Wikipedia talk:Village pump (proposals)/Proposal to require autoconfirmed status in order to create articles

Name of the page
This proposal is misnamed. It is really about Preventing  new users from creating article, not allowing them. Currently, we do allow them. I suggest a neutral title would be Preventing brand-new editors from creating articles.  DGG ( talk ) 22:11, 3 April 2011 (UTC)
 * There's a distinction between "non-neutral" and "obviously incorrect" here ;p. You might want to take a look at which side of the debate I'm on. In any case, it's effectively nothing more than a redirect, so.. Ironholds (talk) 22:20, 3 April 2011 (UTC)
 * I agree with DGG. The page title at present ("Wikipedia:Village pump (proposals)/Allow new editors to create articles") is wrong and misleading. This should be moved. --MZMcBride (talk) 03:08, 4 April 2011 (UTC)


 * What about moving it to Proposal to require WP:AUTOCONFIRM in order to create articles? Basket of Puppies  03:10, 4 April 2011 (UTC)
 * That would be ideal. I thought the title was a bit odd, but this seems accurate and neutral. Rivertorch (talk) 06:20, 4 April 2011 (UTC)
 * Sounds good to me as well. Can we get a sorta instaconsensus on this so we can end the protection, which to me is FAR more onerous to consensus building (a fully protected discussion!?!) than some quibble over the semantics of the title.  The protection is far more offensive than the title to me, so change the title and get it back open again.  -- Jayron  32  14:52, 4 April 2011 (UTC)
 * Seconding what Jayron says. I'm fine with Proposal to require WP:AUTOCONFIRM in order to create articles. I'm a lot less fine with full-protection being forced upon us. A fluffernutter is a sandwich! (talk) 14:59, 4 April 2011 (UTC)

There's a rough consensus for moving but keep in mind that we can't have links in titles, so it should rather be Proposal to require autoconfirmed status in order to create articles. Cenarium (talk) 15:03, 4 April 2011 (UTC)
 * I've renamed the RFC and unprotected. Cenarium (talk) 15:07, 4 April 2011 (UTC)

FYI, the title originated with a different proposal which the current one grew out of (in entirely the opposite direction). Hence a certain confusion at this point. Rd232 talk 16:29, 5 April 2011 (UTC)

Full protection?
I'd like for Cenarium to explain why it was necessary to full-protect this RfC. Their template at the top of the RfC appears to say that because they think the title somehow doesn't meet our standards, this calls for the discussion to be forced to a halt, which does not make a whole lot of sense to me. Can we get a more expansive explanation of what the protecting admin's intention is and why they felt full-protection, with no discussion, was required to handle this? A fluffernutter is a sandwich! (talk) 14:53, 4 April 2011 (UTC)


 * I was about to. A misleading title is a strong prejudice to the debate. A temporary protection is better than letting the debate be prejudiced by the title of the RFC. The question of the title needs to be settled, and this is to enforce the settling of that question that protection is necessary. Cenarium (talk) 14:56, 4 April 2011 (UTC)
 * I've allowed moving. Cenarium (talk) 14:58, 4 April 2011 (UTC)

Was that wheel warring?
I think we may have just had a case of wheel warring. Cenarium fully protected this RfC, followed by IronHolds' unprotection, followed by Cenarium's full protection again. I've had a read of the relevant WW policies and ArbCom decisions. WP:WW says"Do not repeat a reversed administrative action when you know that another administrator opposes it. Do not continue a chain of administrative reversals without discussion. Resolve admin disputes by discussing ." while ArbCom seems to be exceptionally clear on this issue. The most relevant ArbCom decision says"Administrators are expected to act collegially and to respect one another's decisions. If one administrator disagrees with an action taken by another, then unless the situation is an emergency, he or she should seek to discuss the matter with the second administrator or to raise the issue on a noticeboard and seek consensus. 'Wheel warring', in which administrators reverse one another's actions multiple times, is especially inappropriate and may result in substantial sanctions, including desysopping.1" IronHolds seems to be under the impression that Wheel Warring has occurred. Did it? Is this an issue that should be referred to ArbCom? The WW policy seems to indicate that"Wheel warring usually results in an immediate Request for Arbitration. Sanctions for wheel warring have varied from reprimands and cautions, to temporary blocks, to desysopping, even for first time incidents.2"and that even first time incidents are not a reason for giving a pass. I am not suggesting that WW happened here. What I am suggesting is this be investigated by neutral third-parties and subjected to community discussion. Basket of Puppies 16:02, 4 April 2011 (UTC)
 * FYI, I'm not asking for heads to roll, or for this to go to arbcom, or for people to be punished, sanctioned, censured, desysopped, beaten, pummelled, kicked in the 'gnads, GBH'd, assaulted, subject to extrajudicial torture or troutslapped. I was just asking Cenarium to have regard to my actions. That's all. Please don't make a big mess'o'drama out of nothing, BoP. Ironholds (talk) 17:32, 4 April 2011 (UTC)
 * Does this really matter? Let's just move on. Ed [talk] [majestic titan] 17:42, 4 April 2011 (UTC)


 * To prevent this from becoming a drama fest, I am capping this. --Guerillero &#124; My Talk  &#124; Review Me  17:54, 4 April 2011 (UTC)

What was your first edit?
Out of curiousity? what was your first edit to wikipedia? was it a new article? Here I need to answer twice because I edited as an IP for a while before registering for an account. My first edit as an IP was to answer a question at the reference desk, all my first 10 edits were pretty much there. As a proper username my first actual edit was to prod a hoax article on a made-up medical condition. I'm just wondering how many of us, who are obviously all reasoned and active Wikipedians, started with an article. HominidMachinae (talk) 20:04, 4 April 2011 (UTC)
 * My first edit as a registered user was to fix an external link I had added as an IP. My first 19 edits were to my talk page, existing articles and article talk pages. My 20th and 21st edits created redirects. I finally created a new article that wasn't a redirect on my 30th edit. That was 5 1/2 years ago, when there were a lot fewer articles in enWP. -- Donald Albury 20:53, 4 April 2011 (UTC)
 * My first edit was to an article. However three of my first ten edits (which is what should be looked at) were to create redirects. Redirects would also be affected by this proposal.
 * I've looked at the first edits of arbitrators, and at least three have created an article as their first edit, NYB created Arthur Yager, Roger Davis created Patrick Balfour, 3rd Baron Kinross, and Shell Kinney created New York Women's House of Detention (after a sandbox edit). Cenarium (talk) 21:29, 4 April 2011 (UTC)
 * I made 14 edits to existing articles before creating the article ACE Cougar, but since those 14 edits occurred in only two days that would still have been blocked by the proposed change. Interestingly enough, that was tagged for CSD a few hours later, then prodded, and later AfD'd, but it's still with us. Alzarian16 (talk) 21:49, 4 April 2011 (UTC)
 * My first edit was this, the type of copyediting-gnome edit that still typifies me. I didn't create my first article/redirect/from-scratch anything until December 2010, when I had been around for almost three years. It took me about two weeks after I joined to hit the ten-edit autoconfirm limit. A fluffernutter is a sandwich! (talk) 22:31, 4 April 2011 (UTC)
 * My first edit was creating an AfD on a hoax that had been bothering me for a few months. Given the hoop-jumping and instruction reading that required and how motivated I was to make it, I'm fairly certain that I could have entertained myself for some confirmation requirement (whether time or edit count or both). --Danger (talk) 22:44, 4 April 2011 (UTC)
 * My first edit was . My first two mainspace edits was and  respectively. --Guerillero &#124;  My Talk  &#124; Review Me  22:51, 4 April 2011 (UTC)
 * Mine was removing a redlink, because I thought it didn't look good. I quickly moved on to removing the word "perished" from a lot of articles, but my first few edits were removing redlinks.  That was obviously before I knew of WP:REDLINK.  I didn't create my first (and only, to date) page until late January, almost 11 months after I joined (you can see it here if you'd like). The Blade of the Northern Lights  ( 話して下さい ) 23:20, 4 April 2011 (UTC)
 * A copyedit, at least my first reg'd edit. Looks like the first article creation was thirty-some edits in. I probably had some IP edits beforehand, but being as we're talking 2005, it's all lost in the mists of memory. --joe deckertalk to me 23:23, 4 April 2011 (UTC)
 * I created my account in August 2006, and my first registered edit was to answer a question on a talk page in March 2007. I probably corrected spelling errors and made similar minor edits a few times before registering, but I have no recollection of what pages that might have been on—although article pages seem more likely than not.  WhatamIdoing (talk) 01:50, 5 April 2011 (UTC)
 * I registered in Nov 2005 but did not edit until Jan 2008 (can't recall exactly why I waited so long). I was intimidated at first, and my first edits were primarily in Talk space as I began to get comfortable with editing, and I moved to RD/S when I discovered the RefDesks.  I have created close to zero articles.  -- Scray (talk) 03:25, 5 April 2011 (UTC)
 * My first  edit  as a registered user was to  correct  a very  minor spelling  error, and I made one of the biggest  newbie blunders of all - I signed the edit  on  the article space (but  nobody  complained, and I  realosed my  error very  quickly). This proved two  things: Our istructions are not  clear enough (or weren't at that time), and that we should be very  forgiving to serious, good faith new editors. Kudpung กุดผึ้ง (talk) 03:35, 5 April 2011 (UTC)
 * My first edit was this expansion of an article about my high school. As an IP editor I had made several similar edits to the same article, and to articles about my home town.  For the record, the first article I created was the creation of this redirect, which occured 2 months and some 70 edits into my career.  The first non-redirect article I created (that is, one with actual content) was this article, some 5 months and about 1200 edits in my career.  In otherwords, the current proposed changes would have had zero effect on me.  My first interest was in improving existing articles, I didn't even think of creating a new article for several months.  -- Jayron  32  03:55, 5 April 2011 (UTC)
 * My first edit after registering was to a talk page, pointing out an error in an illustration (and I see now that I didn't sign it—oh, the shame!) . I created my first unsourced stub 28 days and 68 edits later, and sweated bullets while anxiously checking my watchlist to see if I'd screwed up and my article had been deleted. It wasn't deleted (although it still needs better sourcing) but if it had been, I would have tried again, not left in a huff. Prior to all that, as an unregistered nobody, I had fixed misspellings and the like and surfed the site avidly, fascinated by the structure and the culture and the fact that something momentous was being formed out of nothing. Rivertorch (talk) 05:08, 5 April 2011 (UTC)
 * Apparently not many have done this, but I actually created Akin Ayodele with my first edit, back in March 2006. I was a complete Myspacer then, but I returned in 2008 and started actually writing articles. Ed [talk] [majestic titan] 06:45, 5 April 2011 (UTC)
 * My first edit was this:, since deleted. I do not believe I have ever started a new article on any of "my" usernames.  Obviously, I have edited work wikis and such.  - Philippe  11:54, 5 April 2011 (UTC)
 * Expanding a video game article (gamecruft mostly), no small edits before. — HELL KNOWZ  ▎TALK 12:48, 5 April 2011 (UTC)
 * Believe it or not, my first edit was a vote. I supported the move of 136199 Eris → Eris (dwarf planet). My account was one month old then, during which I was trying to figure out the weird syntax, and the maze of help pages, with little success. -- Orionist ★ talk  13:22, 5 April 2011 (UTC)
 * My first edit with my old account was adding "digital photography" to the article "Photograph". But that was back in 2002. Promptly thereafter, my first new article was Walker Evans. —Tom Morris (talk) 13:30, 5 April 2011 (UTC)
 * I had over 2000 edits before I created an article (public economics, late Jan 2005) about a month after registering. Rd232 talk 16:35, 5 April 2011 (UTC)
 * My first couple dozen edits (as an IP) were obvious typo fixes. When I decided to actually do some copy-editing, I registered.  The first article I actually did was actually done in my user space, and moved out when another editor noticed it, and suggested to go live.  By that time I was probably several hundred edits in.  No comment on current PC discussion underway (but I'll keep reading. :)) — Ched :  ?  06:48, 7 April 2011 (UTC)
 * My first edit was a disambiguation fix in Modigliani–Miller theorem (still present in the article 6 years later). I created an article as my sixth edit, about an hour after my first edit. That article still exists, and most of the original content is still present in the lede. cmadler (talk) 13:09, 7 April 2011 (UTC)
 * My first edit was this talkpage comment on Talk:David Barton (author), my first mainspace edit, was this tag on David Barton (author) the next day. HrafnTalkStalk(P) 13:33, 7 April 2011 (UTC)
 * My first edit was suggesting an entry in a disambiguation page: . Some 5000 edits later, I still have never created a non-redirect article in my career, and don't intend to in the near future. --M4gnum0n (talk) 16:57, 7 April 2011 (UTC)
 * As a registered user my first edit was an opinion on an AfD for one of Colbert's Wørds (in this case, Wikilobbying, which now redirects to Colbert's page, where the same word links to exactly what I said it should link to - whee, the system works!). Both before and after that I made lots and lots of tiny little spelling/cleanup edits as an IP.  ☯.Zen  Swashbuckler  .☠  20:35, 7 April 2011 (UTC)
 * In dispute of the "conclusion" reached below: If the article hadn't been created and pending deletion already, I would have created it. Assuming that because many of the people who are commenting here did not create an article as their first edit, that therefore the vast majority of IPs and unregistered users who try to create articles will do such an overwhelmingly crappy job that we should take away their power to do so, is dubious at best. ☯.Zen  Swashbuckler  .☠  16:31, 11 April 2011 (UTC)
 * When I started, unregistered users could still create articles. This was my first edit.  The experience of article creation definitely was one of the main things that led me to register and get involved.  Chick Bowen 21:58, 7 April 2011 (UTC)
 * Added some external links, bits of text, etc., didn't create a page until 6 months later.  Ron h jones (Talk) 00:32, 8 April 2011 (UTC)
 * My first edit was telling an IP that they are in the wrong place to talk about spiders Barts1a | Talk to me | Yell at me 01:26, 8 April 2011 (UTC)
 * My first edit shows as this, but I knowthat was not my edit. I usurped this username fairly recently, I think that is the old user's edit. The first edit I can claim as mine (there were probably a few spelling corrections as an IP first, and discounting sandbox edits) is this one. I remember making it and am pretty proud of it. I looked through my old edits (not that old compared to some of yours) and was happy to see that most of them seem constructive and still exist in some form or another. Cliff (talk) 03:30, 8 April 2011 (UTC)
 * My first edit, over 6 years ago, was cleaning up an article on an inventor and correcting misinformation. My second edit was 11 months later. I had done about 500 edits over a period of 16 months before I created my first article, Frederick de Cordova. Reasons for not creating an article sooner: 1) The formatting to put in references seemed very complicated, and it was hard to find the pages which explained how to do it, so mostly I copied what I found in existing articles as to reference formatting. 2) I had seen countless new articles by others deleted, either speedily or through AFD, because of lack of references. 3)It was not clear at the time how to create a personal sandbox for developing an article. Thus I felt the need to create a substantial and referenced draft, with refs to document a claim of notability, before ever hitting "save," with the result that a couple of drafts got most of the contents erased before they were saved. Showing newbies how to create a draft of their first article in their personal sandbox would be a great service. Edison (talk) 15:35, 8 April 2011 (UTC)
 * My was undoing some vandalism on Semi Automatic Ground Environment. - 220.101 User talk:220.101.28.25\ 23:38, 9 April 2011 (UTC)

Conclusions from our edits
90% of us fairly dedicated editors did not start off our careers by creating articles. Or at least, we first made or could have made 10 edits fairly easily before creating an article. -- Nick Penguin ( contribs ) 06:06, 9 April 2011 (UTC)
 * Interesting, especially as older users are logically more likely to have started off creating articles than ones joining in 2011, seeing as back then there were was much more low-hanging fruit available. I mean, in 2005, my first creation was public economics - a major branch of economics which didn't have an article! Rd232 talk 01:02, 10 April 2011 (UTC)

Please don't draw any 'conclusion' from such an unrepresentative sample, because it doesn't mean anything. Some more meaningful results are at User talk:Mr.Z-man/newusers. Also, it's whether a mainspace page (article, redirect) has been created among the first ten edits before 4 days of registration that should be looked at, not just the first edit, since users would be prevented from creating until autoconfirmed. Cenarium (talk) 20:32, 10 April 2011 (UTC)


