Wikipedia talk:Autoconfirmed article creation trial/Archive 5

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1 ← Archive 3 Archive 4 Archive 5 Archive 6

Deletion log not shown

Resolved
Tracked in Phabricator
Task T176100
Resolved

If I log out and for example click Tceqli then I'm taken to https://en.wikipedia.org/w/index.php?title=Wikipedia:New_user_landing_page&page=Tceqli. It doesn't show the deletion log [1] or mention the page has been deleted. I think it should show the deletion log, or at least link to it if it's difficult to show it. PrimeHunter (talk) 11:27, 16 September 2017 (UTC)

Is details about which admin deleted according to which code useful to a brand new editor? Maybe so, maybe not. Perhaps that just adds confusion. Legacypac (talk) 20:00, 16 September 2017 (UTC)

IPs can be experienced. They see the same as registered non-autoconfirmed users. Another issue is that "What links here" shows the useless Special:WhatLinksHere/Wikipedia:New user landing page instead of Special:WhatLinksHere/Tceqli. We are removing information and navigation from users who may want and need it. PrimeHunter (talk) 22:51, 16 September 2017 (UTC)
because IPs can't start a page I guess. Legacypac (talk) 00:47, 17 September 2017 (UTC)
IPs can create draft and talk pages, and modify or remove links to deleted pages if they can find the links. I often use WhatLinksHere on red links when I have no plans to create the page. PrimeHunter (talk) 09:02, 17 September 2017 (UTC)
The landing page doesn't make it easy to find the links to do this stuff, though. I consider it a regression, because the users who are least likely to know their way around MediaWiki are no longer going to have these helpful links readily available to them. RhoLands (talk) 09:26, 17 September 2017 (UTC)

What, if anything, is the page parameter used for? I see a getVal( 'page' ) in WebInstaller.php but that's about it. Do we need a new magic word so that the landing page will have access to the value of the page parameter, making it possible to link to WhatLinksHere, and to the deletion log for the red linked page, etc.? RhoLands (talk) 08:56, 17 September 2017 (UTC)

  • IPs and new users can still use "what links here", they just have to do it manually. Same with the deletion log. There are tradeoffs with any solution here: the blacklist is the most bitey/has the worst UI, edit flitsrs would have caused new users to create new pages wholesale, and this causes IPs and new users to have to manually check logs. The last is the least bad option for new users: the vast majority of whom have no use for everything being talked about here, and given how many recreations we saw, i doubt the deletion log was that useful in preventing recreations. Also, RhoLands, it would be useful if you commented here with your primary account. TonyBallioni (talk) 15:04, 17 September 2017 (UTC)
Really User:RhoLands you created ACTRIAL in mainspace? Legacypac (talk) 16:08, 17 September 2017 (UTC)
Does anyone know what the page parameter (&page=Tceqli) is being used for? Has the code necessary to make use of it been developed yet? As it is now, the landing page is giving inaccurate information when it says "You can create it". It would be better to say, "After you become autoconfirmed, you will be able to create it" to avoid the obvious question, "Okay, how?"
I don't see it as so much a matter of trade offs, as a matter of implementing the necessary features to get the design as good as it can be. Even if you're not going to recreate an article, it's good to be able to quickly navigate to the pages that tell you what links there and what the deletion history is.
The thought crossed my mind that if the landing page were implemented as a special page, the page parameter could be included in the system message as a $1. That would also allow the displayed title of the special page to be something more useful than new user landing page. RhoLands (talk) 16:40, 17 September 2017 (UTC)
I think showing the deletion log would be a net negative for the overwhelming majority of 250k+ users who see this page every day: it would ruin the design and be intimidating. The poweruser IP/new editor does exist, but they are far from the norm. When creating a system like this, thinking about the median end-user is necessary. The what links here might be something the WMF can work on, I'm not tech savvy enough to figure that out, but it is still pretty easy for the poweruser IP to get to that page, it just takes a few extra keystrokes. Again, I'd encourage you to engage with your main account: it doesn't help with goodwill to create a SPA that only exists to comment on ACTRIAL. TonyBallioni (talk) 17:01, 17 September 2017 (UTC)

IPs and new editors can't create pages or become reviewers, I don't see any need that they have for the deletion log. In any case, they can still do it manually if they want to for some reason. For IP editors, the solution is simple. Register an account. — InsertCleverPhraseHere (or here) 18:54, 17 September 2017 (UTC)

@PrimeHunter: This is definitely a bug. Logged out users should see the same thing they've always seen. I filled a bug at T176100. Kaldari (talk) 00:40, 18 September 2017 (UTC)
This should be resolved now. Logged out users should no longer be sent to the landing page and thus should be able to easily access the deletion logs. Kaldari (talk) 00:26, 19 September 2017 (UTC)
Kaldari, does the system discriminate between logged-out users and new users? Or have we just removed the only path to the carefully created ACTRIAL landing page? StarryGrandma (talk) 19:43, 19 September 2017 (UTC)
StarryGrandma, yes, this is only for logged-out users. Non-autoconfirmed users are going to the landing page. -- DannyH (WMF) (talk) 20:30, 19 September 2017 (UTC)

Watchlist banner

resolved (non-admin closure) Legacypac (talk) 08:07, 20 September 2017 (UTC)

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

Why is there a watchlist banner? Every logged in editor will see that including brand new editors. Broadcasting the trial info to brand new editors will only encourage work arounds by sockpuppets and drama from drama creators. An actual new editor would not know there is even a trial - just think it is normal to get a little experience before creating a new page. An autoconfirmed editor not in NPP or AfC should not notice or care about this trial as it has no effect on them. Please remove the Watchlist notice. Legacypac (talk) 11:36, 15 September 2017 (UTC)

