved disastrous. Since the stakes are so high, the closers really must consider every comment. Rivertorch (talk) 05:01, 16 April 2012 (UTC)

To answer your direct questions: it is certain, and I just do. I agree with what you say about the structure of this – I think the co-ordinators will readily concede that the intention of this RfC was to produce a decisive result, and that their skill in interpreting the various shades of grey might well determine whether the process has been a triumph, a waste of time or a total catastrophe. But when push comes to shove, it is beyond dispute that the community is very clear in its wish to see PC available in some form, and beyond question that when this is rubber-stamped, the priority of people from all parts of the opinion spectrum will turn to ensuring PC is a success/minimising the damage it causes to the project before we get rid of it.

I disagree with the rest of what you say on a number of levels. On numbers, only nine people have explicitly gone for option 3 thus far. Yet through a combination of the implicit support of a large proportion from option two, and the fact that we are making strong points which have not been refuted, it is inconceivable that this RfC will be closed without taking into account the community's desire to take one last careful look at policy before the inevitable relaunch (precisely how being a matter for their judgement). I resent your – admittedly skilfully crafted – inference that those outside of the 78 (many of whom did nothing but nod and vote) are not high calibre editors or do not care about the project's future.

As for cogent arguments, you're not even convincing each other. Some argue that PC would be the death of our core principles of instantness and openness; that it will directly harm our ability to attract new editors. Yet among the very same number are those who reject PC on the grounds that we should use semi and full protection instead. And if the arguments are so persuasive, why are an ever dwindling proportion of editors opposed to the idea of PC, despite it being generally accepted that the first trial was a failure? —WFC— 07:48, 16 April 2012 (UTC)

Sorry, but I'm not looking at this in terms of "us versus them"; while I did speak in support of Option #1, the concerns I've expressed here are largely procedural and would still apply if the shoe were on the other foot. I make no apology for being concerned about statements suggesting that the result of this RfC is somehow a foregone conclusion. Such talk seems dismissive of a lot of editors and their reasoning, if not of consensus itself. As for my "inference", I don't agree that my failure to explicitly praise the Option #2 arguments can reasonably be construed as suggesting anything negative about the Wikipedians who made them. If it makes you feel better, I'll state explicitly here and now that I've never drawn any inferences about the "calibre" of my fellow editors or their commitment to the project based on their !votes in a single RfC. Not even "nod and vote" !votes, as you put it, of which there were some on both sides. (Aren't there always?)

There has been no failure to refute anyone's "strong points"; while this RfC doesn't permit direct refutation, I believe that every point, strong or otherwise, has been refuted quite thoroughly. And I don't suppose that Option #1's supporters are trying to "[convince] each other" of anything; the arguments against pending changes are numerous and diverse, as I noted above. To answer your question, I have no idea whether the numbers are "ever dwindling"; they're hard to quantify relative to earlier figures because of the vastly different nature of the different RfCs. If they are dwindling, I suppose there are a number of plausible explanations, none of which involve anyone's having been persuaded of the desirability of implementing pending changes. Rivertorch (talk) 13:59, 16 April 2012 (UTC)

On a positive, I'm grateful for your clarification at the end of the first paragraph. Alas, as with the RfC itself, from my perspective I believe shortcomings of some of your other points are self evident, and from your perspective I suspect you believe that my decision not to respond indicates that you are right. I'm content to see what the closers make of the balance of any relevant arguments.

But I will respond to the last point. While numbers should never be the overriding factor, in practise they are always a factor. The initial straw poll (in which strength of argument was not really taken into account) closed with a ratio of slightly less than two supports per oppose. Since then, we have had a trial which was not deemed successful enough to be kept at the time, and stern criticism from all sides of the PC trial for not abiding by the timeframe which had initially been set out. In spite of that, now that strength of arguments are being taken into account, the ratio of supporters to opposers has risen to over 2.5:1. I draw two main conclusions from this, in relation to people's confidence in their original arguments and the persuasiveness of the argument for PC. I invite everyone else to draw their own conclusions. —WFC— 15:08, 16 April 2012 (UTC)

I was similarly wondering when this RfC shall close. Personally, I voted for option 2. I find this reasoning of "Free editing will be destroyed" to be hugely flawed. Pending changes will not stop any editor from editing, not even the IP adresses. Its just that their edits will have to be reviewed. This is totally unlike semi-protectin and protection where certain sections of editors can't edit at all. According to me, that is the destruction of free editing. In my opinion, pending changes would be an extremely useful level of edit protection that kicks away nobody but still ensures the articles don't get unnecessarily messed up. ~*~AnkitBhatt~*~ 10:29, 23 April 2012 (UTC)

The admin coordinators are discussing a close timeline for the RfC right now. It will almost certainly not be on the 30-day mark, but we'll let everyone know when we've reached a firm decision for when it will be. For the admin coordinators, A fluffernutter is a sandwich! (talk) 12:10, 23 April 2012 (UTC)

http://en.wikipedia.org/w/index.php?title=MediaWiki:Watchlist-details&diff=483558091&oldid=482895665 – The watchlist notice has expired, and this week's Signpost doesn't include a "Discussion report". It'll be difficult for new potential voters to become aware of these discussions. Perhaps we should extend the watchlist notice for another week, or perhaps we should post a "last call for !voters" message on the Village Pump. --Michaeldsuarez (talk) 14:35, 24 April 2012 (UTC)

Closure date set for 22 May. We're working on getting notices back up so the community will be aware of the Rfc and its time frame. A fluffernutter is a sandwich! (talk) 17:00, 26 April 2012 (UTC)

Sounds good. --Michaeldsuarez (talk) 18:48, 27 April 2012 (UTC)

A new watchlist notice has been created. Cheers, theFace 18:49, 28 April 2012 (UTC)

I'm baffled as to why the RfC has been extended, given that discussion is very much static. Please could one of the co-ordinators explain the thinking behind this? The only reason I can think of is that PC is a controversial issue, but I don't see why that in itself would be a relevant factor.

If this RfC is to be closed as an endorsement of PC, the time between now and the end of May would be better spent focussing on a new discussion about PC's re-implementation. I would suggest setting a hard "turn on" date, enabling the community to discuss in the meantime whether to tweak the provisional guidelines or go ahead as-is. If this course of action were adopted today, that turn-on date could quite conceivably be the end of May, which compared to the current timeframe would in no way delay the restoration of PC. —WFC— 08:41, 29 April 2012 (UTC)

I also don't see the benefit in almost a month extension. # was/is there a discussion somewhere that explains why it was done and or what the reasons/benefits are aledged to/going to be? Voting and discussion of this element is now circular - lets get this closed and move on to discussion of what form and when implementation will take. Youreallycan 08:51, 29 April 2012 (UTC)

The situation isn't as simple as you might believe. At the moment, it's pretty difficult to determine whether the current numbers mean "consensus" or "no consensus". Given how controversial Pending Changes is (the opposition to PC is in the triple digits), it's best to wait for more !votes in order to avoid making the wrong decision. I can imagine how difficult being a closing admin for this discussion could be. In addition, many !votes were made in the last 24-hours. --Michaeldsuarez (talk) 02:52, 30 April 2012 (UTC)

I respect the difficulty in closing this. My concern is less about the admins taking time to do so, more the prospect that the closure will be as straightforward as "consensus for PC, it is on with immediate effect" or "no consensus for PC, it stays off". I hope that I'm reading too much into this, but the extension suggests to me that this is the direction of travel. —WFC— 03:04, 30 April 2012 (UTC)

Why not close the "voting" part at this stage (we can all see the numbers) and get down to some serious work on rewriting the Protection Policy to incorporate PC, saying what criteria we think should be taken into account when choosing between PC and other forms of protection. The "discussion" part of this process apparently shows that this question badly needs answering, and people here have put forward what seem like reasonable answers, so it oughtn't to be a difficult task. Victor Yus (talk) 05:45, 30 April 2012 (UTC)

Yes, there is a very clear consensus to move forward in support/switching back on in some manner of reuse of pending protection - Clearly the opposes are opposed to that here but that is to be expected , isn't it - but its no reason to stall or extend a clear support position - Youreallycan 12:59, 30 April 2012 (UTC)

Speaking from my perspective, and I imagine to varying degrees the other coordinators here, we'd rather have our heads chewed off for letting this go longer than necessary as opposed to hearing people say "I didn't get a chance to have my 2 cents". The Blade of the Northern Lights (話して下さい) 14:53, 30 April 2012 (UTC)

The support is currently 68%. No, that is not consensus. The standard threshold for consensus has never been anywhere near that low, certainly not for a permanent change as major as this. --Yair rand (talk) 15:58, 30 April 2012 (UTC)

Allow me to suggest that you allow the admin coordinators to do what we signed up to do; I assume you think we're capable of doing our jobs. The Blade of the Northern Lights (話して下さい) 16:02, 30 April 2012 (UTC)

The support is 67% as of now. I don't see how that's bad. A small example, most political and international organisations opt for a two-third majority policy whenever passing any sort of change, be it far-reaching or temporary. This RfC definitely seems to be the case of that. ~*~AnkitBhatt~*~ 17:46, 30 April 2012 (UTC)

Many legislatures I know of are happy with 51%. I also don't think this is a major change for Wikipedia. So even though I'm on the minority side, I have no problem with accepting the majority standpoint. But it needs to be got right; the questions that have come up in the discussion need to be answered and a proper policy or guideline worked out. I thought Wikipedia was supposed to make its decisions primarily through such discussion rather than by counting heads. Victor Yus (talk) 08:28, 1 May 2012 (UTC)

My math must be getting rusty now. Is it the case that (as of this writing) there are 264 users (give or take a few per the thread below) who support position 2, and there are 125 that support position 1? Then my calculator says: 264/125 = 2.11. In other words for every user supporting position 1, there are at least 2 who support position 2. Is that right? The 67% number probably comes from 264 / (125 + 264). However, when one thinks about it in terms of a "1 to 1" comparison: "for every user who supports position 1, there are at least 2 who support position 2". To me, it looks like position 2 is the winner. History2007 (talk) 14:20, 5 May 2012 (UTC)

Well I took it as percentage on a whole. There are 126 opposes, 269 supports and 9 doubtfuls. That makes 269/(269+126+9) = 66.58%. In any case, we can approximate closely to the number ⅔ for the supports. So yeah, in a way, for every oppose there are two supports. ~*~AnkitBhatt~*~ 04:43, 6 May 2012 (UTC)

I am stunned that anyone could read the support section and conclude that there are only 9 "doubtfuls". From the very, very first support, people acknowledge that there have been and/or are problems with PC. Reading through the comments, it's abundantly clear that many are in the support section not because they believe the proposal is as good as it can be, but because they fear that not publicly supporting PC might kill it off permanently. —WFC— 08:09, 6 May 2012 (UTC)

That would be an interpretation of the comments and as they say "interpretation is often in the mind of the beholder". I am sure others would read that section and see strong support. In any case, the votes are in now, and the 269 people who placed their vote in option 2, probably thought about it before voting. History2007 (talk) 08:32, 6 May 2012 (UTC)

We don't count votes. Far from it, the co-ordinators' job is explicitly to interpret the consensus. —WFC— 08:34, 6 May 2012 (UTC)

Sorry, I had the impression that the number of votes matter. In any case any "We don't do X" statement needs a policy link. I have seen no policy that says: "do not count votes". Have you? History2007 (talk) 08:45, 6 May 2012 (UTC)

As a matter of fact, I have. —WFC— 08:52, 6 May 2012 (UTC)

Yes, I know about that one, the not democracy statement. But the number of editors who support a position does matter, and consensus is not a "decision of the minority", per WP:CON, of course. History2007 (talk) 09:14, 6 May 2012 (UTC)

The number of editors who support a decision matters, yes. But the point that both WP:NOTDEMOCRACY and WP:CON make is that the underlying arguments are the crucial factor, not arbitrary numerical thresholds. My belief is that the support for PC is too great to ignore, and therefore that one way or another PC will return. Fine.

The doubts are over the policy under which it will be used. Due to the design of this RfC and the level of opposition to PC, those who wanted PC back in some form had very little option but to support it as strongly as possible when choosing their options. That's not necessarily a problem – the entire point of this RfC was to determine the 2012 level of support for PC. But it would be entirely wrong for those factors to enable the rubber stamping of a policy which the community has relatively little input on, bearing in mind the inertia you face in trying to change anything on Wikipedia, even when there is broad consensus that something has gone badly wrong.

The other important factor is that the views of those opposed to PC may not have been reflected in the proposed policy. Why would someone who wants to stop PC put significant time in trying to make it seem a more reasonable proposition? You have to assume that these people are more likely to seriously engage in the process once they have been told that PC will happen with or without their input. That would be the time to determine policy. —WFC— 13:49, 6 May 2012 (UTC)

I am sorry, I see serious methodological shortcomings in attempting to read through 300 text fragments attached to votes and attempting to create a mental summary of the opinions expressed within the text fragments. There are serious (and I mean serious) shortcomings in that approach. This type of problem was initially studied in the pioneering work of Kahneman & Tversky, and is now well understood. If someone is reading through 7-9 vote opinions they may have a chance of correctly guessing the intent of the participants, but for 200-300 cases, the pitfalls are just "tremendous". In fact this would make an interesting class project/experiment:

Select several groups of subjects who attempt to interpret and analyze 200-300 responses in two sets of text fragments, say A and B:

• Let one group read the positions in set A then B, see the variations among the groups. Reverse the order, have other groups read B before reading A.
• Reshuffle the order of responses in each set, and have people read them in that order, observe the variations among the groups
• Use stratified partitions among the answers and observe the base norms.

