This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.
I was going to touch base with Barkeep49 directly on an entry we've both dealt with (Leo Garcia / Draft:Leo Garcia (actor)), but as I realize I'm really not sure what to make of the situation, I hope Barkeep won't mind if I ask here for general input--specifically on dealing with NPR-draftified entries if the creator subsequently edits, then moves the entry back to mainspace without waiting for AfC review (as happened here).
Thinking of it from the other side, I would not do this if it were my entry waiting in the AfC, it would definitely feel out of process--but on the other hand, it's true AfC is not strictly obligatory (right? at least for anyone autoconfirmed, which this creator was even when they started the entry), and I sort of do worry we cross into unilateral deletion-by-AfC if we repeatedly revert these moves (here it was a copy-paste which adds further issues, but just staying on the general case for simplicity's sake). It seems to me maybe after a single draftifying, after that it could be best to let something live or die by the regular mainspace deletion processes?
On the other hand, it does seem a pretty straightforward, ah, work-around, of ACREQ happened here: exactly 10 mainspace edits, one week delay and then the new entry posted. But I don't know how/if that should affect our use of draftifying... Thoughts? Thank you all. Innisfree987 (talk) 21:08, 15 August 2018 (UTC)
Realizing there's an additional practical wrinkle in this instance, which is that regardless, this wasn't going to be the main search term for that name, so the redirect had to be restored to point to the dab in any case...and at the same time, the draft is not well-sourced enough that I personally would move it into mainspace (others might, but it's borderline at best), so, in this situation it's a bit hard not move back to draft a second time; the only alternative would be to move something into mainspace myself that I didn't think was suitable... Hrm. Innisfree987 (talk) 21:18, 15 August 2018 (UTC)
As a believer in the collaborative nature of Wikipedia I of course do not mind someone seeking more thoughts on this topic. I am incredibly sympathetic to the long AfC wait currently going on. I feel bad for those people and when asked give advice on how to strengthen sourcing to make it more likely a review will be favorable. However, the long wait doesn't mean I'm looking for workarounds. Given the profile of the editor and that their first attempt was a Copyvio I thought it appropriate in June that this go through AfC followed by NPP. Given that we have no disclosed COI AfC is not strictly required but again given the user profile I do not think I'm assuming bad faith when I say there's a good chance there is an undisclosed COI hence why I've been a little bold on pushing it back into draft space. I welcome the thoughts of others on this and other similar cases. Best, Barkeep49 (talk) 21:29, 15 August 2018 (UTC)
Honestly I see this as a perfect application and an ideal result of draftification. The original version was a copyvio and was Revedled and converted to a redirect (it also could have easily been CSDed, but revdel-redirect is another alternative). it was then recreated, but the creator didn't have an understanding on what kinds of sources were appropriate so Barkeep49 draftified it. The author then re-wrote it, including better sources the second time around and it is probably about as well sourced as it ever is going to be. Notability is probably still a bit borderline, but I think it might survive AfD (depending on what the content of this article is). The original version draftified would certainly not have survived AfD. — Insertcleverphrasehere(or here)21:51, 15 August 2018 (UTC)
Agree entirely about the first draftifying; my question is just about this reversion after the user tried to copy-paste move the improved draft--in effect a second draftifying, on the grounds that Article has not yet been approved in AfC process. Though like I say, I don't have a different suggestion... Innisfree987 (talk) 22:03, 15 August 2018 (UTC)
Ah... I see. I missed that part. Users should generally feel free to opt out of the AfC process, as it is optional, time consuming, and also users should feel free to be WP:BOLD (draftification itself is a matter of WP:BOLD). If the user decides to opt out of AfC, I would generally not recommend re-draftification (as this essentially becomes edit-warring or trying to make AfC mandatory - which it isn't). Instead I'd recommend that the reviewer pursue other means of reviewing the article (tagging/AfD/CSD/etc). Barkeep49 notes above that he has concerns about COI (which if true would necessitate AfC as mandatory), but this has not been brought up with the author and is not the reason stated in his edit summary here. All that being said, the copy-paste move is inappropriate, as the page history really should be maintained for an article like this. I'd recommend to Barkeep49 that we respect the opt-out of Joeysortino and move the draft to mainspace for them, then review it without option to draftify. If his concerns of COI are high, I'd recommend that Barkeep49 follow this up by asking the user if they have an undeclared COI (if they do have one, or fail to reply, then draftifying and running through AfC would be the correct channel). — Insertcleverphrasehere(or here)23:07, 15 August 2018 (UTC)
Yeah, that sounds reasonable enough to me; I'm really not sure what else you could do, barring, I don't know, moving the draft to mainspace and then AfDing yourself, which sure wouldn't be a friendlier option! (And thanks for your understanding, Barkeep; just wanted to emphasize I intended it as a general question rather than criticism!)