 * If your goal is to identify patterns among editors who are highly active in the community (e.g., not merely making a hundred edits to the article about my garage band), then this group is approximately as representative as you could get. WhatamIdoing (talk) 18:50, 11 April 2011 (UTC)

Orionist's view
We have considered all of those options as alternatives (at least, I have), and they simply aren't enough and/or feasible. First of all, getting more NPPers seems to be an impossible task. I've been doing it since June, and I can honestly say I've only regularly encountered about 9 other people doing NPP. It's a job, similar to those who slog through images, that very few people want to do; I've laid those reasons out in my view. Furthermore, there seems to be a very high burnout rate; Kamkek left after a couple of months. As to making a new Article Wizard; yes, that would be a good idea. That would considerably help, but in and of itself it won't stop people. And to throw everyone at AfC under the bus shows you've never participated there or even observed it; things like this (sorry, admins only; it was an AfC submission about a U-12 soccer team in Burlington, Canada) do not need anything more than a generic "declined" tag. And any sort of grace period isn't particularly helpful either; should we have waited any length of to tag/delete the above example when the creator made it a mainspace article? There was no possible way that could have ever become a useful article, and allowing it to sit wouldn't have changed that. It's not any nicer to have your article deleted 12 hours later, and will lead to a dramatic uptick in spam; companies and people will figure out that they can post something for [insert length of period here] to improve their search engine rankings. I would, however, support trying to shorten the talkpage notifications we give editors when an article has been tagged; the helpful information is being somewhat obscured. The Blade of the Northern Lights ( 話して下さい ) 15:49, 5 April 2011 (UTC)
 * Would some semblance of personal notification not be helpful? And you seem to be missing something; if newbies are approached with a better attitude and more encouraged to stick around, the chances are we're going to have a bigger user pool - which means more potential NPPers. Ironholds (talk) 15:58, 5 April 2011 (UTC)
 * Personal notification would be very helpful; the way NPP is right now, there's no way we could possibly leave a personal message for everyone. I think that would help; I also don't consider the two ideas mutually exclusive.  I have conditioned my support on having a trial run first  to see what happens; unlike the ongoing PC drama, this should be much simpler to turn off or on. The Blade of the Northern Lights  ( 話して下さい ) 16:05, 5 April 2011 (UTC)
 * Approaching newbies with a better attitude and encouraging them to stick around is not mutually exclusive with requiring a certain amount of effort to create articles. If anything, I think having fewer U-12 soccer team advocates, CEOs of minor startups, and garage band drummers to deal with will wear less on the patience of NPPers and everyone else who deals with new editors. I'm far less likely to be forgiving if I've already written calm, personalized, and nicely worded messages to 40 people today (4 hours of work if I'm going really fast) and received poorly spelled invective as a result of 5 of those. We are human, not robots and patience is not infinite. Danger (talk) 16:37, 5 April 2011 (UTC)
 * This is true, but the changes do not just hit CEOs of minor startups. Ironholds (talk) 16:49, 5 April 2011 (UTC)
 * No, they don't. They hit those who don't have enough interest in Wikipedia to find and correct ten spelling errors before creating an article, including the CEOs of minor startups who, after an hour or two poking around, may discover that Wikipedia is neither an easy nor appropriate method of advertisement. (As I've said on the project page proper, I don't believe there should a time component here. I think a new user flag dependent only on number of edits should be created for this.) --Danger (talk) 16:57, 5 April 2011 (UTC)
 * Without a time element you encourage people to rack up junk edits in minutes - the worst type of editcountitis. There's a reason autoconfirmed isn't just 10 edits but also 4 days. Also, please recall that making an article without 10 edits/4 days would remain possible, via AFC and possibly Article Wizard. It's not like we're telling non-autoconfirmed users who want to make articles "lalala we're not listening come back when you're autoconfirmed". Rd232 talk 17:09, 5 April 2011 (UTC)
 * Huh. Good point. I didn't think of it that way. As long as there's an obvious link to AFC or the Article Wizard in whatever error message new users get, then okay. I've struck my above. Danger (talk) 17:15, 5 April 2011 (UTC)
 * Has anyone here actually logged out and tried AfC? It's annoying as hell to go through, and I suspect newbies are annoyed and confused while attempting it. Ed [talk] [majestic titan] 21:52, 5 April 2011 (UTC)
 * The group that would be affected by this proposal consists of editors who have registered an account, but have not yet been autoconfirmed. It would not change the experienced of editors who are not logged in. So, are you speaking to the experience of someone who IS NOT logged in, or of the experience of someone who IS logged in? As an aside, I notice that only a little more than 1/3 of the submissions at AfC are accepted. -- Donald Albury 00:58, 6 April 2011 (UTC)
 * Yes, you can't look at how annoying AFC is for experienced users compared to doing what they normally do. You need to get some newbies and try some on AFC and some unassisted and see what their experiences are. I'm sure AFC is improvable (what isn't?). Rd232 talk 15:22, 6 April 2011 (UTC)

I'll try here to reply to the ensemble of the above comments:

I understand that recruiting is not easy, maintenance tasks are never everyone's cup of tea, but some projects seem successful in getting people excited about them, and to great results. It may be time NPPers approached recruitment in a consistent and persistent manner. Because whatever you do you can't run away from the main problem: we need more NPPers. This proposal is only targeting the symptoms of this problem without solving it, and as I detailed in my view, it's going to cause us more problems. Again, what if the numbers of NPPers go down, or if this proposal is applied and it proves to be a minor hurdle, what are we going to do then? Raise the threshold to another arbitrary number? The only sustainable solution is recruitment.

No one is doubting that NPP and AfC are doing a great job in blocking bad articles, it's just that at times they seem rather indiscriminate, and they're throwing the baby with the bathwater. Many good articles are getting blocked. My comment on AfC as being not helpful (sorry if it seemed harsh) came from my observation: the reviewer just accepts or declines an article and, if declining, leaves a short note that newbies have to read several very long policies, guidelines, and help pages just to decipher its meaning. A couple of days ago I went to the declined articles category, and just after a quick scan of the article names at first page, I could recognize 7-8 that should've been accepted. The way AfC works now is like a seemingly stricter version of NPP for unregistered users, although it has the potential of being a great place for editors to get help starting their first articles, but that's another discussion altogether (one that I'd be happy to take part in.)

This proposal won't stop U-12 soccer team advocates or garage band drummers. These are people who have (by definition) too much time in their hands. It won't stop CEOs of minor startups either, as they would do anything to get their page up there: it's good, cheap, has high visibility, and props up their websites SEO rankings. So getting Autoconfirmed would be a minor hindrance for them. And again, why would we want these people to do ten more edits, instead of containing their damage in the new pages area? If you're complaining about their lack of spelling skills, why would you want them to correct ten spelling mistakes?

The only people who are going to be put off by this are knowledgeable, genuinely interested people. These are the volunteers: they have nothing to gain from contributing but self satisfaction, and if that proves to be much trouble for them, they just quit. Here, they have to find their way around baffling wiki syntax, cryptic help pages, and perplexing tangle of policies, and after all that we want to treat them as if they are joining a fraternity? That will only lose us the good, while the bad and the ugly will keep coming back again and again. Regards, -- Orionist ★ talk  07:29, 7 April 2011 (UTC)


 * Funny, I think the opposite. Genuinely knowledgeable and interested people will get used to editing and be less frustrated when it finally comes time to make an article. We're asking them to make 10 ordinary edits... experience you'll need if you want to create a successful article. But rather than taking my word for it (or your word) we should actually put this to a trial or test. There's a consensus to move forward with this proposal. But it doesn't mean we should move forward blindly and recklessly. Shooterwalker (talk) 12:54, 7 April 2011 (UTC)


 * Actually that statement you picked from my long comment, came from my first hand experience in introducing Wikipedia to very interested people, and the results were not that funny. Even if you cross that out, we still have a multitude of issues to solve.


 * In fact, we've already done a trial very similar to the proposal, actually, it's still on, ever since December 2005, when we required registered status in order to create articles, and sent everyone else to AfC, sounds familiar? While that happened for a totally different (and much more convincing) reason, you can still see that it didn't prevent the situation we're in today, and doing the same thing all over again, won't prevent the same situation from recurring in the future, and then we'll have people advocating doing it again and again untill we become a closed club and kill the project we want to protect. That's one of the main concerns I raised, and it's only one amongst many.


 * While the majority seems to support the proposal, that doesn't mean we have consensus. Although I hope that we come out of this with at least some kind of partial agreement. -- Orionist ★ talk  16:53, 7 April 2011 (UTC)
 * Are you a Psychic able to see alternative timelines? No? Than please don't make arguments such as "it didn't prevent the situation we're in today". By definition nothing in the past can ever prevent our current situation, else it would not be our current situation. Yoenit (talk) 17:09, 7 April 2011 (UTC)
 * No need to flare up about the phrasing, I'll put it up differently: If you do the same things you did in the past you'll get the same results. It doesn't need a psychic nor a rocket scientist to tell you that. To apply to our case: raising the threshold for new page creation will have limited effect if the number of NPPers is too small. That's true now and it will be as true in the future. And the only sustainable solution is to work constantly to recruit more NPPers. There's no escaping that. -- Orionist ★ talk  17:53, 7 April 2011 (UTC)
 * Have you ever done NPP before? I've been at it since June, and it'd be nice every once in a while to be able to slow down.  Unless you have a way of getting new NPPers, that solution won't work; people don't want to do NPP.  I can go into detail if you like, but it's the kind of job that doesn't curry much favor with anyone, for reasons I gave in my view.  Furthermore, it won't stop the perceived bitiness; just having more people tell you your article will be deleted won't make a difference. The Blade of the Northern Lights  ( 話して下さい ) 17:59, 7 April 2011 (UTC)
 * Orionist, we should get rid of police too. Can you believe that they created a police force in the 1800s, and over the past 200 years crimes have only increased? What's next, more police? We also created all these environmental protections that don't seem to be working, because the pollution just keeps increasing. The point of my sarcasm isn't to be disrespectful... but to point out that most people agree that limiting anonymous users from creating articles has been a GOOD thing and the problem would be WORSE if we hadn't done so. There's an overwhelming consensus of... what... 80% of editors who want to try something new? The key word being "try". Instead of just assuming what the impact will be... we can take a look at new user retention now and look again at new user retention through the trial and figure out what the impact is. Just as we can look at the impact on new article creation, and new article retention (for those that actually meet policy). Shooterwalker (talk) 00:23, 8 April 2011 (UTC)

Make a watchlist notice for this RFC
Should we seek wider community input for this RFC by requesting a watchlist notification? Yoenit (talk) 15:06, 6 April 2011 (UTC)
 * I would second that. This is an issue that potentially has impact on the entire community; it would be best to encourage the community to weigh in. A fluffernutter is a sandwich! (talk) 15:13, 6 April 2011 (UTC)
 * Consider it thirded, for the same reasons as Fluffernutter. The Blade of the Northern Lights  ( 話して下さい ) 15:16, 6 April 2011 (UTC)

Proposed at MediaWiki talk:Watchlist-details. Rd232 talk 15:34, 6 April 2011 (UTC)
 * ✅ . Amalthea  11:15, 7 April 2011 (UTC)

Actual numbers?
Does anyone have actual numbers about this issue? Not anecdotes (like the section above), but actual data, such as the following:
 * During the last month (March 2011), how many articles were created by non-autoconfirmed editors?
 * By percentage, how many of those articles were
 * Speedy deleted?
 * PROD'd resulting in deletion?
 * AfD'd resulting in deletion?
 * PROD'd, not AfD'd, and retained?
 * AfD'd and retained?
 * Retained with no deletion nomination at all?

I think these numbers would give us a more data-driven view of this issue. cmadler (talk) 12:56, 7 April 2011 (UTC)
 * Besides a little experiment I did (user:Yoenit/CSD research) I am not aware of any experiment into the method of deletion. I tried to get some larger scale experiment of the ground, but got stuck because it would require a custom made bot. One thing I do know for sure is that the vast majority of these articles get deleted with speedy deletion under A7, G11 and/or G12. I would argue that the method of deletion is actually not really relevant as all methods are wp:BITEy, perhaps wp:AFD the most as it gives the new user the idea that he has a chance of protecting his article, only to be ridiculed by veteran editors and overwhelmed with policy based reasons he does not understand. Yoenit (talk) 13:28, 7 April 2011 (UTC)


 * Then, do we at least have numbers on how many articles are created by non-autoconfirmed editors, and what percentage of those are deleted within a month? It seems to me that without at least that much information, all we have to go on are editors' personal feelings as to whether and how much of a problem with is. cmadler (talk) 14:07, 7 April 2011 (UTC)
 * Yes we do, for februari 2010 at User:Mr.Z-man/newusers. Some 12500 articles, of which more than 10000 deleted. Yoenit (talk) 14:11, 7 April 2011 (UTC)


 * Does anyone know how we could get more details on this? Specifically, can we get some idea of the proportions of CSD/PROD/AFD routes to deletion, and could we get some idea of the typical age of the article at the time of deletion?
 * I realize that collecting this data is requires some real work, but if someone knows how to do it, please let me know. WhatamIdoing (talk) 16:56, 7 April 2011 (UTC)


 * Very telling numbers. Has anyone noticed that retention for those who start by creating articles that are kept is 70% more than those who started by editing an existing article that was kept (it's 4.4:2.6)? So judging from these numbers, this proposal is set to reduce retention: we will lose more editors. Regards, -- Orionist ★ talk  17:19, 7 April 2011 (UTC)
 * That merely backs up the quip Mark Twain popularized and a famous book discussing the same concept .  The percentage of editors who start by making a new article is 25%.  Down to 5% overall when we get to those who create an article that's kept.  Then 70% of those means 3.5%.  That's far outweighed by the number of people who leave after their article is deleted; if we can get even a tenth of those people to stay by making them edit existing articles (which looks feasible), then we'd be in the black by plenty.  The Blade of the Northern Lights  ( 話して下さい ) 17:46, 7 April 2011 (UTC)
 * Many of the new editors who leave after their first article was deleted or their first edit was reverted are either pure vandals or conflict of interest editors creating articles to promote their company, their friend, their nonnotable garage band, or their original research scientific theory. They are not all good-faith editors who went off in a huff because their sensible article on a valid topic was deleted, and who otherwise would have been of great value to the project. Spend an hour looking at all "Recent changes-New editor contributions" and you will see what I mean. Edison (talk) 15:50, 8 April 2011 (UTC)


 * Yes, but no: So few editors fall into that category, and so many into the other category, that the "70% increase" produces only 11% as many actual editors.  A small increase in a common outcome is more important than a major increase in a rare outcome; relative risks are frequently misleading.
 * Also, "creates a decent article on a notable subject" isn't something the community can control. The things within our control are "first edit can create an article" or "first edit cannot create an article".   WhatamIdoing (talk) 17:43, 7 April 2011 (UTC)

View from User Jayron32
moved from the main page. Yoenit (talk) 13:51, 7 April 2011 (UTC)
 * Comment Awful. The majority of new users would give up immediately if they could not create. It definitely would stop most people from bothering contributing. And what is the point of making them wait 4 days and 10 edits? How exactly would this help wikipedia?? Silliness. The whole point of wikipedia is that we accept new content from newbies, especially from non-english parts of the world and the good editors here nurture the content, verify it and expand it. This is how wikipedia grows. The new non notable content will always be clear whether is newbie on their first day or their fifth day. You certainly can't assess editorial abilities in four days and 10 edits. Trust and abilities in individual are learned over many months and years. Blatant vandalism and attacks are always caught anyway and speedied. A decent editor will always be a decent editor and they need to learn the ropes as fast as possible. In fact they ar elikely to learn sooner if they create a new badly formatted article on their first day and an established editor wikifies it and they can follow it.♦ Dr. Blofeld  13:39, 7 April 2011 (UTC)
 * Marvellously hyperbolic and unsupported by any data. We need precisely to nurture new editors, and we don't do that by pretending they don't need any assistance in learning the tricky task of creating a new article that meets minimum standards and then angrily cleaning up after them when they make a mess. Rd232 talk 15:35, 7 April 2011 (UTC)

