NPP backlog
>NPP backlog: 8669 as of 06:00, 30 July 2022 (UTC)
1, 2, 3, 4, 5, 6, 7, 8, 9, 10 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 31, 32, 33, 34, 35, 36, 37, 38, 39, 40 41, 42, 43, 44 |
Sections older than 21 days may be automatically archived by lowercase sigmabot III. |
Proposal: increase unreviewed new page search engine NOINDEX duration
Alexandermcnabb had a great idea above. Articles should not be searchable in 90 days automatically, but ONLY AFTER they have passed NPP.
I think this is actionable, it is probably pretty easy to file a Phabricator ticket tagged with Wikimedia-Site-requests to change $wgPageTriageMaxAge = 365;
. Should I create a WP:VPR for this? Is 365 a good number of days to bump it up to? (currently at 90 days) –Novem Linguae (talk) 18:55, 17 June 2022 (UTC)
- This is a great idea; by definition, anything still in draft is not ready to be searchable. Chris Troutman (talk) 19:08, 17 June 2022 (UTC)
- This isn't about drafts, I think: it's about mainspace articles that haven't been patrolled yet. I think drafts already aren't searchable. Extraordinary Writ (talk) 19:11, 17 June 2022 (UTC)
- I agree that this would be a good idea; hopefully it shouldn't be too controversial. Do you know where the 90-day limit was decided? All I could find was this RfC, which seems to suggest that all unpatrolled pages should be noindexed. Extraordinary Writ (talk) 19:11, 17 June 2022 (UTC)
Technical limitations?
- I think there are some technical limitations; I remember a similar discussion where it was revealed that the backlog for redirects is capped at 30 days because the page triage system cannot handle the amount of redirects that would pile up in the full 90 days. signed, Rosguill talk 19:31, 17 June 2022 (UTC)
- @DannyS712, would you be able to weigh in on any technical limitations to raising $wgPageTriageMaxAge? Is there some kind of database issue that makes this not as easy as changing a variable? –Novem Linguae (talk) 20:16, 17 June 2022 (UTC)
- @MusikAnimal, would you be able to weigh in on any technical limitations to raising $wgPageTriageMaxAge? Is there some kind of database issue that makes this not as easy as changing a variable? –Novem Linguae (talk) 23:02, 17 June 2022 (UTC)
- @Novem Linguae There's no technical concern with increasing $wgPageTriageMaxAge as far as I can tell. gerrit:356781 points to this discussion which implies it was created for this very reason. — MusikAnimal talk 05:35, 18 June 2022 (UTC)
- That was an interesting discussion. It gives the reason for 90 days - a concern that someone could "vandalize" an important article by slipping in a NOINDEX. To mitigate that, NOINDEX is ignored on "old" articles and only works on "new" articles. There was an assumption that articles would be patrolled within 90 days. MB 08:01, 18 June 2022 (UTC)
- That could be handled separately by an edit filter that detects edits introducing NOINDEX in mainspace. MarioGom (talk) 08:30, 18 June 2022 (UTC)
- Hmm. I wonder, how tightly coupled are $wgPageTriageMaxAge and __NOINDEX__? I assumed they were separate, each having their own max duration, but it's possible that their durations are controlled by the same setting. Or maybe a wgPageTriageMaxAge = 365 would place a NOINDEX but then the NOINDEX would still be set to 90 and that would defeat the purpose.
- I also wonder if it'd be technically feasible to go over 365 days. One reason I chose that number is I know that the SQL table pagetriage_log_table only keeps 365 days of data,[1] and I wonder if going over that might cause problems. –Novem Linguae (talk) 08:51, 18 June 2022 (UTC)
- Ping MusikAnimal. Was wondering if you could weigh in again, if you know offhand. Thanks! –Novem Linguae (talk) 18:05, 18 June 2022 (UTC)
- @Novem Linguae I'm not aware of a time limit on __NOINDEX__, or at least it doesn't seem to be documented. I didn't read all of the lengthy discussion but it seemed at the time, there were separate, unrelated issues where basically NOINDEX wasn't working as expected. That appeared to have been resolved as well. $wgPageTriageMaxAge acts completely independently, judging by the code. — MusikAnimal talk 21:40, 20 June 2022 (UTC)
- That could be handled separately by an edit filter that detects edits introducing NOINDEX in mainspace. MarioGom (talk) 08:30, 18 June 2022 (UTC)
- That was an interesting discussion. It gives the reason for 90 days - a concern that someone could "vandalize" an important article by slipping in a NOINDEX. To mitigate that, NOINDEX is ignored on "old" articles and only works on "new" articles. There was an assumption that articles would be patrolled within 90 days. MB 08:01, 18 June 2022 (UTC)
- @Novem Linguae There's no technical concern with increasing $wgPageTriageMaxAge as far as I can tell. gerrit:356781 points to this discussion which implies it was created for this very reason. — MusikAnimal talk 05:35, 18 June 2022 (UTC)
- I think there are some technical limitations; I remember a similar discussion where it was revealed that the backlog for redirects is capped at 30 days because the page triage system cannot handle the amount of redirects that would pile up in the full 90 days. signed, Rosguill talk 19:31, 17 June 2022 (UTC)
Sources
|
---|
Break
- This makes sense, if it's technically feasible. MarioGom (talk) 19:43, 17 June 2022 (UTC)
- Is it possible to have a max age based on categories? BLPs, for instance, shouldn't be indexed at all until reviewed. ScottishFinnishRadish (talk) 19:45, 17 June 2022 (UTC)
- The biggest issue may not be technical. This is a quote from Kudpung: "otherwise you'll just have to make an argument at the WMF to extend the un-patrolled period, and believe me, that would be no easy task, even if you have friends there and meet them personally - been there, done that" — Preceding unsigned comment added by MB (talk • contribs) 17:16, June 17, 2022 (UTC)
- I'm in favor, if it's technically feasible. Not sure what to do if WMF is problematic, but I'd hope we could work through any issues. Not sure what the issue might be - if it's "that's the way it's always been," then why hide IP addresses that have always been open, eh? Geoff | Who, me? 21:24, 17 June 2022 (UTC)
- It helps, especially as applicable to PE in that the client can't see the article so no payment will be forthcoming. Atsme 💬 📧 21:30, 17 June 2022 (UTC)
- I agree, rsjaffe, we still need to patch the hole that allows unsourced articles to slip into mainspace, and also work on a more efficient system for handling redirects. Atsme 💬 📧 00:08, 18 June 2022 (UTC)
- Now that I think about it, changing this will actually help. Right now, no one other than the reviewers has "skin in the game". No one else has an incentive to fix this. With NOINDEX, we can motivate those who care about search engines. Might provide a touch of support to meaningful changes. — rsjaffe 🗣️ 00:14, 18 June 2022 (UTC)
- I agree, rsjaffe, we still need to patch the hole that allows unsourced articles to slip into mainspace, and also work on a more efficient system for handling redirects. Atsme 💬 📧 00:08, 18 June 2022 (UTC)
- One of my favorite sayings is that deadlines spur action. So would this mean that some patrolling that happens because people are concerned about a page being indexed wouldn't happen or would it allow a more stable operation of the queue before something gets indexed? Best, Barkeep49 (talk) 00:48, 18 June 2022 (UTC)
- I've given up on deadlines. I patrol and find 1) stuff that belongs in Wikipedia, 2) bad stuff (e.g., UPE, sockpuppetry, obvious non-notable). Anything in the middle, particularly if it is sports related, I just pass on and let it age out. An interesting thing I did was look at the queue starting 31 days out. articles that are not obviously notable in 1) sports category or 2) foreign-language-reference-supported predominate. — rsjaffe 🗣️ 01:25, 18 June 2022 (UTC)
- My intention was that people who WANT their article to get any attention (ie: Search) would ensure the article is in fit condition for NPP to validate them. At that point (and not automatically after 90 days), the article would be searchable and therefore properly 'live'. This both spurs creators to ensure pages are up to par and ensures that pages that aren't are not searchable and therefore limits the need to send articles to draftification (easing the burden on AfC), limits the need to go to AfD except in the most extreme/certain to pass cases and ensures that people aren't searching WP and finding utter rubbish. Additionally, lifting the burden for performing BEFORE on NPP means that the article AS PRESENTED is patrolled - not the notional article that could have been should the creator have actually looked for the sources the article needs - reducing burden on the small number of patrollers and increasing the requirements for the large number of page creators to actually create viable pages. Thanks, Novem Linguae for picking it up. Best Alexandermcnabb (talk) 14:30, 18 June 2022 (UTC)
- I've given up on deadlines. I patrol and find 1) stuff that belongs in Wikipedia, 2) bad stuff (e.g., UPE, sockpuppetry, obvious non-notable). Anything in the middle, particularly if it is sports related, I just pass on and let it age out. An interesting thing I did was look at the queue starting 31 days out. articles that are not obviously notable in 1) sports category or 2) foreign-language-reference-supported predominate. — rsjaffe 🗣️ 01:25, 18 June 2022 (UTC)
- @Barkeep49, I'm thinking that if exposure is the main purpose of creating an article, then it holds to reason that if it isn't being seen (indexed), the creator will be incentivized to fix it. OTH, if it's a BOT creation, or one the creator doesn't give a flip about, then it's highly unlikely that (a) it's worthy of inclusion or (b) that the issues will actually be fixed; both of which warrant CSD or PROD. Any article that is notable and worthy of inclusion will be sourced, but let's not waste time on the symptoms and go straight to the root of the problem. An unsourced article never should have made it into mainspace. If we patch that hole, and prevent unsourced articles from slipping through the cracks, a lot of our problems are solved. I'm not sure where the notion of publish it anyway actually originated, but it defies our core content policies, and taxes the very core of NPP volunteerism without fair representation. As rsjaffe put it, the bulk of those who oppose deletion don't have any "skin in the game", or a "dog in the fight". This whole scenario is one of the reasons we need the stats I've asked about – we need a cost vs benefit analysis, so to speak, as it relates to the arguments for keeping unsourced articles that have been denied deletion at CSD & PROD – where did they finally end up? Did they get stuck in the NPP queue – go to AfD and were deleted – or did they actually get published with/without fixing the problems? Crafty editors will also take advantage of redirects to create articles in mainspace without drawing much attention. The redirects, reverts of redirects and those that lead to AfD are another issue – but I sincerely believe it's one we can fix at the root cause - PREVENTION, keep unsourced articles/stubs from ever seeing the light of day. Atsme 💬 📧 14:37, 18 June 2022 (UTC)