I think I'll second this. The relevant people have been informed via talk pages of various wikiprojects (AfC, NPP). I can't remember where, but I'm pretty sure this was brought up as a suggestion at some point and it was shot down by several editors. I don't think it is a good idea to widely inform new editors about the specifics of ACTRIAL, it will only encourage gaming the system. — InsertCleverPhraseHere (or here) 11:52, 15 September 2017 (UTC)
I'll third it. I was quite taken aback by it when I logged in. ACTRIAL shouldn't interest anyone beyond the people who are managing it. Kudpung กุดผึ้ง (talk) 12:38, 15 September 2017 (UTC)
As an outsider, I must say that I disagree with a few of the viewpoints here. Wikipedia has always told people that they can create articles once registered; a change to requiring autoconfirmed status for article creation, assuming that it becomes permanent down the line, is significant in the image that it sends (that the editor base is starting to aim for quality over quantity). In fact, I wouldn't be surprised if it receives some attention from the media once they figure it out. It's much more important to the site than the latest RFA or how the next ArbCom election will be structured. Add in the past dispute with the WMF over the issue, and this is of wider interest (I think) than the NPP regulars are giving credit for. If you believe this trial is in the best interests of the website, why wouldn't you want the community to know it has started? We're going to !vote on it at some point, and some of us might want to see how the process is working before deciding whether to support it. As for new editors, they'll figure out workarounds to ACTRIAL regardless of whether a watchlist banner is in place. The people wanting to add spam will eventually find weaknesses in the system and tell their spammer friends about them. How ACTRIAL deals with the weaknesses that are raised will determine how successful it is in reducing the number of spammy new articles. Giants2008 (Talk) 13:38, 15 September 2017 (UTC)
I'm in agreement with Giants2008. Thank you for the watchlist notice, which was the first I'd heard of this proposal. It makes me feel less of an outsider and better informed for the next stages. Certes (talk) 15:17, 15 September 2017 (UTC)
Banner removed and noted on the todo list. I also took off the To Do list the idea of a signpost article on ACTRIAL as that has the same problems as the banner. How those work arounds are created will be interesting. The first 4 days of data should not be contaminated by the banner. Wikipedia has the motto "anyone can edit" not "anyone can create an article in mainspace". This is about how we channel new users interested in content creation. It's not a big change at all. Legacypac (talk) 13:44, 15 September 2017 (UTC)
I'd be fine with a Signpost article. As Kudpung has noted, this is a milestone in the history of Wikipedia, so I think having a signpost article on the topic would be fine. The thing is that the signpost isn't that active these days.
I agree on the watchlist notice being overkill. Was that one of the 2011 suggestions that we missed pruning? There'sNoTime, thanks for your good humour in taking it down when people raised objections. TonyBallioni (talk) 14:26, 15 September 2017 (UTC)

I saw the item on the to do list but did not realize what it was really. Signpost is ok, but when there are some results to report. Right now it will just create drama. Legacypac (talk) 14:43, 15 September 2017 (UTC)

In 2011, Watchlist notices had not yet become Wikipedia's headline banner for every piece of tittle-tattle. It was used exclusively for fundraising and the occasional major election such as for stewards and the Board. The death of Signpost is itself indicative of just how disinterested everyone has become in all the meta aspects of Wikipedia and carrying out associated tasks. Kudpung กุดผึ้ง (talk) 16:01, 15 September 2017 (UTC)
  • It seems quite sinister and telling that the enthusiasts for this scheme want to keep it secret. I expect this will be counterproductive – see the Streisand effect. Andrew D. (talk) 18:17, 15 September 2017 (UTC)
    • We've posted it in all the major forums on-wiki, its been sent via bot to anyone subscribed to the admin newsletter or the NPR newsletter, and we're considering writing something about it in the signpost. There is no secrecy here. My only reason for thinking the watchlist notice is overkill is that it'd be the third one we have running, and at this point there is nothing to discuss or do: we've made the change, and its currently running. If you can think of other forums where we could notify people, please let us know or do so yourself. This isn't a hidden thing. TonyBallioni (talk) 18:23, 15 September 2017 (UTC)

We all know Andrew D does not like ACTRIAL. His post is exactly the kind of drama we want to avoid. ACTRIAL does not ikpact existing editors (except in workload at NPP/AfC) so alerting everyone by watchlist is not useful at all. Legacypac (talk) 18:42, 15 September 2017 (UTC)

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

Effect on outreach activities

Hi all

Just to let you know this is going to make it hard for new editor workshops I'm running where new editors want to cover specific topics, which will very often include new articles. I'm going to get them to do some small edits so that they will get permission to create the new articles, however looking at previous conversations on this page that looks like it will be considered suspicious behaviour... Not really sure what to do.... suggestions?

Thanks

--John Cummings (talk) 12:54, 25 September 2017 (UTC)

There's plenty of small tasks in Category:Wikipedia backlog that newbies can do. I suggest correcting {{citation needed}} tags or finding sources for unreferenced/poorly referenced articles in the topic of interest; evaluating source quality and existing article content are both things they need to be able to do anyway. Alternatively, one could expand existing articles. MER-C 13:23, 25 September 2017 (UTC)
You can also simply have them work in draft space and submit their articles to AfC. From there you can review them yourself or get an experienced helper to review them. Having been at an editor meetup, working in draft is the best way anyway, and there is nothing stopping you or anyone else that is autoconfirmed from sidestepping AfC completely and moving the new articles directly from draft to the mainspace for the new users after they have drafted them. Though, I will point out that by moving them you are de-facto endorsing them, so you'll want to review them as you go. — InsertCleverPhraseHere (or here) 13:37, 25 September 2017 (UTC)
@John Cummings:, you could also talk to RexxS about it. If I recall correctly he suggests user sandboxes and not using the draft space/article wizard. Re: small edits: there is nothing wrong or suspicious about that. Part of ACTRIAL is aimed at getting new users accustomed to editing and improving the encyclopedia before they create an article. The discussion above was about users who do this such as give out 10 barnstars or do 10 reversions of blanking their own user page to get AC status. TonyBallioni (talk) 13:43, 25 September 2017 (UTC)
Based on my experience in facilitating workshops, I would say that new editors who have edited their user page to say what they are doing, or trying to do, are not treated with suspicion. If the experienced editor who is running the workshop then also leaves a note on their talkpage to welcome them to the workshop, and shows that they continue to councel the new editor and assist with responding to rejections or reverts, that usualy provides a supportive environment for new editors. Mduvekot (talk) 13:50, 25 September 2017 (UTC)