You will see that depending on the order the items are read and how the responses group together, the impression they will leave will be quite different. Hence an attempt to read through that many text fragments and build a mental summary is inherently rife with errors. History2007 (talk) 14:11, 6 May 2012 (UTC)

And it's because of those pitfalls that four of our finest have volunteered to do the job. They didn't volunteer because it was easy. They volunteered becuase it was hard. —WFC— 14:22, 6 May 2012 (UTC)

I did not even know if anyone was working on it yet - I had not read the discussion carefully, just happened on it. So anyway if there are several people discussing is, that is fine. Did they deliberately select an even number of people rather than five? Anyway, regardless of text interpretations, to my mind "two to one" is a pretty comfortable margin. History2007 (talk) 14:33, 6 May 2012 (UTC)

As noted at the top of the page, there are 4 coordinators that will be closing this RfC. Myself, Fluffernutter, The Blade of the Northern Lights, and TheHelpfulOne. We will all be pouring over it extensively and be hammering out the final proposed decision before closure, probally down to the wording. -- DQ (ʞlɐʇ) 16:27, 6 May 2012 (UTC)

Yes, it is at the top of the page. I did not see it. I will call my optometrist and have a serious talk with him... History2007 (talk) 21:18, 6 May 2012 (UTC)

Question for closing sysop

• http://en.wikipedia.org/w/index.php?title=Wikipedia:Pending_changes/Request_for_Comment_2012/Option_1&diff=489987521&oldid=489930856
• http://en.wikipedia.org/w/index.php?title=Wikipedia:Pending_changes/Request_for_Comment_2012/Option_2&diff=490002610&oldid=489989161
• http://en.wikipedia.org/w/index.php?title=Wikipedia:Pending_changes/Request_for_Comment_2012/Option_3&diff=490003608&oldid=488236200

How will Tylas's three !votes be counted? She seems to reject all three position. --Michaeldsuarez (talk) 11:54, 1 May 2012 (UTC)

LOL. ~*~AnkitBhatt~*~ 12:04, 1 May 2012 (UTC)

They all are problems, but I removed my rejection of #3.~ty (talk) 12:09, 1 May 2012 (UTC)

• http://en.wikipedia.org/w/index.php?title=Wikipedia:Pending_changes/Request_for_Comment_2012/Option_1&diff=490329043&oldid=490255888
• http://en.wikipedia.org/w/index.php?title=Wikipedia:Pending_changes/Request_for_Comment_2012/Option_2&diff=490329857&oldid=490225798
• http://en.wikipedia.org/w/index.php?title=Wikipedia:Pending_changes/Request_for_Comment_2012/Option_3&diff=490112108&oldid=490003608

Tylas now endorses Position #1 and (if I understand this correctly) still rejects Position #2. I guess that the sysops would have to subtract one from the final number in Position #2's ordered list when counting the Position #2 !votes. --Michaeldsuarez (talk) 12:16, 3 May 2012 (UTC)

But those small variations will not make a big difference. As in the thread above, (as of this writing) for "every user who supports position 1, there are at least 2 users who support position 2". Pretty straightforward, in fact. History2007 (talk) 14:24, 5 May 2012 (UTC)

In addition to Tylas's non-endorsement !vote, one user changes his or her mind and there's a !vote from an anon who didn't leave a comment. When calculating the total number of Position #2 supporters, I generally subject three from the last number in the ordered list. (125+9)/[125+(266-3)+9] = 134/397 = 0.3375 = 33.75%, which places the opposition to the proposed policy at over a third. (266-3)/(125+9) = 263/134 = 1.9627, which is actually less than 2:1. You neglected to factor Position #3 into your equation. Position #1 + Position #3 = the opposition to the proposed policy (Position #2). The nine !votes in Position #3 are going to mean the difference between "consensus" and "no consensus", especially if 2:1 is going to be the magic ratio. Every single !vote counts at this point. --Michaeldsuarez (talk) 16:38, 5 May 2012 (UTC)

I am not sure where the policy statements for either calculation may be, and I am not sure if there is a "magic number" that varies between 1.99% and 2.001% etc. Is there a policy page on which your statements were based? History2007 (talk) 16:43, 5 May 2012 (UTC)

No. It's up to the closing sysops to determine how to calculate the numbers and how to close the discussions. I simply disagree with your claim that "those small variations will not make a big difference," and I've provided an explanation for why I disagree. --Michaeldsuarez (talk) 16:50, 5 May 2012 (UTC)

Ok, so your suggestion of the 3-way counting method vs the one-on-one was your own idea, and not policy. Right? And the statement that "2:1 is going to be the magic ratio" is also your own opinion. Right? We should probably not debate it here between the two of us, however, lest it runs into a new RFC on how to add up the votes. History2007 (talk) 16:54, 5 May 2012 (UTC)

The three-way counting method is an idea that I have, but I'm not sure about whether it's my idea alone. Perhaps I'm simply the first to openly share that idea. I also used the word "if", which places my statement in the subjunctive mode. "2:1" / "two thirds" usually doesn't result in consensus. Some users in #Closure_of_this_RfC have suggested using "2:1" / "two thirds"; it isn't my idea, and I never expressed an endorsement of that idea; I simply acknowledged the possibility of the closing sysops using it. --Michaeldsuarez (talk) 17:08, 5 May 2012 (UTC)

Ok, we have both said what we had to say I think. So we should probably not debate it ourselves any more. Thanks. History2007 (talk) 17:11, 5 May 2012 (UTC)

All non-endorsement !votes will be moved to the appropriate sections if there is one when the closing administrators look at it and any duplicates will be nuked. If we are unable to determine where it shall go, personally I think we'd take it more as a comment, and not a !vote. -- DQ (ʞlɐʇ) 16:30, 6 May 2012 (UTC)

My four cents

My first two cents for the closing admins is that this RFC is proceeding similarly to other closely contested RFCs, and I think that's bad news. I believe you guys are in a position to significantly improve the outcome by stepping in now with your thoughts on where this is headed, and by asking the voters a few questions designed to get them to address each other's most salient points and look for compromise.

The longer a closely contested RFC runs, the more it feels like a contest, with "winners" and "losers". That's a lose-lose for Wikipedia; many of the "losers" will feel disconnected from Wikipedia and some will stop editing, and many of the "winners" will suspect that, since there are so many disaffected "losers", they'll have to waste their time staying vigilant over how the process is implemented. The real tragedy is that all this is generally avoidable if the RFC is run correctly, but I don't fault you guys because you're just following established Wikipedian procedure. I think, now that it's clear we've got a tight vote, it's time to let the voters know that neither a total win by one side nor a total loss would be a satisfactory outcome. For one thing, since pro-PC votes are around 65% at the moment, that means that the vote really isn't 65% ... it's 50% in some hypothetical situations (a clear lack of consensus for change) and 80% in others (a clear consensus for change). The odds go way down of chasing away a newbie if an attentive, helpful wikiproject is following up on edit requests; the odds go way up if a page has a history of divisive POV edits. If you share your thoughts with the voters now and insist that both sides drop the "take no prisoners" approach and work together to figure out where PC is most likely to be beneficial and where it's likely to be harmful, they might generate some additional useful ideas before the poll closes. For instance (here's two more cents!) ... I'm guessing that most of the opposition would be relatively content (and possibly the supporters could be persuaded) if the suggested policy is that PC is never used except in cases where you've at least got an argument that it is likely to increase rather than decrease newbie retention. At WP:Milhist, we handle this kind of problem all the time ... if someone is edit-warring over some Balkans issue, we will often try to redirect their attention to another page where they're likely to be more effective. PC could be used the same way, to redirect a frustrated newbie's attention from one article to another, perhaps by suggesting specific articles they might want to work on instead, or specific wikiprojects they might want to ask for help. - Dank (push to talk) 21:37, 10 May 2012 (UTC)

I beg to differ on that. I am not sure if users who voted are still watching this page and I am not even sure if this is running in a way that resembles other RFCs. I am not even sure if this is that closely contested... History2007 (talk)

At least some of us are still watching - and rather more are likely watching the Discussion section of the RFC itself. Allens (talk | contribs) 22:37, 10 May 2012 (UTC)

Yes, given that we both responded. However, unless we have specific numbers on who is watching the percentage of those watching may not be representative of the whole group, introducing a sampling error. So either everyone needs to be informed to watch, and continue discussion, or that should not be assumed a path to further input in order to keep the proportion of responses representative of the entire sample space. History2007 (talk) 22:42, 10 May 2012 (UTC)

I'll say this much. I'm one of the very few admins who's willing to brave AE, and I'm accustomed to making decisions that invariably make people upset. The main differences are 1 the length and 2 the lack of inflammatory, ad hominem remarks that are common to AE threads. I'm not at all concerned with angry reactions, as I knew going in what I was signing up for. Trust me, this will be nowhere near as complex as some of the AE threads I've handled; I already know here what I consider a strong and weak argument, and the numbers are clearly laid out. The Blade of the Northern Lights (話して下さい) 22:55, 10 May 2012 (UTC)

Yes, the numbers are clearly laid out and we should just sit back and let the process work I think. History2007 (talk) 23:01, 10 May 2012 (UTC)

I know you do a great job at AE. (I supported your RFA you know ... they grow up so fast, sniff!) But that didn't address any of my points. - Dank (push to talk) 23:06, 10 May 2012 (UTC)

Basically, what t'e 4 of us have to do is parse somewhat complex arguments and determine consensus, and ultimately get people with ideological differences to work together developing a way forward. A rough analogy, perhaps, but in some ways not so different as they may seem. My point is only that, whatever decision we make, we'll find a way to make things work. It'll certainly be hard, but if we do it right I think we can get those who would be on the fence about working with the other "side" to work. We're never going to get it to work with everyone, but I'm confident we can cause the least amount of angst that you describe. The Blade of the Northern Lights (話して下さい) 23:21, 10 May 2012 (UTC)

Excellent, all I'm asking is that the attempt is made. What I'm saying is that, if I were closing, looking over the current state of the arguments, I would be making a plea that voters spend more time on rationales that have some chance of dealing with the main concerns of both sides. If you want to wait til the 22nd to try to get people to work together, that might work too. I guess my gripe is that people seem not to be noticing, or noticing but not remedying, the negative effects of endless votes on this issue over many years that have wound up generating battle-hardened, "give no quarter" kinds of arguments, rather than attempts to bridge the gaps and take all concerns into account. I'm talking about the voters, not you guys, of course. - Dank (push to talk) 23:43, 10 May 2012 (UTC)

My advice, for whatever that's worth, to the participants on both sides is that a collaborative effort would do everyone a lot of good. I know I may not be the best example of this (c.f. the ACTRIAL fiasco), but I have tried to pitch in a little here and there to help out New Page Triage and Article Creation Workflow (both projects which were hastily thrown together so the WMF could reject ACTRIAL and pretend they had another plan in the works). That's the way forward, not taking shots at the other "side" (which I'll freely admit to having done on several occasions; I've been trying not to with varying degrees of success). But that's only my opinion, do what you will with it. The Blade of the Northern Lights (話して下さい) 02:10, 11 May 2012 (UTC)

Thanks for volunteering on this, I'm glad you're involved. It's not one of the more pleasant jobs. - Dank (push to talk) 02:33, 11 May 2012 (UTC)

I'll second Dank's thanks (!) and add a thought or two. First, I'd like to suggest that there are a couple of critical differences between ACTRIAL and this RfC. One is that the former was about trying something and seeing if it worked, while the latter concerns something that, if implemented, would in all likelihood be with us for the long haul even if it worked poorly. In other words, while ACTRIAL gave us a viable escape clause to return to the status quo if the cons proved to outweigh the pros, this RfC's most !popular option would change the status quo for an indefinite, perhaps permanent, period, with no undo mechanism built in. Now, various editors acting in good faith can argue from now till doomsday over whether PC as implemented in its trial was, on balance, good or bad. What can't really be argued is that, unlike autoconfirmed article creation, PC would fundamentally change the way Wikipedia works for all editors—established ones and newbies alike. I could be wrong, but I suspect those differences make this RfC a lot harder to close than ACTRIAL's RfC because the stakes are way, way higher.

I agree in principle with what Dank says about bridging gaps, taking concerns into account, and trying to be collaborative. In practice, it's a tall order. That may be partly because of the popularity of PC-related discussions and the wordiness of Wikipedians—one is confronted with a wall of text after a discussion has been open only a few days—but I think it may also be because editors choose to leave certain hard questions unanswered, certain misstatements unchallenged, probably for various reasons. For what it's worth, when something along the lines of PC was first proposed several years ago, I asked a fundamental question that hasn't been answered to this day, afaik. In this RfC, it has been unclear to me where one might respond to what another editor says if one were inclined to do so. While the unusual format followed in this RfC has undoubtedly made for cleaner-looking pages and enhanced the efficiency of the process, it has also discouraged meaningful discussion. I don't know how much common ground ardent PC proponents and opponents could find, but I'm fairly sure that this RfC has offered little opportunity for us to find out. Its three-option format (whose options I think of, perhaps exaggerating but only slightly, as "Hell No", "Full Steam Ahead", and "Punt") more likely than not has enhanced polarization over this issue. How could it not? Option 1 means PC is dead with resurrection unlikely in the short term, Option 2 means PC is undead and walking the earth even if one-third of Wikipedians fear and loathe it and it ends up killing the project, and Option 3 just kicks the decision down the road. Good luck distilling consensus from this unholy mess. (I'm not being ironic; I wish Blade et al the best.) Rivertorch (talk) 07:55, 11 May 2012 (UTC)

@Dank: Personally, you have my assurance that I will be reading every single comment and trying to incorporate them into our close. I too encourage people to leave the guns on the ground and actually sit at the table to work things out to the best of their abilities. As TBoTNL (Holy smokes, acronyms...) said I'm not afraid of the negative feedback that could come from this, I may not be as seasoned as TBoTNL but I do have my share of telling people no and them being not happy. (Not to say that no is the response to this RfC). I'm pretty sure we can leave discussion open for the first few days after the !voting close, but i'll check with other admins on that. But yes, I would prefer we get something productive out of this, instead of leaving the RfC with a result going nowhere. -- DQ (ʞlɐʇ) 12:13, 11 May 2012 (UTC)

My understanding is that the four closing admins are going to provide this table first.--Ymblanter (talk) 12:10, 11 May 2012 (UTC)

Table? (Sorry for not signing, as I am a closing admin. -- DQ (ʞlɐʇ) 12:13, 11 May 2012 (UTC)

I was replying to and actually sit at the table to work things out to the best of their abilities --Ymblanter (talk) 12:17, 11 May 2012 (UTC)

I mean, right now it looks more like a battleground, the tables have been destroyed by heavy guns.--Ymblanter (talk) 12:18, 11 May 2012 (UTC)

I agree with the point Ymblanter is trying to make. I see the role of the closing admins to firstly try and determine the community's wishes, and then to try to find a way to ensure that the various "sides" are able to provide constructive input as we move in that direction. —WFC— 13:13, 11 May 2012 (UTC)

What if we all self-vote to keep quiet for a while until it is over...? History2007 (talk) 14:17, 11 May 2012 (UTC)

Ymblanter, the quadriumvirate signed on only to supervise and close this RfC, and I'm afraid that in the sense I think you mean, you may have to provide your own kitchen furniture, because I don't feel it's our place to start setting out compromises for !voters to agree to. Not, mind you, because compromises aren't wonderful things that everyone who commented in this RfC should be working toward, but because for us to set out anything resembling "here are the terms you should agree to" or "here's a compromise you could all agree to" will start to smell a whole lot like us influencing the !vote/discussion. We are deliberately impartial to the options presented in the RfC, and my feeling is that to be able to present !voters with compromises, we need to take a preliminary position on a) how we think the RfC is trending and b) how we think it should end up. Those things would (again, imho) remove our impartiality.

Now, by saying this do I mean that we don't encourage everyone to re-read the RfC and consider and discuss whether there's room for compromise? No. But I do mean that any actual compromise suggestions are going to have to come from you guys, and not us. A fluffernutter is a sandwich! (talk) 14:29, 11 May 2012 (UTC)

Right, but I do not think starting looking for compromise suggestions NOW would make any good. I would rather wait for the conclusions which you guys draw from the discussion. I would normally expect smth like this: "The majority supports PC for reasons ABCD, the sizeable minority opposes for reasons EFG, we think it follows from the discussion that reasons ACF are the most important, please sit down and negociate the issues ACF before the green light can be given". Then we would go and negociate ACF and eventually get green light. Otherwise, it would be difficult if not impossible, with the repetition of the arguments like PC are in contradiction to five pillars and they are evil does not matter what, or that PC is a step forward and we should start using them does not matter what opponent say. We had already these discussions before and they failed miserably. Now, we need a more focused discussion, which will concentrate on really serious issues, which are perceived by the majority. And this can only be done after May 22, not now.--Ymblanter (talk) 14:45, 11 May 2012 (UTC)

If Fluffernutter's statement is reflective of the rest of the closing admins, that strongly suggests that this will be closed on a black-and-white basis: either option 1 in its entirety, option 2 in its entirety, or option 3 in its entirety. Unless there is indisputably consensus to do so, a closure of either "PC is now on with the same policy as before: sort any problems out yourselves" or "PC is now dead in the water" would be an utter disaster. I assume that the former is a more likely outcome than the latter; if the community has shown one thing during the PC debates, it is a complete inability to lead itself. An impartial body has to take the bull by the horns here – create a roadmap if you prefer. If this set of admins isn't willing to do it, I haven't got a clue who will.