- Ultimately if more articles are being made than we have capacity to patrol it's going to be a problem at some point. Pushing out no index would give us more breathing room but not change that underlying issue. So we have to either increase capacity or decrease the number of new articles. So ultimately any increase in the NOINDEX time should be accompanied by some plan to change one of these things. Best, Barkeep49 (talk) 15:20, 18 June 2022 (UTC)
- I'm for decreasing the number of new articles, which is something we can do by heading them off at the pass; i.e. PREVENTION. A BOT could be created to assure each new article going into mainspace has at least 3 citations before it leaves Draft, or is directly created in mainspace, regardless of autopatrolled status. I don't know enough about the "theft" of redirects to address that issue - all I know is that it exists and it's a problem. Wbm1058 can probably explain it far better than I ever could. Adherence to WP:PAG and more CSD & PROD support from our admins will help to self-correct some of the other issues, as will getting more admins trained in NPP reviewing so that fewer CSD & PRODs will be rejected or sent to AfD. Redirects also need attention in an effort to make it more difficult to get a bad article back into mainspace which is another rather substantial prevailing issue. The various discussions at VPP demonstrates where a big part of the issues stem, including ambiguities and misunderstandings of PAGs topping the list. I also noticed that, for whatever reason, some editors don't have a "middle button" - they are either 100% inclusionist or 100% deletionist. What we need are more includelists. Atsme 💬 📧 15:56, 18 June 2022 (UTC)
- @Atsme: I'll give you an example case study – something I caught on my patrols and cleaned up this morning. I won't say exactly how I found it per "beans" though I do have a section on my user page where I list my work queues, and it was something on that list. Untangling the mess there was a time-consuming manual process which would be extremely difficult to automate. My focus has been on occasionally adding more patrols when I stumble across something that wasn't caught by any patrol I try to create a new patrol that finds other cases just like the one I stumbled onto. Aliana was first created 16 December 2007 as an article about an album by Aliana Lohan. This was deleted per Wikipedia:Articles for deletion/Aliana. Aliana was re-created 13 July 2017 as a redirect to Eliana via Articles for Creation. That redirect was usurped by OfTheUsername when they started a new article about Aliana, Texas (OfTheUsername later moved Aliana to Aliana, Texas: Proper naming.). That new article had been shuttled back & forth to draft space. I untangled all the crossed wires, and deemed the Texas planned community to not be a primary topic, so I made Aliana a disambiguation. I'm a "middle button" who generally leaves keep/delete decisions about articles like Aliana, Texas to others. But I did note that AIRIA Development Company might have an interest in promoting their planned community which is unlike the other communities listed on Template:Fort Bend County, Texas. I remember riding my bike through places like Crabb and Clodine on my bicycle rides west of Houston in the early to mid 1980s. OMG, Fulshear, Texas is a city with 16,000+ ppl now? I remember that place as not much more than a country corner with a BBQ joint where we ate lunch either during or after our ride (back then it had a population under 600). – wbm1058 (talk) 17:51, 18 June 2022 (UTC)
It's a can of worms. Atsme 💬 📧 18:27, 18 June 2022 (UTC)
- bro what. I made the Aliana article independently. I don't work for/connected to AIRIA Development company or affiliates or anything. Sorry about the redirection and stuff I tried to make it a draft but then it acted weird. My bad. OfTheUsername (talk) 18:47, 18 June 2022 (UTC)
- @Atsme: I'll give you an example case study – something I caught on my patrols and cleaned up this morning. I won't say exactly how I found it per "beans" though I do have a section on my user page where I list my work queues, and it was something on that list. Untangling the mess there was a time-consuming manual process which would be extremely difficult to automate. My focus has been on occasionally adding more patrols when I stumble across something that wasn't caught by any patrol I try to create a new patrol that finds other cases just like the one I stumbled onto. Aliana was first created 16 December 2007 as an article about an album by Aliana Lohan. This was deleted per Wikipedia:Articles for deletion/Aliana. Aliana was re-created 13 July 2017 as a redirect to Eliana via Articles for Creation. That redirect was usurped by OfTheUsername when they started a new article about Aliana, Texas (OfTheUsername later moved Aliana to Aliana, Texas: Proper naming.). That new article had been shuttled back & forth to draft space. I untangled all the crossed wires, and deemed the Texas planned community to not be a primary topic, so I made Aliana a disambiguation. I'm a "middle button" who generally leaves keep/delete decisions about articles like Aliana, Texas to others. But I did note that AIRIA Development Company might have an interest in promoting their planned community which is unlike the other communities listed on Template:Fort Bend County, Texas. I remember riding my bike through places like Crabb and Clodine on my bicycle rides west of Houston in the early to mid 1980s. OMG, Fulshear, Texas is a city with 16,000+ ppl now? I remember that place as not much more than a country corner with a BBQ joint where we ate lunch either during or after our ride (back then it had a population under 600). – wbm1058 (talk) 17:51, 18 June 2022 (UTC)
- I'm for decreasing the number of new articles, which is something we can do by heading them off at the pass; i.e. PREVENTION. A BOT could be created to assure each new article going into mainspace has at least 3 citations before it leaves Draft, or is directly created in mainspace, regardless of autopatrolled status. I don't know enough about the "theft" of redirects to address that issue - all I know is that it exists and it's a problem. Wbm1058 can probably explain it far better than I ever could. Adherence to WP:PAG and more CSD & PROD support from our admins will help to self-correct some of the other issues, as will getting more admins trained in NPP reviewing so that fewer CSD & PRODs will be rejected or sent to AfD. Redirects also need attention in an effort to make it more difficult to get a bad article back into mainspace which is another rather substantial prevailing issue. The various discussions at VPP demonstrates where a big part of the issues stem, including ambiguities and misunderstandings of PAGs topping the list. I also noticed that, for whatever reason, some editors don't have a "middle button" - they are either 100% inclusionist or 100% deletionist. What we need are more includelists. Atsme 💬 📧 15:56, 18 June 2022 (UTC)
- Is there a page that shows page creation by user metrics (Daily, Weekly, Monthly)? Perhaps some users could be evaluated and "nominated" for Autopatrol by the community.Slywriter (talk) 15:29, 18 June 2022 (UTC)
- @Slywriter: Wikipedia:Database reports/Editors eligible for Autopatrol privilege maybe? – Joe (talk) 18:36, 18 June 2022 (UTC)
- I discovered that there used to exist an edit filter to flag unsourced new articles, though it was decommissioned in 2012 with the comment "disabling, no real use" (?). There's also a fairly recent unanswered query as to whether the filter should be reinstated. While this wouldn't be a universal solution, I believe (in agreement with the recent post) that warning users who are about to publish an unsourced article would either lead them to add sources or refrain from publishing. The filter also got quite a number of hits before it was deleted, but I believe it preceded ACTRIAL. Would reinstating this filter be at least a partial solution? ComplexRational (talk) 17:10, 18 June 2022 (UTC)
- Interesting discovery. I wonder if it will also catch autopatrolled in mainspace or if it only works at AfC draft to mainspace? It should be a universal catch-all filter. No article should be in mainspace without a minimum of cited RS, even if it's just 2. Could it also catch sources that are unreliable by integrating User:Headbomb/unreliable and maybe even User:SuperHamster/CiteUnseen.js would prove useful? Atsme 💬 📧 17:28, 18 June 2022 (UTC)
- From what I see, the filter's old configuration did not target any specific user groups and was set to work for all mainspace pages, with built-in exceptions for pages that aren't supposed to have sources such as redirects and disambiguation pages. I'm unsure about autopatrolled – but in that case, there could be grounds for revocation of the permission. Regarding the scripts, I doubt they could be implemented in a filter because there's enormous potential for false positives/negatives and there are simply too many cases to cover (someone with more advanced programming skills, feel free to correct me though); reactivating the old filter would simply help in catching the most egregious cases. ComplexRational (talk) 17:53, 18 June 2022 (UTC)
- Quick tech note. The edit filter is computationally expensive and it would probably not be performant to look for more than a short list of the dozen or so most egregious unreliable sources. Which we already have some filters for, e.g. filter 869. –Novem Linguae (talk) 18:12, 18 June 2022 (UTC)
- How is the filter catching no sources? Is it using citation format, or reflist, or something of that nature? Atsme 💬 📧 18:32, 18 June 2022 (UTC)
- @Atsme: It flags any new article that doesn't have any of several character strings that would suggest either the presence of sources or a page not needing sources – this means flagging new pages not having any of <ref> tags, http/https, disambiguation in the title, #REDIRECT, etc. It doesn't discriminate types of sources, only whether any could be present. ComplexRational (talk) 18:45, 18 June 2022 (UTC)
- How is the filter catching no sources? Is it using citation format, or reflist, or something of that nature? Atsme 💬 📧 18:32, 18 June 2022 (UTC)
- Interesting discovery. I wonder if it will also catch autopatrolled in mainspace or if it only works at AfC draft to mainspace? It should be a universal catch-all filter. No article should be in mainspace without a minimum of cited RS, even if it's just 2. Could it also catch sources that are unreliable by integrating User:Headbomb/unreliable and maybe even User:SuperHamster/CiteUnseen.js would prove useful? Atsme 💬 📧 17:28, 18 June 2022 (UTC)
- Ultimately if more articles are being made than we have capacity to patrol it's going to be a problem at some point. Pushing out no index would give us more breathing room but not change that underlying issue. So we have to either increase capacity or decrease the number of new articles. So ultimately any increase in the NOINDEX time should be accompanied by some plan to change one of these things. Best, Barkeep49 (talk) 15:20, 18 June 2022 (UTC)
- That 2012 RfC was a result of a couple of meetings I had with Jorm and Erik Möller over the development of New Page Triage, the working name for the new feed and curation software development. Characteristically Oliver Okeyes (WMF), jumped on the bandwaggon before any volunteers could could start the community discussion, and he later continually tried to block development right through to his leaving the community although there was a clear consensus (WereSpielChequers and Scottywong will remember all this).