Further thinking: actually, if the user had moved to mainspace correctly, wouldn't the redirect left behind make it rather difficult to re-draftify, anyway? (Except as sysop/page mover?) Maybe the technical limitations more or less mean it'd have to go to AfD (or not as the case may be), and the question really is just--if this specific thing happens, is this the way to handle or does anyone have a different idea? Innisfree987 (talk) 22:03, 15 August 2018 (UTC)
1 edit redirects can be moved over by anyone with autoconfirmed. Note that when Joeysortino made the copy-paste move, they would not have been able to do an overwrite move, as the redirect in question at Leo Garcia had more than 1 edit. They could have moved it to Leo Garcia (actor) though (which is what I recommend we do now, as the user has shown a desire to opt out of AfC). Experienced NPR reviewers should request the Page Mover user-right anyway, as draftification is difficult without it (has to manually request a CSD R2 of the redirect left behind in main space if there is no reasonable target for the page title). I'll put some information about Page-Mover utility to NPR in the next NPR newsletter. — Insertcleverphrasehere(or here)23:15, 15 August 2018 (UTC)
Based on the logs the page wasn't patrolled at either title. The article hasn't been reviewed and is not indexed. I'm not sure what caused the patrol link to not show up. — JJMC89 (T·C) 19:23, 11 August 2018 (UTC)
If it still shows in the NPP, that's less concerning. Hmm does that mean that it'll however remain there because noone can tag it? (I didn't check if it's in still in the new unpatrolled pages log as I'm leaving, will be back in 5 or so hours). Thanks again, —PaleoNeonate – 20:15, 11 August 2018 (UTC)
Super. Hmm it's possible that the older tools are not being maintained and are now buggy... NPF and curation don't work for me (but NewPages does and the "mark this page reviewed" bottom link is what I can use when present). Thanks again, —PaleoNeonate – 03:00, 12 August 2018 (UTC)
@PaleoNeonate: to clarify; the page curation tools don't come up for you on some unreviewed pages? If not could you describe the issue in more detail? Is there any indication of what is causing this issue? What browser are you using? Does the issue persist when logged in with a different browser? Another thing to check: next time you see the issue, blank your common.js page, then check if the issue persists. If that fixes it, re-enable scripts one at a time until the issue comes back and you might find where there is a conflict. Not sure what you are trying to say here, but it seems that you are familiar with fiddling with scripts... What do you have enabled/disabled that is different to the default that might be causing the issue? — Insertcleverphrasehere(or here)00:36, 13 August 2018 (UTC)
I use a custom browser without javascript (but with very good html/css support). When I did use JS with Wikipedia in the past, some things worked and others not, likely because there was a number of User-agent specific code; following links with anchors would almost never align at the proper page position, pages took much more time to load making patrolling changes tedious, etc. So currently without JS, some relatively new tools don't work at all like Special:NewPagesFeed. However, Special:NewPages works fine. When a page is not yet patrolled, I also usually see a "mark this page as patrolled" link at its bottom right corner, which works to mark a page patrolled when clicked. However, that is the link which was not present for the aforementioned article even if it still was unpatrolled. I have noticed this issue before at times, so I wondered if the page was really unpatrolled (I couldn't see a patrol log entry). It was confirmed above that it was indeed unpatrolled. So at least it's not the type of bug which causes pages to escape patrolling (at least those using NewPagesFeed appear to see it as such). Thanks, —PaleoNeonate – 08:01, 13 August 2018 (UTC)
PaleoNeonate, for well over 20 years JS is as essential to a web pages as html and CSS. Without it, only the simplest static sites will work. To do any kind of maintenance work on Wikipedia, a browser must have the full compliment of what are today's standard features. Web developers assume everyone uses reasonably up to date browsers on both desktop and mobile devices. MediaWiki uses a modern web environment. Kudpung กุดผึ้ง (talk) 23:31, 13 August 2018 (UTC)
That's not completely true: JS was only used for useless gadgets until about 10 years ago and even then for any extended functionality plugins were required (something that only changed recently with HTML5). That said, as someone who knows about DOM and its issues and how sites can gracefully degrade, once WP can no longer work it'll unfortunately join a list and I'll retire. —PaleoNeonate – 23:56, 15 August 2018 (UTC)
I would say feel free to combine those pages if you want. I agree that there isn't much reason to have two separate pages on Page Curation. Their talk pages are already merged. — Insertcleverphrasehere(or here)23:12, 13 August 2018 (UTC)
I was feeling bold and given ICPH's nod I went ahead and did the merge. In the end I could find no content at Wikipedia:Page Curation that wasn't already on Wikipedia:Page Curation/Help so I didn't end up merging any actual content; I hope others check to make sure I did not miss something worth merging. I then changed the shortcuts that went to Curation to Curation/Help, updated the navigation tabs, and put a superseded template on Curation. Best, Barkeep49 (talk) 23:26, 13 August 2018 (UTC)
That a good point. It does appear that everything at Page Curation was already covered on the other page. I renamed the nav bar heading "Page Curation Tools", as this fits wht the function of the page a bit better than the previous header. — Insertcleverphrasehere(or here)23:48, 13 August 2018 (UTC)
Anyone is welcome to help fill out the mysteries in this table. Maybe from here we can figure out a logical way to combine some of these pages, because I'm pretty certain we don't need this many. Natureium (talk) 00:47, 14 August 2018 (UTC)
While some of your descriptors, especially of the pages not in the header, made me chuckle (coordination of coordination) I'm not as convinced they're all as superfluous as you do. To me the R&D page is the one that seems like it should be removed from the header bar as inactive and supplanted by the current work done here and on its talk page. Best, Barkeep49 (talk) 19:09, 15 August 2018 (UTC)
Ok. One of the problems I see with having so many discussion pages is that questions get duplicated because there are multiple places that are appropriate for the same question, so if someone has a suggestion to remedy this, please give a shout. Natureium (talk) 19:32, 15 August 2018 (UTC)
Could the school serve a useful purpose? Especially with time limited PERMS seems like it could be a positive to the project. Best, Barkeep49 (talk) 20:20, 15 August 2018 (UTC)
@Barkeep49: I certainly have considered it. We could remake the school as a place where users that meet the basic requirements of NPR could apply for probationary NPR, then be mentored by an experienced Reviewer. Whether they pass would be dependent on the opinion of the mentor at the end of the probationary period. This would require a rewrite of the NPR School page, and a reworking of its purpose, but advertised properly, it could be very successful in developing new users, and could be used as an alternative to turning down NPR applicants at PERM (e.g. "Have concerns about experience level, please apply to the NPR school instead"). — Insertcleverphrasehere(or here)21:59, 15 August 2018 (UTC)
No changes need to be made to anything. DeltaQuad is aware of this and knows how to use the new time limited rights feature. The auto user talk page message has already been updated. I would not recommend according the right to anyone before they have gradiated. It's a point DQ can make when processing PERM reequests. They will make declines with something like "Not done for now. Please enroll at the NPP School and learn all about it first" Kudpung กุดผึ้ง (talk) 03:09, 16 August 2018 (UTC)