To respond directly to your last comment Fluffernutter, this RfC follows a structure under which gaining consensus for a compromise is literally impossible. —WFC— 15:31, 11 May 2012 (UTC)

Ahem... Fluffer. The Blade of the Northern Lights (話して下さい) 15:37, 11 May 2012 (UTC)

Good point. Tweaked. —WFC— 15:41, 11 May 2012 (UTC)

(edit conflict with Beeblebrox) Sorry, I got pulled away from the computer just as I was finishing my initial comment. To add a bit, based on responses here: First off, I assure you all that we have every intention of thoroughly reading the discussion and addressing what commenters have to say, not just counting heads, when we close the RfC. "Black and white" "headcount vs headcount" closing isn't likely to do the job. Second, the sort of "compromise" (slightly inaccurate label, I guess, but I lack a better word) I think would be most useful at this point in the RfC would be for people to look at their !votes, look at other people's !votes, and give some thought to whether there's any gray area they're willing to give toward the other side, or if they could explain themselves more clearly. For instance, if I were a commenter and my initial vote went under option 1 saying "No PC, ever," I might re-read what's gone on since then and amend it to "Strongly oppose PC because [blah blah this is the issue I care about], which [blah blah PC is bad for that issue because blah]." If I had supported PC under option 2, I might read that and amend my comment from "I <3 PC" to "I <3 PC. In response to [blah concern someone raised in Option 1], I would suggest [blah tweak to PC], which would address [blah] by [blahing]." Mind you, I am not saying anyone in this discussion had good or bad comments - there have been tons of wonderful comments so far, and I'm speaking in simplistic terms here for the purposes of not wasting any more of my vast quantities of hot air. My point is basically that comments that contain more than "yes" or "no" and provide some information about why they feel the way they do, or under what circumstances the commenter would be willing to see their opposed option stumble through, give us closing admins more to work with than "black and white" yes/no votes.A fluffernutter is a sandwich! (talk) 16:45, 11 May 2012 (UTC)

I originally !voted under option 3, then switched to option 1 when I spotted another problem (IMO) with PC, and I concluded that it would be better for it not to be implemented at all rather than eventually implemented with changes/restrictions. However, I would also say that it would be better if it were implemented (now or eventually) with changes/restrictions than if it were implemented without changes/restrictions. Is my splitting my commentary between Option 3 and Option 1 going to be confusing, as in, would it be helpful if I moved my !vote back to Option 3? In other words, my ordering of preferences is 1>3>2, not 1>(3&2), not 3>(1&2), and I've found it hard to express this. Allens (talk | contribs) 21:18, 20 May 2012 (UTC)

The closing admins are right and I'm wrong; the rules for this particular RfC are already in place. Regardless of which way it turns out, there will be another related RfC before long, and after you've closed this one, I'll start talking about the next one with any of you who want to close the next one. I'm really more disappointed in myself than in anyone else for not predicting how long (7 years now?) and draining this issue would turn out to be for the community; I've mostly sat this out, and that was a bad call. - Dank (push to talk) 15:44, 11 May 2012 (UTC)

• I suppose I should go ahead and comment here. Don't blame the closing admins for the structure used here, the design is entirely mine. I've explained the thinking behind it at various previous threads here but I'll sum it up once again. The lack of nuance is quite deliberate. The idea is to avoid the pitfalls of previous large policy discussions that have failed to achieve an actionable result, such as the "de-adminship process" debate in which seventeen seperate proposals duked it out with, big surprise, no clear result, and of course the disaster that was the 2011 PC RFC. The one and only purpose of this RFC is to determine if we are going to use PC at all,. That was supposed to be the only purpose of the 2011 RFC as well, but because it was so open-ended and there were no controls in place we got no usable result from that either, just a tempoaray suspension of use of PC that continues to this day. I waited a full year for someone else to pick up the pieces and take another stab at it, and nobody did, so I designed this deliberately restrictive process to try and force some kind of real result. Option three is there to give those users who feel we still aren't ready to make that decision after all these years a voice. I'm actually rather surprised at the low level of support it has attracted. I have endeavored time and again to make it clear that the policy in option two is a draft. It is basically the policy used in the trial with minor modifications based on what little consensus can be draw from the last RFC. I always expected that it would be challenged or modified more or less imediately after the close of this process, it was never intended to be complete and perfect, so your follow-up RFC was actually already antipated when this was being put together. TLDR version: So, to summarize the summary, the idea here is to decide the one big issue so that we know if there is even a point to trying to create a more comprehensive policy for long-term use. Beeblebrox (talk) 16:37, 11 May 2012 (UTC)

  While Option 2 does mention a draft, it explicitly proposes using the tool in accordance with the terms of that draft and says nothing about a further RfC to amend it. The way I read it, a close in favor of Option 2 would mean that resumption of PC would be authorized immediately. I suspect the fear of that happening led a number of !voters to select Option 1 instead of Option 3. I toyed with the idea of endorsing Option 3 but disliked its first sentence, which unequivocally states that PC should be kept, and was concerned that endorsing it would really just be !voting for a further RfC with only two options instead of three and never any chance for meaningful discussion of the alleged pros and cons of PC. That's really the sticking point for me: serious questions remain unresolved, and unless that changes I'm afraid to say anything that might help even the tiniest bit to open the door to PC. I'd have been more willing to endorse an Option 3 that took no position on the continued existence of PC but instead said something like this: "The benefits and detriments of pending changes are disputed, and much remains unknown about its potential long-term effects on the encyclopedia and the community. Focused discussions or workshops should be set up in a concerted effort to determine whether the premises that PC's supporters and opponents base their opinions on are sound and whether the various recurrent questions raised by both sides can be adequately answered." Protracted? For sure. Messy? No doubt. But sometimes the only road to making an informed decision is a long, hard one. Rivertorch (talk) 00:12, 12 May 2012 (UTC)

I knew going into this that any structure used would upset some people, so I didn't try to please everybody. Discussion of this idea started five-seven years ago, depending on who you ask. It has already been a long hard road. If users don't feel informed enough to make a decision by now that is only because they have not taken the time and effort to be informed. We can't know what the effects of PC really are if we aren't using it. If it is turned back on and is being used the chances of developing a stronger policy that actually works are much higher. That is why the draft policy is based on the one we already used. If instead we alter it based on hypothetical discussion, we will still end up being mistaken about some aspects and have to change it anyway. Despite what some seem to believe, it is possible to make incremental changes to the protection policy in order to make it more in line with community norms. When WP:PP was first created it looked like this. By the time of the "flagged revisions" debates of 2007 it looked like this. During the trial it looked like this. We've changed it before, and we can change it again if we need to. This discussion is the hard one, making minor tweaks to the policy as flaws are discovered by using it will be much simpler and will not require the strict controls being used in this RFC. Beeblebrox (talk) 17:28, 13 May 2012 (UTC)

My intention is not to criticise you for putting your head above the parapit and ensuring that we got on with this. You deserve nothing but praise for that, and I applaud you for your strong attitude in the face of criticism about the structure (including from me). I agree with what you say about minor changes to policy being relatively easy, but I'm not sure we're talking about minor changes. We need wholesale discussion on the very nature of the policy, to ensure that it's consistent with what the community wants PC to do. Tweaking the wording to ensure that we get better at following this framework comes later.

The one thing that is certain is that PC is a big issue. Just like semi protection, and just like the SOPA blackout decision. For a few, it will be the straw that breaks the camel's back as far as Wikipedia in general is concerned, as the previous things were. There is nothing we can (or in my opinion should) do about those whose opposition to it is so strong that they either threaten to or actually leave the project should PC be restored in any form – they should not be allowed to hold us to ransom. But what we can do is ensure that those who are weakly supportive, neutral, skeptical or even deeply unhappy about PC can be confident that we have tried to make it as workable as possible. By implementing it immediately with the draft proposal, we will not have done that. —WFC— 17:58, 13 May 2012 (UTC) EDIT: What this process has done is ensure that we are very confident what the community's opinion on PC generally is. What it has not done is gauge whether the draft policy is how the community wants it to be used. —WFC— 18:01, 13 May 2012 (UTC)

That makes a lot of sense. I'd add that there's a flip side to the coin: those who insist that PC must be implemented shouldn't hold the project hostage either. The encyclopedia has been built, refined, and maintained to date largely without PC, and I don't see any reason why that wouldn't continue if PC were either rejected outright or so narrowly limited in its deployment as to be invisible to most editors the vast majority of the time. @Beeblebrox: I, for one, took a considerable amount of time and effort to inform myself, and I still had nagging questions that I found significant enough to give me pause before I decided how to !vote in this RfC. As for incremental change, it occurs to me that likening PC to PP is something of an apples-and-oranges comparison. PP doesn't significantly affect the user experience; it simply changes "Edit this page" to "view source" for the mopless masses. No biggie. In contrast, PC fundamentally changes the way everyone—from the newest newbie to the rarefied few—interacts with every page it's applied to. While incremental change could occur in terms of deciding when and where it's applied, PC would constitute a fundamentally new and different user experience everywhere it is applied. If its use were to prove deleterious to the project, I wonder very much whether incremental change would be sufficient to undo the damage. Rivertorch (talk) 19:43, 14 May 2012 (UTC)

Just a heads up that I'm encouraging people at WT:MHC#A call to arms to comment on the process, on their honor that they either don't have a position on PC or think the process is more important than which way the call goes. Milhist people are good people, and I think they'll probably be helpful if they can spare the time, but IMO it's completely up to you guys whether to use anything that we come up with for your deliberations. - Dank (push to talk) 19:38, 20 May 2012 (UTC)

I saw that you issued a "call to arms" there. To be fair, you also need to issue a "call to maps" on WikiProject Geography as well, or a "call to charts" on WikiProject Statistics... kidding. But seriously, why just one project? Try physics, economics, etc. as well. They are all good people... But in any case, I think the closing admins should almost entirely focus on what the RFC page itself was about and mostly ignore all the extra lobby-talk here. History2007 (talk) 20:03, 20 May 2012 (UTC)

If you want to invite more people to discuss the RFC, feel free. See WP:BURO and WP:NOTCENSORED: saying something at the wrong time, in the wrong forum or in the wrong way doesn't invalidate what's said, although of course WPians of good faith will generally try to be transparent and well-behaved. I'm not voting myself, and not encouraging others to vote ... for one thing, it's a bit late, but also, it seems to me that good faith has mostly broken down in this RFC; neither side is attempting to rebut, or even engage on, the main points of the other side; instead, each side repeats its own points loudly, fingers firmly planted in their ears. That's going to make the discussion very difficult to close. I've noted over at Milhist that it is of course up to the closers whether they want to pay attention to anything we say, but IMO, anyone who has a difficult close to make is better off engaging the community and getting support for their decisions; even if they don't need the help making up their minds, they'll certainly need help selling their opinions. - Dank (push to talk) 20:41, 20 May 2012 (UTC)

Clearly PC is going to be reintroduced regardless of what anyone says here; anything else would be politically unacceptable. My own position hasn't changed from the initial implementation. I will not accept a "right" that can be removed on the whim of any administrator suffering a bad-hair day, and I will remove from my watchlist any article to which PC is applied. Malleus Fatuorum 20:53, 20 May 2012 (UTC)

Of course you can do that but your worries are undue imo - there was only one user that had his permission removed in a dispute and that was User:Wnt a user that opposes and rejects policy. We are all open to administrative action and if we violate or get on the edge then permission may be removed by an Admin but a discussion will be opened for the community to work it through and only serious violators are likely to have the permission removed - also, you say, "Clearly PC is going to be reintroduced regardless of what anyone says here" - there is a 2 to 1 vocal support for the tool so its not being forced through its being supported by a majority of editors. - the stated position for removal of the user right seems reasonable to me and one that you would never be in violation of or have admin action taken against you for - Reviewer rights are granted by administrators and in cases of misuse of the right or to protect Wikipedia from possible misuse, can be removed by an administrator. The permission can also be removed at the request of the user, the community, or the arbitration committee.Youreallycan 21:06, 20 May 2012 (UTC)

No, only we non-administrators are subject to such sanctions, unless you can persuade me that the PC "right" can be removed from an administrator without all of the hoo-hah of a desyop. ~~

If an Administrator is having his Pending changes rights removed then he is also more or less automatically deysopped - Youreallycan 21:18, 20 May 2012 (UTC)

While in my opinion this issue alone should have little bearing on PC's implementation, Malleus does make a valid point. I once tweaked the rollback policy, to clarify that admins found to have misused the tool could be subject to the same sanctions as non-admins. An extremely respected admin (who at one time or another has had most of the high-level permissions and positions) first attempted to revert entirely, and when s/he realised this would not succeed, tried to neuter the point by surrounding it with hyperbolic language.

So while it is a small factor, it is yet another contributary reason for concluding "Yes, there is consensus for PC, but consensus should be reached on how it should be used and governed by DD-MM-YYYY. If no consensus is reached by that time, PC will be turned on using the draft policy." The first sentence would give the community a sense of direction. The second sentence would show that the RfC was considered decisive enough to turn PC, and that the delay is solely because it is believed that there is likely to be a net benefit in discussing the policy before PC goes live. —WFC— 21:58, 20 May 2012 (UTC)

There appears to be plenty of opining here and in the support/oppose comments that the closing admins will be able assert a position/an interpretation of the consensus position, lets wait for them to close and see where we go from there - Youreallycan 22:02, 20 May 2012 (UTC)

I prefer to think that, instead of clueless, democratic vote counting, they will read points that have been made, read any refutations (where applicable), and determine the balance of the strength of arguments. So instead of what you're suggesting, lets use the time between now and the close to discuss any issues we think are relevant. —WFC— 22:12, 20 May 2012 (UTC)

Rebuttal and engagement from an Option 1 supporter

In response to Dank's comment that "good faith has mostly broken down in this RFC; neither side is attempting to rebut, or even engage on, the main points of the other side; instead, each side repeats its own points loudly, fingers firmly planted in their ears": I don't think the problem is lack of good faith but rather the well-intentioned but flawed format of this RfC, which doesn't lend itself to rebuttal or engagement. However, I'd like to take a stab at rebutting and engaging. I'm not going to singlehandedly respond to every major point made by Option 2 supporters, but I'd hate to think of the RfC closing without at least some of them being addressed. It's unclear to me where I should put this; the project page suggests limiting discussion there because of page length, so apparently it belongs here—or does it? Dynamic trio, please move or collapse as appropriate. Excerpts I've quoted from Option 2 supporters' comments are in italics.