I guess my concern is also about editor retention, one of the main things I try to do when I work with new users in person is to help them leave with a sense of accomplishment so they continue editing, them being able to create a new article is the main source from which they get this sense of accomplishment. If both I and the potential longer term editor invest our time in getting them to a point where they are able to contribute to Wikipedia independently it would be a real shame if they didn't feel a sense of accomplishment and agency that creating a new article often gives them. Whilst Article for Creation is another option "Congratulations, your article has joined the back of a queue of 1600 articles and when it gets to the front (I don't know when) you will be told if it is good enough for Wikipedia" does have quite the same ring to it..... --John Cummings (talk) 13:54, 25 September 2017 (UTC)

Have them create it in their sandbox and move it to mainspace for them when it is ready for publication. Either that or allow them to take it home to work on and then move to mainspace. I know I always create articles as drafts and then publish them after they are no longer stubs. This is a good way to do it to avoid the tagbombers. Showing them how to do this would probably help them longterm IMO. TonyBallioni (talk) 13:59, 25 September 2017 (UTC)
@TonyBallioni: move it to mainspace for them – that works while people are at in-person events but isn't encouraging independence. Richard Nevell (WMUK) (talk) 15:47, 25 September 2017 (UTC)
Understood. Unfortunately the community has decided to oppose any exemption from ACTRIAL for users who attend editathons and outreach events. There are workarounds: have the participants create an account four days beforehand, which should enable them to move their sandbox into the mainspace or move it yourself. I wish there was a better answer, but trying to find a way around the ACTRIAL restrictions for attendees was put to the community twice under two different RfCs and rejected both times. TonyBallioni (talk) 15:57, 25 September 2017 (UTC)
It's always worth advising people to create accounts before events, but uptake is never 100%. I understand there are ways to deal with this situation, but that doesn't get round the fact that it is a barrier to entry. Richard Nevell (WMUK) (talk) 16:14, 25 September 2017 (UTC)
I have found that letting new editors at edit-a-thons create articles directly into mainspace unsupervised often ends in tears. There's always someone who goes ahead and tries it anyway and ends up disillusioned and discouraged, unlike someone who started in userspace and had leave me a note and say: "Hey, can you look at this for me and move to to mainspace if it's OK"? Mduvekot (talk) 16:28, 25 September 2017 (UTC)
There are plenty of successful editathons and workshops, some where all the editing takes place in sandboxes, some where only pre-existing articles are edited, and where no new pages may be created. That's all fine but there will be some people who want to create new pages. At gender gap events for instance one of the reasons people come through the door is to improve coverage of topics which might not even be mentioned on Wikipedia. We've made that harder to achieve. Richard Nevell (WMUK) (talk) 16:55, 25 September 2017 (UTC)
It's not a simple situation, and for editathons much depends on the prior experience (if any) and capability of participants, and the extent to which they have prepared for the event. Event organisers on the wiki-side should be taking this opportunity to liaise early with the local organiser(s) and asking them to encourage participants to register their Wikipedia account at least 5 days in advance. We now have a good reason to push that. The skill-set required to create a new article is considerably greater than that needed to improve existing articles by adding a new section, expanding the text, adding more sources, etc. I will always start new editors off in their sandboxes for obvious reasons, and ensure that they perform more than 10 edits at that point. When they are ready to work in article space, my principal concern is for them to have a good first experience, and expanding an article is far more likely, IMHO, to result in that, than attempting article creation from the outset. Working on an existing article as their first task in mainspace exposes them to the style and format in an existing article, which is much easier to learn from and mimic than starting an article from scratch - let alone all of the considerations of notability they would have to become familiar with. Nevertheless, some participants will have come prepared with sufficient sources and will have the skills to create an new article, but even then, I'll get them to do the draft in their userspace. The AfC system was not designed for our style of training. Our best bet in the case of those new editors is to increase the chance that they will be autoconfirmed by the time they are ready to move their userspace draft into mainspace so they can do it for themselves. if not, then moving it for them is still a good choice. My experience is that the creation of the draft and seeing it grow is the rewarding part; actually moving it into mainspace is less gratifying. YMMV. Cheers --RexxS (talk) 17:02, 25 September 2017 (UTC)
I largely agree, but when people are working on new articles moving it into the mainspace is a big moment for the editors. Richard Nevell (WMUK) (talk) 17:14, 25 September 2017 (UTC)

I'm with John. This is a real detriment to the editing workshops I've been running - successfully - for several years, with new editors creating articles. Yes, I can work around it, but it will take more time and reduce motivation for new editors. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:35, 25 September 2017 (UTC)

  • People running events for new editors should describe putting new articles through AfC as normal - it can make complete sense that somebody should use training wheels or go through drivers ed before they ride a bike or drive on their own. If the presenters describe this as some aberration, it is good for no one. Make the exciting moment pushing the big blue "submit" button ("yes, I really think it is ready now!"). Please don't project your unhappiness with the trial onto people who have no preconceptions or reasons to be unhappy with what should be presented to them as something as normal as drivers ed or training wheels. Jytdog (talk) 19:41, 25 September 2017 (UTC)
    • AFC is an aberration. It is not fit for purpose: I've seen too many publishable articles rejected on bogus grounds. I don't recommend anyone to use it, much less someone with an experienced Wikipedia editor on hand to help them write and publish their first article. There is no "should" about it. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:22, 26 September 2017 (UTC)
If you want to go down that road, Wikipedia isn't fit for purpose full stop. I've seen too many notable articles created in mainspace only to be slapped with {{db-person}} or {{db-spam}}, and too many good faith edits reverted with lots of Twinkle boilerplate. You know who I'm talking about - the sort of people who never ever write anything in mainspace but just mash buttons. I remember you getting quite irate that The Mariposa Trust was deleted, as I did too. I fully appreciate this is going to have a knock-on effect on courses, but we don't have a solution that will solve all problems everywhere. In an analogous way, it's annoying when it's dark and raining, the outside light has blown (again!) and I can't find the right key on my ring to open the front door - but it's still better than having my house burgled. Ritchie333 (talk) (cont) 09:40, 28 September 2017 (UTC)