Thanks for the response Kudpung but I must not have phrased my question clearly. This wasn't about whether you and DeltaQuad are screening appropriately (of which I would hope there's no doubt). The question is if there was a more active NPP School is that something which would help you or which you would direct people towards enrolling in? If the answer's no it's not worth time to think about it. If the answer were yes or maybe it might be worth seeing what's possible. Best, Barkeep49 (talk) 03:16, 16 August 2018 (UTC)
@Barkeep49: Keep in mind i'm jumping in on this discussion, and have not much clue about what is being talked about. I don't know if there is a particular scope I'd want to see an NPP school about except for use of the patroling interfaces. Then a separate guide about how to handle new pages, so people can point to established practice. The rest of the need for the school I feel would kind of be moot, because I screen applicants for bad approvals and good approvals, and that can always be taken up individually. I hope that answers your question. -- Amanda(aka DQ)22:24, 16 August 2018 (UTC)
Yeah, Amanda tends to like to see AfC experience in absence of other patrolling experience, which I kind of agree with. I remember that I actually first sharpened my teeth at AfC before getting involved in new page patrolling, and I think it provides a good environment for practicing and learning the notability guidelines and what constitutes a good page, prior to jumping at NPP. However, AfC doesn't appeal to everyone, and the requirements to do AfC reviewing are not much different than the NPR userright these days, so perhaps it might be useful to still have some sort of mentor-ship program as an alternative to AfC when users are turned down for the NPR user right. Amanda, would you consider commenting on 'failed' NPR applicants suggesting that they find a mentor at Wikipedia:New_pages_patrol/School? — Insertcleverphrasehere(or here)22:54, 16 August 2018 (UTC)
Mentoring is usually done in the process of doing something, so I'm not sure how mentoring someone without the perm would be effective. Natureium (talk) 23:32, 16 August 2018 (UTC)
Not pinging her but Amanda's feedback was definitely helpful. I agree that what I had envisioned was mentoring/schooling of someone with the PERM. As that doesn't seem to be useful to Amanda, it seems like the school might not be the best place to invest energy (as least as I had thought of it). Best, Barkeep49 (talk) 00:11, 17 August 2018 (UTC)
I agree that the school is not the place to invest limited resources. If an editor does not have the knowledge, experience and skill to read the NPP procedures, ask about what they are unsure of and then get on to reviewing in a non-problematic way then they can simply come back and ask for the right once they can. Since the granting criteria are, in my opinion and previous experience, too lax as they are, we should be investing a considerable amount of our more skilled reviewers time in checking the work of other reviewers. Spot checks for the more experiences and more intensive for the less. Specific issues can then be identified and some form of counseling and remediation can be done – that is where a more in depth curriculum would be useful not for 'onboarding'. Jbh Talk00:53, 17 August 2018 (UTC)
@Natureium, people forget that the page curation tools are not the only way to do all of the patrolling processes for articles. Any user without the NPR user rights can feel free to install WP:TWINKLE and then go to Special:NewPagesFeed and start going through and tagging, nominating for deletion, etc. the NPR user right actually only actually gives us one ability: to officially 'review' an article. They also activate the WP:Curation toolbar, and I honestly think it is a shame that the PC tools are hidden behind a userright like that, especially when the twinkle tools have all the same functionality and are available to anyone with autoconfirmed. @Jbhunley: In the old wild west days when I started, there was no restriction, and I just installed Twinkle and then started slow. I would often tag or nominate articles but not mark them as patrolled, for fear that I'd missed something. Users can't really do that anymore, because we are walling them off from the curation toolbar and telling them that they can't be trusted with the tools unless they are already experienced with article review. Where are they supposed to get this experience? I'm not so much in favour of investing time in the 'school', but rather investing time in editors that are keen to get involved but don't know where to start. At present inexperienced reviewers who want to get involved often just get turned down at PERM with no advice on where to go to get the experience needed to attain the user-right. We should be encouraging these users to seek a mentor to watch over them, install twinkle and start having a crack at the New Pages Feed. If they keep a list of all articles they have 'patrolled', then their mentor can simply come along and re-review them all to see if there was anything improper, and if they missed anything. This gives them feedback to improve, and allows them to help out while gaining the experience necessary to become a solid reviewer. We also would avoid any issues of improper reviewing by newbies that isn't checked over again later by a second set of eyes. — Insertcleverphrasehere(or here)02:14, 17 August 2018 (UTC)
I well remember the times of Before :) As you said there is Twinkle (I think it was a huge mistake to split functionality between two software packages but … ) and the logs it keeps and AfD to get a handle on whether something can be improved or not. There is also nothing to stop them from doing everything in the NPP flow except the final acceptance. If fact, given my preference, each applicant for the NPP right should review 20 articles and make a list for the grantor to review before they get the right. A mentor, as you suggest, could critique such a list if they do not qualify for the right immediately. However, none of that really requires development of material much beyond what we have. NPP instruction is more of a practicum and the review, critique cycle is where the learning and improvement is done. I guess I kind of assume that an editor who does not yet have the background to understand and apply the written instructions we have is generally so far down the learning curve the marginal cost of training them is too high considering our limited resources. See User:Jbhunley/Essays/Identifying nonsense at NPP for a worked example. As analogy NPP training should be the equivalent of Uni ie there is a baseline of knowledge (secondary school) that one is simply expected to have before walking in the door. Jbh Talk02:47, 17 August 2018 (UTC)
(edit conflict) (and have not read the past two comments) @Insertcleverphrasehere: If AfC doesn't appeal to someone, neither will NPR really, unless they are looking for fancy buttons... 9_9. The problem I see with school/mentorships is they need people and time. That's one thing I'm maxed out on as is, and have none to contribute. I mean I can help in tad small ways, but not excessively. As for the standards being pretty much the same, maybe on paper they are, but how I would apply the standards is very different. I'm sure i'll be shot for this view point, but with ACTRIAL done, we get more expirienced editors who are more frusturated and leave easier than people at AfC. I'm not saying that AfC people don't matter, they are just less easily upset. NPR also gets less eyes easily, and is a little harder to track. So I mean if you can find the people for a school, i'd consider putting it in my responses, but i'm against putting it in right now as the same reason for CVUA. It's understaffed.