• we need a community commitment to use PC in order to signal the developers that their efforts to improve PC will not be wasted

With more than one-third of the community opposed, I'm not sure that's a realistic expectation. In any event, we're not here to validate the developers' efforts, are we? The reverse doesn't reliably happen, anyway.

• an imperfect PC is better than a perfect editing war

Edit wars can be stopped (and even prevented) without PC. It happens every day.

• As I expect problem edits to get worse, all options should stay open

You provide no reason for expecting problem edits to get worse. Vandalism seems more under control than ever. (There's certainly less of it on my watchlist now than there was, say, five or even three years ago.) I have no idea whether other problem edits, such as POV-pushing ones, are getting worse or not; they remain a long-term challenge, and PC would be an awkward tool, at best, to deal with them.

• It will help IPs Be bold and fix our mistakes, even on controversial pages.

On ths surface, this is an excellent point. PC would allow IPs to make corrections without having to jump through the "edit request" hoops necessitated by protected pages. To my thinking, this might be the most compelling argument in favor of PC. However, I don't think it outweighs PC's many negatives. In any case, while edit requests are cumbersome, they do work—and maybe better than PC. Here's why: either way, the newby proposes text to insert, remove or change. Much, maybe most, of the time, it needs some alteration before that can be done. An established editor responding to an edit requests can engage with the newby (and other editors) right there on the talk page to iron out the kinks before making the change. With PC, it's either approve or reject. I, for one, would never approve a good-faith edit whose wording is unclear or that I suspect contains factual errors, nor would I approve any edit that inserted a grammatical error in an article. WIth edit requests, it's " 'not done for now' and here's why". With PC, the procedure is actually more cumbersome.

• offers a solution to our problems of vandalism without totally shutting out new and unregistered editors

PC wouldn't solve vandalism, simply shift the burden away from ClueBot and vandal patrollers onto a relatively small subset of editors for those articles on which it's enabled. New and unregistered editors aren't "shut out" now—they're merely restricted in what they can do (and rightly so). In fact, with PC they'd be restricted in one more way than they were before.

• the process of removing PC from anyone who abuses it should be efficient and not held back by sentiment.

Does this refer to those who'd abuse reviewer status (i.e., by approving bad edits or rejecting good edits) or actually abuse PC itself (i.e., admins applying it in situations where they're involved)?

• it could use some tweaking, but implement it, then work out the remaining bugs.

Utterly bass-ackwards. Sure, every new tool is going to have some bugs that need to be dealt with after implementation, but a tool with the potential to profoundly change the way all Wikipedians interact with the project should not be implemented with known, major bugs. Yet that is precisely what's being considered in this RfC.

• The minor problems encountered in the trial did not demonstrate the unworkability of the system; instead they demonstrated that it basically did work.

The threshold between minor and major problems is subjective. While it's interesting in the abstract how we can consider the same trial and come to such different conclusions, in practice I find it a little scary. The thing is, I don't hear anyone saying they just can't deal with Wikipedia without PC, but I sincerely wonder whether I could deal with Wikipedia if PC were implemented on any significant scale. I'd certainly consider unwatchlisting any articles to which it was applied because I suspect it would make my user experience difficult and unpleasant. Then again, I'm here long enough to have some clue what I'm doing (most of the time) and am theoretically willing to adapt if I can. How many newer editors who found PC intensely odd-putting would give it their best shot versus just giving up and abandoning us? I really don't know the answer to that, and I suspect no one else does either.

• PC is much more open to editing than the alternatives like semiprotection.

Not really. It can be argued either way, but I think there's little doubt that if PC were implemented it would be applied to considerably more articles than semiprotection currently is. That would actually mean less open editing.

• it would stop IP vandals from rushing rapidly through high profile pages and vandalizing them

The scenario you describe is a minor annoyance that is dealt with quite effectively every day. Currently (i.e., without PC), it results in quick reverting and blocked IPs. It's unclear to me why some consider such episodes dire enough to warrant something like PC, but it's clear enough that PC would prevent them only if it were applied on a massive scale.

• Last but not least, a number of Option 2 endorsers mentioned the issue of PC's supposed particular applicability to BLPs.

One of my greatest concerns about PC is that it would be employed disproportionately on articles with active or long-term disputes and unclear consensus regarding BLP issues. While there is little doubt that PC would thwart most "drive-by" BLP violations (much as it would thwart drive-by vandalism), it would also enable certain editors to prevent negative content from being added to biographical articles even when that content is impeccably sourced, highly relevant, and poses no WP:UNDUE problems. I do not doubt this would happen, nor do I doubt that the result would be increased ANI dramah specifically, as well as a general decline in congeniality among editors who watch BLP articles. Of course, the real victim would be the quality of the articles themselves.

Fwiw. Rivertorch (talk) 08:48, 21 May 2012 (UTC)

Thanks much for making the effort. - Dank (push to talk) 14:22, 21 May 2012 (UTC)

Rivertorch, I think you're right that description of lack of "good faith" wasn't the right way to put it, and I've struck. I meant that most people in this RFC seem not to be responding to points on the other side that are at least reasonable enough to merit a response, even if the voter believes they're outweighed by other concerns. - Dank (push to talk) 16:15, 21 May 2012 (UTC)

The above covers many of my concerns. A couple places for policy changes/clarifications to help regarding those:

• "Not really. It can be argued either way, but I think there's little doubt that if PC were implemented it would be applied to considerably more articles than semiprotection currently is. That would actually mean less open editing."

If it were made clear that PC1 is only to be used if semiprotection would otherwise be used, that might help.

• "Last but not least, a number of Option 2 endorsers mentioned the issue of PC's supposed particular applicability to BLPs. One of my greatest concerns about PC is that it would be employed disproportionately on articles with active or long-term disputes and unclear consensus regarding BLP issues. While there is little doubt that PC would thwart most "drive-by" BLP violations (much as it would thwart drive-by vandalism), it would also enable certain editors to prevent negative content from being added to biographical articles even when that content is impeccably sourced, highly relevant, and poses no WP:UNDUE problems. I do not doubt this would happen, nor do I doubt that the result would be increased ANI dramah specifically, as well as a general decline in congeniality among editors who watch BLP articles. Of course, the real victim would be the quality of the articles themselves."

If it were specified that BLP articles would not have PC1 applied to them, this would take care of that, along with liability concerns for reviewers.

I have to go now - I will comment further later (here or on the discussion page, wherever the admins prefer). Allens (talk | contribs) 15:12, 21 May 2012 (UTC)

A couple of further concerns, with proposed solutions.

• It is possible for a knowledgeable vandal - the most dangerous variety - to first commit vandalism on a PC1 page (especially a subtle form such as incorrect information) using an autoconfirmed account, then edit the page using an IP address, possibly several times. This will block any non-reviewer from repairing the damage.

Putting PC1 into effect on a page should require that someone with the reviewer/sysop flag be involved with that page: able and willing to spot damage quickly. This requirement should also help stop a page from accumulating a long queue for a reviewer to go over.

• PC2 queues for reviewers to go through may well grow significantly before a reviewer gets around to them.

The same policy for PC2 as for PC1 above would take care of this. I also suggest that some listing of pages by how long a change has been waiting in the queue would be helpful - even vital - for both PC1 and PC2.

Seeing nothing from the admins saying otherwise, I'm commenting here. Allens (talk | contribs) 23:48, 21 May 2012 (UTC)

Rebuttal and engagement from a Position 3 supporter

These are responses to some of the comments made in position 2. —WFC— 16:16, 21 May 2012 (UTC)

• We shouldn't let the various problems with the trial prejudice us against the tool itself.

No. But if we acknowledge that there were problems, we should seek to deal with them. While there would appear to be consensus that the positives outweigh the negatives, we should allow those problems to be reintroduced only if after trying to solve them, we conclude that it is not possible to do so.

• we need a community commitment to use PC in order to signal the developers that their efforts to improve PC will not be wasted.

We can make a firm committment to using it, without necessarily turning it on immediately. If necessary, that firm committment can be reinforced with a hard "turn-on" date.

• an imperfect PC is better than a perfect editing war.

PC has many good qualities, but only a form of protection which prevents all sides from editing can prevent an edit war.

• As a community, we should welcome with open arms anything that offers a solution to our problems of vandalism without totally shutting out new and unregistered editors.

This is a compelling argument for ensuring that we make PC work. Introducing PC in a way which irks the community could lead to its downfall later on, which is something I do not want to see.

• I think the answer [...] is [to] make sure, as best we can, that both PC and SP are used where they are most appropriate, and for admins to be flexible in upgrading from PC in cases where editors are experiencing genuine difficulty in administering it.

I agree that there needs to be a degree of admin discretion. But if we don't have some sort of guideline as to the situations in which one or the other might be more appropriate, we will end up with a free for all. That could result in an inefficient form of protection being used, or potentially in wheel wars.

• I do have some concerns, but not enough to put me among the supporters of #3.

Surely it would have been helpful to elabourate on these concerns?

• I am experienced with FlaggedRevs on another project and the most common cause of alienation of new good f aith editors is the perceived lack of trust that a "pending changes" system creates.

Yet another reason to ensure that we are happy with the workability and robustness of PC policy, before going full steam ahead.

• there are too many problems with vandalism on Wiki, and the existing options are poor, especially for the less active pages. Obviously to be used in moderation.

Is it obvious? The original intention was to use it very liberally. If we "obviously" intend to use it conservatively, would it hurt to somehow indicate this in policy?

• A great help [...] on controversial pages.

To an extent, but it depends on the traffic. On a high traffic contentious article (for instance a current Israel/Palestine event), PC would be a terrible solution. Policy gives no guidance whatsoever on this sort of situation, and does not even concede that PC might be a bad idea for very high traffic pages. It only takes one admin to disagree with or not understand this, and stand by his or her decision, for misuse of PC to cause high-profile problems. The solution is to try to get the policy right before that situation arises.

• I'll say this, though - the success or failure of PC will depend in large part on the quality of the reviewers, and that selection process was far too "loose" last time - perhaps partly because it was a trial.
• as long bar for becoming a reviewer is low and Reviewers should themselves be reviewed before their edits can be made visible.

Idealogically I tend to agree with the second point, as I feel there is no problem with us being easy-come, easy-go on userrights in general. But regardless, these are just a sample of posts from the RfC which suggest there is a real split over this issue. See also Malleus' posts above. By far the least divisive and problematic time to determine whether the current reviewer policy is right or needs to be changed is now, before PC is switched on.

• given time to bed in I think it will become so widely adopted we'll wonder how we coped without it.

We need to do everything in our power to ensure that it is able to bed in. The technical side of that is down to WMF and will not happen until we deploy, but the policy side is entirely within our control, and should ideally be sorted out first.

• Support with the caveat that the policy must state that semi-protection is preferable to PC on very heavily edited articles such as current event articles.

Just wanted to highlight this. It is an example of a user who strongly supports PC (as do I) and therefore felt obliged to support option 2, but who in fact wants policy changes made before reintroduction.

• Support with two caveats: First, and obviously, this should be used sparingly. Second, policy should contain a provision to address a backlog overflow.

Another example.

• yes, yes, yes yes yes yes. Such a simple solution.
• Support ~~~~ (numerous examples)

Just thought I'd highlight the quality of some of these rationales.

• Support if only to counter the !voter who said that openness is more important than quality on Wikipedia.

...and that there is plenty of low brow, tit-for-tat "!voting" (or "voting") going on.

• It worked well enough in the trial - we can fix problems as we go.

Cobblers. We know what the problems were, and we know which of those problems were on the policy side. Yet 18 months on we haven't even attempted to fix them.

• The tool was efficient, and still would be, if used appropriately.

How do we ensure that it is used appropriately? By defining "appropriate", which in turn comes through having a more comprehensive policy.

• As a very limited anti-vandalism tool only.
• after seeing its use with little-watched BLP articles, I see the necessity
• Crucially, it should affect only a small number of articles
• Wikipedia should be the encyclopedia that anyone responsible can edit. We do waste too much time reverting vandalism, and the policy can always be adjusted as we learn its limitations from experience. And just in case: if the tool is to be limited only to certain type of articles, BLP and country articles should definitively be a priority, and may I suggest that FA and GA rated articles should also be protected to preserve the effort of the editors who went the extra mile to produce such quality articles.

There are literally dozens of comments along these lines. These comments do not represent all of the different visions, but I've selected them to demonstrate the marked difference in how people have interpreted the same draft policy. A small number of articles? All of our little watched BLPs? Added to articles as and when they have been vandalised? All BLPs, countries, GAs and FAs? The question I would ask is whether people who have supported along these varying lines have read the policy and understand what it would mean? Or if they are supporting the general idea of having PC, in the belief that specifics will be resolved in a future discussion?

• This was useful during the test with no real problems noted

Seriously? Most other supporters are openly acknowledging that there were flaws in the trial (as well as ignoring the fact that we haven't dealt with them).

• it works well when used properly, and if it doesn't work on a particular page it can be removed or replaced

Who determined whether or not it is used properly on a particular page?

...I could go on all day. I've stopped here for now because I've got to go, and for the sanity of those who read this. —WFC— 16:16, 21 May 2012 (UTC)

• The first comment you respond to is mine. You seem to have misinterpreted my intention. The problems I refer to are the problems with the trial itself, hence my point that we shouldn't take it out on PC that the trial was badly mismanaged. Beeblebrox (talk) 17:02, 21 May 2012 (UTC)

Save for the fact that it is intended as a permanent addition, the proposed rollout of PC is being managed under an almost identical policy to the trial. —WFC— 17:19, 21 May 2012 (UTC)

I might have done things differently myself, but "badly mismanaged" is way too harsh. Even though there weren't a lot of disrespectful comments made in this RFC, it's clear that many on both sides felt disrespected ... as if, while they were trying to point out the big, blazing fire, the other side continued merrily chatting about sports and the weather. When people feel that their "people" are being disrespected, they stop attempting to understand and offer compromises, and they start thinking that brute force is going to be the only way to win. Worse: since they fear that the eventual outcome might be some kind of forced negotiated settlement "halfway" between the two positions, they start seeing any attempt to compromise as shameful behavior that might move the eventual settlement in the wrong direction. Then it becomes a point of honor to avoid any attempt to reason or compromise, and you can see this instinct in action in every intractable conflict in the world. Generally, once things get to this point, there's only one way to get them to put down the guns and start talking: both sides have to lose hope that brute force is going to work. So, even if all that happens here is that both sides "lose" ... the Option 1 guys fail to stop a new RFC, and the Option 2 guys fail to get immediate PC as the draft suggests ... it's just possible that that failure is a necessary precursor to a successful RFC (or other compromise-seeking process). - Dank (push to talk) 17:26, 21 May 2012 (UTC)

I think that's a very reasonable position Dank; let's hope that others listen to your words of wisdom and act on them Malleus Fatuorum 20:32, 21 May 2012 (UTC)

Thanks Malleus. I liked the reactions to the Option 2 and Option 3 rationales above, and don't have much to add to those. I'll give a response to the Option 1 rationales a shot tonight, unless anyone would like me to wait another day (or forever!) - Dank (push to talk) 21:06, 21 May 2012 (UTC)

Critique of Option 1 rationales

There are critiques of the Option 2 and Option 3 rationales just above. Before I start with my critique of the Option 1 rationales, I want to express appreciation for the Option 1 voters, because I usually work with "high-end" history articles and with with very nice people who are committed to Wikipedia and very committed to their articles. You guys are taking on a much less pleasant task, attempting to make sure that the things that go wrong on Wikipedia don't go so insanely out of control that they kill off the project. Better you than me :) I'm attempting to cover all the main Option 1 points (given in quotes), but I don't necessarily quote the fullest or best presentation of each point, just an excerpt from the first voter who made the point.