- 'No Index until patrolled' was much like Jumpback which Oliver also promised and did nothing about and wasn’t resolved until TonyBallioni stepped in years later and it finally got boxed through at the WishList in 2019. Anyway, Extraordinary Writ, you can all blame me for the 90 days. When I was asked, it's what I suggested thinking it would be enough. A few years later we got ACTRIAL done which greatly reduced the flood of effluent but it wasn't long before the problems with patrollers started.
- Per Barkeep49:
Ultimately if more articles are being made than we have capacity to patrol it's going to be a problem at some point.
Yes, pushing out the 90 days would certainly just be a palliative. First off - and I know you'll all hate me for this - some stats are needed over a sample period: how many articles are kept immediately, how many are draftified, how many are PRODed, how many are CSDd and AfDd? Or did someone do that already and I missed it? Kudpung กุดผึ้ง (talk) 10:23, 19 June 2022 (UTC)
I think that extending the 90 days is beyond logical....knowing what we now know, the short 90 day number defeats the whole purpose of that (good) feature and makes it pointless. North8000 (talk) 10:33, 19 June 2022 (UTC)
- The suggestion is not to push 90 days out, Kudpung, but institute 'No review = No Search' as a flat deal and make that public, possibly with some guidance on notability and sourcing that lets new page creators know that their page will not be searchable until it reaches a minimum standard of notability/sourcing and has been patrolled. I'm not sure where stats help in this - we can all see the scale of the problem and this would apply some systemic leverage for creators to focus on sourcing. This also means sub-standard articles aren't 'rewarded' by being searchable after 90 days whether reviewed or not - when right now we have little hope of getting to articles in 90 days. The best stat I can see is the NPP queue and how little we are managing to reduce it! Best Alexandermcnabb (talk) 13:12, 19 June 2022 (UTC)
- I think it's quite clear, Alex, that at least Barkeep49 certainly concur that a palliative would not have much effect. One advantage, as the potted NPP history I wrote above is intended to demonstrate, is that the en.Wiki and a few others have since asserted their maturity vis-à-vis the WMF and can now instigate their own controls and policies.
'No review = No Search' as a flat deal
is not only an excellent suggestion, but where no amount whipping all the 750 reviewers into action is ever going to work, it's also the only solution. Just do it. It's technically a doddle so just ask at Phab for the switch to be thrown. OTOH, if the Grand Masters of Phab decide you need a community consensus first, you'll have get one. ACTRIAL is a seminal example of changing outdated 'founding principle' policy; we had some very heavy participation and extremely convincing consensus each time we ran a debate for it, but only because our mission statements were armed with a lot of significant stats and very carefully crafted proposals. Now its time for some action. Kudpung กุดผึ้ง (talk) 15:21, 19 June 2022 (UTC)- Now, I can just about get my head around NPP and AfD, and I can even create the odd page (bless him, but it was User:TonyBallioni wot granted me autopatrolled, as you mention him) but this Phab switchy stuff is, to be honest, a bit beyond me. Have you SEEN what a mess I can make with source editing? I'm not sure I'm the person to actually take this one forward, but am perfectly happy to have made the suggestion if others want to activate it - and just as happy to support. But as for anything beyond tinkering with content, I'm really sure it's not my long suit... Best Alexandermcnabb (talk) 15:56, 19 June 2022 (UTC)
- I think it's quite clear, Alex, that at least Barkeep49 certainly concur that a palliative would not have much effect. One advantage, as the potted NPP history I wrote above is intended to demonstrate, is that the en.Wiki and a few others have since asserted their maturity vis-à-vis the WMF and can now instigate their own controls and policies.
- The suggestion is not to push 90 days out, Kudpung, but institute 'No review = No Search' as a flat deal and make that public, possibly with some guidance on notability and sourcing that lets new page creators know that their page will not be searchable until it reaches a minimum standard of notability/sourcing and has been patrolled. I'm not sure where stats help in this - we can all see the scale of the problem and this would apply some systemic leverage for creators to focus on sourcing. This also means sub-standard articles aren't 'rewarded' by being searchable after 90 days whether reviewed or not - when right now we have little hope of getting to articles in 90 days. The best stat I can see is the NPP queue and how little we are managing to reduce it! Best Alexandermcnabb (talk) 13:12, 19 June 2022 (UTC)
- Yes, thank you MB for taking the initiative. Personally, I would have asked for indefinite 'NO INDEX' and see what the WMF offered. It's always better to ask for more and then negotiate down if held against the wall by the throat. That said, if you get 365 days it might just do the trick but I'm wary of being back here again in 5 years and asking for more. Anyway, by then Wikipedia might be dead, and so might I ;) Kudpung กุดผึ้ง (talk) 11:46, 20 June 2022 (UTC)
- I did ask for INDEFINITE (as you know now that you have been to the phab ticket, just clarifying here for others). There is already a suggestion there for 365 days, and the task was renamed to reflect that :( MB 13:46, 20 June 2022 (UTC)
- Yes, thank you MB for taking the initiative. Personally, I would have asked for indefinite 'NO INDEX' and see what the WMF offered. It's always better to ask for more and then negotiate down if held against the wall by the throat. That said, if you get 365 days it might just do the trick but I'm wary of being back here again in 5 years and asking for more. Anyway, by then Wikipedia might be dead, and so might I ;) Kudpung กุดผึ้ง (talk) 11:46, 20 June 2022 (UTC)
Yes, I realise that now. All the fun of the fair, just as I predicted.
The Phab request has now been marked as 'stalled'. I hope enough people are following this because the devs are going to need a lot of convincing this is necessary. Kudpung กุดผึ้ง (talk) 14:03, 20 June 2022 (UTC)
- It also says a "local discussion" is underway. If that means a private discussion among devs, that is not transparent. All our discussions on this are public. MB 14:21, 20 June 2022 (UTC)
- That is classic WMF dev-speak. There is nothing for them to discuss. It's their way of finding another ruse for stalling. They are pretending - like I said above - that the whole thing needs yet another RfC. Oh, I know Phab so well - and Bugzilla, its predecessor. So do Scottywong and The Blade of the Northern Lights. Kudpung กุดผึ้ง (talk) 14:54, 20 June 2022 (UTC)
- Well, to be fair, the phab ticket has only been open for a day or two, and it doesn't look like anyone has said anything like "we're not going to do this." So, I'd cut them some slack and give them a reasonable amount of time to figure out how to make the change without introducing any unexpected consequences. If they do eventually refuse to make the change, there are probably ways for us to make this happen on our own, without help from the devs. It wouldn't be nearly as efficient or elegant, but you could have a bot automatically add a template to all new articles that ensures they're not indexed, and then removes that template once the article is patrolled. It would be a silly way to do it, but it's possible if the devs put up a brick wall. But from what I can tell, we're not yet anywhere near the point of needing to contemplate such actions. Give them a week or so to figure out what they're gonna do. —ScottyWong— 18:04, 20 June 2022 (UTC)
- That is classic WMF dev-speak. There is nothing for them to discuss. It's their way of finding another ruse for stalling. They are pretending - like I said above - that the whole thing needs yet another RfC. Oh, I know Phab so well - and Bugzilla, its predecessor. So do Scottywong and The Blade of the Northern Lights. Kudpung กุดผึ้ง (talk) 14:54, 20 June 2022 (UTC)
Picking a specific # of days for NOINDEX proposal at Village Pump
I opened the phab and asked that the 2012 RFC, which called for unpatrolled articles to be NOINDEXED (indefinitely, until patrolled) be implemented. There was some pretty strong support for that here. In follow-up discussion at Wikipedia:Page Curation/Suggested improvements#Extend NOINDEX beyond 90 days, Novem Linguae and MusikAnimal are suggesting we not go that far and start with 6 months or a year instead of indefinite, and suggest a broader discussion since the RFC was 10 years ago. I don't think more discussion is necessary here, but how about a poll to see if we have a clear local consensus before discussing at VPP: MB 18:50, 21 June 2022 (UTC)
- 180 days
- 365 days
- indefinite
- Anything is fine with me, as long as we as NPPs coalesce around a particular number before we take this to WP:VPR. MusikAnimal recommends 180 days. I originally recommended 365 days.