Re JBH's comment, I agree that the standards should be higher. I would want to remove guidelines for requirements and just add admin discretion. Also I'd want to codify the need for AFC/CSD/XFD. But idk how popular that opinion is. I also agree random reviews should be a thing, but I don't know about the logistical hoops on that. -- Amanda(aka DQ)02:57, 17 August 2018 (UTC)
I very much like the idea of presenting a log of 20 NPP type patrols as part of the request. It does seem like mentorship could be valuable here (I know I would have been eager for some mentoring when starting and still appreciate the chance to learn now). I also know that if I were applying today Amanda wouldn't approve me. As to the review maybe two editors (or more ideally three so it's slightly less awkward) could peer review some sample of each others patrolling? Best, Barkeep49 (talk) 03:04, 17 August 2018 (UTC)
@DeltaQuad: with regards to the appeal of AfC and NPR; while they might require similar knowledge and skills, they are very different in terms of the reviewer experience. I've done a lot of both, but especially these days AfC is clogged with promotional garbage from COI editors. I have little interest investing my time in a sewage pit like that, especially for an optional process that really doesn't have much impact on the encyclopedia. COI articles in AfC draft space don't matter, so why should I waste my time reviewing them? NPR however is an essential firewall for Wikipedia, and COI articles submitted onto mainspace must be dealt with. When it comes to well crafted articles that I want to read, there are very few that arrive via AfC; anyone who knows what they are doing and can write an article well either never bothered with AfC, or used it once and never again due to the terrible user experience. In short, well crafted articles almost always get submitted via mainspace directly. I actually like to read these, and like helping to clean up an otherwise decently crafted article. Worthwhile topics show up overwhelmingly more at NPR, especially after ACTRIAL. I know there are many other editors that share my views after having done both processes. — Insertcleverphrasehere(or here)03:41, 17 August 2018 (UTC)
As I said above, after 7 years of nursing them, I really do not want to have anything more to do with NPP, NPR, and PERM, so do bear with me for a moment. Admins such as DeltaQuad, xaosflux, Beeblebrox, and Alex Shih have PERM well in control and together with the clerking bot created and helper scripts by MusicAnimal, PERM is now working well to sort the wheat from the chaff - any admin who has ever worked at PERM is fully aware that even when assuming good faith, some applications are clearly hat collecting. Back in 2010 we realised that the standard of patrolling new pages was dismal and there was no way of controlling it, especially due to the fact that the process was open to everyone and his dog, particularly new and inexperienced users for whom all maintenance areas are a magnet.
Addressing this in several steps, we had the page curation system developed, then a user group created, and then turned NPR into this vibrant project community which it had never been before; being a member of a recognised group gives one a sense of responsibilty and recognition for the work. ACTRIAL was finally achieved and rolled out permanently for the specific purposes of reducing the load of junk on qualified reviewers, and with the intention of reducing a huge backlog to manageable proportions (which is unfortunately not panning out so well).
Through ignorance of the WP:NPP tutorial, tagging for maintenance and deletion by inexperienced users through Twinkle still exposes articles to incorrect tagging and the creators to being bitten. A retrogressive deprecation of any of those three important steps would be, IMHO, a disaster. We have demonstrated through a recent research that the vast majority of a registered reviewers are largely inactive or have never even used the right they were granted. This high number of 650 reviewers gives a totally false impression of the available resources. ICPH is correct in his comparison of the difference between NPR and AfC, the latter being an optional gesture for some users and Wizard users.
As far as I can see, the majority of drafts will never make it to mainspace and should be able to be processed, and deleted if necessary, reasonably quickly even if more articles from the new pages feed are moved to draft. Undeclared paid editing is on the increase and taking more subtle and sinister forms, the reduced number of articles in the feed should enable UPE to be more easily identified, and to allow more thorough reviews for all new articles. Adminship is not granted without sufficient demonstration of prior knowledge of notability and deletion, and nor should NPR which also needs an excellent knowledge of these policies and guidelines - we regularly see for example, many AFD being logged by users who are not versed with these prerequisites. The entire purpose of these discussions is therefore twofold: to reduce workload and at the same time increase efficiency. But it won't be done by finding reasons under any pretext to send out newsletters all the time, people will just ignore them and/or strike their names from the mailing list as they have been doing. Check out the history of the mailing list - anyone who has struck themselves is obviously not interested and can be struck from the group.
Only by addressing these issues organically and intelligently can Wikipedia continue to offer the world a reliable and trusted knowledge base.
What do you think about an editor who started editing years ago, has done virtually nothing since 2011 and suddenly produces a 15KB article Brian C. O'Neill with perfect formatting and referencing in a single edit, followed by no further activity? I have seen similar contributions from other editors and I think to myself, "paid editor". I wondered if there is someone orchestrating a system of contacting lapsed editors and paying them to present articles as their own, or perhaps managing to take over lapsed editors' accounts by getting a new password? Cwmhiraeth (talk) 13:06, 18 August 2018 (UTC)
"taking over" the account never came to my mind. But what I always thought was/is: such accounts are created as sleepers by content writers/marketing/PR firms. And when they get a client, these accounts are activated. Within past two weeks, while patrolling special:log/newusers, I have come across at least 10-12 accounts who did nothing but adding their name to the userpage, ranging from "Cwmhiraeth", "hi! I am Cwmhiraeth" or some similar variety. No edits at all after that. I have also seen accounts with same pattern from 2-3 years ago. Just one or two edits. And a few articles created by such accounts. I think they do this to avoid CU. I can imagine what they are thinking/doing, but i shouldnt say it out loud cuz of WP:BEANS. Smartse, and Tony have very good experience in that field. —usernamekiran(talk)13:35, 18 August 2018 (UTC)
I know I've been less active here of late than normal, but I was in the new pages feed today for the first time today in a while, and I was thrilled to see that we've gotten the number of pages needing review down to less than a month: I never dreamed we would get to this point when we were making ACPERM a thing, and now we really can say that the backlog of pages on the English Wikipedia needing review before hitting a search engine is non-existent: you all have successfully eliminated the backlog completely and gone far beyond it (the backlog point is 90 days, which is when the page hits Google.)