• "everyone should be able to edit equally (while semi-protection for example blocks IPs, those users can easily get the status that allows them to edit regardless - PC on the other hand would restrict editing in those cases to a small group of users).": You're assuming that PC supporters are generally hostile to the idea of letting everyone edit a PC page immediately, even on brand-new accounts ... the current draft proposal requires autoconfirmed (4 days and 10 edits), but I don't see anything in Option 2 that makes me believe that the supporters couldn't agree to this (and this problem of treating new editors differently is probably the main sticking point between the two sides, as I read it). Then, it's just a matter of getting potential editors who don't have an account (or don't want to use their current account) to take a few seconds to give a username and password ... this can be done in a very quick and non-threatening way ... just say "You can see and edit the current working version of this page immediately just by choosing a username and password." ... don't say "register", people expect that to be a chore. Then there won't be any confusion or "second-class citizen" status ... every logged-in user sees the same page and believes they're working on the "real" page. A system is hardly discriminatory if it allows everyone to immediately edit 99% of our pages without logging in, and only requires that they take a few seconds to log in or pick a username to edit the other 1%. Furthermore, PC will probably reduce, rather than increase, the thing that's most likely to make newbies actually feel like second-class citizens ... quick reversions .... because the page doesn't need to be "fixed" right away, only at some time before it gets "published" to non-logged-in users.

I wasn't active at the time of the first PC trial. Can someone who's not logged in not see the "working" version at all, even if they actually want to? I'd call that problematic, as in not allowing full anonymnity of reading (an important version of) Wikipedia. The above also may be targeted at the worry that PC2 will be overapplied; if PC is to go into effect, policies do need to be put in place to prevent this. Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

First, on the question of "IP rights", it doesn't oppress anyone to allow them to edit without choosing a username on 99% of our articles, and ask them to take a few seconds to choose an anonymous username for the other 1% (or enter one they already have). If they object, the explanation is quite simple: edit wars develop on some of our pages, and when that happens, we institute some form of protection, in the hope that people will start talking with each other and stop edit warring. You can't talk to an IP address; it's impersonal, and for all you know, the person at that address now is different than the one who made the edit. If the version of PC sketched above is the form of protection chosen, that means that logged-in users see a different page than non-logged in users; that's another reason you need to pick a username. This in no way stops a person from editing as an IP, it just identifies a few pages where edit warring is preventing IP editing from getting the job done.

Second, everyone involved in publishing outside of wikis thinks of the working version they're putting together for the next edition as the "real" version; that's the version they read, edit and argue about. Their readership only has access to the published version and thinks of that as the "real" version. I'm not seeing the argument that this is such a novel concept that it will confuse everyone, or make people feel like they've been shut out. Everyone who logs in will see and work on what they think of as the "real" version of the Wikipedia page, and everyone who has never edited will be perfectly happy to think of the (older) version they're looking at as the "real" page. - Dank (push to talk) 18:29, 23 May 2012 (UTC)

With regard to the first, I just realized that it's probable than an IP would be able to see the latest version via the Page history tab, unless they're blocked from requesting certain revids (and I can't see why).

With regard to the second, I suspect that some are attracted to Wikipedia precisely because of differences from (other types of) publishing. Allens (talk | contribs) 19:16, 23 May 2012 (UTC)

Yes, I think that's probably true. When I first began editing (as an IP), I found it uniquely satisfying to be able to immediately see the improvements I had made. I don't know of any other serious reference site on the Web where that was possible. Waiting for my edits to be approved would have made for a very different user experience—more like posting to a discussion board where posts have to be moderated before going live. Instead, I realized that I was being entrusted to help create something worthwhile, and that created an immediate impression of investment in the project.

About "[taking] a few seconds to give a username and password": let's be realistic. Registration involves can involve a little more than giving a username and password—there is a further step further, optional step which is not always glitch-free (it wasn't for me)—and that shouldn't be glossed over. (There are sites that don't require email confirmation, where one just picks a username and password to gain full entry. Wikipedia isn't one of them. ask for email address upon registration.) Rivertorch (talk) 19:49, 23 May 2012 (UTC)

I should have been explicit: I agree with you, I'm suggesting that PC would work best if it allowed a 3-second login procedure. Tell me you want to be called "Joe", give me a password, and you're ready to edit with the username Joe142 (or whatever). I realize that opinions on what's needed for a "real" login are a matter of religious fervor, and I'm not saying that we have to abandon the email and the 3 checkboxes, only that it would be best to put those questions off to a future edit (for IPs whose first edit ever happens to be to a PC page), for the purposes of getting PC to work the way everyone wants it to. - Dank (push to talk) 20:03, 23 May 2012 (UTC)

Hmm. I'm not so sure the sort of "Registration Lite" you're suggesting would be much better than no registration at all, but I haven't thought that through and could be wrong. I do think that re-enabling PC with such pesky little details unresolved wouldn't be such a good idea. Rivertorch (talk) 20:23, 23 May 2012 (UTC)

As far as I know, Wikipedia's sign up form doesn't require an email address to be entered. isaacl (talk) 22:40, 23 May 2012 (UTC)

My bad. That was a page I hadn't looked at in years! Rivertorch (talk) 05:42, 24 May 2012 (UTC)

• "create huge mountains of work ... most suggested edits to pages under PC were not constructive, and the edits which were useful did not justify the expense of editor effort to weed out the problematic edits.": See the metrics from the trial; nearly all the suggested edits were accepted, so they couldn't have been that bad.
• "does active harm to Wikipedia and its reputation if understaffed/ignored": So, don't understaff it; don't put more pages under PC than we can comfortably handle. (This and other rationales carry a theme of "It could work in theory, but it won't work because the people making the decisions will screw it up somehow." If you believe that, then help us develop guidelines to make it work properly.)

I suggested above a principle of making sure that there's at least one reviewer (ideally, more than one) who's monitoring a page to be put under PC. Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

I'm not sure how that would work. Editors take breaks, planned and unplanned, and items drop off of long watchlists rather quickly. Unless there's a large enough pool of reviewers watching each page, coverage would be spotty, to put it mildly. It's easy to say "don't understaff it" but not so easy to do. Even if strict criteria had to be met to apply PC to a given page, I don't suppose there'd be anything to stop the total number of PC-protected pages from growing eventually into an unmanageable burden. We frequently have backlogs at various noticeboards, not because there aren't enough editors nominally watching but because those editors aren't here around the clock, and when they are here they're distracted by other tasks. I can't count the number of times I've seen something where I could have gotten involved and made a positive difference but chose not to because it would have meant not doing something else that I thought was more important. If PC is turned on again, I have to wonder if I'll take the time to review pending changes or simply go straight to other tasks. Rivertorch (talk) 20:14, 23 May 2012 (UTC)

I think the best way to organize this would be centralizing PC1,2 pages: For instance, they would show up on a special page when they get unreviewed edits, pages with the oldest edits first. Reviewers pick up pages from the list (preferrably the oldest ones) and review the edits. If there are not so many PC-ed pages, it should be doable.--Ymblanter (talk) 20:37, 23 May 2012 (UTC)

Agreed, although I can also see a page ordered by number of different accounts/IP addresses with unreviewed edits. One of each? Allens (talk | contribs) 00:18, 24 May 2012 (UTC)

One can actually do both, for instance, on two different pages.--Ymblanter (talk) 05:33, 24 May 2012 (UTC)

• "we can reasonably anticipate that PC will be used more liberally, e.g. every BLP.": This is where I diverge from just about everyone ... I don't think it makes a bit of sense to apply PC to BLP pages before we even know if or how it will work. BLP pages are the hardest pages to maintain in the whole project; using them as a testbed for PC is setting PC up for failure. Make it work on less troublesome pages first.

As previously mentioned, I'd say not to apply it to BLP at all - and have this as part of the policy. Or, if it is applied at some point in the future, make sure that part of the policy is that anyone not approving a referenced addition (negative or positive, but particularly the former) to a BLP page gets reviewership removed - including if this requires desysopping. Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

• "It's okay if a 'bad' revision is public for a bit-- that's an opportunity the readers to learn just how democratic WP is.": The 9,999 out of 10,000 pages not under PC will be sufficient to show how democratic we are, I think. Also, mob rule (the kind that leads to page protection) isn't democracy.
• "In some ways, "reviewing" could make things worse-- implying that an article has been 'screened for quality' in some way, when in fact, it's only screened for obvious vandalism.": Readers won't be able to tell from looking at the article that it's been "reviewed" at a previous time. But for users ... yes, I completely agree, no matter how simple you try to make the rules for reviewing, people will have expectations that, at the very least, editors were treated fairly and consensus was respected. PC can't work unless we can find reviewers with those skills ... that's one of my assumptions above, one that not everyone will agree with, but I see no way around it.

With regard to readers, you're saying that there won't be a lock symbol applied, with an explanation about reviewing? Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

I clarified by adding "at a previous time"; correct me if I'm wrong, but I believe we're all agreed that a special lock symbol would appear on pages currently under active PC protection. - Dank (push to talk) 15:35, 23 May 2012 (UTC)

• "Wikipedia is already complex enough, especially for newbie editors.": See my response to the first rationale above; it won't be complicated at all, they won't even be able to tell a difference from normal editing. If you're saying that editing something now so that it gets published for readers later is too complex ... well, this is the way publishing works everywhere except on wikis ... I think people get the idea.
• "doesn't even work on high-traffic pages, which seem like they'd need it the most": Again, see my response to the first rationale; a version of PC that lets users edit the working version normally and only publishes the page to readers after the edits are approved will require fewer updates to the page readers see, so it's hard to see why it wouldn't speed up rather than slow down the servers.
• "*despite being active in page protection for two years, I've never seen a single situation where I thought PC would be a better solution than just semi'ing [semiprotecting] the page ... Witness the Russian Wikipedia,": Well, assuming we're talking about a version of PC that allows anyone to edit, many of your Option 1 comrades would disagree.
• "The often-obscure change leaves a reviewer scratching their head as the editor who made the edit has long gone.": OTOH, the people who are more diligent and care more are more likely to stick around before the page goes "live"; at the least, the possibility that PC will bias editing in a more responsible direction is a hypothesis that needs testing.
• "increases the ability of anonymous editors to introduce errors (unless the reviewer is 'all knowing').": It's true that readers can't fix what they can't see ... but all readers will see it, now if they try to edit and then take a few seconds to pick a username, and later if they don't.
• "the responsibility for accepting edits is currently on the entire community (implicitly, by not reverting them) - transferring that onto a single individual (by explicitly approving an edit) does not seem like a desirable situation to me." Agreed, but sometimes "the entire community", while nice in theory, means no one in practice ... and that's exactly what we're talking about here, pages that are being considered for protection because whatever we were doing wasn't working. It's not unreasonable for either an individual or a community of reviewers to then take some measure of responsibility that things are done right; that happens all over the place on Wikipedia.
• "Any official control of content by a select group is completely antithetical to the spirit of WP": There are several possible answers here; one is that reviewers need to be applying local and project-wide consensus rather than imposing their own opinion, which is what we ask of people who "close" any kind of discussion on Wikipedia.

I'm getting the impression that you see as good the ability of IPs to effectively block, until a reviewer approves, autoconfirmed-user edits to a PC1 page? Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

See the first rationale. - Dank (push to talk) 18:29, 23 May 2012 (UTC)

• "It risks producing inadvertant content forks": Actually, the current situation creates content forks all the time, when different mirror sites grab different versions of the page. Not that I expect PC to be widely used, but if it were, by reducing the rate at which changes appeared to readers, it would reduce content forks. Web spiders don't generally grab page versions that can only be seen by logged-in users.

It, along with disagreements with the Wikimedia foundation on things like "principle of least surprise" or however they put it, could result in a deliberate content fork at some point - as in a split. Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

If someone says "You have to do it this way or I'll go join the competition", then I would invite them to go ahead and leave. But you're right, not everyone is happy with the same set of rules, and that's a tough problem, and new wikis are sprouting up all the time. None of them seem to be a real threat to Wikipedia, at the moment ... in fact, all the new content, including academic content, that has popped up on the web and will pop up over the next couple of years seems to be helping us on balance rather than out-competing us. - Dank (push to talk) 17:55, 23 May 2012 (UTC)

• "potential legal burden PC puts on page patrollers": I mentioned this in my assumptions above; in 11 years, there's never to my knowledge been a legal problem for anyone else who approved edits or caused edits to go "live", such as article reviewers and people who undeleted pages.

I know of a case (about which I can't say anything more - I know about it via a close off-Wikipedia relationship) of an implied legal threat against someone who reverted deletions (of referenced material) on a politician's BLP page. It wasn't anything more than implied, since the person doing the reversion lives in the US, but if the person lived in a more libel-lawsuit-friendly place like the UK... Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

First, it takes legal training and experience to know which threats are so significant that they merit abandoning an otherwise promising path. The layman's argument is: we're now up to around half a billion different readers and 4 billion hits per month, and we've been around 11 years, so if there were a huge problem, you'd think we'd know by now ... but I'd rather we heard an opinion from an attorney with solid credentials in this field. Second, any lawyer will tell you that the best way to avoid getting sued is to avoid pissing people off in the first place. See my response to the first rationale for why I think the model described there is less likely than full protection to piss people off. - Dank (push to talk) 17:41, 23 May 2012 (UTC)

• "it seems widely agreed that it fails to adequately protect BLPs": It's widely agreed that anything will fail to adequately protect BLPs, so that's not a good way to judge the merits of PC.
• "Not until this reviewer "right" can be taken away from administrators just as easily as it will undoubtedly be taken away from regular editors": Leave it to Malleus to come up with a point I don't have a quick answer for :) Yes, everyone agrees that PC means a new userright, and PC-reviewer recalls are sure to create the same headaches that other kinds of recall do ... another reason we'd better treat reviewing seriously and get good reviewers in the first place. On the bright side: this is a userright that has more to do with judging article quality and less to do with WP:ANI ... if we have to have more bureaucracy, that's the kind to have :)

Exactly what is the justification for making admins all be reviewers? It may make sense for PC2, but for PC1? About the only justification I can see is that it increases the pool of reviewers to prevent backlogs, etc. Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

I'm not dismissing that point, but on what I see as the main question ... did the rationales make a convincing case that that PC won't work ... I don't think that argument does that (and I don't remember seeing it, I'll read back through in a minute). - Dank (push to talk) 17:41, 23 May 2012 (UTC)

• "*Although I had originally supported pending changes, the reality was that it turned out to be a pain in the ass. It was overused as a whole and frequently used on pages that required specialist knowledge to correctly review. This created a situation where the number of reviewers capable of approving the changes with confidence was proportionately low to the number of people trying to edit the page. Combined with the logged use of forcing people to either approve or decline a change, it also made it problematic for people wanting to help: if they approve a change that turned out to be "wrong," then their name is just as much on it as the person making the change. Similarly, should they decline a change that turned out to be "right," they opened themselves up to accusations of not assuming good faith or biting the newbies. In either case, it could be something used against you in the future, so instead of making a choice, people would choose not to choose. Because of this, changes would sit in the queue for long periods of time.": Nicely argued, and that influenced my assumptions above. I guess all I can say is: we can't create more PC pages than we have PC-reviewers to deal with, and if no PC reviewer is ready to make a call yet, then they shouldn't make a call. As long as everyone can continue to edit the page, that won't be a disaster, it just means the readers won't see the results yet. If there's some kind of breaking news or important correction the readers really need to see, I'm sure PC-reviewers will take that into account.