There is an increase in the ease of the patch if we use an integer number, since indefinite would require custom code.–Novem Linguae (talk) 18:59, 21 June 2022 (UTC)- On Jun 23, someone submitted a patch to allow wgPageTriageMaxAge to be set to infinite easily, so there is no longer a technical barrier to implementing infinite if we desire. –Novem Linguae (talk) 07:52, 23 June 2022 (UTC)
@Chris troutman, Extraordinary Writ, MarioGom, Atsme, Barkeep49, Rsjaffe, Kudpung, North8000, Alexandermcnabb, and Scottywong: Pinging other participants in the discussion, please state your preference as to the time period. MB 19:31, 22 June 2022 (UTC)
- I was going to suggest 180 days when I saw the section title; 365 also seems fine to me and adds more of a buffer in case our backlog situation gets worse. Given that indefinite is apparently more technically difficult, I think it makes sense to pick a cutoff time, even if it's a bit arbitrary. As a sidebar: do we know how the cutoff currently affects articles created from redirects that were created more than 90 days ago? They sit in the back of the queue and it's not clear to me if they get deindexed or not. signed, Rosguill talk 19:43, 22 June 2022 (UTC)
- 365, though I prefer indefinite. I’m thinking about how that would affect the article writers, as I want to see a strong incentive to add notability sources.
- on the other hand, if we fix the workflow, the noindex duration would become less important. — rsjaffe 🗣️ 19:52, 22 June 2022 (UTC)
- 180 is enough - adding that I'm ok with 365 or even indef 03:04, 23 June 2022 (UTC) And I also have questions. Is there a way to check for dupes from time to time while an article is in draft? I've actually come across a few dupes that were in main space and also in draft. Typically, truly notable topics don't just disappear because one attempt failed. I don't see WP running out of articles because we had to draftify a few problematic or unfinished stubs/articles. I was also wondering if there's a way we can promote those types of articles to educators who want to teach the WP editing process, and can make good use of them to show students where the problems are, and how to make them main space ready? Atsme 💬 📧 20:19, 22 June 2022 (UTC)
- 365 Anything less than indefinite creates a perverse incentive for article authors to overwhelm our ability to patrol but I think a calendar year is sufficient time to patrol, especially since implementation of 365 days is apparently technically easier. Chris Troutman (talk) 20:23, 22 June 2022 (UTC)
- 90 days is a big problem.....large amounts of unreviewed articles go past that. Indefinite is best and sort of needed if it's looking at the creation date of the page because the page creation date on articles created by converting a redirect or moving from a draft space is the date of creation of the draft or redirect which can be 10 years ago for a page that just showed up as an article space yesterday. Beyond 180 days you still also have a lot of completely new articles in the NPP cue.....usually ones several NPP'ers looked at and avoided. So indefinite is best, 365 is second best, 180 days is third best. 90 days is far far too short. North8000 (talk) 20:35, 22 June 2022 (UTC)
- I'm opposed to indefinite for philosophical reasons (I think deadlines are helpful to spurring volunteer action and I have a soft spot for our "anyone can contribute" roots) and practical (I don't think the foundation would go for it). No preference between 180 or 365. Best, Barkeep49 (talk) 20:42, 22 June 2022 (UTC)
- 180 with adjustment to filter or a separate report that shows articles over 90 days (actual 90 days, not un-redirects so 2005 articles appear). Between the two, should be able to keep anything egregious out of indexing. Slywriter (talk) 20:53, 22 June 2022 (UTC)
- Slight preference for 365. If the backlog gets beyond that, then the deadline would be the least of our problems for NPP. MarioGom (talk) 21:45, 22 June 2022 (UTC)
- Indefinite or 365 as a second choice if the former is not possible. Someone, sorry I can't recall who, made the excellent point much earlier that article creators - especially UPE, who won't get paid until it goes 'live' - have a much bigger vested interest in creating better articles - I echo User:North8000's concerns about redirects - is the timer indeed article creation or is it reset by queue addition? Sorry, me no teknikal... Two 'sorries' in one sentence - spot the Brit. Best Alexandermcnabb (talk) 04:37, 23 June 2022 (UTC)
- Indefinite or 365 as a second choice if the former is not possible. We should not be offering articles to google if they have not been patrolled, since so much stuff I come across violates core content policies. (t · c) buidhe 05:22, 23 June 2022 (UTC)
- Indefinite or 365 as a second choice if the former is not possible. As there are no chances whatsoever of
'spurring volunteer action'
with or without backlog drives or canvassing for new reviewers. This is a time to be pragmatic and not wax philosophical. 'Indefinite' is not a hurdle and it would avoid having to go back and ask for more next year. As I said earlier, we should not be suggesting anything alternative to 'indefinite', but this RfC unfortunately now opens up the decision by the 'gatekeepers' at the WMF and allow them to beat us down to something that can no longer be negotiated down to when they hold the volunteer community by the throat against the wall - and they will. Kudpung กุดผึ้ง (talk) 07:34, 23 June 2022 (UTC) - Update: it looks as if the devs are going to do 'indefinite' after all. Kudpung กุดผึ้ง (talk) 07:59, 23 June 2022 (UTC)
- yay! >little indef dance< Best Alexandermcnabb (talk) 08:19, 23 June 2022 (UTC)
- Indefinite. I never understood why we had a limit. 90 days, 180 days, a year... it doesn't really make a difference. Articles should not be searchable until they've been checked by at least one other human being. – Joe (talk) 08:55, 23 June 2022 (UTC)
- Indefinite in light of the report that its implementation is technically possible. Unpatrolled pages with potential issues should not be indexed under any circumstances, though I also support a separate report (suggested by Slywriter) for pages older than 90 days so that they don't get lost in the queue and can easily draw attention from multiple reviewers if necessary. ComplexRational (talk) 12:09, 23 June 2022 (UTC)
- BTW I did some looking and also from my sometimes work at the back end of the que ....it looks like that back end of the "regular article" cue is somewhere around 9 months. There it looks like mostly articles that several NPP'ers looked at and decided not to handle. Lot's of what looks like "edge case that should probably go to AFD but I don't want to be the one to decide and do that". Back around a year (or older) it becomes more articles that have a old birth date from their birth as a redirect or draft, but where the actual article is much newer. North8000 (talk) 13:20, 23 June 2022 (UTC)
- Off topic but agree, User:North8000, I've spent the last 10 days at the back end of the horse and the resulting AfDs are at times attracting controversy and a lively debate! Best Alexandermcnabb (talk) 05:28, 24 June 2022 (UTC)
- BTW I did some looking and also from my sometimes work at the back end of the que ....it looks like that back end of the "regular article" cue is somewhere around 9 months. There it looks like mostly articles that several NPP'ers looked at and decided not to handle. Lot's of what looks like "edge case that should probably go to AFD but I don't want to be the one to decide and do that". Back around a year (or older) it becomes more articles that have a old birth date from their birth as a redirect or draft, but where the actual article is much newer. North8000 (talk) 13:20, 23 June 2022 (UTC)
- Indefinite or 365 would encapsulate the maximum size of the NPP queue that I've seen to date. Either one would suit me down to the ground. scope_creepTalk 22:53, 29 June 2022 (UTC)
- 180 days would align with the 6 month limit of draft space. The word Wiki means quick and so talk of years is absurd. And an indefinite limit would be a surrender, encouraging the idea that it's safe to let backlogs climb to infinity. The lack of any sense of urgency would tend to kill motivation so the patrol process would fail even more than it does currently. Andrew🐉(talk) 22:19, 1 July 2022 (UTC)
- 365 days . which will be long enough to deal with the material. "Infinite is more likely to lead to complications. DGG ( talk ) 23:34, 3 July 2022 (UTC)
By my count, no limit (indefinite) has clear consensus (nine including those that said any of the choices were OK, three said 180/365, and no one objected to increasing from 90). I will put a notification at WP:VPR. MB 19:47, 23 June 2022 (UTC)
- No additional comments. Will update the Phab ticket. MB 04:54, 26 June 2022 (UTC)
- I think "indefinite" is a bad idea, slight preference for 180, but OK with 365. Creating and distributing content is in our core mission - that it could be indefinitely hampered because of slow or insufficient volunteers is sort of anti-wiki. If we are going to be fine hosting bad content for an entire year, then asking some search engines to please look the other way past that time is diminishing. — xaosflux Talk 13:13, 29 June 2022 (UTC)
- Follow up. So here is a use-case scenario:
- An editor, Alice, notices we are lacking an article on something she read about Amanita muscaria var. flavivolvata
- Alice writes up a new short article on this, (assuming good faith it is a fine start-class article)
- It sits noindexed - making it harder for the general public to find
- This noindex stays forever until some other volunteer volunteers to approve the page
- Follow up. So here is a use-case scenario:
- Also, as this is a change that impacts almost all editors - as I noted at the just recently listed VP posing, this doesn't seem to be well advertised. — xaosflux Talk 13:14, 29 June 2022 (UTC)
- I thought the idea of this discussion was to workshop a proposal to take to the village pump, at which point it would be widely advertised. – Joe (talk) 13:23, 29 June 2022 (UTC)