Just wanted to drop in with a cheery message since I know some people have seen the number rise and are a bit concerned. Really, the day count is much more important that the actual number of pages, and on that metric, we're rocking it. Great job, guys. TonyBallioni (talk) 15:47, 19 August 2018 (UTC)
Well, yes, but the 'length' of the backlog was down to nearly a week at the end of June, and is now almost back to a month long. Very soon there will be articles in the backlog that are older than a month if nothing changes. The trend seen in the graph at the top of this page is very concerning. — Insertcleverphrasehere(or here)20:31, 19 August 2018 (UTC)
Well, for better or worse, we (collectively) are not doing enough. Look at the chart at the top of this page and you will see the trend. But there is also a trend that the back of the backlog is getting farther and farther away. Tony's comment didn't really reflect this, and I'm not a fan of us just backslapping each other when there isn't really call for it. Tony is correct that the backlog is currently in ok shape, but it could be better, and the current trend isn't making me optimistic, that's all. — Insertcleverphrasehere(or here)00:07, 20 August 2018 (UTC)
I share ICPH concern about the growing backlog. The time to figure out how to ensure we don't get to 90 days is now when we're at ~30 days not when we're at ~85 days. However the "we" in we're not doing enough is probably not any reviewer reading this - it's we as an encyclopedia. It's some of of the ~550 reviewers who in a given month don't seem to be doing any reviews (this I think was Kudpung's point). There are two ways to get more reviews. Get more reviewers - this seems hard given that the most likely candidates have the permission or don't want it. Or we can get existing reviewers to do more reviews. This seems hard (motivating volunteers can be) but also seems like an area that less time has been invested in as an approach. I for one would support an RfC that says anyone who hasn't reviewed any articles with-in a X months (6 seems like the right number to me) of getting the PERM loses it. And that anyone who hasn't used it in X time frame also loses it (because they might not have been keeping up with changes in policies, procedures, consensus at AfD, etc). If we had 100 additional reviewers doing just a single review each week that solves about half of the increasing backlog right there. Best, Barkeep49 (talk) 00:28, 20 August 2018 (UTC)
(edit conflict)Also, the reduction of the backlog was almost entirely due to action by Onel5969, who reviewed some twenty-one thousand articles in the period of about two months. He is now taking a break, but even before that the backlog had begun to raise due to the rest of us getting complacent. Some of us have been busy, and not everyone has been slacking, and to those that haven't; thank you. But it doesn't take much to do a few reviews here and there, and we are only falling behind by about 30 articles a day (on average) at the moment. — Insertcleverphrasehere(or here)22:20, 19 August 2018 (UTC)
I've thought about it before, and its a good idea. There are a lot of admin maintenance activities that no-one else can do, which is why I have been reluctant in the past. A neutrally worded message indicating that we could use a bit of help if they have any spare time would probably be appropriate. — Insertcleverphrasehere(or here)04:29, 21 August 2018 (UTC)
Index Point. 90 Days or 30 days.
The index point used to be 30 days. It was increased to 90 days sometime in early 2016 in an attempt to give patrollers more time to catch up on the backlog. This was disastrous, as the backlog eventually grew so large that it became unmanageable.
I propose that we reduce the index point back to 30 days; so that we won't be tempted to let the backlog grow to an unmanageable point again in the future (three months worth of articles is simply too large a number to feel like you are making a dent, which made fighting the backlog in 2017 feel often feel like a hopeless exercise). While it can sometimes be advantageous to wait a while before reviewing an article, to me there seems no reason that an article should be reviewed more than 30 days after being created. — Insertcleverphrasehere(or here)00:20, 20 August 2018 (UTC)
I would support this with a phase-in. So we drop index to 60 days and then 45 days and then 30 days (alternatively it could drop to 30 days the first time we get to under 30 once the change has been approved). Creating some urgency among reviewers is another way to spur people into action, especially when the problem seems manageable. Best, Barkeep49 (talk) 00:30, 20 August 2018 (UTC)
The back of the backlog is currently at about 26th of July, or 25 days ago (disregarding a handful of stragglers and articles that have become unreviewed due to blanking, unredirecting, etc.), so there isn't too much of a need for a phase in at present. — Insertcleverphrasehere(or here)00:35, 20 August 2018 (UTC)
No. Please stop creating this false crisis. There is no backlog and there hasn’t been for virtually all of 2018. We have it at a manageable pace now. I’m all for people working harder, but the end goal here is making sure all content is indexed before it hits Google. 90 days allows that with breathing room. This is a horrible idea meant to drum up participation in this behind the scenes project at a detriment to the one audience that actually matters: our readers. To be honest, I would prefer more articles in the feed and a longer last article date because that would mean we aren’t burning people out and that people are hopefully taking their time reviewing. There is no problem here. Lets stop trying to invent one. The entire point of my post above was to try to stop the toxic idea that the volunteers who have been working here have been slacking, not to manufacture an emergency so we could recruit more people or whatever. TonyBallioni (talk) 00:37, 20 August 2018 (UTC)
@TonyBallioni: As I wrote above I don't think the people here are the issue. However, by the time we got to a 90 day backlog absent a new person like Onel the backlog would be intractable. Some sense of urgency to keep articles being reviewed promptly does help the readers of the encyclopedia. Best, Barkeep49 (talk) 00:45, 20 August 2018 (UTC)