I am guessing this would also be your response to my worry about "subtle vandalism while logged in followed by editing as an IP"? Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

I have to go back to: all I have to show here is that the Option 1 arguments didn't defeat PC, although they may have successfully rebutted one version of it ... and I thought this argument of yours in the Discussion section was an especially good one (but I didn't address that section in my critique). - Dank (push to talk)

• "By the way, we already have pending changes waiting at Category:Wikipedia protected edit requests. It, too, is backlogged, and I'd like to see supporters of proposal 2 or 3 work on that before endorsing a plan to expand it even further.": Excellent point ... but it's also more proof that we need to rely on normal editing as well, not just reviewers responding to talk page requests, to fix problems on protected pages, and that's exactly what PC allows that full protection doesn't.

The question is whether we really need another layer of bureaucracy. You may wish to take a look at the paper I referenced below. Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

There's a reason I waited until the RFC was over before attemping to present the case that arguments of the Option 1 guys didn't succeed in ruling out any future form of PC. (But they did IMO make many good arguments against the draft proposal. And by the way, when I started writing this, I was personally on the fence; I thought both sides made very good arguments that neither side fully addressed. In the course of trying to act like a lawyer and present one side as persuasively as I could, I think I convinced myself that another PC trial under different guidelines is at least worth a shot. I think it's best to be honest, because my own views probably came through in my text. But my views aren't relevant, of course; it's up to the closers to decide which arguments have weight.). During an RFC, arguments are like hydra heads ... kill one and five more sprout. The question here isn't whether all attempts to treat reviewers and reviewership seriously are bound to fail ... the question is whether that case was successfully made in this RFC. It was not. It's a great question for the next consensus-making process or RFC. - Dank (push to talk) 14:11, 23 May 2012 (UTC)

• "Because I was automatically granted 'reviewer' status, any of my edits automatically implied approval of the prior/pending edit, whether I intended to or not. As a gnome, I only have a fleeting interest in quite a large number of articles, my gnoming actions would actually create a false sense of security.": Yet another good argument for why reviewership shouldn't be handed out to, or expected of, everyone, or approached in an automatic or gnomish way. It should be a serious attempt to help resolve a conflict that was serious enough to require a form of page protection.

It's also an argument for an improvement to the technology so that someone with the reviewer flag has a choice over whether to be treated differently. Someone may well have subject-specific knowledge making them a good reviewer on some pages, but "only" be doing copyediting on others. (I'd be a horrible reviewer for any page requiring knowledge of other languages - e.g., for confirming references - beyond what Google Translate can give...) Allens (talk | contribs) 07:13, 23 May 2012 (UTC)

Agreed. - Dank (push to talk) 14:11, 23 May 2012 (UTC)

• "Seeing their edit seem to disappear will put off newcomers.": Totally agreed. See my answer to the first rationale.

Okay, I'm done. I realize there's a great danger here of coming across as dismissive of serious concerns; my intent is only to attempt to make the case that it's really not impossible to imagine that at least some of the Option 1 guys can craft a version of PC that would work for them and for others. Feel free to comment, disagree, whatever, inline or below. - Dank (push to talk) 05:35, 23 May 2012 (UTC)

Thanks for your thoughts! Your portion that was formerly above (now at User:Dank/PC) was also of interest, but harder to respond to. Allens (talk | contribs) 07:31, 25 May 2012 (UTC)

Your responses were great. Thanks, I forgot to add the link here. Are there any "big picture" points now missing? The only one I think people really need to get is: even though some Option 1 guys have implied that they "won" because the format of this RFC required them to consider just one PC proposal, and they made a lot of points that got no successful rebuttal during the RFC ... in fact, we all know that the Option 1 guys didn't just want to win a battle against one PC proposal, they wanted to win a war against all PC proposals. I'm hypothesizing various forms of PC and trying to show that the arguments were not sufficient to counter all of them. - Dank (push to talk) 12:47, 25 May 2012 (UTC)

Rivertorch has objected to this as polarizing, and I guess he's right; what I meant to say was that I'd be surprised if anyone arguing Option 1 would be satisfied if they succeeded in defeating one proposal, only to have a similar one take its place, and I support their effort, actually. - Dank (push to talk) 20:35, 25 May 2012 (UTC)

I see what you're saying now, and you make a good point. The process has seemed to have a Whack-a-Mole aspect to it at times. Rivertorch (talk) 23:21, 25 May 2012 (UTC)

Btw, Beeble was commiserating that he might have picked the wrong format for this RFC ... I think he was thinking that the voters didn't engage with each other sufficiently, and it looks to me like the voters were saying in some ways that this was the wrong RFC. Nevertheless ... I'm coming around to the idea that he picked exactly the right format for this RFC, because the job of anyone who's evaluating the rationales becomes impossibly open-ended unless you can say: this was the job they had to do, and IMO they did/didn't do it. Yes, the format meant that we couldn't get to the promised land of a definitive answer in one step, but that wasn't going to happen anyway. For example: some Option 1 voters stated that PC is unacceptable, that unlike semi-protection, it treats editors differently. My reply was that the requirement of 10 edits and 4 days inherent in the draft proposal was just a proposal, that there's nothing in the Option 2 votes that suggests that they're dead-set against letting brand-new accounts edit, as long as a reviewer evaluates the edits and makes a call. Encouraging IPs to log in or create an account just to edit a PC page will inevitably invite some socking, so we'd probably at least want to start with the question "Do you have an account?", so that if they answer "no" but they do have an account, we at least have the opportunity to catch them in a lie, if the lie becomes apparent. There are devils in many of the details, which is one reason this question has been so hard over the last 6 or 7 years. (Another reason is, we haven't done a great job of it IMO, but we're remedying that now.) If we could never close any RFC until we resolved details, we'd never close an RFC. I realize my critique is unsatisfying when it "grades" the Option 1 guys by a standard that wasn't present in the RFC ... but I think if people know this is the way it's going to go in the next RFC or consensus-gathering process, they'll be more engaged. - Dank (push to talk) 14:41, 25 May 2012 (UTC)

How is this a 'request for comment'?

I apologise if someone has already asked this, but looking over some of the thing itself and the discussion, I haven't seen anything that really explains the matter. The thing is formatted as a vote, actively discouraging discussion, directed rebuttal and any other suggestions - vote for this, or this, or this, and don't put too much on the discussion page, or it will get too big. While I appreciate the reasoning for doing so, this approach does have its downsides, as some folks have mentioned, but what I'm wondering about is the label itself. Seeing as WP:RFC explains RFCs as an informal process by which Wikipedians discuss things, and also that they are explicitly not votes, just why is this particular thing is labelled as a 'request for comment' when it goes against that? -— Isarra ༆ 18:12, 21 May 2012 (UTC)

I agree, this is more of a plebiscite than an RFC. At the start of the process I tried to open it up into more of an RFC style attempt to find a solution we could all live with, but my proposed option 4 was removed. This "RFC" may be used to justify a majority decision of some sort, but it isn't exactly an attempt to bring the community together. I'm still convinced that it would be possible to come out of this with a more effective defence against vandals whilst retaining open editing. But we need a process that involves people seeking to understand and accommodate each others concerns, and I don't see this as being that. ϢereSpielChequers 20:47, 21 May 2012 (UTC)

I think the manner of closure will make or break this process. Any PC discussion was going to be painful, however collegiate we tried to make it. But from where we are now, the key to minimising this pain will be whether the closers recognise that getting PC right is more important than getting PC right now. If they do, and they close in a way that reflects that recognition, this might well have been the best structure. —WFC— 20:59, 21 May 2012 (UTC)

I think when this started, we all knew that not everyone was going to be happy. That happens in every community. This situation is not an exact case of Arrow's impossibility theorem but the basic idea is the same. In many societies, some people want higher taxes, some want lower ones, etc. In the end most people are not happy with the processes that run that community, but they all live with it. The same should apply here. Let the process work, and let cooler heads prevail. There is no need to get worked up on these if you think in terms of variants of Arrow's result. History2007 (talk) 13:44, 22 May 2012 (UTC)

Also, we won't know if we've got PC "right" until it's live and people can see working. We're never going to get it perfect on paper; it has to go live at some point, and I doubt the debate will ever be over. — The Hand That Feeds You:^Bite 14:28, 22 May 2012 (UTC)

Of course we will never get it "perfect", but there is a gaping chasm between "perfect" and "having a degree of confidence that the framework is not a total train wreck".

The arrows theorum stuff in particular is not applicable here, as it is pretty obvious that if PC were to be introduced, all sides would want it to be used in a "sensible" way. Yet no attempt has been made to define "sensible".

I'll respond to the one substantive, relevant point ("In the end most people are not happy with the processes that run that community, but they all live with it."). The difference here is that our policies for Arbcom, Adminship, In the news etc were carefully thought through before being implemented. Most people now think most of them are in need of overhaul, but all of them served us well for years as a result of the thought that went into them before they were implemented. It is also worth noting how difficult it has proven to change these processes, and therefore how much thought we need to put into getting relevant policy and guidelines more-or-less right before it is given the level of protection/inertia that live policies enjoy/suffer from. —WFC— 14:58, 22 May 2012 (UTC)

Actually the basic structure of Arrow's result applies to this, as well as many other situations. It has been discussed for long in many academic circles. Do give it further thought please. History2007 (talk) 15:59, 22 May 2012 (UTC)

It's not applicable, because for the purposes of that theorum option 2 and option 3 cannot be considered distinct. Both groups want to see PC introduced, and both groups have raised concerns about its reintroduction. —WFC— 16:47, 22 May 2012 (UTC)

Actually if you read the proof part will see that their contents do not have to be mutually exclusive and they can be distinct choices even if they are overlapping. By distinct it refers to voting for just one of them, even if their contents are not mutually exclusive. But we are diverging here now.... Kenneth Arrow is still breathing, so maybe we should get him to come ever, if that might not prove to be impossible... pun intended. History2007 (talk) 17:00, 22 May 2012 (UTC)

I somehow managed to edit the below out by accident (edit conflict that I THOUGHT I'd resolved properly...); sorry! Allens (talk | contribs) 18:56, 22 May 2012 (UTC)

"Yes" and "Yes soon" simply are not distinct in any way, shape or form, and options two and three were explicitly set up to ensure that option 3 would not be entirely distinct from option two. You come across as someone intelligent enough to realise this, so I can only assume that you are knowingly trying to misapplying the theorum in an attempt to strengthen the position you support.

Anyway, let's put away our academic anatomy and get back to discussing the matter at hand. Option 3 exists purely as a fig leaf to the fact that this was set up as a binary straw poll. The result of that binary straw poll (in my personal opinion) would be a "yes". Thus the question turns to the matter of implementation. (The following is admittedly still opinion, but an opinion which not one single person has attempted to directly dispute) Implementation has not been resolved to the extent that the closing admins can confidently conclude that the community wants PC to be implemented straight away, in anything resembling the manner set out in the draft proposal. Taking the inertia of Wikipedia into account, they have an obligation to be confident that the draft proposal is a good reflection of how the community wants PC to be used. If they cannot be confident of this, they have an obligation to ensure that they get this answer. Tying those together, the logical answer is either a hard turn-on date, leaving the community to its own devices, or a true RfC which determines the policy, at the end of which PC will be turned on (defaulting to the draft policy if there is no consensus of any kind). —WFC— 18:38, 22 May 2012 (UTC)

I think I will stop now before I talk myself to death. History2007 (talk) 19:14, 22 May 2012 (UTC)

• "RfCs are not votes. Discussion controls the outcome; it is not a matter of counting up the number of votes." —WP:RFC guidelines
• "Refrain from making this a discussion page, take it to the talkpage. The page is already long enough with the !votes." —WP:PC/RFC2012 guidelines   — C M B J   12:53, 24 May 2012 (UTC)

German Wikipedia

I think it is very important in this debate that a similar system is already in use in the german Wikipedia and that most of the users didn't find it useful. A lot of useful edits are withhold too long and sometimes there are about 15 or more edits that are in the "queue" for one article... Also it's not very nice to work with when you edit an article with unreviewed changes and causes sometimes confusion to new users. Maybe a few other german users can also tell a bit about it? -- Laber□^T 03:08, 22 May 2012 (UTC)

Doesn't the German Wikipedia use Flagged Revisions (called geprüfte Versionen there) on all articles? Just about everyone here is talking about using Pending Changes on a tiny fraction of our articles; it's not the same thing. - Dank (push to talk) 03:20, 22 May 2012 (UTC)

There is also the differences that, at least as of not very long ago, it apparently wasn't very hard to get reviewership on the German Wikipedia (minimum # of edits and time on the system), which may be an argument against PC if they're still having backlog problems; on the other hand, even registered users on the German wikipedia apparently have their edits reviewed until they pass (a lesser level of) experience, which (along with having all articles effectively under PC1 at least) would increase the backlogs. Allens (talk | contribs) 17:57, 22 May 2012 (UTC)

Are the demographic type distributions of German and English Wikipedias in any way similar? I have seen no data on that and so would have to be guessing. Unless that type of information is known, it will be a pure guess to try to generalize from one community to the other. History2007 (talk) 13:49, 22 May 2012 (UTC)

My (very limited) experience tells me that German Wikipedia editors are on average more mature than English Wikipedia editors.--Ymblanter (talk) 13:59, 22 May 2012 (UTC)

The French Wikipedia is also an interesting case; they examined both flagged revisions and PC, and rejected both. Allens (talk | contribs) 17:13, 22 May 2012 (UTC)

There is, incidentally, a much lower rate of vandalism on the German and French Wikipedias (here, 11.5%; the others, 3.5%), according to the article I note below (which in turn is citing a study by User:West.andrew.g - who's also the author of WP:STiki, incidentally). Allens (talk | contribs) 17:51, 22 May 2012 (UTC)

At last, some data! That is an interesting item now: 3 times less vandalism on those than here. If we can now get those by time of day, day of week etc. may start to say something. And then break it by rough project e.g. arts, science, BLP etc. (Local time of day for IPs is easy, for others will need some work). If that type of data can be charted, it will say a lot. Most people do not know that their car insurance rates are often set by the time of day their last car accident happened.... That type of info can often say a lot. History2007 (talk) 18:02, 22 May 2012 (UTC)