- Oh I see, it was taken to VPP. @MB: I think Xaosflux is right, your post at VPP doesn't show sufficient consensus for this idea. It should have been formatted as a clear proposal and advertised at WP:CENT, etc. Instead it reads like you're informing VPP of a decision that's already been made (which isn't the case – the heading of this section is "Picking a specific # of days for NOINDEX proposal at Village Pump"). – Joe (talk) 13:34, 29 June 2022 (UTC)
- Xaosflux and Joe Roe, I echo what Kudpung said below. There was clear consensus at NPP here to reaffirm the prior RFC. I didn't think this had sufficient likelihood of being controversial enough to warrant forking into a separate VPP discussion and/or new RFC. I placed a neutrally worded notice there about this discussion and asked for further comments, of which there have been almost none. This change does not impact editors; people who are concerned about whether their article is indexed are usually trying to promote something which is clearly against our core policies. They should not be rewarded in getting added visibility via search engines after some arbitrary time period because NPP volunteeers are overwhelmed by the quantity of poorly sourced and difficult to review articles. NPP is trying to address the backlog and I do not believe will be any less motivated to do so when this is implemented. We are just trying to improve the encyclopedia by closing this way to get unvetteted articles fully into mainspace. I also note that any newbie who follows our suggestions and uses AFC is blocked from publishing trash, but someone a little more sophisticated bypasses AFC and has a good chance of having their trash live and indexed. This is a step towards equalizing these processes. MB 23:56, 29 June 2022 (UTC)
- I think "indefinite" is a bad idea, slight preference for 180, but OK with 365. Creating and distributing content is in our core mission - that it could be indefinitely hampered because of slow or insufficient volunteers is sort of anti-wiki. If we are going to be fine hosting bad content for an entire year, then asking some search engines to please look the other way past that time is diminishing. — xaosflux Talk 13:13, 29 June 2022 (UTC)
- @Xaosflux and Joe Roe:, IMO this is a purely local NPP issue, and an essential software request. I do not see how
this is a change that impacts almost all editors
; it is directly related to the workload of the NPP process and its inability to keep up with the stream of mostly inadmissible, or at best, articles possibly of relative unimportance, that nowadays make up the bulk of new submissions. ACPERM (also an NPP initiative) greatly reduced the tide of effluent a couple of years ago but it's already grown again to its previous proportions. Indexing for search engines may be a coveted bonus for some creators, but it is not a right. I do not understand why challenging improvements to NPP would be particularly helpful, or why the few genuinely active NPPers should be constantly be made to feel they don't do enough. Prolific reviewers are being lost already through burnout. Kudpung กุดผึ้ง (talk) 22:14, 29 June 2022 (UTC)- Am I reading that currently volunteers are encouraged to quickly patrol new pages due to the current settings, yet by extending the setting they will be less encouraged to quickly process new submissions - especially by extending the setting to forever? Reviewing new pages is indeed important and should be done regardless of the request to external indexing engines - if we are hosting bad information it should be dealt with as soon as possible. — xaosflux Talk 22:57, 29 June 2022 (UTC)
- @Xaosflux: I think you are most likely reading this wrong. Your comment:
'Creating and distributing content is in our core mission '
is correct, but hampering it due to preventing NPP from doing its work by challenging its requests for improvement to the process is, IMHO, decidedly 'anti-Wiki'. When I created the NPP user right in October 2016, it was with the express intent of introducing experienced and high quality patrolling into the system and for no other reason - it was certainly not intended to make it slower! I fail to understand why the WMF is determined to undermine its own objectives. - Hence the local Wikipedias and their task-force volunteers are obliged to take matters into their own hands like they would have done for ACTRIAL if the WMF had not acquiesced after 10 years of bitter wrangling. IMO, WMF employees should not even be participating in these community discussions. Clearly the number of articles in the corpus is a far more important boast for the WMF than the quality and reliability of the content in them. The reason that so many inappropriate or totally unsuitable new pages are submitted is directly due to the WMF's refusal to do anything about it - despite the constant begging for a proper new user welcome page. Compared to some idipendent projects using MediaWiki, the Wikipedia is totally antiquated - a Model T Ford in terms of progress in information technology - and based on ideologies that now, after 20 years have little in common with today's reality: that the English Wikipedia is no longer short of content and the mediocre stuff can either wait or be stopped before it is created. Anyone who cares for Wikipedia and recognisess that NPP is a foundering process, is invited to come up with effective solutions rather than impede them, and to get cracking on implementing them.Kudpung กุดผึ้ง (talk) 23:47, 29 June 2022 (UTC)
- @Xaosflux: I think you are most likely reading this wrong. Your comment:
- Update: I'm not sure if this was what the NPP community really wanted but the rollout that was scheduled by the WMF for next week appears has been successfully blocked. Kudpung กุดผึ้ง (talk) 22:59, 29 June 2022 (UTC)
- Atsme, I don't generally trust the Trustees much more than the WMF. I always got the impressions that the Trustees do the WMF's bidding and their main tool is a rubber stamp. I may be wrong, though, times may have changed. Question of which tail is wagging which dog. I don't follow what goes on there, and I have no time for tedious Zoom meetings. Kudpung กุดผึ้ง (talk) 00:17, 30 June 2022 (UTC)
- @Kudpung: <sigh> Can you expand a bit? What happened and where? Geoff | Who, me? 23:09, 29 June 2022 (UTC)
- @Glane23:, I can understand if not everyone knows what or where Phabricator is (it used to be called Bugzilla), so, <sigh>, here is the link to the Phab task again. Everything you need to know is there → Kudpung กุดผึ้ง (talk) 00:17, 30 June 2022 (UTC)
What's happening now?
- @Chris troutman, Extraordinary Writ, MarioGom, Atsme, Barkeep49, Rsjaffe, MB, North8000, Alexandermcnabb, Scottywong, Novem Linguae, and Buidhe: so what's happening now? Has everything stalled? With or without a backlog drive, 100s of totally unsourced articles in the feed are getting dangerously close to the current 90-day limit. Backlog drives never have a permanent impact and constantly need to be repeated. Maybe it's time to do something else - the ACTRIAL and its ACPERM were a resounding successes - perhaps now pushing 'autoconfirmed' out to XCON with a mandatory use of the Article Wizard might be an option to go for. Anyone can still edit Wikipedia, no one needs a PhD to do it, but as DGG says:
the purpose is not just getting articles; it's teaching editors. This is much more difficult and time consuming, and the existing templates do a notably poor job of it
, which clearly echoes what I was saying above at:The reason that so many inappropriate or totally unsuitable new pages are submitted is directly due to the WMF's refusal to do anything about it - despite the constant begging for a proper new user welcome page. [...] the English Wikipedia is no longer short of content and the mediocre stuff can either wait or be stopped before it is created.
Kudpung กุดผึ้ง (talk) 09:28, 3 July 2022 (UTC)- In addition to the backlog drive, the last two newsletters and Buidhe's recruitment efforts have attracted a number of a new patrollers, and I've been working through this list to find candidates for autopatrolled (I know we both have reservations about that right, but if nothing else it's effective at getting the backlog down). The backlog is going down now: by about 200 articles a day for the last ten days, which if sustained will get us under 10000 again by mid-July and theoretically to zero in a couple of months. As for permanent solutions... extended confirmed is a massive hurdle compared to autoconfirmed and I highly doubt you'd find consensus for an ACXCON. What I'm curious about is why we seemingly get cyclical backlogs like this every two years or so. We talk a lot about the number of reviewers and how many reviews they're doing, but there are other variables that could have just as much of an influence on the backlog: rate of article creation, proportion of articles autopatrolled, time taken to review each article, regularity of reviewing, etc. Maybe getting data on these would generate new ideas. – Joe (talk) 09:57, 3 July 2022 (UTC)
- One issue that confuses us is assuming that the backlog number (which is very essential and useful) indicates more than it does. At any given moment, there are only about 10-20 days worth of manual reviews sitting in backlog. The tiniest shift in our overall "equation" (of incomming articles vs. reviews getting done) causes large changes in the backlog. But it's basically all that we have. Many of the the indicators that would also be very useful are not available. North8000 (talk) 10:38, 3 July 2022 (UTC)
- The swings in the backlog over the last year or two directly correspond to some top reviewers quitting NPP or coming back. The top reviewers I am thinking of are Onel5969 and John B123. Their efforts are appreciated and they are missed. Not sure about trends over a longer time period. –Novem Linguae (talk) 11:01, 3 July 2022 (UTC)
- You are getting to the main huge topic which I avoided trying to get into here.North8000 (talk) 11:37, 3 July 2022 (UTC)