The timeline given above is factually wrong. The backlog was well over 30 days when we upped it to 90, and it may have been over 90 days. What caused it to increase were that specific reviewers who had been doing the lions share of the work stopped around the same time we had an increase in monthly articles created. The 90 days was a lifeline, not some impossible burden to make life easier. Finally, I’ve seen no evidence ever that the “sense of urgency” argument has ever worked to motivate people to review. If anything, it makes people want to work less because it seems impossible to meet. TonyBallioni (talk) 00:51, 20 August 2018 (UTC)
@TonyBallioni: How would more articles in the feed and a longer review time help us? I agree that more reviewer energy spent on each article would be great, but it is currently a zero-sum-game. The less reviewing we do today, the more we have to do tomorrow. With regards to NPP, a 'crisis' develops slowly. My point about 90 days is that if we let the 'backlog' grow to 90 days it is already too late. The number of articles in a 3 month pile is too many for any individual reviewer to reasonably expect to be able to make a large dent by themselves. This becomes discouraging and is a detriment to the project. The index point should be about as far as should be necessary and no further. Perhaps the 'sense of urgency' argument doesn't work for everyone, then why do we have an index point at all? Per your rationale above, should we then not just put it 6 months away? A year? An indefinite index point? At what point do reviewers actually decide to review the articles that are piling up? How do we tackle a backlog of that many month worth of articles without a massive collective effort like the New Year Backlog drive? It took us a year of whittling away at the 7+ month backlog that had built up in the early part of last year, and I don't want the same thing to happen again. In terms of defining 'backlog'("an accumulation of uncompleted work or matters needing to be dealt with."): in absence of any reason for being unreviewed, any article that isn't 'reviewed' is part of a backlog of unreviewed articles. We've gotten so used to having thousands of articles being unreviewed at any given point that we have lost sight of that. — Insertcleverphrasehere(or here)01:09, 20 August 2018 (UTC)
It would show people are taking the time to actually read what they were reviewing. 90 days is a reasonable time frame that is long enough to account for ebbs and flows of work, but short enough to make sure everything hits google eventually. The backlog is any article that is past 90 days: things that are beyond a reasonable working period. Not every article that hasn’t been reviewed. It isn’t bad at all that we have unreviewed articles. It’s bad if they hit Google without review. You are losing sight of the reason this project exists: to provide a reasonable check on articles before the general public sees them. It does not exist to feed itself or exist to get to zero unrevised articles.Your suggestion here would solve absolutely nothing while having the potential to cause harm to the public within a week. You have all of your facts and theories backwards from how they actually happened: we had one or two reviewers stop reviewing or get blocked who were doing a lot of work. Around the same time, article creations increased slightly, but enough to have a noticeable impact on the backlog. In response to this, we upped the index point to 90 days: I think it was past that at the time, but if it wasn’t, it was very close and upping the index point had exactly zero impact on the backlog except to prevent BLP violations, copyvios, and spam from hitting Google for 90 days. ACTRIAL came along and steadied the rate of creations to be manageable. We were able to motivate people to decrease it through drives. I don’t think the 7 month figure had much impact on depressing people.I’m sorry to come out so hard against this, but you are suggesting that we manufacture a crisis for no reason, have your facts wrong, and that we open up the possibility of going back to an era where the worst of the new content gets sent to Google because someone goes on vacation. This would be a disaster. TonyBallioni (talk) 01:25, 20 August 2018 (UTC)
I'm reconsidering my stance on this based on your comments, but have a serious question. With ACPERM does the way we got to 90 days matter? It seems to me that ACPERM substantially changed the nature of NPP reviewing such that good ideas/bad ideas or historical examples from before then should carry different weight than they would in its absence. Best, Barkeep49 (talk) 01:34, 20 August 2018 (UTC)
ACPERM/ACTRIAL lowered the amount of articles created so where it was at a sustainable pace. What it does is prevent the high school kids, mom and pop shop owners, and autobiographers from creating articles. It also greatly reduced the admin burden in CAT:CSD. What it did not do was deal with the spammers who wait, COI editors who edit existing articles first, or good faith editors who don’t know our policies. NPP still exists to catch a lot of stuff in those areas: this in effect means ACPERM got rid of the easiest part of NPP and left the harder part. Keeping the index point at 90 gives us more time to catch commissioned works, check that BLPs are verifiable, check for cooyvio, etc. I think having the cushion is a good thing. TonyBallioni (talk) 01:42, 20 August 2018 (UTC)
(edit conflict)I've struck the part of my comment above that Tony has pointed out as misleading. Apologies. I'm not trying to manufacture a crisis though; we have a legitimate issue maintaining the 'unreviewed article pile' (which I will continue to refer to as the 'backlog') at a reasonable level. The current trend of the increasing backlog should be arrested to below 30 days if possible. I certainly don't want to cause any friction within the reviewer community, so if this proposal is decisive, I'll drop it if nobody else thinks it would be helpful. Perhaps 'urgency' works for me as motivation (serial procrastinator here), but not for most of our reviewers. I agree that ACTRIAL did away with what was essentially the easy part of NPP, and actually did not have as much of an impact on reducing workload of reviewers as some have claimed. — Insertcleverphrasehere(or here)01:50, 20 August 2018 (UTC)
Oppose (the original proposal). The NOINDEX period shouldn't be used as a motivational tool for reviewers. It should make sure that unpatrolled and therefore potentially damaging articles never show up in search results. 90 days gives us a comfortable margin for error and shouldn't stop us aiming to make sure that all articles are reviewed within a month (or a week, or a day).