The idea of ACTRIAL is very sound and it's working. As Jtdog suggests - using AfC should be presented as the normal way to go. The direct to mainspace way still ran the page through NPP (multiple reviewers checking it) and subjected the pages to CSD PROD etc which is much harsher than AfC. Nothing stops a facilitator from becoming an AfC reviewer as well. Don't be scared by the backlog message (which we should change actually). Most pages get reviewed within hours. Legacypac (talk) 19:48, 25 September 2017 (UTC)

And yet ...
I don't want you to think I'm being rude about the good work that volunteer reviewers do, but I simply don't want participants at events I run to risk having that sort of wait before their article gets published. I can't see why any experienced editor should not be capable of judging when a userspace draft is ready to be moved into mainspace. Although I'm happy for other trainers to use the AfC process, I simply disagree that it should be presented as the normal way to go at events. --RexxS (talk) 20:51, 25 September 2017 (UTC)
those stats are correct but don't tell you the older pages are usually either resubmissions no one has gotten around to declining or overly complex topics that need a specialist to review. I work abandoned drafts and submit a few pages a day to AfC. The real wait is usually hours. Further any autoconfirmed user can move the page out of AfC to mainspace, including you. If you sign up for WP:AfC you can use the AfC tool that makes the move much easier. Legacypac (talk) 22:24, 25 September 2017 (UTC)
I already move drafts into mainspace when I judge them to be ready, and have always done so, but thanks for reminding me. There's no sign-up at WP:AfC, but anyway I'm not sure I need a tool to help me (1) click the move button and (2) remove any template clutter. I've not found a problem doing that manually so far. --RexxS (talk) 23:29, 25 September 2017 (UTC)
"The real wait is usually hours." - as regards AfC, this is very much not the case from my experience. A turnout of less than two days is extraordinary, and most submissions wait months, if not years. It doesn't scale. Ritchie333 (talk) (cont) 10:18, 28 September 2017 (UTC)
CommentRichard Nevell (WMUK), Andy Mabbett, John Cummings, you guys are overlooking the easiest solutions. They can't move from draft/sandbox to main space, but you can. It takes 30 sec to shift it for them if you feel like bypassing ACTRIAL. Why is this a big deal? I also think that we should be encouraging new editors at meetups to use AfC as normal, and you can simultaneously facilitate that 'instant gratification' of article creation by having the organisers review their articles for them on the day (you don't have to rely on the guys over at AfC and wait in line, you can just join them for the day and review the AfC submissions of your group yourself). Why is this hard?
This attitude seems to stem from either "I want to do it the way I always have" or "I don't want the responsibility to have to review the submitted articles". Either way, tough luck. You can't do it the way you always have; this is a feature, not a bug, and there are easy solutions.
Overall ACTRIAL is helping SO MUCH to improve the quality of new page submissions, and I have a sneaky suspicion that the overall higher quality for new articles means that New page patrol are responding to the higher quality of submissions by getting a little less bitey as a result. So this is good for new users too. — InsertCleverPhraseHere (or here) 20:00, 25 September 2017 (UTC)
Not overlooked at all. I am fully aware it's an option, indeed responding to a comment along those lines above. It doesn't change the fact the trial will make matters more complicated. Richard Nevell (WMUK) (talk) 22:01, 25 September 2017 (UTC)
Not overlooked at all here, either. Indeed, I wrote: "Yes, I can work around it, but it will take more time and reduce motivation for new editors. ". AfC is not the "normal" way of writing new Wikipedia articles. Your assertions about our reasons for concern are an egregious failure to assume good faith. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:30, 26 September 2017 (UTC)
  • All of this substantiates my belief that by making the "article" the unit of analysis and fetishizing the creation of new articles, we encourage a lot of POV warriors that are only participating in Wikipedia in order to write their own narrative in a space they can feel they own. Everyone knows this is a bad idea for any number of reasons. Beyond the excellent explanation given by InsertCleverPhraseHere, I'd like to point out that expanding existing articles and providing much-needed sources doesn't require autoconfirmed permissions and actually accomplishes much more. These sentiments lay bare the fact that editors aren't joining so that they can contribute to an encyclopedia and I, for one, am glad if they become disheartened before I visit their talk page. Chris Troutman (talk) 20:10, 25 September 2017 (UTC)
    • " These sentiments lay bare the fact that editors aren't joining so that they can contribute to an encyclopedia" Utter poppycock. No-one evry left one of my training sessions having been given the impression that they could "write their own narrative in a space they can feel they own". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:34, 26 September 2017 (UTC)
  • Wouldn't it be possible to simply request the confirmed user right for the workshop participants (at Wikipedia:Requests_for_permissions/Confirmed) a day or two before the workshop? If that is currently not a valid reason to request that user right, then we should work to add it as a valid reason, at least for the duration of ACTRIAL. ‑Scottywong| comment _ 20:18, 25 September 2017 (UTC)