- That's certainly a recurring issue. I forget who corresponds to which spike, but the departure of SwisterTwister, for example, way back in 2016, eventually led to a backlog of more than 22,000 articles. But I think it raises more questions than it answers. Why do we end up depending so heavily on one or two reviewers, who inevitably burn out? When the backlog is low, for example, does one person taking all the low-hanging fruit lessen the engagement of other reviewers and set us up for a crisis down the line? – Joe (talk) 11:38, 3 July 2022 (UTC)
- OK, so I'll dive in. Wikipedia runs on volunteers doing what they enjoy doing, (with some "doing it for a good cause" thrown into that "enjoy" equation) For most reviewers, due to the nature of things, reviewing is an extremely painful, slow process. Then after they hit a certain threshold, where they have fluency in the zillion words of guidelines, policies, venues, have learned that it's a big fuzzy system where they have to make judgement calls, where they need to have a thick skin, and where they realize that they don't have to feel guilty about not reviewing for all of the areas where the article needs fixing / development then it becomes less slow and painful and the evolve into one of the reviewers with bigger numbers who numerically get most of this week's work done. And once in a while they evolve further in all of theses areas and also put in more time they become one of the huge-number rock stars....a handful of them could knock a 14k backlog down to 4k in a week. And then when they leave the opposite can happen. Aside from getting a few things elsewhere in Wikipedia fixed, our best approach would be to get good experienced people started, and then help them develop into the category where it becomes less painful and slow and guilt-ridden to review articles. It's not just about getting the bunch of articles that is in the backlog done, it's about attaining & maintaining the horsepower to keep it in check painlessly. North8000 (talk) 12:21, 3 July 2022 (UTC)
- It's cyclical in that BOTs are being used (they get caught & stopped, but over time, new ones are introduced and resume the spamming), plus we're getting translations from other Wikis by the 100s, and we're getting a lot of new article submissions from South Asia and the Middle East, which creates more redirects and AfDs. Add to that, seasonal sports, new movies and lists. Atsme 💬 📧 15:12, 3 July 2022 (UTC)
- OK, so I'll dive in. Wikipedia runs on volunteers doing what they enjoy doing, (with some "doing it for a good cause" thrown into that "enjoy" equation) For most reviewers, due to the nature of things, reviewing is an extremely painful, slow process. Then after they hit a certain threshold, where they have fluency in the zillion words of guidelines, policies, venues, have learned that it's a big fuzzy system where they have to make judgement calls, where they need to have a thick skin, and where they realize that they don't have to feel guilty about not reviewing for all of the areas where the article needs fixing / development then it becomes less slow and painful and the evolve into one of the reviewers with bigger numbers who numerically get most of this week's work done. And once in a while they evolve further in all of theses areas and also put in more time they become one of the huge-number rock stars....a handful of them could knock a 14k backlog down to 4k in a week. And then when they leave the opposite can happen. Aside from getting a few things elsewhere in Wikipedia fixed, our best approach would be to get good experienced people started, and then help them develop into the category where it becomes less painful and slow and guilt-ridden to review articles. It's not just about getting the bunch of articles that is in the backlog done, it's about attaining & maintaining the horsepower to keep it in check painlessly. North8000 (talk) 12:21, 3 July 2022 (UTC)
- That's certainly a recurring issue. I forget who corresponds to which spike, but the departure of SwisterTwister, for example, way back in 2016, eventually led to a backlog of more than 22,000 articles. But I think it raises more questions than it answers. Why do we end up depending so heavily on one or two reviewers, who inevitably burn out? When the backlog is low, for example, does one person taking all the low-hanging fruit lessen the engagement of other reviewers and set us up for a crisis down the line? – Joe (talk) 11:38, 3 July 2022 (UTC)
- You are getting to the main huge topic which I avoided trying to get into here.North8000 (talk) 11:37, 3 July 2022 (UTC)
- A proper new user welcome page would probably help a bit. Something that is displayed to a user when they're about to create their first new article, pointing them in all the right directions and setting their expectations. Perhaps it's worthwhile for someone to put together a page where we can brainstorm all the requirements for such a page, and maybe even put together a draft of the page itself and present it either to the wider community or WMF. Or maybe this has been done already? I think it's a bit unrealistic to just tell the WMF to "make a better welcome page" without even telling them what problems we're trying to solve or giving any suggestions for what should be on that page. —ScottyWong— 20:45, 3 July 2022 (UTC)
- I did it Scottywong, several times, that why I keep mentioning it. But it's got swept under the carpet and lost in the annals of time and on dead Mac computers, just like the Article Wizard that I painstakingly rewrote and a newbie just went and reverted it all. I have worked successfully with the WMF to get several things done for NPP, but my main complaint is that the volunteers are expected be doing the leg work (i.e. coding) on things that should be paid for, but AFAIK, no one among the 550 paid staff has skills in developing things within the scope of UX and they don't appreciate it very much when someone hands them a beautiful GUI and says, now code this in MediaWiki for us. Kudpung กุดผึ้ง (talk) 21:05, 3 July 2022 (UTC)
- Do any of these drafts still exist somewhere on WP? —ScottyWong— 10:56, 4 July 2022 (UTC)
- @Scottywong: 'swept under the carpet and lost in the annals of time', but I expect I could recreate them easily enough - reluctantly. Kudpung กุดผึ้ง Kudpung กุดผึ้ง (talk) 17:35, 4 July 2022 (UTC)
- Do any of these drafts still exist somewhere on WP? —ScottyWong— 10:56, 4 July 2022 (UTC)
- I did it Scottywong, several times, that why I keep mentioning it. But it's got swept under the carpet and lost in the annals of time and on dead Mac computers, just like the Article Wizard that I painstakingly rewrote and a newbie just went and reverted it all. I have worked successfully with the WMF to get several things done for NPP, but my main complaint is that the volunteers are expected be doing the leg work (i.e. coding) on things that should be paid for, but AFAIK, no one among the 550 paid staff has skills in developing things within the scope of UX and they don't appreciate it very much when someone hands them a beautiful GUI and says, now code this in MediaWiki for us. Kudpung กุดผึ้ง (talk) 21:05, 3 July 2022 (UTC)
- If we follow exactly the NPP instructions, we will never be able to handle things; when we have come near doing so is by using judgement for which articles are fundamentally ok and which aren't, and proceeding accordingly. Judging this takes experience, not just good understanding of the rules. Similarly with editors--it makes not minutes but hours to really teach someone, and I try it more than most, but I have never attempted more than 1 or 2 per week. (I've mostly worked at AfC not NPP the last few years, but I consider that almost identical, except that at NPP its safe to assume 90% of the material is coi.
- I have lately become much less willing to continue. In a few special areas I am the only one handling them fluently, or even handling them at all, is extremely discouraging--and I'm sure many experienced reviewers find themselves similar trying to cope with their own areas of interest almost unaided. The likely result is that many more articles will be mistakenly accepted or rejected. AfD is no substitute, for it requires the at least equal judgment. We will end by discouraging good contributors in unusual areas; we will also be letting through much more junk, but at least those can be dealt with later along with the half million similar junk accumulated over 21 years. Finding new contributors to replace those who get discouraged is much more difficult. I never expected to do this work as long as I've been doing, which is now 15 years. I had earlier expected that by now we would have new people who could take over--and there are a few, but not enough to replace those of us who are leaving--and the ones who do, mostly need further experience and background to work accurately enough. For people like me or Kudpung, WP is perhaps the most worthwhile work we've done in our lives, but we need to gradually stop before we find ourselves forced to. DGG ( talk ) 03:03, 4 July 2022 (UTC)
- Thank you for that DGG, I'm sure it will strike a chord with our long-time reviewers and NPP/AfC activists.Wikipedia has matured and most of the traditional encyclopedic areas are covered and maintained by topic experts who quietly get on with their work. As the Internet becomes more accessible in developing regions and smartphones can be bought for a few dollars, the vast majority of today's new article submissions mainly comprises football (soccer) bios, other sports people and events, Bollywood, hardly intelligible English, vanity pages, spam, and pure junk. This makes the patrolling of new pages a tedious, boring, and soul destroying task and an increasing number of articles being pushed into draftspace.
- Most new users (and some older ones too) resent being taught - believing it's their right to dump what they like in Wikipedia with the expectation that someone will clean their stuff up for them. It might be The Encyclopedia Anyone Can Edit, but editing here is still a privilege and there are rules to be followed. It's hardly a wonder that those who apply for the NPP user right give up so soon, leaving 90% of the work to 10% of the patrollers who then burn out and leave anyway. Like DGG, I'm sure that most patrollers go for the low-hanging fruit and/or topics in their own knowledge areas; I know I do.