Take a look at User:west.andrew.g's papers. I note that STiki, as well as being able to use CluebotNG's probabilities, also has its own vandalism detection algorithm. That algorithm does use (for IPs only) time of day and day of week, but those turn out not to be that indicative in most cases, at least for the English and (according to the papers) German wikipedias (the Spanish one looks like it's different for some reason) - with the exception that they help pick up schoolkid vandals, who seem to be a pretty high proportion of the lot, judging from most vandalism being during school/work hours (plus a lot of the vandalism I revert uses rather immature terminology even considering its content). In terms of overall subjects, good question - something that needs further investigation. Particular articles tend to have significantly higher and lower rates of vandalism (at least significant compared with time-of-day and day-of-week), and another paper appears to indicate that FA status is correlated with higher proportions of edits being vandalism. Allens (talk | contribs) 18:39, 22 May 2012 (UTC)

Thanks. But I have to look at them next month. I will not have time to do it justice in the next few weeks. I would be interested in how they did their analysis. And what they mean "indicative". I am not sure if time of day or week can be indicative of a single event, but if that data is available in anonymous format than its analysis is bound to show something, specially if they do more than univariate analysis. I will try to leave him a message within the next couple of months after reading the material. History2007 (talk) 19:12, 22 May 2012 (UTC)

IMHO it is not true that "most of the users don't find it useful" on German Wikipedia. Although there are some users with a very critical opinion about "Gesichtete Versionen", the system was confirmed in two votes with a large majority. BTW: 15 edits in the queue is an absolute exception. Regards --Iste (D) 19:37, 22 May 2012 (UTC)

Article on this debate

There's an interesting academic article on the debates, and the differences between different-language Wikipedias, on this topic called "Coercion or empowerment? Moderation of content in Wikipedia as ‘essentially contested’ bureaucratic rules". I'm reading over it right now. Allens (talk | contribs) 17:13, 22 May 2012 (UTC)

Thanks very much for posting this. A fascinating overview of the opposing "control" and "enabling" schools of thought, with particular attention paid to the divergent ways they have manifested themselves on the three largest Wikipedias. While I think de Laat's attempt to distill the disagreement down into these two camps is something of an oversimplification when it comes to en.wp, I can't help thinking he's onto something. My opposition to PC has largely involved pragmatic concerns, but I do wonder if fundamental philosophical differences in how different editors view Wikipedia—and the larger world—may be playing a significant role here. One word that has repeatedly come to mind when I've considered this debate is "authoritarianism". I'd like to see vandalism and other unconstructive edits prevented just as much as anyone here, but I am deeply worried that we may end up throwing the baby out with the bathwater. PC might usher in a repressive environment where legitimate (albeit often imperfect) edits are suppressed, leading to stagnant articles and a shrinking pool of new contributors. Even worse, to my way of thinking, PC might enable groups of like-minded established editors to stifle ideas with which they personally disagree, making it much harder for the diversity of thought and experience that has fueled the building of this incredible resource to continue to manifest itself.

For all its flaws, many of them serious, Wikipedia has become the most comprehensive encyclopedia in the world, with astounding breadth and depth of coverage. This has occurred in little more than a decade, and it never would have happened had a strict gatekeeper approach been in effect in the early years. Obviously we want to safeguard what we've got, but are we willing to risk slowing further progress to a trickle—or even halting it—in a bid to make protection of our current assets easier or more reliable? I don't think there are any easy answers to such questions, but I do think we should be asking them. Otherwise, we may fall victim to our own shortsightedness. Rivertorch (talk) 21:06, 22 May 2012 (UTC)

The long and short of it is that no one in this debate is pushing any specific agenda, except that they all want the best for the encyclopedia. But their perceptions of the future happen to differ, as the perceptions of many economists differ on what is best for the economy. So debates take place, just as they do among economists (even those with Nobel prizes), and the discussions do shed light on policy, although in the end no one can predict the future. So we can have debate but should be laid back about it anyway. History2007 (talk) 21:58, 22 May 2012 (UTC)

Quick apology to the closers

Regarding my long "critique" above ... I started out determined not to tread on any ground that the closers had indicated they wanted to work on, including the exploration of possible consensus, but I really had to attack that question to critique the rationales. I really hope that no one reads into this any lack of confidence in the closers to do their job. Lawyers can argue cases, but it's entirely up to the judges and juries what works and who wins. Apologies if I gave any other impression. And thanks very much for the opportunity, I really enjoyed wrestling with this one. - Dank (push to talk) 19:38, 23 May 2012 (UTC)

Not at all, your comments were very insightful and certainly gave me pause to think about what I was doing. The Blade of the Northern Lights (話して下さい) 16:51, 26 May 2012 (UTC)

Scope of PC

Hey all. I have largely stayed out of this one because it seemed to be a straight vote, which I am not really interested in. I heartily approve of the rebuttal and engagement that has gone on recently. If I could throw in one thought on something:

Original point	PC is much more open to editing than the alternatives like semiprotection.
Response by Rivertorch	Not really. It can be argued either way, but I think there's little doubt that if PC were implemented it would be applied to considerably more articles than semiprotection currently is. That would actually mean less open editing
Response by Allens	If it were made clear that PC1 is only to be used if semi-protection would otherwise be used, that might help.
Response by me (Yaris678)	I agree that confining the scope of pending changes could help. However, I think that part of the problem with the trial was that it was used on pages that were suitable for SP but not PCP. It would be better to confine PC1 to articles that are on the borderline between SP and unprotected. If there is a concern that this will be interpreted too broadly and we'll get loads of pages PC protected, perhaps we should also say that the number of PC-protected pages should not exceed 25% of the number of semi-protected pages.

BTW. I am neutral on whether or not PC does actually happen. I am just hoping to contribute to reaching a consensus.

Yaris678 (talk) 11:15, 24 May 2012 (UTC)

Expanding on my above point... maybe it would make sense to say that SP+PC2 should be used on articles (and templates etc.) that are borderline between SP and full protection.

I'm not sure whether that point is a weird esoteric one that no one cares about or something that is easy to get consensus on, given that people tend not to talk much about PC2... still less the SP+PC2 combination.

Yaris678 (talk) 18:51, 24 May 2012 (UTC)

SP+PC2? Odd combination. About the only reason I can imagine someone using PC2 is an edit war that doesn't have any reviewers on either side. If that page was also being attacked by IP vandals... Where can we find some statistics on how many pages are SP and FP, in each namespace? Ah; Wikipedia:Database reports#Protections. 49 pages currently FP indefinitely, and 2 other pages FP for over a year. 3812 pages currently SP indefinitely, and 80 more SP for over a year. No idea on the numbers of temporary FP/SP. Allens (talk | contribs) 21:50, 24 May 2012 (UTC)

Hi Allens,

Useful link.

Two pages listed as indefinitely FP, which may be more appropriate for SP+PC2 are Gazi Mahabubul Alam and List of Hannah Montana episodes. I don't know the history of these articles, but the rationales given look borderline to me so I'd be prepared to give SP+PC2 a go... after a bit of discussion... if we were to do the whole PC thing. Yaris678 (talk) 22:20, 24 May 2012 (UTC)

A statistical analysis for the inquisitive and the uninitiated

Metric	Opt.1: total	Opt.1: moral-only support (conservative estimate)	Opt.1: net	Opt.2: total	Opt.2: moral-only support (conservative estimate)	Opt.2: net	Opt.3: total	Opt.3: moral-only support (conservative estimate)	Opt.3: net
Participants	179	26	153	309	55	254	17	0	17
Words (total)	11255	53	11202	9462	152	9310	1851	0	1851
Words (per person)	62.9	2	73.2	30.6	2.8	36.7	108.9	0	108.9
Characters (total)	66638	321	66317	55112	841	54271	10920	0	10920
Characters (per person)	372.3	12.3	433.4	178.4	15.3	213.7	642.4	0	642.4
1 "Moral-only support" includes lone signatures and other expressions that unequivocally lack any original input; for example, "Support per X" is included, but "Support - overdue" is not. The community's working definition of a "!vote" in the context of most consensus debates is considerably more expansive. 2 One individual withdrew support from option 2, reducing the total number of associated participants from 310 to 309. 3 Dataset available upon request.

  — C M B J   12:05, 24 May 2012 (UTC)

I guess this confirms the ideas of "vocal minority" vs. "silent majority" and "vocal opposition" vs. "silent proponents". --Michaeldsuarez (talk) 23:21, 24 May 2012 (UTC)

It could just as easily confirm that the reasons for unconditionally supporting option two are very weak. Of course, I would not expect a firm supporter of option two to reach that conclusion.

It's also worth noting that many of the commentators from section two have actually expressed a degree of reservation (support with caveat, support despite the flaws etc). —WFC— 00:03, 25 May 2012 (UTC)

If one used criteria like this at RFA, nobody would ever pass. Largest number of words≠strongest argument. ~Adjwilley (talk) 19:07, 5 June 2012 (UTC)

More stats

I too have been calculating a few things over the past two days, and I hope the following will complement CMJB's fascinating contribution above. I draw no particular conclusions from any of these numbers but find them intriguing nonetheless. My totals may be slightly different from those currently shown on the project page, as I have subtracted several duplicate !votes from my figures (and reported them to one of the closers). None of my figures should be taken as absolutely precise, although I believe there are no significant inaccuracies.

---

Option 1

Out of 178 users (35% of RfC participants) who endorsed this option,

84 users (47%), including 35 administrators (20%), currently hold reviewer user rights;

36 users (20%) have made fewer than 1000 edits;¹

19 users (11%) offered no substantive comment² along with their !vote.

Option 2

Out of 308 users (61% of RfC participants) who endorsed this option,

194 users (63%), including 71 administrators (23%), currently hold reviewer user rights;

45 users who (15%) have made fewer than 1000 edits;^{1, 3}

55 users (18%) offered no substantive comment² along with their !vote.

Option 3

Out of 17 users (3% of RfC participants) who endorsed this option,

9 users (53%), including 5 administrators (29%), currently hold reviewer user rights;

3 users (18%) have made fewer than 1000 edits;¹

0 users (0%) offered no substantive comment² along with their !vote.

---

Notes

¹ as of 25 May 2012

² "No substantive comment" sounds subjective, but it isn't really. Having decided not to distinguish between users who typed "Support" before signing and those who didn't, I further decided that certain slightly longer comments (e.g., "strong support", "yes please", and "support - no further comment") contained no additional substance, only additional words.

³ One IP user endorsed Option 2. The number of edits an IP user has made generally cannot be determined.

I hope someone finds the above helpful. Rivertorch (talk) 10:27, 25 May 2012 (UTC)

The last analysis that breaks the votes by number of edits, rights etc. is interesting. You put a lot of work into this.... Yet I should say that the text analysis and how long the sentence was etc. is not standard practice. How long a user types may depend on a number of factors (personal verbosity level, daily schedule, etc.) Many organizations do that on user feedback, but most acknowledge that part as very soft/fuzzy data, unless each message is read and categorized by 3 different people given a predetermined rating scale, etc. History2007 (talk) 15:03, 26 May 2012 (UTC)

True. My standard wasn't how much they said; it was whether they said anything that gave a hint of why they were endorsing a given option. In other words, I was trying to distinguish between those who simply !voted and those who provided a reason for their !votes or an opinion on the option. For instance (these are hypothetical, of course):

856. Endorse because I hate vandalism. —Jane Doe

is substantive, if not especially helpful, while

857. Endorse. —Jane Doe

is no more substantive than a signature alone, such as

858. —Jane Doe

because the second example tells us no more about Jane Doe's views than the third does. Rivertorch (talk) 20:21, 26 May 2012 (UTC)

Well, yes. But there are inherent limits to extracting the intentions of people from text fragments. In market research these things quickly translate to product success vs luke warm sales. So it is a problem that many industries face.History2007 (talk) 21:11, 2 June 2012 (UTC)

I'm sure you're right. But I wasn't trying to extract intentions, just distinguish between those who essentially just checked one of three boxes and those who made a comment as well—a very simple and basic distinction, it seems to me, although perhaps not revealing of anything in particular. Rivertorch (talk) 23:21, 2 June 2012 (UTC)

My friend this is an old story: humans have tried to gain understanding from "clues" for centuries - including determining someone's character from the shape of their head. It is just hard to do... History2007 (talk) 19:57, 3 June 2012 (UTC)

The format was done to allow those who believed this has been discussed enough to simply make their position known without further discussion, while at the same time giving those who felt more discussion was needed an opportunity to do so. I can't say I am surprised at the number of users from every side of the issue who decided they was no need for more discussion before making the core decision regarding whether to use PC at all or not. We've already discussed it for five years after all. Because the structure made it inherently permissible to simply add an endorsement and be done with it I don't think they should be given less weight than remarks that had a sentence or two attached to them, but that's up to the closers to determine. Beeblebrox (talk) 21:14, 4 June 2012 (UTC)

It is indeed up to the closers to decide, yes. It's interesting that if one subtracts those who didn't make a substantial comment, one gets the following: Option 1 37% (159); Option 2 59% (253); Option 3 4% (17). Allens (talk | contribs) 00:54, 5 June 2012 (UTC)

Too late but....

here's an example of a situation where pending changes worked well; for literally years Wisconsin Hoofers and some related low-traffic articles have been subject to an ongoing, BLP-violating campaign against someone who has worked at the club. When pending changes were used, the violations stopped, but positive IP contributions were possible. Since PC was removed, the violations have resumed, and recently some negative allegations about a living person remained in the article for a month before they were taken out. My only choice now is to semiprotect, and so the IPs and new users who have been contributing usefully to the article are shut out. I hope pending changes becomes a choice again, because in my view it has a clear and useful place in certain situations. Slp1 (talk) 22:38, 26 May 2012 (UTC)

Hi Slp1,

If we are going to do PC then Wisconsin Hoofers is just the sort of article that it would be good on. It is what I would say is on the borderline between semi and no protection. See #Scope of PC, above. Furthermore, it has a low edit rate, which is also good.

Unfortunately, some people want a much broader scope for PC protection. People say things like "put it on all BLPs". Anyone who knows anything about the administrative burden associated with PC is scared by this idea. They can't see how they can stop PC going above a small number of appropriate articles so they argue against PC per se. That's my analysis of the current stand off.

Yaris678 (talk) 12:11, 28 May 2012 (UTC)

I agree that would be a good use of PC, and note that no opponent of PC has yet taken the opportunity to suggest otherwise.