On the subject of the 'backlog', I also find it a bit depressing that the queue is climbing again. But let's not panic. Backlogs go up and down for any number of reasons, and clearly relying on one editor doing massive amounts of reviews was never sustainable (in fact, the same thing is what led to the mega-backlog of 2016/17 in the first place). Ultimately, if the backlog is not stable, it means that we don't have a big enough pool of reviewers to handle the number of articles entering the queue. Cajoling on this talk page is unlikely to change that and backlog drives are by definition temporary measures. The only way to solve it is to keep building the reviewer team – and I hope you'll continue your vital efforts there, Insertcleverphrasehere. – Joe (talk) 06:03, 20 August 2018 (UTC)
It kind of confuses me; why have an index point at all? Still, I guess if we can't have 'no index point', the next best thing is as long an index point as possible. I understand your sentiments of longer = better, just in case. I've struck the proposal above, and I'd rather this didn't turn into a !voting gallery. I will of course continue inviting users. If anyone else wants to help with invitations, check out this page: User:Insertcleverphrasehere/NPR invite list and you can help me manually check the pre-vetted list (all instructions are on the page). — Insertcleverphrasehere(or here)07:07, 20 August 2018 (UTC)
This should not develop into a voting gallery, but you started it. I agree with your statement that The number of articles in a 3 month pile is too many for any individual reviewer to reasonably expect to be able to make a large dent by themselves. This becomes discouraging and is a detriment to the project. I do not agree entirely that the reviewers have been slacking - I will simply stress for the Nth time that among 650 reviewers there are some who do hardly anything or even nothing at all. That figure falsely imparts the notion that we have enough reviewers. The 90 day cut-off was not introduced as as a buffer for a lack of reviewing, it was intended as a safeguard that unreviewed articles beyond the low hanging fruit do not get indexed. It was considered at the time that it would be about right, and experience has shown that it is. I'm not saying that my way of doing things was necessarily the best way, but while I was the de facto caretaker of NPP/NPR for years, new things got introduced and established by consensus - and worked. NPR is becoming a fiasco like AfC - what this process needs is some genuine coordination rather than creating monumental and frequent news letters and backlog drives, and invites to apply for the right, for some things that could be addressed by more objective coordination. I thank Tony for taking some initiative here. Kudpung กุดผึ้ง (talk) 00:44, 21 August 2018 (UTC)
@Kudpung: I'm getting whiplash here; On one hand I have people saying we should invite more users, and on the other I have you complaining about it. You say that the 650 figure "falsely imparts the notion that we have enough reviewers", but that is exactly my goal of inviting more users: getting more reviewers. If your goal is to damage my morale and make me feel unwelcome, you are succeeding. — Insertcleverphrasehere(or here)01:32, 21 August 2018 (UTC)
I am not, and that would be my least intention. I have always appreciated your initiatives and hard work for NPR. However, I still maintain that there may be some better, or alternative ways for going about things. Kudpung กุดผึ้ง (talk) 01:35, 21 August 2018 (UTC)
(edit conflict) @Kudpung: It sounds like what you're saying is that no one can measure up to the way you ran NPP. You've mentioned multiple times on this noticeboard that you no longer want to be involved in NPP, but you're still here telling everyone what we're doing wrong. You may not realize that what you've been saying offends people. Maybe you should consider actually taking a break from NPP entirely. Natureium (talk) 01:40, 21 August 2018 (UTC)
That's not what I'm saying at all, try not to read between the lines. I did spend years gettiing NPP/NPR set up and running, but I don't own it. I would nevertheless be sad to see it not performing well. I t certainly does not have to be continued my way of course (but it did work), but it should be run with considered efficiency and well managed by others. I tried to do an election for coordinators last year if you remember, but there were no takers. I'll might continue to offer my opinions occasionally, but apart from reiterating what I still think are reasonable sugestions for ways to go, I'm not compelling anyone to follow in my footsteps or to exert any undue influence. That said, if you had spend so much energy for 8 years on such a project, including the battles with the WMF over it, you would also be disappointed to see it degenerate - NPR is one of the most important process on Wipedia, if not 'the' most important. I'm going to take a back seat now that Tony has commented. So relax a bit, please, without offending me.Kudpung กุดผึ้ง (talk) 06:02, 21 August 2018 (UTC)
Patrol log query
I reviewed my first couple of articles yesterday but I don't see any entries in my patrol log. The two that I ticked do show up under public logs, so the actions must have been registered. I just can't figure out how to filter them. Is the patrol log deprecated or otherwise unreliable? The tutorial says to check it frequently. Similarly, I found Operation Jackstay in Category:Unreviewed new articles from August 2018 and went to look: it's showing as reviewed (the tick in the review toolbar on the right is highlighted green, mouseover reads "mark this page as unreviewed", but I'm not seeing anything in its patrol log either. That makes sense, since the creator has the "autoreviewer" right, but in that case perhaps it shouldn't be showing up in an 'unreviewed' category? › Morteetalk11:50, 22 August 2018 (UTC)
And Operation Jackstay shows up in the Category:Unreviewed new articles from August 2018 because the Template:New unreviewed article is still in the article. You can remove the template since its been reviewed. Bennv3771 (talk) 11:59, 22 August 2018 (UTC)
Ah, thank you. Is there an alternative to the toolbar, and is one method preferred over another? The tutorial also mentioned that "In some editing contexts, authorised editors will see a 'mark this page as patrolled' link". I haven't yet seen any link like that, including on pages I'm looking to review, so I've just been using the toolbar. I'll leave the template on Operation Jackstay for now since the creator seems to have placed it there intentionally, presumably to get a second pair of eyes, and I haven't reviewed it yet myself. › Morteetalk12:09, 22 August 2018 (UTC)
@Mortee: You will see the "mark this page as patrolled' link" at the bottom right of the article if you close the curation toolbar. If you patrol the article without using the toolbar it will show up in your patrol log. Bennv3771 (talk) 12:12, 22 August 2018 (UTC)
@Bennv3771: Thanks again! Sorry to ask so many basic questions, but one more: is there an advantage to using one method over the other, or is it just a matter of preference? They seem to be two ways to achieve the same goal, but since they're logged separately I wonder if there's an important distinction I'm missing. › Morteetalk12:24, 22 August 2018 (UTC)