As one editor put it "The only new users willing to commit themselves to the extra effort will be the determined, ideological POV-pushers". That is exactly what will happen. Given that new visitors or not paid or compelled to contribute if they bother to create an account they should at least be able to have the chance to contribute what they want.♦ Dr. Blofeld  16:24, 7 April 2011 (UTC)


 * Even if "what they want" turns out to be non-compliant with wikipedia policy and is immediately deleted because of this? Which improves their chances of sticking around how? Like Rd232, I'm getting rather sick of the argument by assertion -- just because "one editor put[s] it", doesn't make it true. HrafnTalkStalk(P) 16:31, 7 April 2011 (UTC)

This editor is exactly why it is a lame idea. Some newbies are highly intelligent and capable of constructing some worthwhile and useful immediately. It is only likely to be the POV pushers or advertisers who would really feel compelled to make ten edits and wait days anyway.♦ Dr. Blofeld  16:33, 7 April 2011 (UTC)
 * More unsupported assertion, and also fundamentally illogical; why should the "good editors" we want who are willing to put in the effort to write something non-deletable be put off by not being able to put it live immediately without assistance? The fraction of new editors (as opposed to new accounts) who can create articles immediately without assistance is tiny. And as explained in my View, the proposal should include mechanisms to ensure that such users are not put off, by having the option of immediate creation (via Wizard), or the option of getting feedback (via AFC), or the option of userspace draft + request to move. The key thing is not actually immediate creation, it is making clear expectations of standards and providing feedback; very few things are bitier than having something deleted that you the wrote; if we want to up the numbers of substantial contributors, we need to stop biting them and start helping them. It's broke, so we should fix it. Rd232 talk 16:48, 7 April 2011 (UTC)


 * About the "majority of editors" claim: Three-quarters of new editors do not create an article immediately, so this is simply false.  Whether a minority of editors might give up immediately, or whether a majority of the one-quarter (i.e., as few as 13%) might give up immediately is unknown, but we do know that three-quarters of new editors will be unaffected by this.  WhatamIdoing (talk) 16:59, 7 April 2011 (UTC)
 * Dr. Blofeld, the editor you gave as an example sort of argues against your point. You chose an editor whose first contributions seem to have been creating articles on specific brands of a product created by a company, and whose edits consist of adding that company to other articles and adding trademark symbols, as an example of someone who is the opposite of a "POV pushers or advertiser"?--Yaksar (let's chat) 22:52, 8 April 2011 (UTC)

I mean the majority of new editors who want to create an article and which was the reason they created their account.♦ Dr. Blofeld  17:04, 7 April 2011 (UTC)


 * Too bad. An encyclopedia shouldn't be for the "I have a cool idea so I'm gonna write about it" crowd.  That's what blogs are for.  I'd rather see a WWII buff come in and edit existing articles in the subject area for a bit, then if he/she sees something or thinks of something in the topic that doesn't have an article yet, they're better prepared to create a quality one. Tarc (talk) 17:24, 7 April 2011 (UTC)
 * Perhaps Dr Blofeld believes that writing articles about cool topics would be just fine, or at least better than the alternatives. After all, the English Wikipedia has taken that basic approach since the beginning, and we haven't died yet.  It seems to me that the approach you reject here worked very well in the early years, and that it's working maybe sorta okay now (at the cost of huge numbers of newbies getting smacked around for being imperfect or having a bad idea for an article), but this is fundamentally a matter of opinion, and reasonable people are entitled to hold different opinions on that point.  WhatamIdoing (talk) 18:15, 7 April 2011 (UTC)
 * It worked at the beginning, but the older WP gets, the less it works. Organizations which don't adapt ossify and die. Rd232 talk 19:40, 7 April 2011 (UTC)

Sadly I'd have the agree that the majority of newbies don't produce anything decent. And I don't mean formatting or reference problems. I mean subject choice. Most newbies I see producing new articles tend to be either childish personal attacks on peers, pure vanity COI, crappy Myspace bands or fictional sub sub characters. Probably less than 20% actually create anything remotely encyclopedic or with potential. I'm not sure though by them waiting 4 days and making 10 edits is suddenly going to change their way of thinking... Its not about cool idea or cool topics. Newbies from developing world countries have the potential to identify missing notable topics, sadly few of them really do.♦ Dr. Blofeld  22:10, 7 April 2011 (UTC)


 * There's no one-size-fits-all solution. I've fought in the past to help new editors keep an article in the face of repeated deletion when it seemed clear to me that the subject was notable, and I'm always particularly careful if it seems that the editor's first language isn't English. But the raw facts are as you say; most articles written by new editors are not worth the bytes to store them. Malleus Fatuorum 04:48, 8 April 2011 (UTC)
 * I'd add that in every case I can remember the new editor never edited again, or even bothered to thank me for my efforts. It's my very definite impression that most new editors regard their wikipedia article as a one-off. Malleus Fatuorum 05:15, 8 April 2011 (UTC)
 * I'll agree with this second part completely. Perhaps the techo-utopian promise of wikipedia is a bit to blame.  We tend to advertise the editing process on wikipedia as characterized by collegial improvement and collaboration where the actual median number of editors making substantive improvements to an article is very close to 1. Protonk (talk) 20:19, 8 April 2011 (UTC)

Mmm. I do agree that the argument "This will help make the encyclopedia a bit less of a target for trash edits." is a valid one. I hate seeing the stupid autobiographies, attacks and not notable cruft appearing from newbies too. But some new editors do produce something useful. Maybe if it is really important to them they will stick around and wait the 4 days. But it also potentially could be damaging in that people refuse to bother who might have something encyclopedic to add. Does this poll though have any chance of taking effect? I mentioned it to Jimbo and he said that a trial for anything is the best way forward rather than a final decision made straight away.♦ Dr. Blofeld  17:09, 10 April 2011 (UTC)

View from Kudpung
See: User:Snottywong/Patrollers. It offers great  insight  into  the profile of the average and not  so  average NPPer. I'll leave the community  to  decide: In spite of the extensive work and research that  some experienced editors have put  into  the problem of NPP, I feel the time is right  to  decide: Either clean up  NPP as a task, or install any  of the proposed methods for preventing  new articles from  going  live immediately. No article is so  urgent  it  must  go  online immediately; no  serious creator is going  to  be put  off by  waiting  24 hours or even longer. A decent corps of patrollers could review all  these new articles and do  it  without  the incredible speed and superficiality  that  appears to be customary. Kudpung กุดผึ้ง (talk) 04:13, 8 April 2011 (UTC) Kudpung กุดผึ้ง (talk) 04:13, 8 April 2011 (UTC)
 * Whether we have enough NPPers.
 * Whether NPPers have read WP:NPP before they start patrolling.
 * Whether NPPers have read WP:DELETION and WP:CSD before they  start  patrolling.
 * Whether NPPers are drawn from the ranks of experienced users, or whether they are rank beginners themselves.
 * Why so many  newbies and VYE want  to  be page patrollers.
 * Whether New Page Patroller should be a right that  should be applied for.
 * Whether establishing NPP as a 'right' would lend more kudos tho  the job, and consequently raise the quality  of it.


 * Interesting stats. All I can say is that if NPP became another "right", then it would be another "right" that I would refuse to have. Malleus Fatuorum 04:39, 8 April 2011 (UTC)
 * I oppose the idea of patroller as yet another right. The existing rights are already sometimes used as a kind of editor hierarchy, and this would only make it worse. Also, it would imply some kind of individual authority to judge the worth of particular pages, giving undue weight to what should be a fast and informal process. Feezo <FONT SIZE="-2">(send a signal | watch the sky)</FONT> 05:50, 8 April 2011 (UTC)
 * Wouldn't that be the same as asking permission to contribute to wikipedia and asking somebody to authorise it who probably knows less about the article than you do. Bad idea.. People should be free to create whatever they feel like, if its not notable they are usually caught anyway♦ Dr. Blofeld  15:23, 24 April 2011 (UTC)
 * I've occasionally seen people say that only people who have the WP:Autopatrolled userright ought to be able to do NPP work. I think this is a bad idea.
 * Basically, the purpose of NPP is to get a second set of eyes on any new page. Not "a highly skilled set of eyes" or "a carefully trained set of eyes", or "a very trusted set of eyes", but any second person, because any second person is highly likely to object to a page whose contents is outright vandalism ("I love cheeseburgers!!!") or blatant libel ("John has AIDS!!!!").  That, rather than spamming clean up tags into brand-new articles, is our primary goal in creating Special:NewPages.  Pretty much anybody can do it, just like pretty much anybody can remove such vandalism and libel when added into an existing article.
 * The thought behind the userright is "We trust that whatever page you create will be good enough that we don't need any second person to look it over." It's not necessary for us to trust a person to create good articles "sight unseen", to trust them enough to be the second person for a complete stranger's article.  WhatamIdoing (talk) 15:58, 8 April 2011 (UTC)

Categorisation of views
I hope this isn't too controversial; I've categorised views as best I could, to reduce the WP:TLDR effect. That is important partly because the effect encourages the further proliferation of pretty duplicative views (naming no names... but there are some where I was tempted to ask the author to remove them), creating a vicious circle. Rd232 talk 14:48, 8 April 2011 (UTC)


 * Good move. That's far more usable. AGK  [</nowikI>&bull; ] 15:07, 8 April 2011 (UTC)
 * Controversial? It is just a bad faith attempt to influence the result by downplaying and silencing the opposing view. I am not going to consider the result of this so called "rfc" anything but illegitimate. Ruslik_ Zero 15:19, 8 April 2011 (UTC)
 * "bad faith" - lovely, from the same person who was so constructively engaged in the VP discussion. I actually considered not doing the categorisation because it makes it easy for the kneejerk opposers to skip the arguments in favour and sign a couple of opposing views. But I decided the risk of that was a price worth paying for more organisation/usability. Rd232 talk 15:41, 8 April 2011 (UTC)


 * Really, Ruslik? If the RFC closes as no consensus because of the fairly strong opposition, will you still consider the RFC illegitimate?
 * Or will you only consider the result illegitimate if you don't like the outcome? WhatamIdoing (talk) 16:11, 8 April 2011 (UTC)


 * I'm assuming good faith but seeing highly questionable judgment. Boldness is not necessarily a good thing on a widely watched RfC in project space. Before making such a major change to the structure of the page, it would have been entirely appropriate to seek consensus. You might well have found it—and quickly. More troubling, I think, is the order you placed the views within each category, which doesn't appear to correspond to the order that the views were added by their respective editors in the first place. I'd be curious to know what method you used to sort them. Rivertorch (talk) 16:36, 8 April 2011 (UTC)
 * On the boldness - perhaps. On the sort order - I was going for the order the views were added (within each category), but it's a bit tricky editing such a large page, and I know I didn't get it 100% right. I could have gone back to the pre-categorised page to check and move stuff about further to match, but I didn't think it worth the effort. I certainly have no objection to anyone else doing that though. Rd232 talk 16:53, 8 April 2011 (UTC)
 * Let me get this straight: it's tricky editing a large page, so you'll cheerfully leave others to clean up the mess that you created? With all respect, that beggars belief. Leaving aside general questions of the wisdom of reordering the page, if you choose to take on such an arduous chore then you're obligated to do it right. The reality of long pages like this, however they're structured, is that views nearer the top are more likely to be read in their entirety and thus more likely to be considered fully. For that reason, your sloppy rearrangement has the unintended effect of shifting the advantage away from those who put their oar in early in the game—and that calls into question the integrity of the whole RfC. Rivertorch (talk) 18:25, 8 April 2011 (UTC)
 * Erm, what? That's totally illogical. Not only does the categorisation much reduce the effect of chronology (first mover advantage), but any errors in maintaining that chronology reduce the effect further. Rd232 talk 18:47, 8 April 2011 (UTC)
 * I was less than clear; my apologies. I was bemoaning not the effect of chronology, which is a natural order, but the substitution of other, arbitrary effects which may be less easy to identify and which could create an appearance of tampering. (I do not think you were tampering, and I never did.) All a moot point perhaps now, but for future archaeologists. . . . Thanks for restoring it. Rivertorch (talk) 23:28, 8 April 2011 (UTC)
 * Thanks for your reply. Rd232 talk 11:02, 9 April 2011 (UTC)
 * I will assume good faith here too, but I would kindly request that Rd232 return the discussion to the state it was in. Ease of use is only one consideration in doing this sort of thing, and I think that the disruption to the discussion, by removing chronologization of the views, removes some vital information.  Many of the later views are in response to, or reference, views that were added earlier, and the current scheme destroys that.  Furthermore, even moves in boldness and in good faith should be made prudently; any large scale and unexpected change is likely to piss someone off, and that potential for pissedoffedness needs to be weighed against one users sense of aesthetics.  Please put the page back the way it was.  It doesn't matter much if you were trying to help, the results do not match the intent, so this needs to be put back.  -- Jayron  32  16:50, 8 April 2011 (UTC)
 * I don't think cross-referencing between views is enough of an issue to outweigh the helpfulness of the minimal degree of organisation introduced. Rd232 talk 16:53, 8 April 2011 (UTC)
 * I don't see this as helpful. You basically polarized the page by splitting the viewpoints into different factions. A change like this needed consensus. Although bold, your unilateral decision lacked wisdom. Before the categorization, participates were forced to find a viewpoint to endorse by reading through all of the viewpoints; now they could just endorse all of the viewpoints in one particular section and neglect to read the other sections. It'll just change how the closing admin(s) perceive the results. They'll could now only take a glance at the arguments, count the "for's" and "against's", and create a draft analysis based on that glance. Before then, the closing admin(s) would've been forced to search and read through the entire page thoroughly, count, double-count, and draft an analysis based on that thorough reading. I prefer the situation before the categorization. --Michaeldsuarez (talk) 17:10, 8 April 2011 (UTC)
 * "polarized the page by splitting the viewpoints into different factions"? The polarised views exist anyway, and there is a third section for views that don't definitely fall into either camp or which go substantially beyond merely arguing for/against. I'm not worried about a closing admin not doing their job; I think generally with RFCs of this size someone ends up doing it pretty well. And whilst a closing admin would surely read everything (they'll get stick if they don't do a good job, and their close will be heavily scrutinised), I'm concerned with editors contributing to the RFC not reading things carefully. The TLDR effect was starting to cause real problems. The RFC should really have been categorised this way from the beginning, but better late than never. PS I've long argued that RFCs of this type need collaborative views in support of a particular position in order to remain manageable and not lead to the proliferation of similar and overlapping views we see here again. Oddly enough, this wiki approach to RFCs has never gained much support when I suggested it. Rd232 talk 17:21, 8 April 2011 (UTC)
 * I personally think that this makes this RFC more readable. We passed the saturation point as far as different views on this subject back at view 16. Now at 23 this is the best way to make some meaning out of this --Guerillero &#124; My Talk  &#124; <font color="Dark Orange">Review Me  17:12, 8 April 2011 (UTC)


 * I'm assuming good faith but I don't think that was a good idea. Categorizing the views in support/oppose/other categories carries the inherent risk of drive-by !voting, i.e. people supporting views based on the category they are in, without actually reading them. Having a mixed bunch of views of both kinds means that such people are forced to actually read those views first and thus could possibly be convinced by another viewpoint than the one they already have. I know that in an ideal world people would first read all views and then !vote on them but I'm realistic enough to know that many won't. Regards  So Why  17:41, 8 April 2011 (UTC)
 * Yes, it's a concern. But the alternative seemed to increasingly be that people were adding new views instead of reading the existing ones and supporting as appropriate. You can't win - but at least this is less messy. Plus, I do think the reduction of the TLDR effect makes it more likely that people will at least look at the section with views they disagree with, rather than looking at the wall of text and saying teee ellll deeee arrrrrrrr where's the first view I agree with, can't find it, add my own.... Rd232 talk 17:47, 8 April 2011 (UTC)

Well, I've reverted it, in favour of a "collaborative views" attempt to summarise which would be much better if it worked, but will probably be ignored. Rd232 talk 19:56, 8 April 2011 (UTC)

Before anyone takes the choice of system too seriously, the level of support for smoothing the learning curve has been had a 75-80% consensus since the RFC started. There hasn't been much variation with new layouts, new opinions, new forums, and new notifications. Short of some kind of heroic WP:CANVASSing, I expect this to close a consensus to at least try this proposal... probably finding a careful way to try it out, gather data, and empirically evaluate if the dissent's concerns are real or imagined. Shooterwalker (talk) 23:38, 8 April 2011 (UTC)