    • I think that particular proposal might have gotten shot down in the recent pair of RfCs on the topic (not explicitly, but it is implicit in many of the comments in the opposition to giving the ability for event organisers to grant confirmed status. I think the general consensus was that editathon editors are not necessarily any more capable than normal editors. I repeat: there are easy solutions within the current framework that take just a little adjustment of the current editathon modus operandi. There is no reason to need to game the system to 'get out' of ACTRIAL, or to make special rules for users starting out in editathons. — InsertCleverPhraseHere (or here) 20:38, 25 September 2017 (UTC)
      • The recent RfCs shot down the idea that non-admins should be able to confirm users. I think admins could be able to do it in some circumstances, and it was floated as an idea before those RfCs for editathons. I'll ping Xaosflux since he is the most active person on that PERM board.
        I generally agree with the idea that it is not horrible for people to be forced to wait, but there are some editathon coordinators who are telling us they see this as an issue. I think we should be sensitive to their needs, while still recognizing that ACTRIAL is going to run for 6 months, and that there is not going to be a way for non-admins to confirm users. Creating a set of best practices for how to run ediathons during ACTRIAL is a positive thing: ACTRIAL is not simply about preventing the creation of articles. We are trying to also help new users learn Wikipedia so that when they create their first article it is more likely to be kept. I don't think it is all doom-and-gloom on the outreach/editathon side, but I also think we need to listen to them and shouldn't be so defensive about ACTRIAL. It is just a trial after all: it looks to be working out very well, but its purpose is to figure out how the autoconfirmed restriction will affect en.wiki. Working with people who coordinate editathons will only be a positive, because it will help when formulating the followup RfC.
        I want ACTRIAL and its eventual outcome to be something where the English Wikipedia wins, not just one group of editors, and we can only do that if we take into account all the feedback we are getting while it is running. TonyBallioni (talk) 21:01, 25 September 2017 (UTC)
      • (edit conflict) @Scottywong: You suggestion would require me to collect a username from every participant some time before the event. Then I would have to make a request for "Confirmed" for a group of people whom I did not know. Leaving aside the point that if they have a username four days before an event, they will be autoconfirmed once they've done 10 edits with me, that scheme would effectively require me to vouch for those people's bona fides which I'm not prepared to do until I've met them and learned a little more about them. I don't think that's unreasonable. --RexxS (talk) 21:07, 25 September 2017 (UTC)
        • Yeah, that's a problem. Requesting confirmed status and getting an administrator to give it to a bunch of random users isn't really any different than non-admins being able to confirm users (in the sense that neither the admin nor the event coordinator can vouch for those individuals). It seems like everyone is looking for a solution that will make it so there is zero impact on event coordinators, but realistically, ACTRIAL is all about changing the process of new users creating new articles; there really wasn't any chance of it having zero impact on events focused on new users creating new articles. Having been involved in a IRL meetup for Women in Red, I have to say that I don't see an issue with having the event coordinator simply tell the new users that all they have to do is put their hand up when they think their article is ready to be moved to mainspace and the event coordinator (or another autoconfirmed helper) can simply look it over and move it for them. I feel like a broken record here, but can someone please point out to me why this isn't the most practical solution? — InsertCleverPhraseHere (or here) 21:19, 25 September 2017 (UTC)
          • No, you're exactly right: it only needs one computer with a logged in autoconfirmed user to do all of the moving once you're happy with the draft. Although, having spent a lifetime in the classroom, the last thing I want at events now that I'm retired is to have people putting their hands up. I'd recommend to all helpers that they patrol around the participants, helping and advising - and at the same time evaluating any new draft articles being developed. I'd feel I'd been slacking if I wasn't aware of when a draft was ready before the editors themselves. --RexxS (talk) 21:38, 25 September 2017 (UTC)
    • or even a smart phone to do a move. Legacypac (talk) 22:32, 25 September 2017 (UTC)
Reply to @TonyBallioni: above: historically, 'confirmed' was liberally applied to editors that could be identified to be real people - as the only real impact was related to uploading images and captcha avoidance. To determine how to deal with it during ACTRIAL depends on what is actually trying to be measured: Article creation by pure vandals-external bots-etc, or more broadly article creation by "new" editors. If the later, then bypassing it by passing out confirmed is missing the point. As far as mass in-person editing events go, if their purpose is to create articles, I'm personally inclined to support InsertCleverPhraseHere's views above: create them in draft/sandbox then have anyone there do the "go live" if they are ready during ACTRIAL. — xaosflux Talk 23:01, 25 September 2017 (UTC)