However, it is important for the closers to take into account the can of worms that would be opened if we implement PC immediately, without clearly defining its scope. Specifically, the collateral damage that disagreement on its usage would cause, both directly on the affected articles, and indirectly in the form of the admins and editors that the resulting drama could cost us. The main argument for going ahead anyway is the assumption that PC would be used "sensibly", but there are dozens if not hundreds of different opinions of what "sensible" means, and there is no workable mechanism to override an individual admin's judgement of "sensible". The Germans probably consider their equivalent (flagged revisions) to be "sensible". The big Wikipedias which have chosen not to use PC or Flagged Revs probably consider their system to be "sensible". And while from my limited experience most of the big Wikipedia are "sensible" communities, the majority of English Wikipedians do not consider either of those approaches to be "sensible". —WFC— 13:25, 28 May 2012 (UTC)

Please allow me to rephrase that as I understand it: "Closers, don't do it.. And if you do, do it later". I am sorry, but talk page should not influence the closing, the actual process, the votes, comments etc. should. As the section title suggests, it is too late for debate now, the Rfc has concluded. History2007 (talk) 20:00, 3 June 2012 (UTC)

Please allow me to put my response to you in a similar style. "WE ARE NOT A GODDAMN DEMOCRACY, AND WE ARE NOT A GODDAMN BUREAUCRACY." I'm sorry for the tone, and even more sorry if I have misjudged the intention of your comments on this page. But you repeatedly refer to this as a vote, and anything that so much as looks like an attempt to bypass qualitative discussion should be resisted, even if that flies in the face of Wikipedia's increasing tick box culture. —WFC— 03:58, 5 June 2012 (UTC)

Don't shout in all caps if you want to be taken seriously. — The Hand That Feeds You:^Bite 13:51, 5 June 2012 (UTC)

If you had taken the time to read my previous comments, you would have seen that the above was perfectly proportionate. I have repeatedly explained in terms that anyone can understand that we are not a democracy, and that decisions here are unlikely to be swayed by bureaucratic technicalities. History2007 clearly cannot grasp this concept.

If you have something to contribute to the discussion other than critiquing the tone of isolated posts, then we're all keen to see it. —WFC— 14:58, 5 June 2012 (UTC)

• Here's what I don't think I'm getting about the idea that we should "fix the policy" before we use it. How do we know what is wrong with it if it we aren't using it? If I buy a used car without test driving it and then service the transmission, I may find that what it really needed was new tires and an oil change, and I've just wasted my effort messing with the transmission instead. So my question would be, where, specifically is the problem with the policy? It is at least as complete as the other sections of WP:PP, which it would become a part of. As with all other forms of protection, only admins would be applying it and we expect them to be able to interpret policies without having every last detail made perfectly explicit.

There is no ironclad definition of when to apply semi or full protection, when to block a user, when to delete a page, when to unblock a user, what exactly constitutes an edit war, etc, and yet the vast majority of such decisions are made, hundreds of times every day, without problems. I would estimate that only about 1-5% of all admin actions are reviewed at ANI or other forums, so I find it hard to understand why there seems to be this belief that with this tool, admins will just run off the rails and put aside the common sense and good judgement we normally expect from them. Beeblebrox (talk) 21:28, 4 June 2012 (UTC)

Several examples have already been given: The process for allocating the tool, the standard for which users will be held in approving edits (does it have to be vandalism? can they use it to refuse any edit they don't think should be in the article?), the types of articles where it should not be used (very large articles, very active articles, one with a history of severe BLP violations). Those are just the ones I came up with in 30 seconds. Just because Wikipedia has a habit of doing things bassackwards doesn't mean that we need to perpetuate this. Bottom line, though, this RFC isn't actually about turning on this tool, it is about making fundamental changes to the protection policy. I'm not sure whether or not that was your intention, Beeblebrox, but that is certainly the effect. If this is approved, we have changed the protection policy without actually having an RFC on the protection policy. Risker (talk) 21:42, 4 June 2012 (UTC)

Well, ok, those are some specific issues and I thank you for clarifying them. However, I think the answers are a lot simpler than you make them out to be. How the tool would be allocated is explicitly stated in the draft policy. What problems it is to be used to prevent is also fairly explicit. Really, it's not that big of a change. rejecting a proposed edit is roughly equivalent to rolling back an edit that has already been made, hence the "similar level of trust" statement in the policy. If you wouldn't rollback an edit in the case where there was no protection, you shouldn't reject it under PC. That is just common sense and it could easily be made more explicit without having another prolonged discussion, adding what I just said would do it. Article size is admittedly not addressed, and as I recall that was an issue during the trial, and it isn't an issue we can expect a fix for while the tool remains in limbo. Again, I would suggest a very simple solution, such as adding "PC should generally not be used on very large articles at this time, this may be rectified in the future." I really don't think this is something we should delay over, the policy is a draft, if and when it goes live simple clarifications such as the ones I have just proposed can be added right away by anyone. I can't imagine why anyone would disagree that you shouldn't reject an edit you wouldn't rollback if it were live, we shouldn't need to discuss that, just clarify it. Beeblebrox (talk) 22:11, 4 June 2012 (UTC)

I realise that you're trying to be conciliatory and reassuring; however, it's coming across not quite in the way you're intending it. Article size issues require a major rewrite of the software; that was discussed during and after the trial. The problem, Beeblebrox, is that when you read the actual votes supporting turning on pending changes, a very significant number of them talk about applying pending changes to articles that are specifically excluded from the current policy, such as little-watched BLPs with no recent history of vandalism, featured and good articles (there's that article size problem again, and once again many GA/FA editors do not want to play the "special rights" game so these articles will no longer be watched by the people who are most knowledgeable about them), medical articles "just in case", and so on.[For the record, my practice when reading RfCs is that I do not count a support vote when its commentary contradicts the proposal; that vote is supporting something *other* than the proposal. I hope the closers consider doing that.] Even during the trial, it was very difficult to keep administrators on track using pending changes only on the selected target group of articles; even then, we never got past 2000 articles on pending changes instead of the target 10,000. And unless it is in black and white that only an edit that would otherwise qualify for rollback should be rejected as a change, you *will* have most edits rejected if they're not perfect. I cannot count the number of edits I see reverted as "vandalism" or "test edit" when they're clearly good (or at least good faith) edits already. Keep in mind that all kinds of editors are using scripts to do the automatic equivalent of rollback, except that it comes with a little edit summary so it doesn't count as rollback. We can count on Huggle and Twinkle being modified to permit that very same thing to happen on a PC-protected article; it wouldn't surprise me if their respective developers are already writing the code. In other words, we already have issues with inappropriate reversion of edits, and this is more likely to exacerbate the existing problem. But now, the editors will have to have special permissions to do it, and instead of being corrected, they'll be humiliated by having tools stripped from them for doing exactly what they've been doing for years without correction. So yes, it has to be written into the policy. Risker (talk) 22:59, 4 June 2012 (UTC)

So, if and when the tool goes live and the policy is added to WP:PP, you or I or anyone else can come along and immediately add the specific wording we just discussed. As I said I can't imagine that anyone would disagree with it. A follow-up RFC solely for those interested in continuing to tweak the policy is also something that has already been discussed and which is probably on the drawing board right now waiting for a close here.

I would also anticipate that discussion would open back up at WP:RVW. Reviewing seems to be the aspect of this that is causing the majority of the concern, but at present all we have as far as an explicit reviewing guide is the failed policy proposal there, which seems to have simply petered out because we weren't using PC so what was the point of perfecting the guideline for reviewing edits with it? It actually looks like it would specifically address most of the concerns we've heard about reviewing as it stands now, it just needs to be "officially" discussed in order to bump it up to being an active guideline or policy.

This is something a lot of folks seem to be missing, a lot of work has already been done here, well above and beyond the draft policy here, but there's no point, in my opinion anyway, in re-opening and discussing all these other things if we don't even know if we would use PC at all. Now that it does seem more less certain that it will be back, any and all of the related discussions, proposals, etc can be re-activated as needed. You could open an RFC at at RVW talk page right now if you wanted to, I'm sure that would making the closers of this discussion feel more confident in their duties. Beeblebrox (talk) 23:22, 4 June 2012 (UTC)

This is a hard fight to have before the official close. - Dank (push to talk) 02:23, 5 June 2012 (UTC)

(e/c)There was insufficient support the first time around to keep pending changes enabled, but it was never actually turned off—just removed from the pages it was on. I wondered why it stayed that way for over a year, because usually when there is "no consensus" for a major change to anything (e.g., article, a blocked user, a policy change), the change isn't left sitting around—it's reverted. It was never reverted here. On top of this, it seemed that there wasn't even an attempt or RFC requesting a second trial. Instead, it was basically just a repeat of the end-of-trial RFC, only with fewer options. Not surprisingly, the preliminary numbers reveal that nothing's changed since the first RFC. There's still no consensus for implementing pending changes. Realistically, the whole RFC was out-of-process in the first place, because there probably should have been a !vote on whether there should even be a second trial, much less a permanent implementation. --slakr^\ talk / 02:50, 5 June 2012 (UTC)

Dank, the RFC's been officially closed for a long time, we're just waiting for the reading of the entrails. (I do sympathize with the closers, it's not an easy one to interpret.) And Beeblebrox, I'm really sorry that you think it would be simple to make the modifications you propose; the fact is that much of what I have pointed out is clearly disputed even on this RFC, so I wouldn't count on it being a cakewalk. Unfortunately, I have other responsibilities to the project and am not in a position to invest a few hundred hours into proposing modifications to a policy proposal that essentially failed because the community decided a year ago that the tool it discusses should not be used. I do pretty much agree with Slakr, I would much rather have seen a second, more carefully organized trial; we had plenty of time to develop it, and even at the close of the first trial there were many of us who proposed processes for doing so. Slakr, just for the record, the Engineering Department decided not to remove the software, but left it to the community to decide whether or not to use it, just as we have many other software features that are built in but just not in use due to community decision. Risker (talk) 03:10, 5 June 2012 (UTC)

@ Beeblebrox: "If you wouldn't rollback an edit in the case where there was no protection, you shouldn't reject it under PC". So what should you do with it? Just leave it sitting there? I'm not about to deliberately admit a harmful edit into an article, no matter whether it is made in good faith or not. On the other hand, the time I devote to WP is sporadic and often fleeting, and I cannot fix everything I see that's wrong, whether it's tortured syntax that needs to be rewritten, a partial url that needs to be googled, bad markup that will mess up the page, or plausible but somewhat farfetched content sourced to something that needs a careful check to see if it's reliable . . . the list is endless. If PC is applied to an article and one runs across such a situation (as I did repeatedly during the trial), what exactly is the best practice? This has never been explained to my satisfaction, which is one of several reasons why I don't think PC is anywhere near ready for prime time. PC has no "good faith reject" button, yet in the way it operates it isn't analogous to rollback so much as it is to undo—broadly construed. Rivertorch (talk) 05:03, 5 June 2012 (UTC)

Well, there's a lot here to respond to, I will touch on a few points made in the last few posts:

• Regarding why there was no second trial period: because nobody ever made a proposal for one. I don't happen to think that would have been a productive use of our resources so I didn't propose it. Why nobody who did think it was a good idea bothered to propose it in the intervening year between RFCs is something only those persons could answer. As to the related statement that this RFc is "out of process" what process would that be? There is no guideline, policy, or other process that dictates that we were obligated to do anything in particular. There was no mandate for a new trial then, and I seriously doubt you could get one now, but if that's something you want to see now would probably be the time to raise the issue. Two months ago when the RFC opened and was advertised all over the site probably would have been a an ever better time to raise that idea actually.

• I'm surprised that there are still comments being made that suggest that are even now some who do not understand the results of the last RFC. The idea that there was not sufficient support for keeping the tool and that is why it was turned off is incorrect. We were not permitted to make that decision during last year's RFC, the attempt to force a clear decision at that time was derailed due to concerns over the fact that the supposedly temporary trial period had never actually ended. Instead of just declaring it over and moving on, for some reason a small but extremely vocal group of users fillibustered the debate until they got a new phase of the RFC that left behind the original purpose and concentrated solely on the short-term question. That is the exact reason another RFC was needed, the long term question of whether to use it or not has never been allowed to be answered.

• As to what one would do if they saw a proposed edit they did not agree with but which was not made in bad faith: The limited guidance we have at this time for reviewers, along with comments during the course of the two RFCs, indicates that the community expects reviewers to reject bad faith edits only. What they could do instead is another aspect that needs clarification. If it was up to me I would say best practice would be to speak to the user making the submission explaining why you think their edit should not go live. If they see your point, then rejecting it is not a problem. If they don't then you need to go to the article's talk page and discuss it like any other edit. If that is too much trouble for you to bother with then leave the edit unreviewed and let another of the 5000 or so reviewers take care of it. Here's something not everyone knows: admins do this all the time because they know if they don't want to mess with a particular issue someone else who does will happen by soon enough. There is also the distinct possibilty that if and when we do resume using PC, the developers could give us the "good faith reject button" you mention, but so long as we aren't using it that is not going to happen, along with any other bug fixes or upgrades. Beeblebrox (talk) 09:29, 5 June 2012 (UTC)

  As for the last point, every reviewer is also an editor. If they see inappropriate good faith edits, nobody can prevent them undoing these edits as a usual editor, with explanation in the edit summary and/or on the talk page of the user who first introduced the edit.--Ymblanter (talk) 10:03, 5 June 2012 (UTC)

So we approve the inappropriate edit only to immediately turn around and undo it? What a clumsy way to maintain the integrity of an encyclopedia! Aside from requiring an unnecessary extra step, there's something deeply troubling about green-lighting a bad edit even if one intends to revert it presently. By accepting a pending change, a reviewer arguably takes some of the responsibility for that change's appearing in the article. Maybe that shouldn't bother me, but it does. (I have an awkward tendency to get hung up on principles sometimes.) Rivertorch (talk) 19:07, 5 June 2012 (UTC)

It's not inconceivable that the closing statement for this RFC will say something like "insufficient support, and given the discussion, we don't recommend another RFC". If so, and if in fact no future RFC happens, then we're wasting our time here, so I'd prefer to wait either for the closing statement or for at least a statement from one of the closers that they're not headed in that direction. But this much I can say now: what I continue to see in the comments from all sides is the conviction that the process is going sideways because people aren't keeping up with the relevant points ... and that's a realistic fear, actually, because there have been so many points. It's not the usual solution, but this particular debate, it seems to me, needs one or more working groups that could be joined by anyone at any time, where continued membership is subject to the condition that you have to preface any comments you make by linking to, referring to or restating what seems to you to be the current state of the opposing or supporting arguments that you're responding to, unless you're just replying to the current thread. (And hopefully, some kind souls will go through from time to time and organize the arguments, so that readers can focus on only the parts they want to focus on. For instance, I'd love to be able to skip over the process arguments ... "You shouldn't have been allowed to say that, because ...") Without at least some clues, it's going to be impossible for anyone to keep up, and comments will probably continue to have kind of a "shouting into the void" feel to them. - Dank (push to talk) 12:02, 5 June 2012 (UTC)

Better still the closing statement could say "This plebiscite was inconclusive, at best this experiment reaffirms that there are good reasons why RFCs are supposed to be discussions not votes. Needless to say the closing of this pseudo RFC should not dissuade those who want to seek a consensus from formulating compromise proposals and filing a proper RFC to discuss them and attempt to build consensus by both sides trying to accommodate each others concerns". ϢereSpielChequers 11:04, 17 June 2012 (UTC)

Another trial

There are some comments above about the possibility of doing another trial. After the first trial, this was being discussed seriously at Wikipedia talk:Pending changes/Straw poll#Closure - Compromise language. I would be happy to see that idea revitalised now. Specifically, I like the wording by Ocaasi at Wikipedia talk:Pending changes/Straw poll#Alternative phrasing. Yaris678 (talk) 15:34, 6 June 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.