 * On the meta-issue: Perhaps on larger RFCs like this we should encourage people to provide more descriptive section headings, like ==View by User:Example:  A simple solution== or ==View by User:Example:  Not quite==.  It would be easier to scan, and more memorable.  WhatamIdoing (talk) 00:05, 9 April 2011 (UTC)


 * So what happens now? Are we all supposed to go back and endorse the "collaborative view" that is closest to what we want and treat that as a straw poll? Or do we need to explicitly create a straw poll? I do think we need to wait a couple more days before taking another step. Even overnight (US time) we're still getting a couple of posts an hour, and I think it is important to give ample time for users to respond. I do not want to see this proposal derailed over claims of failure to follow process. -- Donald Albury 11:15, 9 April 2011 (UTC)
 * I don't think the collaborative views can be endorsed per se, since they're changeable. They're more intended to summarise the arguments and help people think about the issues. Now if the RFC were set up that way from the beginning, we could have a straight "Support"/"Oppose" list which the collaborative views would more clearly relate to, but doing that at this point wouldn't work because it would need 100+ people to come back and sign again (or else somebody to transpose signatures, which seems redundant). Rd232 talk 12:18, 9 April 2011 (UTC)
 * It's not a democracy. We don't base decisions on polling. We base it on discussion. What people agree on. The broad agreement is pretty obvious when you look at Jayron's proposal, and consider the overwhelming number of participants. Shooterwalker (talk) 13:19, 9 April 2011 (UTC)
 * " We don't base decisions on polling. We base it on discussion. What people agree on. " - and how do we find out what people agree on? Really, all we mean by "not a democracy" is that as far as possible decisions should be based on evidence. But that still leaves plenty of room for judgement, and which ever way you cut it, if you aggregate that judgement in a halfway sensible way, it's a form of democracy. Rd232 talk 13:27, 9 April 2011 (UTC)
 * I don't disagree. I was only trying to make the point that we don't need to organize the poll neatly around yay or nay. If anything that would just polarize the whole discussion. (Polling is evil.) We had a discussion, and there just happens to be one proposal that dominates the discussion, and a lot of other discussion around proposals that enhance OR moderate the proposal. That's better than flat support for "yay", because it shows people thinking critically and supporting it overall with various different caveats. Shooterwalker (talk) 13:38, 9 April 2011 (UTC)


 * "It's not a democracy. We don't base decisions on polling. We base it on discussion. What people agree on. [look at option A] and consider the overwhelming number of participants". Do they do an "unintentional irony" barnstar? Ironholds (talk) 19:03, 9 April 2011 (UTC)
 * I don't appreciate the sarcasm. I'm taking issue with those who are arguing about whether we need some kind of regimented polling system, when we should just read the discussion and look at what nearly all editors agree on. Shooterwalker (talk) 19:15, 9 April 2011 (UTC)
 * Except both are polling :P. Ironholds (talk) 19:49, 9 April 2011 (UTC)
 * I guess it depends on your definition of a poll. To me it's reading a discussion. Looking at the aggregate of comments across multiple proposals, on a spectrum from strong opposition to strong support, understanding compromises and outside of the box solutions, looking at the reasons for the opposition, looking at the caveats of the supporters, and looking for a consensus. But if you want to call that a poll... I guess it could be. It's just unlike any democratic system in human history. But now we're getting into metapedia and semantics. Shooterwalker (talk) 19:56, 9 April 2011 (UTC)
 * I see where you're coming from, and I do agree there is a key distinction between mere polling and serious public deliberation (that distinction being particularly that in the latter there is the expectation that people will examine the arguments and evidence and be open to changing their minds - the process does not merely reveal set opinions). However, both may be forms of democracy - see deliberative democracy. Rd232 talk 21:22, 9 April 2011 (UTC)


 * I see that put this section back in at the bottom. As it stands at the moment it is ridiculous there are huge amounts of space given over to supporting arguments and the vast majority of what has been said on the opposition side (even ignoring the old VP discussion just the stuff on this page) is left out. I'm going to try and spend some time distilling it down but welcome any help as well. As it's currently written it is incredibly biased to a level that I don't think it can possibly be left on the page (just look at the 'responses' section under the supporting views which is far more of a response to the responses section). I'm removing it again for now as I see no consensus above to have it in and the effect of it being there is,while well intentioned, inappropriate. For now I'm transcluding below though I'm not even really sure if it belongs here in it's current state. James of UR (talk) 05:41, 11 April 2011 (UTC)
 * Just a note. Moved the sections below and left a message for Rd232. If we want to include them (and maybe even if we don't) I think it would be great to work on them here instead of on the main page. James of UR (talk) 06:14, 11 April 2011 (UTC)
 * I've moved them back. It's been a while since I've heard something more ridiculous than applying WP:NPOV to views expressed in an RFC. As for length: (a) in general it is entirely expected that the "For" view for any proposal will be longer - the onus is on proponents to make a case for the change (b) in this specific instance, if the opponents haven't put as much effort into this collaborative view, so be it (there are a lot fewer of them...). Or perhaps their points are simply more concisely put; or they just have less to say. Rd232 talk 13:03, 11 April 2011 (UTC)
 * When you are doing it with an implied and stated intention of making it easier to follow the walls of text on the page I think you would totally expect it to generally be as balanced as possible. I'm not going to remove it again but I stand by what I said, it is in no way a collaborative view at this point and is just being used as another argument that leaves little room for rebuttal but pretends too. It's a little frustrating when you try and say that "maybe there are just less arguments" when you've been following this as long as you have since I assume you've read the opposition arguments and you know that is far from the case. I will try and pull stats and views into the opposition section (and the responses section of the proposals) but given the amount of work I have today it is unlikely it will happen until later tonight :( James of UR (talk) 21:28, 11 April 2011 (UTC)

Level of participation
Considering that this is a major structural change of proportion precedented only by the restriction of page creation to registered users, a particularly large level of participation seems needed to ensure that there is consensus of the community at large. I am not certain if people think that the RFC should be evaluated directly to interpret consensus or if a poll/discussion will follow. It seems clear that a large majority supports a restriction but there is no consensus as to how it should be implemented, trial or not. So it would make sense to make a proposal for a trial, with a poll/discussion, which would be advertised on the sitenotice for registered users. If however people think that the RFC should be evaluated directly (however difficult that might be), then we'd need to add it now to the sitenotice for registered users. Cenarium (talk) 14:53, 11 April 2011 (UTC)
 * The RFC is already on the sitenotice. It is prominently displayed on the watchlist notice, and has been for some time.  Regardless of where we go from here, this RFC is as prominently availible on Wikipedia (transcluded to the Village Pump, listed at Centralized Discussions, listed on the watchlist site notices in the banner at the top of watchlists).  Short of droping a personal invitiation to the usertalk page of every registerred user and IP address inviting them to participate, I am not sure there is much left to do.  I agree that the RFC is not ready to close and judge yet, and I do believe that this RFC represents a first step, and not a last step, in implementing this change.  However, I think all we need to do right now is a) let the RFC run for a little while longer to make sure that every editor who wishes has a chance to make their voice heard and b) brainstorm potential ways to implement the idea in a successful manner should the RFC ultimately decide that the community wishes to go ahead with the proposal.  This isn't ready to just "close now and flip the switch".  We have some work to do yet.  -- Jayron  32  14:59, 11 April 2011 (UTC)
 * By sitenotice I mean the MediaWiki:Sitenotice, not MediaWiki:Watchlist-details. If this is only the first step and the definitive proposal will come later, then that proposal should logically be added to the sitenotice. Cenarium (talk) 15:24, 11 April 2011 (UTC)
 * No, it should not. This is an important policy change, but we have never featured anything policy related on sitenotice before and I don't see a reason to start now. It will just annoy the hell out of people. Yoenit (talk) 19:56, 11 April 2011 (UTC)
 * We also never had a proposal for a structural change of this importance. It will provide the feedback necessary to legitimate the move. Cenarium (talk) 20:23, 11 April 2011 (UTC)
 * No, it belongs as a watchlist notice, and not a site notice, because it affects editors, not readers. The different notices exist for different reasons.  The vast majority (see Pareto principle) of people who use Wikipedia do it solely to read the articles, and have no interest in actually editing anything.  This proposal, while it represents a huge change for editors of Wikipedia, has next to zero effect for readers of Wikipedia.  This is one of those "laws and sausages" issues, and making this a site-wide notice will only serve to confuse the readers, who probably don't care much about internal Wikipedia politics.  -- Jayron  32  21:31, 11 April 2011 (UTC)
 * It certainly does not belong as an anonymous site notice (the spam would be bad) but it is incredibly easy to have a logged in only site notice and I assume that was what was intended. Those are generally editors, not just readers, and have every right to know about this discussion. There are an awful lot of people who do not read their watchlist frequently. James of UR (talk) 21:35, 11 April 2011 (UTC)
 * Indeed. The sitenotice is for registered users, and I had mentioned this above. It's the mediawiki:anonnotice that is for unregistered users. Cenarium (talk) 21:40, 11 April 2011 (UTC)
 * Ah, if this could be restricted to user accounts only, I could see the difference. I'm not sure I support that move entirely, but I oppose it much less now.  -- Jayron  32  21:45, 11 April 2011 (UTC)
 * Registered users only might be OK. Rd232 talk 00:08, 12 April 2011 (UTC)
 * I'm still not sure if the RFC is supposed to be final in determining if there is consensus for a trial. The page is also getting a little long, could not the proposal be discussed on another page ? Cenarium (talk) 11:25, 12 April 2011 (UTC)
 * I'd particularly support this as this discussion is the third message in the watchlist notice, it's not that likely that even people usually consulting their watchlists would note this. Cenarium (talk) 16:51, 12 April 2011 (UTC)

Article wizard extension
It seems that many users would support this, and this can be considered independently of any restriction on article creation. So as I understand it, this would be an extension which when a user gets to the create page screen for articles, and maybe other namespaces, is directed to a special page which forms a wizard similarly to the article wizard, with each page being editable by admins as a mediawiki page. There should of course be a way for users to bypass this (at least when autoconfirmed), and to no longer show it (it could also be modified as a preference). This could be of great help for new users, and would likely fall within the competence of the usability team. As an intermediary between the current status and full restriction, new users could be forced to use the article wizard for creating articles, yet without having to put them to AFC. I suppose that this would considerably reduce the number of bad articles and at the same time not massively increase the workload for AFC. Cenarium (talk) 15:17, 11 April 2011 (UTC)


 * Wizard-by-default is an option I've punted around as just a user option, with no question of auto-confirmed. Have it on by default, with explicit instructions on how to disable it on the first page.  Not really a NPP issue, more of a "tutorial mode."  SDY (talk) 23:47, 11 April 2011 (UTC)


 * That could work. Rd232 talk 00:06, 12 April 2011 (UTC)

On a related note, I've been messing around with User:Rd232/mysandbox.js to try and create a "my sandbox" link next to the "log out" link, which would point to Special:MyPage/MySandbox as a place for userspace drafts. I'm sure the Javascript could detect whether a page exists there and provide a helpful Article-Wizard-like preload and editintro if it doesn't. This approach would, basically, make the userspace draft approach much, much friendlier. However, I don't really know Javascript; perhaps someone who does could hack this together in a few minutes to prove the concept? Rd232 talk 00:06, 12 April 2011 (UTC)


 * Something to keep in mind... AFC and Wizards are nice. But part of the point of this proposal isn't to limit them or make them use some kind of special interface to create an article. It's to give them a chance to get used to ordinary editing for a while. Someone who makes 10 edits will be better equipped to create an appropriate article than someone with no edits. It's about the learning curve. Don't let that stop people from making a great Wizard, but it will be easier AND more effective to give some advice about your first ten edits and put it in a nutshell. Shooterwalker (talk) 00:13, 12 April 2011 (UTC)


 * I know that I made ~50 edits before creating an account. If creating an article was really the very first thing that someone did on coming across the site, sure, they'd need more than a little help, but I'm guessing that most people have edited quite a bit before making an account.  I mostly created an account because some guy I shared an IP with kept getting blocked for vandalizing railroad articles.  SDY (talk) 00:25, 12 April 2011 (UTC)


 * On general principle, I'd hate anyone to be forced to go through the wizard, but I have to admit that after trying it out today I rather like our wizard. It's attractive graphically, easy to follow, and damned informative. I'm sure there's always room for improvement, but I was left with the impression that anyone of near-average intelligence and reading skills who uses it and emerges with a totally horrible speedy-worthy article must be either not paying attention to what they're doing or acting in bad faith. I could be wrong. Rivertorch (talk) 05:03, 12 April 2011 (UTC)

So we could build consensus on the article wizard extension independently of this discussion ? If so, we could make a separate proposal at WP:VPR. Cenarium (talk) 11:26, 12 April 2011 (UTC)
 * It would probably be best to separate those issues out, whilst signposting them clearly here, given the size of this discussion. VPR makes sense - would someone with a clearer idea of what's being proposed in relation to this kick it off there? Rd232 talk 14:11, 12 April 2011 (UTC)
 * I spelled out the proposal here. Cenarium (talk) 20:59, 12 April 2011 (UTC)

Canvassing and forumshopping by oppose voters?
Seriously, this is getting annoying. After somebody tried to canvass Jimbo earlier we now have somebody who decided to go wp:forum shopping at meta. I understand they may feel very strongly about this matter, but this is getting disruptive. Yoenit (talk) 08:27, 12 April 2011 (UTC)
 * It has already been advertised on the watchlist banner, so you can expect that many are already aware or it, apart from the new users that it would affect. Graeme Bartlett (talk) 08:54, 12 April 2011 (UTC)
 * No, many people from other projects will not be aware of this. If there is one thing you are not gonna find at meta it is new users. Yoenit (talk) 09:11, 12 April 2011 (UTC)


 * I would also point out that the tone of the 'meta' posting is clearly in violation of WP:CANVAS, which requires that notifications be neutral in order to be appropriate. <span style="font-family:Antiqua, serif;">HrafnTalkStalk(P) 09:24, 12 April 2011 (UTC)
 * The support has hovered close to 80% since the beginning. It was there when the RFC started. It was there after there was a watchlist notice. ArbCom has made rulings on this before and they're prepared to discount sudden shifts in opinion due to canvassing. That's in addition to treating the editor as a disruption, which I would sadly have to support if someone decided to start a discussion at WP:AN/I. Shooterwalker (talk) 12:40, 12 April 2011 (UTC)


 * Oh, wow. That's an astoundingly axe-grindy canvass. BarkingFish didn't even make an attempt to present the issue in a nonpartisan manner. Whether his posting actually brings in new editors to support his view or not, that's really, really not an ok thing to have done. A fluffernutter is a sandwich! (talk) 12:54, 12 April 2011 (UTC)


 * Agreed. to clarify, Shooterwalker, you mean "arbcom thinks it is appropriate for those evaluating consensus to discount votes made due to canvassing" and not "arbcom's job is to decide what consensus is and discount things personally", right? Ironholds (talk) 16:49, 12 April 2011 (UTC)
 * Of course. No one in their right mind thinks ArbCom decides what consensus is. They just call out misconduct as they see it, including various battleground/filibuster tactics. Shooterwalker (talk) 23:04, 14 April 2011 (UTC)
 * Right, now I've had it properly explained to me, I concede that I am in violation of WP:CANVAS. I have collapsed the comments I made at meta, with clear notification of this in the collapse header, and placed a clear, simple, non-partisan message underneath it. Now I understand what I did wrong (thanks to someone on my talk page who explained it to me, instead of simply picking apart my behaviour).  I would ask that if you feel my behaviour is inappropriate in future, please tell me, or raise it at WP:AN/I if you feel it is warranted. <font color="red" size="2" face="Tahoma">Fish <font color="blue" size="2" face="Tahoma">Barking?  11:26, 15 April 2011 (UTC)
 * An apology is always helpful. People are entitled to mistakes which is probably why no one raised it at WP:AN/I. At this point... I don't think anyone is concerned with what to do about you so much as what to do about closing this RFC fairly. Shooterwalker (talk) 12:43, 15 April 2011 (UTC)