Arbitrary Break

  • I agree that there are simple solutions here, and I think that a bigger deal is being made of this than it should. My posts are mainly that we should work with editathon coordinators to figure out solutions to their concerns: they are going to be impacted. The community consensus on the matter is that editathon participants should not be automatically exempted from ACTRIAL. What are the best ways forward for them given these two concerns? I think userspace creations that are then moved if ready is the best way forward if a participant really wants to write an article. The other way is to emphasize improving current articles (we have plenty from the early days that desperately need it). TonyBallioni (talk) 21:26, 25 September 2017 (UTC)
I've mentioned several times that one solution is to have an admin on call to do the confirming during the editathon. I have yet to meet an editathon participant who came along with the intention of learning how to vandalise Wikipedia. Kudpung กุดผึ้ง (talk) 01:10, 26 September 2017 (UTC)
As someone who has run dozens of Edit-a-thons with new users i would just like to support the thoughts of John Cummings, Andy Mabbett and others. New users, often lack confidence and understanding of Wikipedia's complex governance. They don't want to be faced with work-arounds, they want, and expect, a simple and intuitive interface for contribution. I am sure that these changes will negatively effect edit-a-thon output and longer term user retention. Also, I hear lots of people here suggesting the the trainer simply moves articles to the main space, however for many trainers and Wikipedians in Residence it is very important to have good metrics demonstrating the impact of events. Publishing articles on behalf of new users will skew these metrics and make it even harder to measure user behavior and the success of events. For me, some method for event trainers to auto confirm new users would be greatly welcomed, and as these users will be working in a controlled and supervised environment i cannot see that this would be anything other than beneficial. Jason.nlw (talk) 08:13, 26 September 2017 (UTC)
Jason.nlw, this has been put to the community twice. Both times there was a strong consensus that event participsnts should not be able to be autoconfirmed by event coordinators (I've linked to the close above). This is unlikely to change. The best we can do is help think through how to operate in this new environment. The community has been very clear that it does intend for ACTRIAL to apply to outreach events, however. I personally don't agree with that, but it is what the current consensus is. TonyBallioni (talk) 11:36, 26 September 2017 (UTC
How many event coordinators were involved in that discussion (I certainly was not aware of it, or of what was planned)? And since when did the wiki way of working mean that bad and harmful decisions cannot be reversed? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:14, 26 September 2017 (UTC)
Unsure how many event coordinators took part in the RfCs: they took place at the village pump, both were advertised on CENT, and the first proposal was advertised as a watchlist notice. Consensus can change, but it is unlikely that a third RfC in a two month period would have any different result than the first two, especially since the recent consensus was pretty strong. TonyBallioni (talk) 18:01, 26 September 2017 (UTC)
You did comment in the first related Request for Comments discussion. I appreciate, though, that it may have been unclear what the followup would be, given that a number of possibilities were under discussion. isaacl (talk) 16:24, 27 September 2017 (UTC)
Your link is to an August 2017 discussion, about mitigating the harm of the trial. The RfCs about whether to hold one, and discussed above, were in 2011 (and, frankly, anything discussed that long ago and not acted upon until now should trigger a new discussion). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:10, 27 September 2017 (UTC)
My apologies; in Tony's comment to which you responded, he said "Both times there was a strong consensus that event participsnts should not be able to be autoconfirmed by event coordinators (I've linked to the close above)," and he linked to the second of two RfCs in August, so I thought you were referring to them. I understand your feelings regarding the discussion for the trial having been discussed many years ago. isaacl (talk) 02:02, 28 September 2017 (UTC)
Jason.nlw You wrote: Publishing articles on behalf of new users will skew these metrics and make it even harder to measure user behavior and the success of events I don't know how you do your metrics, but for the events that I have been involved in, moving articles into mainspace on behalf of new users would not have skewed them. It has never been a problem for us. Mduvekot (talk) 12:17, 26 September 2017 (UTC)
Whatever solution is found must apply equally to the trial and to whatever enviroment will be imposed on new editors thereafter. A fair trial cannot allow special measures to allow the creation of articles during the trial which would have been lost to Wikipedia had the proposed changes which you're testing already been in force. Certes (talk) 14:07, 26 September 2017 (UTC)

New editors don't know there is a trial or that had they started a month back they could put their creations in mainspace unless someone tells them. It used to be an IP could put stuff in mainspace but not now. Everyone expects some escalation of user rights over time. You don't start at level 200 on Candy Crush. User:Pigsonthewing's point about AfC is correct in that there are too many declines of valid pages. I'd urge them to join AfC and review declines for such pages to accept. Then when they run an event they can use their AfC status to approve good submissions. Legacypac (talk) 16:47, 26 September 2017 (UTC)

Another event coordinator here, adding their concerns, & support for points made by John Cummings, Andy Mabbett & Jason.nlw. I wish that I'd seen the original discussions. Whilst I wholeheartedly support anything which increases the quality of our encyclopedia, I'm very concerned about the additional barriers to entry that this presents. I work mostly in sector-wide advocacy (museums in 2015-2016, and now public libraries), and I've observed new-article-pride to be a great motivator in creating a positive disposition to further institutional work for GLAMs. Yes, of course there are workarounds, but it's not making life any easier when you're bringing good faith / legitimate users on board - especially when, in my case, these are new users who you hope to take the good wiki-word back to their organisations. I do appreciate the frustration when this has already gone to RfC; I've been out of the loop for a little while so I'm not surprised that I didn't see it, but to see other experienced event coordinators here expressing concern (and surprise) would suggest to me that the consultation process could stand some improvement - perhaps this is something which could be taken into consideration in the future? I'll be interested to see the effect of this trial on my own work and that of others for the trial duration... Lirazelf (talk) 10:24, 27 September 2017 (UTC)

  • Lirazelf, I get the frustration on your all's end too. Out of curiousity, what could have been done to make this more visible (serious question, not sarcasm)? The first RfC on the topic received a watchlist notice, and the other was advertised on WP:CENT for a month. I'm not sure if Kaldari advertised it in other ways, but this is our typical way of advertising sitewide discussions. Any feedback would be appreciated. TonyBallioni (talk) 12:34, 27 September 2017 (UTC)
TonyBallioni Heya, thanks for this. Very much appreciate the understanding, and I do appreciate the amount of work that gets done by those working in these arenas! So: I can't think that I saw any mention of the original discussion on any of the mailing lists or FB groups that I'm on (GLAMWiki, WMUK, Scotwiki, Libraries etc) Some of those are small/UK-centric, of course, and there are a few of them, so I don't imagine for a second that hitting every possible mailing list or FB group would be appropriate for every topic in WP:CENT, it would get ridiculous. But where the outcomes of particular policy discussions would likely have a big impact on the work of a specific group of editors (in this case, Outreach / WiR), I wonder if there's a way that they could be flagged for more targeted approaches for feedback? (Perhaps if a particular issue is deemed to be important enough to merit a watchlist notice, for example?) Would an approach to appropriate local/national chapters to ask them to solicit feedback from relevant stakeholders have been feasible in this instance? Again, I'm aware that I've been out of the loop for a little while... Lirazelf (talk) 14:18, 27 September 2017 (UTC)

Lirazelf - out of the loop for a little while? With only 386 mainspace edits since you joined in 2014 and almost totally absent for the last year, you've never been in it. Your valuable outreach work is is of course important, but editathons are also done by people like RexxS, for example, whose contributions demonstrate that they not only know what is going on and who have contributed to these discussions, but who actively contribute to content and the management of it. Since a row of RfC with very high participation of 500 users and impressive consensus, the trial, which is not irreversible at the end of its run, has never ceased to be a topic of discussion - a discussion about why the WMF initially refused to implement it under a fake pretext. The en.Wiki has undergone a mutation and is in a far worse state now than it was in 2012 and urgent measures need to be undertaken and possible solutions have to be examined and experimented. The discussions for the roll out have been broadly participated, in depth, since I put pressure on it in May 2016, and more so since it's implementation has been heavily discussed and planned for the past four months since the WMF finally expressed their wish to conduct it.