@Mortee: No problem. AFAIK it's up to reviewers which method they prefer. The curation toolbar just makes it much easier to tag new pages and leave messages for their creators. Bennv3771 (talk) 12:26, 22 August 2018 (UTC)
Another newbie question: is it intentional that using the curation toolbar to tag an article with maintenance templates marks it as reviewed? What other actions do the same? The example is Battle of Suchodoł. I hadn't finished considering it, and it might have benefited from another reviewer; I assumed you had to actually tick it, but it seems it went off the list when I tagged it. Just checking that's expected and what other actions I might take have extra weight now. › Morteetalk01:48, 24 August 2018 (UTC)
Yeah, if you use the curation tool for tagging, it also marks it as reviewed. If you don't want this, you can just click to un-review it. Natureium (talk) 02:09, 24 August 2018 (UTC)
I'm not sure what the status of the NPP toolset is at the moment with the WMF taking on improvements to the toolset to help AfC, but if we do want to see some other improvements, now might be a good start thinking about how we present those requests. Perhaps User:MMiller (WMF) can clarify what the status of the curation tools is? Are we expected to use the Community Wishlist Survey? --Vexations (talk) 21:15, 20 August 2018 (UTC)
Perhaps User:DannyH (WMF) could provide more information for the above question (he runs the community tech team). While NPR is a core function of Wikipedia, and shouldn't have to resort to begging for someone to improve our tools, DannyH has previously indicated that we need to use the wishlist. If so, we will need to put together a list of stuff from the 'Suggestions' page to include in our wishlist entry. — Insertcleverphrasehere(or here)21:41, 20 August 2018 (UTC)
A core part of the Wishlist entry should be making the Page Curation tools editable by us, so that we don't have to rely on the WMF for changes and updates that will be invariably needed. I do think that the page curation tools are relatively user friendly, and I like that about them, and they have a lot of great functionality (like the 'info' button, and the wikilove functionality). But when you compare the lack of customisation to the Twinkle preferences panel, it is a bit laughable. I know that some users prefer to use Twinkle for all their reviewing, and I for one use it for all of my Deletion nominations (due to a lack of CSD/PROD log functionality for the Page curation tools). There are loads of errors that no one has bothered reporting because we know they won't be fixed. For example: it's not possible to make second or third AfD nominations with the page curation tools, as it will try to paste it over the first nomination page (why I use Twinkle for AfD too). — Insertcleverphrasehere(or here)22:01, 20 August 2018 (UTC)
The items on that list have absolutely nothing to do with Horn's Xmas list for Santa.
The items on that list comprise a single upgrade request - not individual requests.
The wishlist is for comfort and convenience gadgets such as what your watchlist looks likes or syntax coloring.
The developers spend their time creating new gadgets such as new blocking criteria for admins (I don't recall where that was asked for, I find it confusing and time consuming).
The required items for the Curation tool set are not comfort & convenience gadgets. They are part of a core MediaWiki extension that Wikipedia users are not able to further develop or modify - nor should they, strictly they are content creators and maintainers, not computer programmers. The Curation system was engineered by the WMF and it is their responsibility and obligation to ensure that it fulfills its intended task as fully as possible. But that was in the days where the community was involved in development, and regular Skype conferences took place about its development between the WMF Deputy CEO (not even such a position exists nowadays where the CEO is not involved in management and spends 200 days a year in the air), the senior engineers, and the community.
While we are grateful that part of the work being done to the Feed for the use of AfC may possibly enhance some of the tasks for New Page Reviewers, it is a fallacy for the WMF to be doing this while persistently refusing to recognize the other work that needs to be done. It's popped the WMF's bubble about it needing to be requested through the Wishlist. The only other person who could possibly be need to be convinced is Toby Negrin, who appears to be in overall chsarge of develoment, but he does not have a superior.but the WMF staff structure is now so complex that they don't know themselves who is in charge, many development issues overlap different departments and as a result some things get overlooked, or even cancelled at Phab by non-WMF volunteer devs who who write MediaWiki code but don't understamd the needs of its prime user: the Wikimedia movement..
The more important an urgent a critical software requirement is, the more difficult it is to get the WMF to address it - it's a long standing WMF attitude, we've seen it with ACTRIAL.. They will continue to develop top down bits and pieces that are not asked for and that are not like NPR which is one of the most important maintenance tasks in Wikipedia - if not the most important. It needs serious attention. The WMF being at loggerheads with the community over this issue is totally counter productive and inadmissible.
The volunteers who feel they would like to take on some responsibility for the maintenance of NPR need to be aware of the history, and avoid anything that would simply create setbacks. Kudpung กุดผึ้ง (talk) 00:11, 21 August 2018 (UTC)
Thanks for bringing this up, Vexations, Insertcleverphrasehere, and Kudpung. DannyH (WMF) is my manager, and he has been following along closely with the project I've been working on to add AfC, ORES scores, and copyvio detection to the New Pages Feed. That project is narrowly scoped to just those three improvements (though we are trying to tackle a bug or two while we're in the code) because the Growth team needs to move on to our work on new editor retention in mid-size Wikipedias.
The Community Wishlist is the right place to bring up additional improvements to Page Curation and the New Pages Feed. The wishlist is definitely meant to be a way for important tools to be built and improved. I think that a wishlist proposal that incorporates many of the items on the Suggested improvements page would make a lot of sense, and would hopefully get a lot of votes from the NPP community (and others who care about the quality of new pages).
We do know how the wishlist process works, and disagree with it on the simple premise that popular, urgent and necessary aren't the same thing, and the only thing that counts is being popular. MER-C16:11, 22 August 2018 (UTC)
I see that Natureium archived that this morning. Is there a reason that the project subscribes to this particular newsletter? Unless I'm mistaken we don't even subscribe to our own. If it's useful for people maybe there's a different of our many talk pages for them to go on rather than our main one here. Best, Barkeep49 (talk) 14:44, 27 August 2018 (UTC)
I think that if editors want to read it, they can subscribe on their userpage. It's just clutter here. Many wikiprojects create articles, and we don't follow them. Natureium (talk) 21:56, 27 August 2018 (UTC)