Removal of comments by me and several other users
At 18.10 on 8 April I added a comment - No 114 under User:Jayron32's proposal. Another comment was added after mine, but at 19.55 that day User:Rd232 removed comments 110-115 inclusive, with the edit summary "d categorisation, retaining recent votes" as can be seen at line 124 of this diff. I am happy to AGF, rather than an attempt to fiddle the responses, but could someone reinstate these comments, and any others that may have been deleted. Arjayay (talk) 15:11, 12 April 2011 (UTC)
 * I'm assuming that this was a mistake related to the undoing of his prior action (see above). —DoRD (talk) 15:21, 12 April 2011 (UTC)
 * I was told to look at this at a comment on my talk page. After investigating, I didn't remove these comments, User:Rd232 did.  Arjayay, you should contact User:Rd232 to request that he help with the problem and undo his removal of your comments.  -- Jayron  32  15:27, 12 April 2011 (UTC)
 * I will note that Blade of the northern lights later added another support (nr 127), so be careful not to include him twice. Yoenit (talk) 15:42, 12 April 2011 (UTC)


 * whoops, sorry. I have a feeling I left the Jayron view endorsements that needed transposing to the end, as the easiest to handle... and then forgot. I think the rest are in order. On the plus side, the consensus is easily strong enough to survive such small blips. Rd232 talk 16:18, 12 April 2011 (UTC)

The bikeshed must be painted blue
I'm not happy about all of the "how to run a trial" items and would like to suggest that we split those off on to a separate page. Here are my reasons:


 * We do not have consensus to run a trial at all.
 * No, really. About 310 different accounts have edited the page, presumably to endorse something.  Only about 11% of them endorsed Shootwalker's proposal to have a trial (the most popular trial proposal).
 * For every person who endorsed Shooterwalker's proposal to have a trial, there are five people who endorsed Jayron's proposal to make the change.
 * More people oppose doing this at all than support having a trial. If that minority doesn't constitute a consensus against this change, then surely even fewer people doesn't constitute a consensus to run a trial.
 * What we have is a couple of people who are enthusiastic about a trial. Enthusiasm isn't the same thing as a consensus.
 * Now is the not the best time to design a trial.
 * Now is the best time to decide whether a trial ought to be run, but not how to do it. If we decide to have a trial, then we should design it.  Decide first.  Design second.
 * Trials are run to find out things that we don't know. They are not the religious rituals of the scientific community.  If you know everything you need to know to make this decision, a trial is a completely unnecessary waste of time and resources.  If you don't know enough to make a decision, then the decision is "We need to find out ____, and the only way to do that is to hold a trial", not "Let's hold a trial!"
 * Smart people never design a trial without having some idea of what the ____ in "We need to find out ____" is.
 * This is possibly the worst place to design a trial.
 * Trial design by an ad hoc committee of people whose primary qualifications are enthusiasm for designing a trial is practically a guarantee of an idiotic design.
 * It is obvious to me that at least half the folks trying (sincerely) to help with the trial design don't actually have the skills or knowledge to do this. Even more haven't actually spent enough time thinking about the practical mechanics to make useful contributions.
 * A thousand minor variations lead to confusion, fragmentation, and disagreements. In short, we'll be arguing about the color of the bikeshed, and no design will ever have consensus.

I think that the best approach is to stop trying to design the trial at all: We don't currently have a consensus to run a trial or any agreement about what information that trial is supposed to produce. I recommend archiving all those discussions and letting it sit until there's some indication that we need to have those discussions at all.

IMO second-best approach is to move the trial design stuff to a separate page and let interested folks have fun designing a possibly unnecessary and probably irrelevant trial without further disrupting the overall RFC. If a trial is needed, then the community can call for proposals of trial designs, and choose whichever proposal seems most likely to provide the information that the community wants to have. WhatamIdoing (talk) 17:55, 12 April 2011 (UTC)


 * We don't currently have a consensus for anything, and a trial is part of forming a consensus, not a decision in itself. I don't see any other way forward - we're just kicking around anecdotes at the moment, and a trial (assuming a meaningful design) would generate actual data that might provide a resolution of sorts.  The only way to have a meaningful design is to kick around ideas until we know what information we want to collect.  At this point, this is just turning into dueling filibusters of "we have a majority so we can't let the proposal die" and "we won't let you do this."  Do you have any other suggestions, other than a trial, on how to break the stalemate?  Failing the proposal due to bureaucratic train wreck isn't a good option.  SDY (talk) 18:24, 12 April 2011 (UTC)


 * I wouldn't say we don't have a consensus for anything, SDY, unless you are disputing that there is overwhelming support for jayron's proposal, summarized as "Restricting article creation to autoconfirmed users is a change we should make." Are you intending to argue that that is not a consensus for the flipping of the switch? A fluffernutter is a sandwich! (talk) 18:31, 12 April 2011 (UTC)


 * I'm basing this on the "just do it" proposal section, which hasn't seen much activity. I think there's definitely a consensus to do something, but how to implement it isn't so clear.  In particular, there are concerns about ACW and AfC that I think should be addressed: if we're going to force people to use them, they should be viable.  SDY (talk) 18:43, 12 April 2011 (UTC)


 * I broadly agree that the implementation issues (of which whether and how to trial it are just a part) are starting to be a problem on this page, and it would actually be best to split them off. This page should focus on whether to go ahead with the basic idea. However the current implementation discussions are messy enough that moving them elsewhere won't be very helpful; better to restart discussion after reaching a clear conclusion on this RFC after 30 days (which seems highly likely to be Yes to the basic idea). So I'd suggest, leave things as they are for now, and plan for a second RFC on the how of implementation. Rd232 talk 18:45, 12 April 2011 (UTC)


 * That seems like a reasonable summary, Rd232. SDY (talk) 18:51, 12 April 2011 (UTC)
 * I too agree with Rd232's conclusions here. It seems reasonable to wait a full 30 days before making any headway on implementation.  Let this part of the process work itself out, then work on the specifics of how to make it work right.  -- Jayron  32  19:03, 12 April 2011 (UTC)

There cannot be any consensus at this point, primarily because the level of participation has been quite small to legitimate a change of this magnitude. Moreover, support to Jayron's proposal doesn't imply a support for immediate implementation. This is a first step. Cenarium (talk) 22:04, 12 April 2011 (UTC)
 * Just seeing this now. I agree. Talking about a trial this soon is sort of like talking about how long to block an editor before you've even decided if they've done anything to warrant it. We have 160+ people who've weighed in support of this idea, but literally handfulls of editors moving onto implementation before we've even closed the basic discussion about whether to even proceed. If someone wants to move it, go ahead. But I think any admin who decides to close this discussion will end up ignoring the discussion at the bottom anyway, since the number of editors focusing on implementation is so small compared to the bigger question. Shooterwalker (talk) 00:59, 14 April 2011 (UTC)

Quick question..
If and when this proposal goes through, hands up all those supporting it who plan to help out at Articles for Creation? Ironholds (talk) 20:57, 12 April 2011 (UTC)
 * Absolutely. I actually look forward to it. The Blade of the Northern Lights  ( 話して下さい ) 21:04, 12 April 2011 (UTC)
 * yes Yoenit (talk) 21:38, 12 April 2011 (UTC)


 * It crossed my mind that AFC is one of those tasks which could potentially use an invitation model similar to the Comment request service to parcel out support requests to a wider audience. I think a lot of people would be happy to sign up for an occasional "there's an AFC backlog, you're 1 of X people being invited to help reduce it today" (with full user control over frequency of requests, eg max 1 per week). The CRS system could be adapted for this (except that CRS is actually not live yet, as the putative bot creator got too busy. Don't suppose anyone would like to help?). Rd232 talk 21:56, 12 April 2011 (UTC)
 * hand up* I expect less to do at NNP, so yeah... Choyoołʼįįhí:Seb az86556 > haneʼ 22:02, 12 April 2011 (UTC)


 * I'm already trying to do my part as an Online Ambassador, and I don't want to spread myself too thin. I think there are several ways to help new editors besides at AfC. -- Donald Albury 13:22, 15 April 2011 (UTC)

How to close this RFC
Jimbo has asked that we send a song signal to the Wikimedia Foundation to at least test this proposal. Thus far, the proposal has been hovering close to 80% support. There's been some canvassing on meta but I don't expect it to do more than a couple points of damage. Some people have gotten ahead of themselves talking about how to design a trial, when this discussion hasn't been closed.

On April 30th someone should approach an uninvolved administrator to close this discussion and sum it up -- aside from the mess of a discussion about a trial. At that point, we should pass it onto Jimbo and the foundation, and let them decide how to proceed. I'm sure that 80% will be good enough for them. But since even a trial would require changes to the technical platform, we're going to have to rely on the foundation either way. Shooterwalker (talk) 23:50, 12 April 2011 (UTC)
 * This is getting annoying. If you read my reply to the comments I posted on the Wikimedia Forum, you will see that canvassing is entirely not my intention, and I would appreciate people not accusing me of it. This is an important change to editing process, and one which I don't see anything wrong in throwing open to the wider community.  Meta is a co-ordination hub, and I felt the forum was the best place to put a note about it and ask for the views of others. <font color="red" size="2" face="Tahoma">Fish <font color="blue" size="2" face="Tahoma">Barking?  13:10, 14 April 2011 (UTC)
 * You appreciate Jimbo is, as far as we're concerned, just another editor? Saying Jimbo sez it carries no more weight than saying Shooterwalker sez it; less, actually, because statistically those areas Jimbo gets involved in tend to be the scene of screaming and wikidrama before long. Ironholds (talk) 12:06, 13 April 2011 (UTC)
 * It's not so much what Jimbo said so much as what Jimbo asked, on behalf of the foundation. Again, this is going to require a technical change, which means it's going to have to go through the foundation. Really... to close this discussion we're going to need a neutral admin anyway. So that's the first step, before deciding how to implement/try the proposal. Shooterwalker (talk) 12:29, 13 April 2011 (UTC)


 * Jimbo was not asking on behalf of the foundation. Jimbo was asking on behalf of himself, an editor. Ironholds (talk) 12:39, 13 April 2011 (UTC)
 * Shooterwalker, is it your assertion that Jimbo was asking the community here (on behalf of the Foundation) to send a strong message to the Foundation? Afraid I'm with Ironholds on this one... Jimmy was fairly clearly speaking in an individual capacity.  His voice is one that should be listened to, absolutely ... he's shown a great deal of careful leadership through the years... but I think he'd be the first to tell you that a comment from him is not necessarily on behalf of the Foundation.  - <span style="font-family:Papyrus, sans-serif; color:#775ca8;">Philippe  12:42, 13 April 2011 (UTC)


 * And if your "on behalf of the foundation" comment means "he was asking the foundation to do X", then again, he's just an editor. Ironholds (talk) 12:48, 13 April 2011 (UTC)
 * The RFC didn't get anywhere close the level of participation needed to legitimate such a major structural change. The participation of 200 users is common in policy discussions, as noted here this should be advertised on a sitenotice. This RFC also can't be used to determine consensus on implementation (trial or not), by its format. So we've discussed at the bottom of the RFC on making a trial proposal on a new page. I think the RFC should be closed as 'moving to trial proposal'. Anyway, the way to ask devs to 'turn the flip on' is by making a bugzilla request, it doesn't involve Jimbo. Cenarium (talk) 13:12, 13 April 2011 (UTC)


 * Cenarium,
 * You say here we need at least 200 editors to participate in the RFC for it to be reasonably indicative of the community's views. I find that so far 329 (yes, three hundred twenty-nine) unique, non-bot registered editors have participated.  Count for yourself.  We're well beyond the minimum standard of participation that you set.  WhatamIdoing (talk) 17:33, 14 April 2011 (UTC)
 * No, I say that 200 is common, and such a major structural change requires much more. Cenarium (talk) 23:03, 15 April 2011 (UTC)

People are getting hung up on what they think I meant by contacting Jimbo, when my real point was to say that there's no point on talking about a trial until we've closed the existing discussion. Running it by the foundation is going to be a necessary step when we go from "we've decided we want to do this" to "how do we do this" (including the question of a trial), because it's a pretty big technical change. Shooterwalker (talk) 00:55, 14 April 2011 (UTC)

Since it is clear that there are strong objections to declaring a consensus in support of the proposal, and because it will need to go to the foundation for implementation, I suggest that a bureaucrat (rather than an admin) be asked to close this discussion and determine whether or not to report a consensus to the foundation. It may not yet be the time to close the discussion, but I think it is time to settle how to close it. -- Donald Albury 13:43, 15 April 2011 (UTC)
 * No again, this RFC cannot be used to solely determine whether there is consensus. We need a contradictory debate on a clear proposal. This format of RFC is primarily for dispute resolution, it's not for determining consensus on a proposal, it doesn't allow for enough contradictory debate, it just provides for editors to express views and endorse expressed views. It is completely false to assume that because there are two views which seem broadly in opposition, that there are they are their exact opposite, that those who endorse Jayron's support going ahead and vice-versa, and that those who endorse Ironhold's oppose going ahead and vice-versa. Not to mention that going ahead isn't clear, many would only support first a trial. Clearly the next step should be to put a trial proposal for consideration, as suggested at Village pump (proposals)/Proposal to require autoconfirmed status in order to create articles/Trial. The current phase is a first step, we should go on the next phase by putting the proposal clearly up for consideration, so that we can evaluate consensus on it, then we ask for a closure. Cenarium (talk) 23:03, 15 April 2011 (UTC)
 * I don't think it's productive to argue about a trial until we establish if there's a consensus to move forward in some capacity (which I think is obvious from reading the various opinions on the RFC, but should really be assessed by a neutral outsider rather than the 5 or 6 people who are still around). Shooterwalker (talk) 23:15, 15 April 2011 (UTC)
 * "It is completely false to assume that because there are two views which seem broadly in opposition, that there are they are their exact opposite, that those who endorse Jayron's support going ahead and vice-versa, and that those who endorse Ironhold's oppose going ahead and vice-versa." say what? Those views clearly relate to a reasonably specific proposal, and we have to assume that people have read it and the views they endorse. So unless you can show substantial numbers of people saying self-contradictory things (by endorsing views that contradict each other), the conclusion is quite clear. Some details remain unanswered (notably the issue of a trial), but saying things like "cannot show consensus" and "many would only support first a trial" is bizarre. Rd232 talk 00:54, 16 April 2011 (UTC)
 * So let a bureaucrat determine what this discussion means. You can't just dismiss it as meaningless. -- Donald Albury 11:23, 16 April 2011 (UTC)
 * Close it as tl;dr and get it off my watchlist.--RaptorHunter (talk) 05:34, 27 April 2011 (UTC)

Trial proposal
The next step imo should be to make a trial proposal, as discussed here. There's no consensus in the RFC on how to implement, but the trial idea seems to have considerable support, and there's been not enough participation to legitimate an implementation yet, so I think making a trial proposal which will be advertized on the sitenotice should be the next step. I've suggested a poll at Village pump (proposals)/Restriction of article creation to autoconfirmed users. Please tell me what you think of this. Cenarium (talk) 13:22, 13 April 2011 (UTC)
 * As stated above... I think it's important to close the existing discussion first before deciding how (and if) to implement a trial. I know people are getting impatient. But some amount of step-by-step process is needed to keep things organized. Shooterwalker (talk) 00:53, 14 April 2011 (UTC)
 * "Proposal to require autoconfirmed status in order to create articles" v. "Restriction of article creation to autoconfirmed users"? A clean, separate page for an outline of a trial is a good idea, but let's not confuse the two. Simpler: "/Trial" ˉˉanetode╦╩ 01:01, 14 April 2011 (UTC)
 * Ok, moved to Village pump (proposals)/Proposal to require autoconfirmed status in order to create articles/Trial. Cenarium (talk) 11:17, 14 April 2011 (UTC)

What are you going to do for new users?
So, I've been thinking - whether this goes through or whether it doesn't, people clearly have strong opinions about new users. Everyone, on every side, appreciates that attracting those who are interested and competent is necessary; the distinction is only on what constitutes "interested" and "competent", and what balance we should strike between attracting them and keeping The Great Unwashed Masses out. To that end, here are a few projects people might want to get involved in to help out new users - and maybe get an appreciation of both what they have to offer and how difficult their initial experience can sometimes be, and that we have users who turn up not to contribute productively but to cause harm.