The old RfC was carefully discussed and crafted. It was not a question of yes or no, it was an ensemble of viewpoints that fully showed each participant's view on the topic. (...) ACTRIAL is not new, it is unfinished business that is years behind schedule (...) —Esquivalience

We don't go back and discuss every policy after 5 years because someone who is out of touch decides they don't like it. Please consider keeping up to date with the product you are selling - your are many models out of date. You can get in tune by thoroughly reading here, and here and their archives. It would help you understand today's challenges, and then you might probably even want to help. Kudpung กุดผึ้ง (talk) 01:31, 28 September 2017 (UTC)

I think we would do well to remember that people's experience and value cannot be reduced to an edit count. Richard Nevell (WMUK) (talk) 09:30, 29 September 2017 (UTC)
You need to tell that to half the people who turn up to RfA. Actually, you need to staple it to their heads. Ritchie333 (talk) (cont) 10:19, 29 September 2017 (UTC)
@Lirazelf: Respectfully, targetted notification of groups likely to be disproportionately affected by tangentially related RFCs is certain to be interpreted as inappropriate canvassing. It is an issue. Your work in outreach is appreciated, irrespective of your activity levels here at enwiki. Snuge purveyor (talk) 18:24, 30 September 2017 (UTC)
  • I guess my assumption was that if an editathon host requested confirmed status for his/her participants in advance of the event, then that host would be expected to supervise the participants while they edit, and ultimately be responsible for the articles they create. If the editathon produces a bunch of crappy articles, then the host would be blamed. And if that happened repeatedly, then their future requests for confirmed status would be denied. Seems like a reasonable system of checks and balances to prevent abuse. ‑Scottywong| confess _ 05:05, 28 September 2017 (UTC)
  • I think we're missing another trick, albeit one that most people here will recoil at with horror. Above, RexxS says he doesn't want to be a gopher going round moving all his class' userspace drafts into mainspace. (Out of interest, how many are we talking about here? 10? 100? 1,000?) If somebody asked me to come along to the British Library for an afternoon and do the "gopher" work for the course, and pays me, say, £500 (plus lunch and travel expenses) for it, hell yeah I'll turn up and do it. I realise this sounds like "paid editing", but it's much easier to sell time-consuming work when there's money involved. Ritchie333 (talk) (cont) 10:24, 28 September 2017 (UTC)
That's too cheap. I charge that much for sitting in front of my computer all night (UTC+7) at home to do it. Kudpung กุดผึ้ง (talk) 12:49, 28 September 2017 (UTC)
I'm sorry I gave you the wrong impression, Ritchie333. For the record, I always go around helping the event participants, and if any have created user drafts of new articles that I think will survive in mainspace, I'm more than happy to move them myself – or even better, get the helper who is operating the computer linked to the projector to do the job, as the whole room can share the moment when the new article is published.
You sell yourself too cheaply – when I did paid IT training work for big companies 20 years ago, I wouldn't get out of bed for less than £1,200 per day plus expenses. Unfortunately, neither the WMF nor WMUK can afford to pay my rates, so I have to do it just for the joy of taking part. Getting rich is overrated, anyway. --RexxS (talk) 17:39, 28 September 2017 (UTC)
Well yes, but the point wasn't so much that I want to make lots of money (I'm sure I've told enough people at meetups now about the time I did some weekend IT work for Goldman Sachs (in its pre-2008 crash days) and got enough to go straight from the job down to Denmark Street and buy a Rickenbacker 4003 bass sitting in the shop window), rather that I can easily justify spending a day in London away from my family if they know I'm going to come back with some cash we can then spend on "fun" stuff. Ritchie333 (talk) (cont) 10:18, 29 September 2017 (UTC)

Grants application

A grant has been applied for to develop a solution that may or may not assist in the way new pages are controlled. The community is invited to comment on the request at metaKudpung กุดผึ้ง (talk) 22:54, 3 October 2017 (UTC).

Special sandbox functionality

Have a look here: User:ManosHacker/sandbox, all of the issues mentioned above have a single solution. Over 900 hours on systematic courses to beginners in Wikipedia since 2014 produced this magic in code, after their needs. Have you ever seen a user sandbox created or moved to wrong space? The new editor will do it on his/her own, after the restriction expires. Foolproof 100%.

  • 8-10% of new articles in Greek Wikipedia in 2016 were produced after my students, who had their first steps using this tool.
  • A class of 8-year-olds created 11 articles in 2017.
  • A cherry in the pie is editathons, see how you can prepare an article template palette for people who can only fill in content.
  • There is custom page template functionality for advanced users, too.

Click New page, Help and visit a child sandbox, to see the change in functionality and adaptation of the template.

  • In the bottom of the page version of the sandbox call, click on medical condition to create an article sandbox on this template.
  • Have a look here, too, for real article example: User:Janstee/sandbox. When the sandbox becomes an article, it clears itself from the sandbox list.