- Unfortunately users like DGG and me are no longer spring chickens and very few of the current active reviewers have a long institutional memory or solid experience. As the curation tools get better (and they are a vast improvement on what we were using 12 or 15 years ago) the rate of detection of abuse of privileges gets better and this has led to the alarming discovery that even after the Orangemoody affair a few years ago, we are probably still only scratching the surface. . Kudpung กุดผึ้ง (talk) 05:04, 4 July 2022 (UTC)
July 5 update
It looks like there isn't clear consensus on how much notification/discussion is needed to determine the consensus on this issue. The hangup really is about indefinite, which is favored philosophically by most NPPers to ensure nothing is externally visible by without review. Since there is little to no opposition to extending from 90 days, I just asked at the phab to implement 365 days at least as an interim step. I think we should take that for now if we can get it to give some immediate relief. MB 17:54, 5 July 2022 (UTC)
- Perhaps we can get someone uninvolved to do a compromise closing - it can always be revisited. From above there seem to be 2 very-related issues that are going on: (a) Pages that aren't reviewed for a 90 days fall off the reviewing tool view. I don't think there is anyone that objects to making that longer of itself. (b) Pages that aren't reviewed for 90 days become indexible - that is what has more differing opinions. Now, these appear to be very linked in that (a) seems to just be a view of pages in (b), but if they were divorced from each other (a) being indefinite seems to be fine while values for (b) get hashed out. I don't think the system currently supports that though? — xaosflux Talk 19:04, 5 July 2022 (UTC)
- (a) shouldn't be an issue. Special:NewPagesFeed shows all unreviewed pages regardless of age. I believe this discussion has been entirely about (b). –Novem Linguae (talk) 21:22, 5 July 2022 (UTC)
- @Novem Linguae so there is nothing that will hinder the patrollers from continuing to patrol pages and have a queue of unpatrolled pages? Some of the comments above suggested that this would be hampered. — xaosflux Talk 21:32, 5 July 2022 (UTC)
- Xaosflux, as NL said above, this is entirely about (b). Articles never fall off the reviewing tool view. For example, I just looked at the NPP feed and there are articles going back to 2007 - but those are usually (or probably always) redirects newly turned into articles that still carry the original creation date. The oldest "new" article in the queue is from January 22 of this year. I assure you that in your scenario above, Alice's "fine start-class article" would be reviewed promptly. There are always some reviewers who look at the newest articles and pick off the "lowest hanging fruit". I do that myself in a few topic areas whenever I have a few spare minutes. The old articles that are still unreviewed (and those that we want to remain no-indexed) have usually been looked at by several reviewers, are often stubs, and usually have tags for notability, possible UPE, possible copyvio, etc. These get held up until someone comes along who is willing to spend the time looking for additional sources, translating foreign languages sources, etc. and determine it is fine, or become confident enough to send it to AFD. There has been a lot of VP discussion on shifting the burden for proving notability to the article creator and allowing more liberal draftification of articles that fall short based on their present state (the sources in the article), but that is unlikely to gain consensus. So for the time being, we are stuck here with a lot of difficult articles to review. As having a WP article is so valuable today, there is an endless stream of new articles on NN people and companies. These promotional type articles especially should not be indexed until they have passed NPP. Hope that helps explain things better. MB 23:29, 5 July 2022 (UTC)
- Yes, so there is nothing that is preventing patrollers from patrolling, and this is all about if we should extend the auto-indexing-of-unpatrolled value. That is what I expected was going on - but was trying to be sure to not miss anything in the comments above suggesting that this proposal only impacts reviewers, while it actually has very little impact on their own volunteer workflow - it impacts authors and readers. I contributed above and am personally fine with some of he proposed extensions. — xaosflux Talk 23:39, 5 July 2022 (UTC)
- This is a matter of perspective. This does not impact reviewers' workflow, but it diminishes their work if there is a backdoor way to get a fully-visible WP article. This is another reason (and there are others) for people to say why bother with NPP and go do something else. The only "impact" to readers and authors is that borderline articles are harder to find. IMO, people who care about seeing their article appear in a google search are probably trying to promote something - and that shouldn't be a priority to us. MB 00:23, 6 July 2022 (UTC)
- I already supported extending it above, was just making sure we didn't have something else going on that breaks the workflow unnecessarily. For example, we have a cutoff on other RCP workflows - but we have a much larger technical challenge trying to extend the entire RCP tables much further as the incoming data is much higher. — xaosflux Talk 01:09, 6 July 2022 (UTC)
- You supported extending it only to 180 or 365, and if I am not mistaken, are one of only 3-4, and probably the most vocal, opponents of going indefinite. The phab is on hold pending some indication that there is a "community consensus". Although indefinite is still a topic for discussion, I hope extending to 365 is recognized as having clear support and this, at least, is implemented ASAP. MB 01:42, 6 July 2022 (UTC)
- @MB to be clear, my main being "vocal" part is that there should be current community support for this change, whatever the value. Yes, I expressed a preference for something not-indefinite, however that's just my quiet voice :) The discussion above does seem to be trending to saying 365 is an acceptable first step, and why I suggested someone uninvolved may be able to find a compromise closure. — xaosflux Talk 10:08, 6 July 2022 (UTC)
- You supported extending it only to 180 or 365, and if I am not mistaken, are one of only 3-4, and probably the most vocal, opponents of going indefinite. The phab is on hold pending some indication that there is a "community consensus". Although indefinite is still a topic for discussion, I hope extending to 365 is recognized as having clear support and this, at least, is implemented ASAP. MB 01:42, 6 July 2022 (UTC)
- I already supported extending it above, was just making sure we didn't have something else going on that breaks the workflow unnecessarily. For example, we have a cutoff on other RCP workflows - but we have a much larger technical challenge trying to extend the entire RCP tables much further as the incoming data is much higher. — xaosflux Talk 01:09, 6 July 2022 (UTC)
- This is a matter of perspective. This does not impact reviewers' workflow, but it diminishes their work if there is a backdoor way to get a fully-visible WP article. This is another reason (and there are others) for people to say why bother with NPP and go do something else. The only "impact" to readers and authors is that borderline articles are harder to find. IMO, people who care about seeing their article appear in a google search are probably trying to promote something - and that shouldn't be a priority to us. MB 00:23, 6 July 2022 (UTC)
- Yes, so there is nothing that is preventing patrollers from patrolling, and this is all about if we should extend the auto-indexing-of-unpatrolled value. That is what I expected was going on - but was trying to be sure to not miss anything in the comments above suggesting that this proposal only impacts reviewers, while it actually has very little impact on their own volunteer workflow - it impacts authors and readers. I contributed above and am personally fine with some of he proposed extensions. — xaosflux Talk 23:39, 5 July 2022 (UTC)
- Xaosflux, as NL said above, this is entirely about (b). Articles never fall off the reviewing tool view. For example, I just looked at the NPP feed and there are articles going back to 2007 - but those are usually (or probably always) redirects newly turned into articles that still carry the original creation date. The oldest "new" article in the queue is from January 22 of this year. I assure you that in your scenario above, Alice's "fine start-class article" would be reviewed promptly. There are always some reviewers who look at the newest articles and pick off the "lowest hanging fruit". I do that myself in a few topic areas whenever I have a few spare minutes. The old articles that are still unreviewed (and those that we want to remain no-indexed) have usually been looked at by several reviewers, are often stubs, and usually have tags for notability, possible UPE, possible copyvio, etc. These get held up until someone comes along who is willing to spend the time looking for additional sources, translating foreign languages sources, etc. and determine it is fine, or become confident enough to send it to AFD. There has been a lot of VP discussion on shifting the burden for proving notability to the article creator and allowing more liberal draftification of articles that fall short based on their present state (the sources in the article), but that is unlikely to gain consensus. So for the time being, we are stuck here with a lot of difficult articles to review. As having a WP article is so valuable today, there is an endless stream of new articles on NN people and companies. These promotional type articles especially should not be indexed until they have passed NPP. Hope that helps explain things better. MB 23:29, 5 July 2022 (UTC)
- @Novem Linguae so there is nothing that will hinder the patrollers from continuing to patrol pages and have a queue of unpatrolled pages? Some of the comments above suggested that this would be hampered. — xaosflux Talk 21:32, 5 July 2022 (UTC)
- (a) shouldn't be an issue. Special:NewPagesFeed shows all unreviewed pages regardless of age. I believe this discussion has been entirely about (b). –Novem Linguae (talk) 21:22, 5 July 2022 (UTC)
- With respect, I think you're making a bit of a mountain out of a molehill MB. You asked people to "[pick] a specific # of days for NOINDEX proposal at Village Pump" and we decided on indefinite. So let's just go ahead and make the proposal at the village pump. I think it's very likely to be supported (to be safe we could propose 365 days as an option B per xaosflux) and it's much easier than trying to wikilawyer or badger the MW devs into doing it without a clear, recent consensus. I'd be happy to make the proposal myself, if it's not stepping on your toes. – Joe (talk) 11:08, 6 July 2022 (UTC)
- It may not be realised by everyone here, but 'NoIndex - Indefinite' was an original feature of the Page Curation development process. IMO this doesn't even need debating. It just needs implementing. The people working at, or in charge of Phab may not have been around at the time. Kudpung กุดผึ้ง (talk) 22:30, 7 July 2022 (UTC)
Concerns over AI are warranted
I cannot overemphasize the AI/Bot threat to WP. Perhaps we need WMF to create a tool that can detect such submissions? See the following articles: New AI Generates Horrifyingly Plausible Fake News, This Site Uses AI to Generate Fake News Articles, but there may be hope THIS AI CAN HELP HUMANS SPOT AI-GENERATED FAKE NEWS. An acquaintance emailed an article titled SOMEONE SLIPPED HUNDREDS OF FAKE ARTICLES ABOUT THINGS THAT NEVER HAPPENED ONTO WIKIPEDIA published in The Byte. Atsme 💬 📧 13:30, 6 July 2022 (UTC)
- Regarding the Wikipedia articles (which require a lot more than just text. e.g. references) that was a human that did that. But the threat of fake articles is real. I once uncovered a set of dozens of them. Very technical looking, heavily referenced and very wikipedian looking, but which were actually gibberish. North8000 (talk) 14:09, 6 July 2022 (UTC)
- Oops - should've made it a separate ¶. I miscalculated thinking "someone" in the headline would be the clue along with the username in the article. Regardless, my point was that if a human can slip-in hundreds, imagine what a Bot can do. In the interim, we're here and at a few other venues trying to determine whether to include/exclude (a) imperfect articles that have -0- RS or inadequately sourced, (b) moving them to Draft, or nominating them for deletion, and (c) what is or isn't notable. It appears minor when one considers the hazards of being swamped beyond human capability and the damage fraudulent (and flawed, unsourced, or poorly sourced) articles could do to the project's credibility if we don't work on prevention. j/s ¯\_(ツ)_/¯ Atsme 💬 📧 15:01, 6 July 2022 (UTC)