 * The 'Wiki Guides guide new users from all over, regardless of how they got involved in the project - the intention is to get new contributors involved with a mentor right at the get-go, and gather some hard statistics on the issue. This is probably the most useful place, when thinking about long-term gain and the sheer breadth of what you can cover, to work at.


 * The Online Ambassadors help out new users from an academic background who are joining as part of the public policy initiatives; the intent is "to keep the newcomers from being "bitten" and help them learn both the how and the why of editing Wikipedia and participating in the community". Contributors are expected to mentor individual newbies, help them understand key policies and how to edit, and co-ordinate with the Campus Ambassadors and Wikimedia Foundation.


 * Articles for creation is.. well, everyone knows AfC. It's got a backlog, and that's a constant - if anyone wishes to help out there, it'd be much appreciated. Ironholds (talk) 05:40, 14 April 2011 (UTC)


 * Given the persistent issues with AfC, I'm thinking that the most "useful" alternative for new users will be the creation wizard. It forces them to at least have the relevant policies pass in front of their eyes (reading them is of course still something that they could avoid), so if their article does get deleted it's much harder for them to say that they were surprised.  It's a little clunky, but it doesn't require another volunteer to be involved.  WP:WikiOtters are just too rare of a breed to rely on sufficient supply.  SDY (talk) 12:27, 14 April 2011 (UTC)


 * Okay. First, the creation wizard is an opt-in system which requires the ability to create articles. Secondly, that wasn't what I was actually asking ;p. Ironholds (talk) 17:01, 14 April 2011 (UTC)

Remove the notices clutter on the top of the page
I removed two notices from the top of the proposal page as clutter. This page is way too long as it is, last thing we need to is to distract editors with canvassing attempts on meta and a only somewhat related proposal on expanding the wizard, before even telling them what this proposal is about. My edit was reverted by Barkingfish, who stated the related discussions are important. I obviously disagree and hereby seek other opinions on whether these notices should be kept at the top of the article. Yoenit (talk) 12:37, 14 April 2011 (UTC)


 * In reply to this, I would state that the discussion on meta is not canvassing, as I've already stated there.  I was accused of forum shopping on meta too, and gave this as a reply -


 * -<tt>The reason I posted this here was that since this is the coordination hub for Wiki(m/p)edia projects, I felt that it would be a good idea to let them know of a significant change and one which as I say, could breach at least one of the Founding princples, which are essentially the footings we were built on. In that respect and in others, I felt it right and proper to open the discussion to the wider meta and WMF community. I don't see anything wrong in that.</tt>


 * -It is courtesy when opening a discussion elsewhere to notify people involved in one already, that it's up somewhere else too, or that's my understanding anyhow. These notices are not distractions or clutter, they're signposts to related items and should be left in place, which is why I reverted you.  <font color="red" size="2" face="Tahoma">Fish <font color="blue" size="2" face="Tahoma">Barking?  12:46, 14 April 2011 (UTC)

Too much clutter
There are now 55 views. Many of them are supported by no one but the author, a few have a dozen signatures. Compare this to the current leader in signatures which has over 150. We should collapse and move to the bottom all of the ones except for those of Ironholds (60), Jayron32 (176), HominidMachinae (56), DGG (50), and Rivertorch (51), in other words, all the ones with less than 50. There happen to be six others with more than 20 and less than 50. That would make for either five or eleven instead of 55.

Of course, the simplest option would have been to do a straight up or down vote on this, i.e. two threads instead of 55, but it's too late for that.  S ven M anguard  Wha?  03:44, 17 April 2011 (UTC)
 * The page is now half a megabyte. It's enough to keep people away from visiting the page. But I think we should err towards more views rather than less. It's worth noting that there are more than 300 participants, but Jayron and Ironholds (the most obvious "yay" or "nay" !votes) account for only 236. So it might be better to leave in more of the middle ground views, since a lot of editors are looking for some other way forward.
 * It would be a huge benefit just to migrate the 1/3 of the discussion at the bottom that discusses how to conduct a trial. It really was premature, and a bad idea to start discussing implementation before we even figured out consensus on the overall idea. If you talk about implementation without agreeing on the main idea, there is no chance of getting a consensus. Plus the numbers will never work out in practice: there's only about 20 or so participants down there (including myself), compared to the hundreds above. So there really is no chance of that discussion bearing fruit. Just migrating that to the talk page would be a step in the right direction. Shooterwalker (talk) 04:36, 17 April 2011 (UTC)
 * We could cut the page in half at least by cutting anything with less than 5 support votes (which includes all the discussions)  S ven M anguard   Wha?  04:43, 17 April 2011 (UTC)
 * I'd support that, providing you make it clear on the page that other discussions exist, and you direct people as to where to find them. <font color="red" size="2" face="Tahoma">Fish <font color="blue" size="2" face="Tahoma">Barking?  12:06, 17 April 2011 (UTC)
 * Yeah, I'd support it too. Maybe the talk page isn't suitable... so put them on some kind of subpage? Explain why and where at the top... Shooterwalker (talk) 14:00, 17 April 2011 (UTC)
 * Definitely put some on a subpage; I suggest a threshold of ten replies. <font style="font-family: Palatino Linotype, Book Antiqua, Palatino, serif;"><font color="#BBAED0">Feezo <font size="-2">(send a signal | watch the sky)  07:39, 19 April 2011 (UTC)
 * I support this as well. I tried to be bold and actually move all minor views to Village pump (proposals)/Proposal to require autoconfirmed status in order to create articles/Minor views, but the pc I am on now lags for several seconds when I try to edit the page. Would somebody with better hardware please be bold and move every view with less than 10 votes and posted at least a week ago to this or another subpage. Yoenit (talk) 08:40, 19 April 2011 (UTC)
 * I went ahead and migrated the side discussions, including the summary of views (collaborative views) and the discussion of a trial. I'm reluctant to actually start migrating the individual views, no matter how minor... just because there is a good 50-100 people who can't be accounted for in the top 2 views. The discussion is back down to the size that it was more than one week ago (and about 300 contributions ago). I think most of the editors who saw the watchlist notice and actually care have participated, plus whatever WP:CAMPAIGNING we saw on Meta. So I wouldn't anticipate that many more views coming in anyway. Shooterwalker (talk) 02:20, 20 April 2011 (UTC)


 * Thank you for moving the trial discussion. WhatamIdoing (talk) 06:12, 24 April 2011 (UTC)

Even More Clutter
There are now 88 opinions. Many of the bottom half, or even the bottom three fourths, have noting new to say, and didn't read the "don't duplicate comments" notice. I'll ask again, can we cut out all the ones with little to no support?  S ven M anguard  Wha?  21:13, 30 April 2011 (UTC)
 * I don't think users who have participated in the RfC should remove the views. I suggest we make this RfC more manageable by moving every view, howsoever minor, to its own page (in the nature of Wikipedia :Village pump (proposals)/Proposal to require autoconfirmed status in order to create articles/View from User:Example) and put up a template for accessing already existing views as well as an inputbox for creating new ones.--   Forty two  22:21, 30 April 2011 (UTC)
 * How would that help? I think the use of subpages would just confuse matters more, but it should not be done anyway as this RFC is supposed to close in a few days. Yoenit (talk) 23:06, 30 April 2011 (UTC)
 * It would make the page(s) easier to read and edit and probably make it possible for users with slow internet connections to participate in the RfC.--  Forty two  00:09, 1 May 2011 (UTC)

Spotting consensus or the lack of it
The community needs to agree upon a method for identifying consensus before we engage in a large scale discussion such as this one. I am afraid this is going to kick the bucket just like the Pending changes proposal. Any ideas as to how and when this proposal should be closed?--  Forty two  08:20, 26 April 2011 (UTC)
 * It should be closed 30 days after the discussion started by somebody uninvolved, preferably a bureaucrat. As to how it should be closed: that is up to the closer. At the least I would expect a followup RFC with a clear yes/no structure, at most I would expect acceptance that this proposal should be implemented in some form, leading to a followup RFC on whether we should have a trial first or implement it directly. In neither case would this proposal kick the bucket, nor would it lead to immediate implementation. Yoenit (talk) 08:42, 26 April 2011 (UTC)
 * I'd leave it up to the bureaucrat to decide what the next RFC should entail. We had a watchlisted proposal with 450+ participants. There's no sense on asking "so should we do this?" again unless it's truly inconclusive. The next RFC should proceed from whatever consensus we can find here, assuming there is one. If there is a consensus in principle to proceed, then the next RFC should discuss the parameters of implementing a trial. Shooterwalker (talk) 12:53, 26 April 2011 (UTC)
 * I feel this is a bit of a mess. The page is so long its becoming unreadable. I have my ideas about this but so many other people do it hardly seems worth me giving my two pennies/cents worth. Still someone much more patient than me will I'm sure :-) Cls14 (talk) 21:22, 29 April 2011 (UTC)
 * I've proposed for the next phase Village pump (proposals)/Proposal to require autoconfirmed status in order to create articles/Trial. I still do not think that this RFC can serve as a basis for determining consensus on implementing anything, because there's no clear proposal, indefinite or trial implementation, and there's been no contradictory debate on any proposal. Adding and endorsing views can't be a basis for consensus, there needs to be a clear proposal, then proper discussion of it to assess consensus. The number of people endorsing one view is not an adequate way of measuring consensus, people can oppose a proposal while not endorsing a specific view, and vice-versa. If we were to consider this numerically, as of writing more than 500 people edited the page, of which 223 endorsed the most endorsed view, so no view gained a majority of adherence. We've never made consensus decisions by considering which view has the most support, it's not a way to establish consensus. Cenarium (talk) 00:24, 30 April 2011 (UTC)
 * A relic of the "voting" mentality is that we can't read an RFC without tallying up the most popular proposals. You just have to take a closer read. Look at the 500 or so participants, including the ones who voted for less popular proposals, and try to parse what their position is. I think you'll find a consensus against just implementing fully, and a consensus for doing some kind of trial. And if I'm wrong, I'm wrong. But we didn't do a watchlisted RFC to just throw it away and say "it doesn't mean anything". Shooterwalker (talk) 00:36, 4 May 2011 (UTC)
 * I'm not suggesting that we throw it away, but that we go to the next step. Cenarium (talk) 09:05, 6 May 2011 (UTC)

This RfC has officially "expired" -- could somebody please close it
The bot says so. Or is it now undead and needs special treatment like a stake through its heart? <span style="font-family:Antiqua, serif;">HrafnTalkStalk(P) 20:17, 3 May 2011 (UTC)
 * That's not how you kill the undead, that's how you kill vampires. Cliff (talk) 15:23, 4 May 2011 (UTC)
 * Last I checked vampires were undead -- but whatever. :) <span style="font-family:Antiqua, serif;">HrafnTalkStalk(P) 15:42, 4 May 2011 (UTC)
 * Vampires, ghouls, zombies (of the celluloid but not the Haitian persuasion)—all undead, and each requiring a different method of extirpation. Would a request at WP:AN be the silver bullet (damn, no, that's werewolves) a good idea? It's a complicated close and will need not only Van Helsing an uninvolved party but someone with a fair bit of time and patience. Rivertorch (talk) 04:11, 5 May 2011 (UTC)
 * I'm fairly sure fire works against all forms of undead (though that may count as WP:OR or WP:SYN). But maybe we need an RfC to decide the best way of closing this RfC -- wouldn't that be fun? >:) <span style="font-family:Antiqua, serif;">HrafnTalkStalk(P) 04:35, 5 May 2011 (UTC)
 * We have people still editing and commenting on the RfC. I went ahead and requested page protection, which was denied. Something needs to be done asap! --  Forty two  08:35, 5 May 2011 (UTC)


 * Nothing needs to be done ASAP. If folks are still discussing the proposal, then the only thing that needs to be done is that somebody needs to change the timestamp under the RFC tag, to let the bot know that we're still busy here.  There is no bot-enforced 30-day maximum for RFC discussions.  WhatamIdoing (talk) 17:31, 10 May 2011 (UTC)
 * The discussion is however getting increasingly unfocused (with the last 2/3 of the views garnering little in the way of support), so is adding little in terms of providing a basis for a way forward. <span style="font-family:Antiqua, serif;">HrafnTalkStalk(P) 18:30, 10 May 2011 (UTC)

comment
Over the period of this rfc I have mentioned it to numerous prospective editors. About half said its adoption would greatly reduce their desire to contribute here. The other half said it would prevent it altogether, sometimes adding they they would never bother with a place that had such an overtly hostile attitude to newcomers. Myself, I often think we already have a overtly hostile attitude to newcomers, so I suppose they mean  a place that goes to the length of making it an official   attitude.  DGG ( talk ) 04:32, 4 May 2011 (UTC)
 * I concur - and I fear that if we do this not as a "test" but rather as a "let's change it, we can always change it back" policy, the word will spread as Wikipedia having closed the door on contributions from newcomers, and that will cause lasting damage. --Nat Gertler (talk) 04:43, 4 May 2011 (UTC)
 * Which prospective editors? In the interests of evidence-based decisionmaking, I'd like to see those numerous responses; it could be really valuable information. bobrayner (talk) 12:15, 4 May 2011 (UTC)


 * I must say I concur with DGG here. After 3 months of mentoring 26 undergraduate and graduate students as a Campus Ambassador on Wikipedia, I was truly amazed at how intimidating the whole thing is to new comers.  It really doesn’t matter what is perceived as an obstacle by a newcomer or not, but when they encounter it, they generally get discouraged.  This Autoconfirmed proposal is just another of many obstacles.  Even the name Autoconfirmed is intimidating and likely to cause angst among someone who isn’t familiar with WP.  I think my real concern here is strategic.  I’ve seen nothing related to this proposed tactic that would seriously improve WP quality (a Wikimedia strategic goal).  But, if even one prospective editor is intimidated enough to forego continued participation, then this tactic definitely impedes progress on the strategic participation goals.  And in my view, any tactic that impedes progress on any strategic goal, is a poorly chosen one. --Mike Cline (talk) 13:14, 4 May 2011 (UTC)
 * AfC could definitely use some streamlining, and this would be a great time to do that. The person who does easily the most work there is Chzz, so he might have some suggestions on how to do that.  Even if we don't end up going through with this, it'd still be a good idea. The Blade of the Northern Lights  ( 話して下さい ) 13:36, 4 May 2011 (UTC)


 * It's easy enough to get just about any answer you want from a survey by asking the right question. For example, if you present it as "you can't create articles until you're auto-confirmed" vs "for the first few days of having an account you need to choose 1 of 3 assistance options in order to create articles". And the context of the discussion matters too; prefix the issue with a bunch of "Wikipedia isn't as open as it used to be..." and it's different from prefixing it with "Most articles created by Wikipedia newcomers are deleted, because they don't know enough about how to meet a bare minimum standard of acceptability, and then get frustrated and leave". Rd232 talk 22:38, 4 May 2011 (UTC)
 * I don't think we're presenting a new obstacle at all. We're removing the presentation of an option that doesn't exist in practice. In practice, most of the people who start swimming in the deep end will drown. I see nothing wrong with making sure they have to enter through the shallow end, and swim for a while (4 days and 10 edits) to reach the deep end. Designing the waters that way isn't an obstacle. Drowning is the obstacle. Shooterwalker (talk) 23:37, 4 May 2011 (UTC)
 * Whether or not you think it's an actual obstacle, will it appear to be an obstacle to those who we wish to encourage to be involved? (And in practice, most people who start swimming in the shallow end will drown as well. But if you force those who would enter at the deep end to enter at the shallow end, then they'll be both dead and annoyed.) --Nat Gertler (talk) 23:43, 4 May 2011 (UTC)
 * I imagine it wouldn't "appear" at all. You join Wikipedia. You start editing. You don't see advanced options until you're an advanced user. Shooterwalker (talk) 23:56, 4 May 2011 (UTC)
 * You go to wikipedia to create an article. You find that, despite the claims about Wikipedia, you cannot create an article because you haven't gone down the secret pathway. Sure sounds like an obstacle to me. --Nat Gertler (talk) 00:15, 5 May 2011 (UTC)
 * "You find that, despite the claims about Wikipedia, you cannot create an article because you haven't gone down the secret pathway." - I think you're confusing the proposal with Dungeons & Dragons or something. We already have plenty of situations now where editors try to do so something and find they can't; we do our best to provide them with an array of relevant options (eg the edit request system for editing semi or fully protected pages). Likewise, non-autoconfirmed users won't get a YOU CAN'T CREATE AN ARTICLE YOU DAMNED IGNORANT NEWBIE HOW DARE YOU THINK YOU COULD, but "here's three ways you can create an article: AFC, Article Wizard; userspace draft + request someone to move it. Also, later on you'll be able to create articles a fourth way, without any assistance." Rd232 talk 01:19, 5 May 2011 (UTC)
 * Newsflash: Wikipedia has obstacles. It has policies. Guidelines. Standards. Editors. Consensus. Thinking that you improve the user experience by letting them run off into the wildnerness is a nice idea, until they run off a cliff. We're never going to agree on this, you and me. What we can do is actually gather data. Do a trial, and see if putting up a guard rail scares people away, or saves lives. Shooterwalker (talk) 00:22, 5 May 2011 (UTC)