   ManosHacker talk 07:15, 1 October 2017 (UTC)

This is of special interest to RexxS and anyone else who facilitates editathons. Kudpung กุดผึ้ง (talk) 23:02, 3 October 2017 (UTC)

Listing the Backlog

How do I list the backlog of pages that were created by non-autoconfirmed editors? Is there a filter that I can use for the purpose? Robert McClenon (talk) 05:27, 19 September 2017 (UTC)

Robert McClenon, you can see it by going to Special:NewPagesFeed, clicking "set filters", and then clicking "were created by new editors", and then setting the filters. TonyBallioni (talk) 05:34, 19 September 2017 (UTC)
Reporting problems
I get "Showing: unreviewed, new editors" 13906 pages - maybe we can occasionally log the count here to track how fast it disappears. Legacypac (talk) 08:10, 20 September 2017 (UTC)
To be clear, it should not disappear because for many pages we should allow hours before they are ready to be reviewed. I would say ideally we are down to about a hundred and should stay at that level.--Ymblanter (talk) 08:32, 20 September 2017 (UTC)
Ymblanter this is only pages by non-autoconfirmed editors, so there will be no more coming in and we can get it to zero. Overall backlog - yes it would be wonderful to get to a few hundred pages. Legacypac (talk) 09:30, 20 September 2017 (UTC)
I believe 14K is the total backlog (I see there pages created after the ACTRIAL started), or am I am missing something?--Ymblanter (talk) 09:33, 20 September 2017 (UTC)
I'm missing that the unreviewed, new editor filter lied to me. It now shows "unreviewed, new editors 334 pages in your filtered list". That number should get to zero and hopefully the 14,000 drops to a few hundred. Legacypac (talk) 09:49, 20 September 2017 (UTC)
Yes, this was indeed my point.--Ymblanter (talk) 09:51, 20 September 2017 (UTC)
  • 227 unreviewed from New Editors. 13,955 total unreviewed Legacypac (talk) 05:40, 22 September 2017 (UTC)
  • 201 unreviewed from New Editors 13,934 total unreviewed Legacypac (talk) 15:49, 22 September 2017 (UTC)
  • 128 unreviewed from New Editors. 13,583 total unreviewed pages Legacypac (talk) 10:53, 25 September 2017 (UTC)
  • 1 unreviewed from New Editors 12,936 total unreviewed pages. The pre-ACTRIAL new user backlog is gone and the overall backlog is slowly dropping Legacypac (talk) 06:17, 6 October 2017 (UTC)

Trivial editing detection

I've been wondering if any work was put into enhancing (or creating) trivial edit detection? To achieve article creation many accounts are likely to do fast trivial editing. Detecting this and blocking such with the least effort possible may be ideal (edit filters helping, etc). Thanks, —PaleoNeonate – 05:19, 14 September 2017 (UTC)

MusikAnimal, I think you may have mentioned a filter for this in the past (or I could be making shit up). Regardless, it is actually pretty easy to spot manually. TonyBallioni (talk) 05:22, 14 September 2017 (UTC)
This could be an interesting thing to track, as part of the research for this trial. Is there a way to distinguish bad-faith trivial editing (specifically done to get around the block) from a good-faith effort to follow the instructions on the landing page, and get some editing experience before you try creating an article? -- DannyH (WMF) (talk) 14:52, 14 September 2017 (UTC)
Special:AbuseFilter/806 (private) sort of detects "(extended-)confirmed dancing", as I like to call it, but there's no 100% foolproof way to detect innocent minor edits from bad-faith. We could introduce a filter to detect if they created an article on their 11th edit, as soon as they are autoconfirmed. That might be interesting to see — MusikAnimal talk 15:00, 14 September 2017 (UTC)
Thanks for the replies. Logging new article creations at 11th (or 11-20th) edit is also an interesting idea... In any case, I'll be following the results/stats of this trial too. —PaleoNeonate – 18:12, 14 September 2017 (UTC)
11-20 edits article creation would be really interesting to see. Legacypac (talk) 19:07, 14 September 2017 (UTC)
I'm working on doing some analysis along those lines, we're interested in understanding how article creation relates to user experience level. Not sure when I'll have it ready, but I'll make a note that I should post something here when I do. Cheers, Nettrom (talk) 00:31, 15 September 2017 (UTC)
It would be neat if autoconfirmed would not be conferred automatically if the respective user's talk page had vandalism warnings. If we spot a series of "test edits" or other nonsense we could defeat those edits counting and force new users to at least be constructive. Chris Troutman (talk) 19:15, 14 September 2017 (UTC)

I am concerned by the wish to deter a new editor, who fixes ten typos then makes a decent attempt at starting an article, from making further contributions. You've got your trial; let's see how much new blood still flows in before discussing any further measures. Certes (talk) 10:52, 15 September 2017 (UTC)

Fixing typos occasionally is different from adding and deleting spaces or making rapid fire minor edits to run up an edit count. We can tell the difference. Legacypac (talk) 11:24, 15 September 2017 (UTC)

There could be legit reasons why experienced users would make 10 minor edits just to get autoconfirmed. It could, for instance, be someone who has a main account and now wants to create some articles on sensitive topics they don't want linked to their real identity, so now they want to hurry up and get autoconfirmed. RhoLands (talk) 18:15, 17 September 2017 (UTC)

I could be mistaken, but I question that this is legitimate use: it would probably be better to abandon the old account and create a long term anonymous one... —PaleoNeonate – 18:30, 17 September 2017 (UTC)
WP:SOCK#LEGIT explicitly allows creating an alternate account to work on a controversial topic, so long as the person is careful not run afoul of any other sock-issues. However I would expect that user to make 10 reasonably productive edits when building auto-confirmed status. Alsee (talk) 07:00, 20 September 2017 (UTC)
I now see what you mean. Thanks, —PaleoNeonate – 07:17, 20 September 2017 (UTC)
  • There are very few legitimate reasons for making 11 minor edits just before then posting a spammy article with an immaculately formatted list of sources. I have come across many such articles since ACTRIAL was rolld out - and I do not patrol as often as I used to, so I assume I'm only seeing the tip of the iceberg. Kudpung กุดผึ้ง (talk) 00:23, 12 October 2017 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Autoconfirmed_article_creation_trial/Archive_5&oldid=1076155851"