(unindent)  This is not a big, scary change that will ruin Wikipedia; it's a natural step in the evolution of the project, and a rather minor one at that. This is about saying to new editors, "We're glad you're here, but we want to make sure you know the basics of editing Wikipedia before you waste a lot of time and effort—yours and ours—posting something for Google-able public view that could make our encyclopedia the object of ridicule around the world." What on earth is the big deal about that? Do we really need a formal trial to see if such a reasonable message (more carefully worded, of course) is going to scare off potential good editors or bring article creation to a screeching halt? A trial would be vastly more complex, more confusing to newbies, and potentially much more contentious than simply implementing the proposal to change the standard to require autoconfirmation to create articles. The effects of the change can still be studied (as well they should be) and the change reversed in the unlikely event that evidence shows an alarming drop in viable new articles or the retention of productive new editors. (I posted the preceding elsewhere three weeks ago, and it was removed as premature or something. It seems to fit this discussion just as well.) Rivertorch (talk) 03:55, 5 May 2011 (UTC)
 * I support this change, but I do think this is a significant change. While it isn't the dominant route for new users, a significant number of new users do start by creating an article. The questions are: Where will they go? Are those the users we actually want? A trial doesn't have to be confusing to newbies, we don't even have to tell them it's a trial (it would probably be better if we didn't, actually; they might act differently if they know the change isn't permanent). While we don't necessarily have to do a trial, we should at least establish some criteria for success or failure before we start analyzing any data, otherwise it becomes too easy for people to interpret the data to suit their pre-determined conclusions. We have to consider what things are deal-breakers and what we can live with. If new user retention increases, but AFC becomes even more backlogged or the workload at NPP doesn't decrease significantly, is that a success? If the number of new articles drastically decreases, but the number of edits to existing ones significantly increases, is that a failure? Mr.Z-man.sock (talk) 18:11, 5 May 2011 (UTC)
 * I think we need to has that out as part of a second-stage RFC on implementation, not on the talk page here. Personally, I'm focussed on increasing the number of what one might call "serious contributors" or perhaps "Wikipedians", that is, editors with a modicum of clue and experience. I'd define that as (perhaps) 100+ edits to >5 articles and >5 edits to talk pages - that sort of ballpark. If we get the numbers of such editors up substantially (compared to prior trend), then I'm willing to take quite a hit on other fronts, like fewer new accounts, bigger backlogs, that sort of thing (within reason). Rd232 talk 02:58, 6 May 2011 (UTC)


 * I agree Rivertorch, but sadly wikipedia's decision-making processes are incapable of making rational decisions, or indeed any decisions at all in most cases. Malleus Fatuorum 03:20, 6 May 2011 (UTC)
 * Processes? I don't see no processes. ;) Rivertorch (talk) 15:08, 6 May 2011 (UTC)
 * I have to say that, were I a newcomer here to create an article, I would be much less frustrated if I was told the process of creating an article, or requesting it be made, than if I spent time creating an article, and then had it deleted because it was not considered noteworthy,etc... If my effort were wasted in such a way, I'd be pissed, and I'd leave. If I was told about a process, and went through the process, I'd learn more about WP and its standards and even if my article request was declined I'd be more likely to stay because I would understand the reasoning. Cliff (talk) 18:25, 6 May 2011 (UTC)

Implementation
Some thoughts on implementation: The current MediaWiki:Nocreatetext would be moved to MediaWiki:Nocreatetext-anon, and a new version of it created. It would be much the same, but replacing the "log in /create" line with "click here to start an article". That link would go a page looking something like this mockup: User:Rd232/creationdraft. Or the text there might be incorporated directly into MediaWiki:Nocreatetext - but at this point, for discussion purposes, it's a bit clearer to do it this way (especially considering how complex the innards of that page are).

To some this might seem jumping the gun, but actually I wish I'd done this right at the beginning of the RFC, to show just how easy it is to make it look helpful and welcoming. Rd232 talk 03:38, 6 May 2011 (UTC)
 * It looks good to me; gets the message across and explains everything without being harsh. The Blade of the Northern Lights  ( 話して下さい ) 05:02, 7 May 2011 (UTC)
 * Definitely clean for that page which is good. We should have something more like that I think for anons compared to our current one which is a bit more cluttered. If it comes to this you need to remember to note the days/edits requirements, just noting a couple days is going to confuse a lot of people (most editors have not made 10 edits by their 4th day and so will not become auto-confirmed). I also don't think we can possibly implement something like this until we have 1. Technical changes (currently flipping the switch would mean those users couldn't even create user space pages) and 2. a huge update/redo of the Article Wizard/AfC etc. The current system is no where near ready to be used and would be worse especially when added to the psychological side effects of sending people through templates or to pages they know need to hurry up and wait to get approved. Obviously I'm bias since I still think this is a horrible idea but if it ends up happening I want to make sure it happens right; I think way too much damage could be done in 3-5 months for me just to "wait" for some trial period to be over. James of UR (talk) 06:53, 8 May 2011 (UTC) I also think this is going to become a PR nightmare in the press if it happens... but we can't really make policy on that. Said as a concerned community member and not as a WMF staff member


 * There's a lot of implementation work that will have to be done if we do this. It's not a matter of deciding on May 31st to restrict page creation, and on June 1st having the devs flip a magic switch.  If we do this properly, it will probably take a month to line up the changes.  If we do a proper study—as opposed to an unstructured, uncontrolled, possibly temporary implementation under a fifty-dollar name—it will take at least two months to get ready for that change.  WhatamIdoing (talk) 18:02, 10 May 2011 (UTC)


 * If it's to be done properly, it needs a clear task list drawing up something like this so that the devs know what do:
 * 1. New page creation in mainspace by autoconfirmed users only.
 * Their new pages are patrolled in the normal way.
 * 2. Non autoconfirmed can start creating immediately in their userspace using the wizard:
 * They click a 'create page' button
 * When prompted, they fill in a field for the name of their new page, and click 'Go'
 * They are sent to a Wizard page
 * When they think they have finished, they click another button, which creates it in their user space as  'User:Foo/article name (draft)', and queues it for New Page Patrol.
 * Patrollers either reject it based on any of the standard CSD criteria, or tag it with any of the usual pagetop or inline maintenance templates.
 * Creator can use the remaining time waiting to become autopatrolled autoconfirmed to address any issues with theier creation.
 * This addresses the four points of the proposal, but is not to be confused with PFC although slightly similar in concept. The most important achievement is that it prevents all the vandal, hoax, spam, autobio, and garage band from going live immediately, being ignored too long and getting referenced in Google.  Technically, it's easy enough to implement in the board software. The recently abolished right of Reviewer could be reinstated as 'New Page Reviewer', and we could probably then dispense with NPP altogether - it's still in a mess despite the efforts of Snottywong and myself over the last 7 months to encourage the patrollers to tag more accurately, and not to pass pages that should be tagged. Kudpung กุดผึ้ง (talk) 14:29, 21 May 2011 (UTC)
 * Autoconfirmed or autopatrolled? Just want to make sure. The Blade of the Northern Lights  ( 話して下さい ) 17:14, 22 May 2011 (UTC)
 * Autoconfirmed - thanks for pointing out my genuine slip (The 'preview' does not currently load due to server timeouts, so I can't review my posts, or even get back to the page see what I saved). Kudpung กุดผึ้ง (talk) 05:30, 23 May 2011 (UTC)

Incidentally, the mockup at User:Rd232/creationdraft lists 3 options. In theory we could also use (or test) a fourth: use the editintro/preload of the userspace draft option (appropriately amended) and make the results create an article immediately. (That fourth option could be flagged in the list as "the most at risk of deletion" or some such.) This would be something of a compromise, because with this option available those who oppose the proposal really couldn't claim "autoconfirmed users won't be able to create articles anymore!" Just a thought. Rd232 talk 14:10, 23 May 2011 (UTC)

Going ahead with trial proposal
Are there objections to going ahead with the trial proposal ? As I commented earlier, this RFC is not amenable to determining consensus on the issue at hand. If no admin has closed it yet, it's because it's too convoluted to permit any kind of consensus to manifest. As we had discussed previously, a trial proposal has been set up on this page. Are there any objections to move on to this next step ? Cenarium (talk) 02:00, 23 May 2011 (UTC)


 * Yes: I object to the questions.  It proposes "a trial" without providing any information at all about what that trial is or does.  Is this "trial", meaning "we turn it on for a while, and then spend the next year fussing about whether to turn it off again"?  Is this "trial", meaning "we want to see whether more accounts reach autoconfirmed status"?  Is this "trial", meaning "we want to see whether fewer CSD-worthy articles are created"?  Is this "trial", meaning "software beta test"?


 * "Should we have a trial?" is not an answerable question. "Should we have a trial to find out ____?" is answerable.
 * If I've guessed correctly at your point, the questions ought to be these:


 * {| class="wikitable"


 * 1. There is very strong, but not overwhelming, support for restricting unassisted article creation in the main namespace to auto-confirmed editors (any account more than 4 days old plus more than 10 edits). The following facts are known:
 * Three-quarters of new accounts do not attempt to create new articles immediately, and so would be unaffected by this change.
 * One-quarter of new accounts attempt to create new articles. More than 80% of these articles are promptly deleted, and more than 99% of those editors leave Wikipedia within a few months after deletion.
 * The change as previously discussed would permit unlimited creation of user pages, including a userspace draft, along with access to WP:Articles for creation and possibly immediate article creation in the main namespace through the WP:Article wizard.
 * The change as previously discussed would permit unlimited creation of user pages, including a userspace draft, along with access to WP:Articles for creation and possibly immediate article creation in the main namespace through the WP:Article wizard.

Do we have enough information to make a decision about whether to make this change, or do we need more information (for example, information gathered in a temporary or experimental implementation)?
 * 2. If we do not have enough information to make a decision for or against this change, then what other information do we need?
 * }
 * }


 * As you can see, the questions proposed do not decide whether to make this change; they only determine whether a trial is necessary. The questions currently listed at /Trial will only result in people saying "yes, let's do a trial, because all the cool people run trials".  WhatamIdoing (talk) 03:04, 23 May 2011 (UTC)


 * I loathe voting. It would probably be more productive to just invite people over to design the proposal through the WP:BRD process. Asking "do you support a trial" is going to invite different responses unless we actually work out what kind of trial it is. Shooterwalker (talk) 03:06, 23 May 2011 (UTC)
 * Of course, if we ask them to design a trial, then we're telling them that some sort of trial must be done, perhaps involving a chicken and a full moon. The need for a trial (as something more than a superstitious ritual) is not apparent to me.  WhatamIdoing (talk) 03:25, 23 May 2011 (UTC)

What about turning the question on its head? Get those who think the idea is a bad one to come up with "failure criteria" that would prove them right - criteria that would mark the idea as having failed. that would certainly be a starting point for discussion, and it might lead to the easier process of not having a formal trial, but instead a regular monitoring and scheduled evaluation of those criteria. If the failure criteria are breached, there can be a discussion on what to do, which might default to "turning it off" if the discussion reaches no conclusion. Rd232 talk 04:10, 23 May 2011 (UTC)


 * (Replying to Cenarium's original question) If I'm gauging consensus accurately here, I doubt there's much point in going forward with the trial proposal. A large number of contributors to the discussion—perhaps a plurality—appear to be disinclined to restrict article creation, even on a trial basis. Then there are oddballs like me (or maybe I'm the only one—I've lost track), who support the original proposal but dislike the whole idea of a trial. I suspect that attempting to move this forward right now will result in nothing but a lot more talk, at best. Honestly, I think this proposal is dead in the water for the time being. I suggest the establishment of some simple method of quantifying new article creation by permissions status. (This could be done informally, either by the brave souls guarding the gates or at least in conjunction with them.) Then, using the data collected, let's try again in a while. Rivertorch (talk) 07:43, 23 May 2011 (UTC)
 * "A large number of contributors to the discussion—perhaps a plurality—appear to be disinclined to restrict article creation, even on a trial basis." - I don't know where you get that from. Right now the endorsements of Ironholds' view vs Jayron32's view are 100 (excluding #2 there for obvious reasons) to 231. I make that 70% support which is the sort of level we call "consensus". Besides, the arguments are much stronger in favour; too much of the opposition rests on handwaving which greatly overstates the significance of the change, and some of it is quasi-apocalyptic and less than entirely rational. It has to be remembered that getting support for implementing a site-wide change of any significance is always an incredible achievement; given the inbuilt resistance to change in the Wikipedia community, especially anything that has even a vague sense of impinging on founding principles. In this context, the RFC cannot be interpreted as anything other than a resounding endorsement of the idea of implementation. Support/opposition for a trial is much weaker (many fewer have expressed an opinion), which is why I suggest the burden should be on the opposers to come up with criteria by which success/failure might be measured (which those supporting can then comment on). That way, if the opposers can't come up with measurable criteria in a reasonable time frame, we can agree to just do it, and have an informal review scheduled after say 3 and 6 months. Rd232 talk 09:46, 23 May 2011 (UTC)

The brave souls are unfortunately not doing a very  good job of it. Most of the patrollers are extremely  new and inexperienced and do  not  themselves understand what  the requirements for new articles are. Seven months of research by  Snottywong  and myself has proved this so  I'll not  go  into  detail  here. Furthermore there is a backlog of several thousand articles that they  find to difficult to asses, and leave to pass after thirty days unpatrolled and untagged into the vast  expanse of the encyclopedia. Then there are the articles they  pass that  should have been seriously  tagged or even CSD'd,

I personally see that there is a clear consensus for preventing new pages, especially those made by SPA to promote themselves, their band, or their company from going live immediately they press the 'Save' button. It would also stop the vandal, hoax, attack, and nonsense pages dead in their tracks. It would thus also address a lot of issues that BLPPROD has failed to prevent. It would dramatically reduce the need (or even abolish it completely) for NPP to be done by inexperienced but keen editors, who could then apply for the now redundant right of Reviewer. Ultimately it would also significantly  reduce the number of articles to  be speedied, and/or sent  to  AfD, allowing  editors and admins to get on  more with  the job  of content  building,, SPI,  civility  issues,  and blocking vandals.

The Wizard does not prevent poor articles from going live - its recommendations do not technically prevent the creator from going forward and posting live. Do we need to provide diffs for the dozens of unreferenced "Article name is..." articles that the NPPers have to cope with daily to prove yet another point? As I suggested above, clicking 'save' in the Wizard should queue the article for the New Article Reviewers who would pass, CSD or tag in the same way that is done at NPP. If the article is OK, it can go online, if not, the creators can use their 4 days/10 edits to smarten it up and publish it when they are autoconfirmed. It still has to pass the eyes of the New Article Reviewers.

I think this poll is unnecessary. Asking for a poll for a trial, especially this one with its 12 voting criteria, would be to return to 'Go' and re-debate the whole proposal again - so is the nature of Wikipedia discussions -  for one thing, as it did at BLPPROD, it will attract more users who were not around from the start, who  haven't  read the 1,000s of words of debate so  far, who will again make endless new proposals. Even straw pools that are supposed to be a simple Y/N affair get bogged down with new threads started within the voting columns. Debate, re-debate, and counter-debate, is a counter-productive process. We need to draw a line and say when a proposal has reached maturity. I believe this one has reached that point, and it should now be implemented for as long as it takes to come up with conclusive data that it is seriously affecting the number of quality articles that are created, or driving  new editors away  from  the project.

I realise this is all TL;DR  for some but I'll reiterate again for those who missed it: a great many popular and serious web forums, blogs,  and collaborative web sites require a moderator's approval before posts go live, and it doesn't suffocate the creative will. For something as serious as an encyclopedia it's essential and a major feature that  was simply omitted at Wikipedia's inception. Kudpung กุดผึ้ง (talk) 08:24, 23 May 2011 (UTC)


 * I don't disagree with anything you say. But if I was wrong, and the proposal isn't dead, how can it be moved forward now? Rivertorch (talk) 08:43, 23 May 2011 (UTC)


 * By closing the thing  off with  a summary  that  establishes the  very strong, but not overwhelming, support for restricting unassisted article creation in the main namespace consensus for it . (I can't  close  it  because I'm involved). I  suggest  that  there is to  be no  trial and that  it  goes ahead broadly  as suggested by Whatamidoing, by  getting  the technical  aspects sorted out and tested, and then implementing, with  the proviso that  it  continues for as long as it takes to come up with conclusive  data that it is seriously affecting the number of quality articles that are created, or driving  new editors away  from  the project,  and needs to  be reverted. Kudpung กุดผึ้ง (talk) 08:58, 23 May 2011 (UTC)


 * I asked to close the discussion at User talk:RL0919. Because of the discussion's length, complexity, and format, he may or may not be able to close it. Cunard (talk) 09:04, 23 May 2011 (UTC)
 * Kudpung, let's be serious a minute. If this 'feature' had been in place at Wikipedia's inception, then no user could have ever created an article and no user could have ever edited anything. Cenarium (talk) 14:33, 24 May 2011 (UTC)
 * Cenarium, let's be 'serious' a minute. If Wikipedia had existed in 2000 BC, no user could have ever created an article and no user could have ever edited anything, because computers didn't exist. Rd232 talk 14:41, 24 May 2011 (UTC)
 * Above Kudpung wrote "For something as serious as an encyclopedia it's essential and a major feature that was simply omitted at Wikipedia's inception.", this was a response to that. Cenarium (talk) 15:00, 24 May 2011 (UTC)
 * Ah, well, bad indenting on your part. I agree the feature couldn't have been a very early one (though it could have been done a while ago). Rd232 talk 15:58, 24 May 2011 (UTC)

In response to WhatamIdoing, it is the purpose of the next questions (automatic reversion, trial effectiveness and trial safety) to clarify the specifics of the trial. It asks if we turn it off immediately after, and which criteria we should check before making further decisions. If we have consensus for the trial, then we'll discuss the specifics within the conditions set forth by the additional questions. The questions which you suggest are open ended and will not result in anything clear or definitive. In response to Kudpung, this is an important issue and needs to be properly debated, the WMF recently asked us to take particular care on this issue. This RFC has just accumulated views with varying level of endorsement, this didn't produce any kind of consensus. There has been more than 500 participants, no view gained a majority of endorsement; we don't look at the most supported view to gauge consensus, and the users who endorsed the most supported view do not agree on how to implement this. We need a clear proposal presented to the community. Users who wish to see the implementation permanent without trial can vote against an automatic reversion. Cenarium (talk) 12:36, 23 May 2011 (UTC)
 * "didn't produce any kind of consensus" - um, pardon my French, but, vraiment?? :P See my comment above, 09.46. Rd232 talk 13:10, 23 May 2011 (UTC)
 * And see the sentence which immediately follows the sentence which you cite. Cenarium (talk) 13:37, 23 May 2011 (UTC)
 * ? what? "no view gained a majority of endorsement" isn't true: there are two opposing positions exemplified by Ironholds and Jayron; the other views are variations thereof (or weakly supported complements), so we can focus on those, and that gives a 30/70% split which is considered consensus. "we don't look at the most supported view to gauge consensus" isn't true unless you add "in isolation", and if you do, it doesn't mean much. Finally, the arguments on one side are much stronger. PS Arguably, all the !votes by people who think the proposal would prevent non-autoconfirmed users creating pages at all should be discounted as invalid, since there is wide agreement that there should be mechanisms for them to do so with some form of assistance. Rd232 talk 13:57, 23 May 2011 (UTC)
 * It is incorrect to assume that those two views represent the whole of the discussion. You are disenfranchising the views of more than 150 editors. Consensus is a deliberative process, working on discussion, contradictory debate, compromise, driving towards an outcome taking into account as much of the stated concerns as possible; we never gauge consensus on a policy issue by having different views presented and counting how many support them. Consensus building on a proposal is totally different of dispute resolution through RFCs. Here I can't see any kind of consensus which emerged. I don't think that you are uninvolved enough to analyze the strength of arguments. Cenarium (talk) 14:18, 23 May 2011 (UTC)
 * I'm quite happy to discount the views of those 150 editors (how'd you get that number?) who didn't bother to endorse either one of the core views, because prima facie they weren't paying enough attention (which is why so many of the later views particularly are pretty pointlessly duplicative). I'm also willing to bet that a conscientious closer wading through them all would be able to easily assign most of those 150 to either camp, and that the balance of opinion would fall in the same 1/3 to 2/3 ratio in favour of supporting the proposal which has consistently been there from the beginning. And you can think what you like, but most of the opposition is justified with false assumptions, handwaving about supposedly breached principles and refuted arguments; it all more or less boils down to (a) "I don't like it" and (b) "it'll put people off". A isn't an argument, and the responses to B are well covered so I won't repeat them. PS I assume you're not confusing editors with Views, because that would produce the sort of ridiculousness which a while back (2003?) saw the House of Commons spend a day voting for variations of how to elect the House of Lords, and despite overwhelming support for at least 50% election, couldn't agree anything so the Lords remains completely unelected to this day! Rd232 talk 14:31, 23 May 2011 (UTC)
 * There has been more than 500 participants, of which 332 voted for one of the two most supported proposals, so more than 168 users didn't vote for either. There has been no clear proposal on which users were asked to weigh in, just an accumulation of diverging views with varied level of support. There is no basis for coming to the conclusion that those supporting Jayron's view want an immediate implementation without further discussion. And even if they did, that didn't get a majority of support. If the other users wanted to support Jayron's view, then they would have. If you want to make an analysis by percentages, then that ought to rather be 44% for Jayron's view, 20% for Ironholds' view and 36% for other views. There is clearly no consensus that emerged, further discussion is needed. Cenarium (talk) 00:56, 24 May 2011 (UTC)

I suggest a small group of editors drafts a trial proposal and presents that as a yes/no proposal to the community. Trying to draft a proposal in an open RFC is likely to produce several mb of discussion with no solution. There will always be disagreement about how to do the trial, so it is better to lay out a well defined proposal and hope people accept an imperfect trial when the alternative is no implementation at all. Yoenit (talk) 13:55, 23 May 2011 (UTC)
 * Since the supporters by and large are OK with just doing it, the burden of drafting should be on opposers, at least initially, and if they fail to come up with anything workable, it should be done without a formal trial, but with an informal review scheduled for 3 months and 6 months. Rd232 talk 13:57, 23 May 2011 (UTC)
 * That would only work if this RFC is closed as consensus to implement, but I strongly doubt if it will ever be closed at all. Yoenit (talk) 14:09, 23 May 2011 (UTC)


 * There is a consensus in principle to move forward. 70% in favor of moving forward with no preconditions, even if you account for the canvassing on meta against it. Then there's a slew of other compromise proposals involving various kinds of preconditions and trials. The problem is there isn't an obvious consensus on how to implement it. But again, there is a consensus in principle that this is a good idea if done correctly. Shooterwalker (talk) 15:10, 23 May 2011 (UTC)


 * Jimbo Wales stated today that he prefers data supported rationale over emotional arguments. If  applied, - because the WMF has asked for special attention  to  paid to  this proposal -  it  may discount a significant number  of the votes that  are not  in  favour of the proposal. Kudpung กุดผึ้ง (talk) 11:02, 24 May 2011 (UTC)


 * Don't we all?
 * But the point is that we already have data. The question is only whether we have enough data and relevant data.
 * IMO the answer to that question—Do we have enough of the right kind of data?—is determined by directly asking that question (and its natural follow-up question, "If not, then what data do we need to get?"), not by asking, "Do you want to have some kind of a trial?" WhatamIdoing (talk) 18:47, 24 May 2011 (UTC)
 * Data we would need: the number of articles created, the number of new articles kept after 30 days, the number of new editors, the number of new editors still active after 30 days. And definitely, some kind of before/after picture, so we can look at numbers and percentages to see how this proposal changes things. Any other data we need? Shooterwalker (talk) 21:56, 25 May 2011 (UTC)

Closing
I have volunteered to close this discussion, with the caveat that it will take several days (among my other life commitments) to review the lengthy discussion and properly summarize it. My preliminary estimate is Friday, which isn't all that far in the future in the grand scheme of things, although I realize folks involved with this proposal are eager to move forward or move on, depending on the outcome. I will post here if there is any change in that expectation.

Please note that I will be making offline copies of the relevant pages to facilitate review when I'm not online, so any comments added after this posting are likely not to be considered in my closing. --RL0919 (talk) 17:02, 23 May 2011 (UTC)
 * I put together this table which is an accurate organization of the RFC data, to the best of my ability and knowledge. So far, we've been using Ironholds vs. Jayron32 as a proxy for the level of support for this proposal. When you factor in the other comments, the %'s basically don't change overall, which is surprisingly convenient. But it's worth looking over regardless. We're still back to what quantity and quality of support would be considered a consensus, and how specific or vague that consensus is on the details of this proposal. Shooterwalker (talk) 22:14, 25 May 2011 (UTC)

Where is a discussion of the trial going to take place? Protonk (talk) 20:46, 29 May 2011 (UTC)
 * User:Kudpung/RfC for trial (draft) was mentioned at WP:VPR. See also the related discussion at my user talk page, about whether or not to include an exemption for creation via the Article Wizard, as was discussed during the RFC and (I thought) had widespread support. Rd232 talk 20:55, 29 May 2011 (UTC)
 * Ah. I should put that on my watch list.  My deep worry is that we will half-ass this trial and not gather relevant data on article not created.  Unfortunately that data and the ability to design a trial which will appropriately provide it resides with the developers.  I sincerely hope we do not just redirect all new accounts to AFC/AFC wizard for three months and call that a trial.  Not for something of this scale. Protonk (talk) 21:04, 29 May 2011 (UTC)


 * Lest there be any misunderstanding: That link was put on the VP in GF in error. It's caused some people to think that those rough scratchings in my user space are the real thing. That said, there are no suggestions that all new accounts should be sent to AFC or Wzard. All that the proposal and consensus concerned was that users must have autoconfirmed status to be able to create articles (that will go live online). Any other ideas and or exceptions to that  rule should be treated separately and proposed and ciscussed in  the normal  way - consensus was quite clear on  this. As for the trial, the duration  of which  is the reason for the draft  of a draft  for a suggestion  for a proposal  for a discussion, what criteria needs to be gathered during the trial? What metrics should be applied? What scripts/bots need to be written that current stats gatherings don't cover? What  suggestions are there for counting  articles that  are not  made  and what  they  weren't  about? Kudpung กุดผึ้ง (talk) 22:12, 29 May 2011 (UTC)


 * This is why I want some clarification. First, if the system changes to allow only autoconfirmed editors the ability to create new articles, what happens when a non-autoconfirmed editor follows a red-link.  right now they have the option to create the page.  Should the trial go live, would they have the option to go to AFC, be immediately redirected to AFC, or simply be shown a page that says "you need X edits to make a new article"?  And if we make a trial, developer input is necessary because the downside to this change is reduced participation from new editors.  If we collect statistics available only to editors, we will only see raw article creation rate, article retention and user retention for users who found the AFC or made enough edits to get autoconfirmed before creating the article.  What we need to see as well are what percentage of users reach a page that says "you cannot create an account" and just stop editing.  I would prefer to see a randomized trial, but I can outline that at a trial talk page (which is why I am looking for one). Protonk (talk) 22:18, 29 May 2011 (UTC)
 * On the tech aspect, I suggested long ago that MediaWiki:Nocreatetext would be moved to MediaWiki:Nocreatetext-anon, and a new version of it created. It would be much the same, but replacing the "log in /create" line with "click here to start an article". That link would go a page looking something like this mockup: User:Rd232/creationdraft (though that would obviously need amending if the Wizard doesn't grant a creation exemption). See also the tech discussion in Kudpung's userspace draft. Rd232 talk 22:32, 29 May 2011 (UTC)
 * I have added an alternate proposal here. Protonk (talk) 23:09, 29 May 2011 (UTC)

Draft RFC on implementation
OK, we now have User:Rd232/RfC for trial (draft phase) as a venue for kicking around thoughts on implementation, which should then become a formal RFC. Let's see how we do with that for now, and in particular how much we can firm up details like what statistics to collect. I'm hoping visitors to this page will help lick this into shape before we present an RFC to a wider audience. Rd232 talk 00:21, 30 May 2011 (UTC)

bugzilla 30208
30208 ("Trial for restricting non-autoconfirmed users from creating new articles on enwiki") was created on August 3, and has been resolved as "WONTFIX". The bugzilla page isnt a good location to vent or comment about this, and etiquette on bugzilla recommends not adding "me too" type comments. Bugzilla allows voting on a bug, which is a good way to vent, especially as the vote tally is typically ignored by the devs anyway. There are software discussions continuing at Article creation workflow, and a process related discussion at Ignoring community consensus. John Vandenberg (chat) 04:02, 17 September 2011 (UTC)

There is some related discussion at User talk:Philippe (WMF). It would be good to have a discussion around the next steps now that the bug has been marked WONTFIX. This proposal isnt going to go away because the WMF has concerns with how it aligns with their strategy (maybe the strategy is wrong?! ;-)), but we need to take in the feedback from WMF. We don't want this proposal to be merged into Perennial proposal.

It seems the WMF are interested in developing something else to complement the requested site configuration change, so it will be helpful to know their proposed timeline so the community can work with that. Also, can the community help the WMF by collating more data in order to assist explaining this change to the public in a positive way?

I think this proposal will have more success if we reduce the impact of the trial.

We could get a reasonable amount of data from running this configuration alteration for only one day of the week, for only four weeks. That would avoid anyone getting the impression that this is a permanent change.

Also, we could run this trial without any changes to the site permissions, by using an WP:Edit filter rule to implement the new restriction. We could also change the local site messages (e.g. MediaWiki:Noarticletext) and have no software enforcement of the new rules. John Vandenberg (chat) 06:20, 17 September 2011 (UTC)
 * The idea of using the abuse filter for this purpose is quite subversive to what the WMF and devs have said. Not to say I'm against it, but it makes one wonder what their reaction would be toward it. Killiondude (talk) 05:00, 18 September 2011 (UTC)
 * Using an edit filter has subsequently been mentioned on the bug, and people who oppose this limit have threatened that anyone who does it using the edit filter will be desysoped. The desysop would be doing it on ideological grounds rather than technical grounds, as the required edit filter would be quite efficient. '(article_articleid == 0) & (article_namespace == 0) & !("autoconfirmed" in user_groups)'  All of those filter components are already calculated in other filters.
 * IMO enabling a edit filter for a very limited period (one or two days) will allow some real data to be obtained. John Vandenberg (chat) 15:07, 18 September 2011 (UTC)
 * To respond to Killiondude (and thank you, John, for pointing me toward this talk... it is appreciated).... I don't, honestly, have a strong feeling for how the WMF and devs would feel about using the editfilter for this purpose. I'll be very honest and say that I'm not as "learned up" on the edit filter as I should be.  I'd appreciate pointers that would fill me in, Killion?  Philippe Beaudette, Wikimedia Foundation (talk) 05:07, 19 September 2011 (UTC)

I support the change (I hope that, although fewer people might start as an editor, far more of them will be retained), but I think it will require a lot of work. It's not just the code, which requires scarce dev resources. We also have to change all the documentation and warn the community.

I do think that a six-month trial is much longer than necessary, but I'm not sure that four days spread over a month is feasible. WhatamIdoing (talk) 23:26, 19 September 2011 (UTC)