- We're probably particularly vulnerable on imported articles where there is less scrutiny in the wiki where it was developed and where all of the sources are non-english languages and non-arabic character sets. North8000 (talk) 17:16, 6 July 2022 (UTC)
- North, some of our vulnerabilities are insurmountable without the proper automation/tools. If it takes a decade to discover a single troll editor, how many more have gone undetected? Mark my words - the bots are coming - in fact, they've already been here, and it's just beginning. (Sounds more like the title for a horror movie.) Atsme 💬 📧 17:28, 6 July 2022 (UTC)
- We're probably particularly vulnerable on imported articles where there is less scrutiny in the wiki where it was developed and where all of the sources are non-english languages and non-arabic character sets. North8000 (talk) 17:16, 6 July 2022 (UTC)
- Oops - should've made it a separate ¶. I miscalculated thinking "someone" in the headline would be the clue along with the username in the article. Regardless, my point was that if a human can slip-in hundreds, imagine what a Bot can do. In the interim, we're here and at a few other venues trying to determine whether to include/exclude (a) imperfect articles that have -0- RS or inadequately sourced, (b) moving them to Draft, or nominating them for deletion, and (c) what is or isn't notable. It appears minor when one considers the hazards of being swamped beyond human capability and the damage fraudulent (and flawed, unsourced, or poorly sourced) articles could do to the project's credibility if we don't work on prevention. j/s ¯\_(ツ)_/¯ Atsme 💬 📧 15:01, 6 July 2022 (UTC)
- True that this is very concerning, but I don't see it as a separate problem as much as an extension of the current problem. We're always gonna be on the backfoot in the war against vandalism, whether human or AI; the determinant for us is just our capability to react. It's pretty easy to figure out historical or biographical hoaxes, but extremely difficult with regards to advanced sciences and maths. As North said, it's very easy for a vandal to portray themselves as a someone knowledgeable in a technical field (like a mathematician) and then create nonsense. Some reviewers might assume they're credible because they seem to know what they're talking about, and the topics involved are more complex than 95% of us can understand. I can only imagine how many of our articles on mathematics are complete fiction. Curbon7 (talk) 17:41, 6 July 2022 (UTC)
- @Curbon7: Have you come across any highly technical hoax articles? Almost all the entries on Wikipedia:List of hoaxes on Wikipedia just seem obscure, not technical to the point where most people would throw up their hands, consider it abstract nonsense, and move on. For mathematics in particular I think there are enough keen-eyed editors and readers who can tell when something is bogus, like Wikipedia:Articles for deletion/Smarandache number. I think a more likely place of disruption is, like, obscure species; it'd be interesting to see if any such hoaxes have cropped up. Ovinus (talk) 02:08, 18 July 2022 (UTC)
- Luckily, I don't think a hoax species article would get through. Part of the work species gnomes and species NPPers do is check databases such as Catalogue of Life, WoRMS, NCBI, Index Fungorum, iNaturalist, etc. Every species has a scientific binomial name that should be in one of these databases. If I ran across a species not in these databases while NPPing or running my SpeciesHelper script, I'd scrutinize the citations super carefully, and likely take action. –Novem Linguae (talk) 03:11, 18 July 2022 (UTC)
- I didn't think of that, that's really interesting... I wonder if other things with clear categorizations can be checked automatically. There was a book hoax AfD recently (can't remember the name, unfortunately), so it's conceivable that ISBNs could be checked automatically, although I'm not sure if there exists a free-to-access database of ISBNs. Ovinus (talk) 05:08, 18 July 2022 (UTC)
- Luckily, I don't think a hoax species article would get through. Part of the work species gnomes and species NPPers do is check databases such as Catalogue of Life, WoRMS, NCBI, Index Fungorum, iNaturalist, etc. Every species has a scientific binomial name that should be in one of these databases. If I ran across a species not in these databases while NPPing or running my SpeciesHelper script, I'd scrutinize the citations super carefully, and likely take action. –Novem Linguae (talk) 03:11, 18 July 2022 (UTC)
- @Curbon7: Have you come across any highly technical hoax articles? Almost all the entries on Wikipedia:List of hoaxes on Wikipedia just seem obscure, not technical to the point where most people would throw up their hands, consider it abstract nonsense, and move on. For mathematics in particular I think there are enough keen-eyed editors and readers who can tell when something is bogus, like Wikipedia:Articles for deletion/Smarandache number. I think a more likely place of disruption is, like, obscure species; it'd be interesting to see if any such hoaxes have cropped up. Ovinus (talk) 02:08, 18 July 2022 (UTC)
The big batch that I caught consisted of random pieces of copied text from random sources. So no one item was identifiable as bogus, but the whole article was random gibberish. Most readers would just figure that it was their own weakness that they couldn't understand what was being said. In technical areas outside of my fields I have some Platte river type expertise in many, and my "inch deep" knowledge was enough to spot it. North8000 (talk) 17:58, 6 July 2022 (UTC)
- Interesting topic! Wikipedia's defenses against vandalism and hoaxes (and pretty much any bad edits) are multi-layered, like an onion. The layers from outer to inner are something like ClueBot -> Huggle -> recent changes patrollers -> new page patrollers -> watchlisters -> subject matter experts. The good news is that very few bad edits are able to get through all the outer layers. So the amount of bad edits that need a subject expert to remove them is (in my opinion) pretty low. Considering that we work with millions of articles yet Wikipedia:List of hoaxes on Wikipedia only has around 450 entries, and also that the scandal mentioned in the above newspaper article is from the Chinese Wikipedia, we may be doing OK. –Novem Linguae (talk) 21:42, 6 July 2022 (UTC)
- It took 10 years to discover the 100+/- articles by a single editor. We've had the wool pulled over our eyes more than a few times. I'm not quite as confident that (1) it isn't serious enough to be too concerned over, (2) that it isn't happening as I type this response, or (3) Bots
aren'tmay be busy creating hundreds of articles in main space that we may not be aware of now or for quite a while into the future, if ever – we need certain tools to do the job. I realize that it's hard to be concerned over something that isn't right there staring back at us, or may be taking place in some other part of the project without our knowledge. Granted, luck and good editing skills have been our allies, but why discount the fact that we have serious problems that need automated tools to fix? The WMF is quite capable of getting techs together to write scripts for a crawler or other type of hound dog algorithm that can sniff out some of the problems. It's better for us to have 5,500,000 quality articles than 6,500,000 that include a million garbage articles. Funders donate because they are expecting some level of quality in what we publish. Either way, it's certainly worth pursuing, keeping in mind that an ounce of prevention is worth a pound of cure. Atsme 💬 📧 00:22, 7 July 2022 (UTC)- We could have a CAPTCHA as a solution to the AI problem when a user creates a new article. The counter-vandalism effort will probably go on just as long as this conflict, but there are plenty of bots and users willing to help out. In regards to hoaxes, they will eventually be discovered. Sure, there are probably some 20-year-old hoaxes that still linger out there, but nothing that gets created will go unnoticed forever. CollectiveSolidarity (talk) 21:15, 7 July 2022 (UTC)
- I've suggested captcha in the past, but the flood of articles tend to come from registered users who are proficient Bot users, and/or mass producing translations from already created articles in other languages. I've also learned that AI is capable of citing sources, paraphrasing, fact-checking and creating legible articles. See this article if you haven't already. As for hoaxes - maybe WP could have a "treasure hunt" to find the hoaxes and offer a cash reward. Atsme 💬 📧 16:07, 17 July 2022 (UTC)
- We could have a CAPTCHA as a solution to the AI problem when a user creates a new article. The counter-vandalism effort will probably go on just as long as this conflict, but there are plenty of bots and users willing to help out. In regards to hoaxes, they will eventually be discovered. Sure, there are probably some 20-year-old hoaxes that still linger out there, but nothing that gets created will go unnoticed forever. CollectiveSolidarity (talk) 21:15, 7 July 2022 (UTC)
- It took 10 years to discover the 100+/- articles by a single editor. We've had the wool pulled over our eyes more than a few times. I'm not quite as confident that (1) it isn't serious enough to be too concerned over, (2) that it isn't happening as I type this response, or (3) Bots
- While patrolling, I caught and CSD flagged an article yesterday, Atsme, that I immediately recognised as a hoax (if I had admin tools I could have deleted it on the spot because it was potentially also a negative BLP). I am sure that it would not have been detected by a new(ish) reviewer. This intuition only comes from years of solid experience just like an older soldier's instinct for self-preservation on active service in an urban or jungle war zone, or and older doctor's skill for diagnosis rather than relying on modern med-tech. The problems as far as begging for AI solutions are concerned, is that the WMF (or more exactly the boss of Phab) is still claiming only yesterday, lack of resources for addressing even simple but urgent software bugs. I contend that as the WMF is awash with money, there is no excuse for not hiring more software engineers. The dilemma is knowing whom to escalate to in the vague hierarchy of the WMF when requests are denied or shunted into a holding pen by Phab. Kudpung กุดผึ้ง (talk) 22:49, 7 July 2022 (UTC)
- Kudz, I plan to attend the Trustee meeting next week with bells on, and a positive attitude. Rest assured that I am cautiously optimistic about making our concerns known and getting them escalated to priority status. I am also waiting for the next in-person WikiCon and Wikimania event where I can meet the people I need to meet to get things done. I recently shared some ideas with Bluerasberry re: a special project I was thinking about submitting to WMF, and he was receptive to the idea. I just need to put a team together to help me get it implemented. Oh, and I already submitted some of my written concerns with diffs for the WMF Trustees to review ahead of next week's meeting so they will be prepared for my questions, and if I'm lucky, they'll add it to the agenda. It certainly can't hurt.
- A summary thought in response to what you described above: I do try to teach my trainees what to look for when reviewing articles, especially WP:NCORP and related BLPs, and to use their critical thinking skills when reviewing rather than simply approaching articles from a linear PAG approach. I am very pleased with the last few rounds of trainees I've had the absolute pleasure of working with - they are sharp, know what to look for, know PAGs, and while I haven't kept them under a microscope, I do try to read some of their discussions as reviewers from time to time. But like you, I'm concerned they will get burned out if we don't get the tools we all need to automate some of the mundane tasks. Atsme 💬 📧 23:48, 7 July 2022 (UTC)
- @Atsme, CollectiveSolidarity, Novem Linguae, North8000, and Curbon7: Maybe the NPP School I created 10 years ago is being underused. Perhaps most reviewers don't think they need further training but I remain convinced that many of them never bothered to read the comprehensive tutorial before asking for the right. Maybe others seriously need a refresher course. Around half the 750+ reviewers have never made a patrol while less than 100 are truly active. The tools we most urgently need are listed in the table on the Suggestions page and reviewers are being asked what they think the priorities are (see new tread below). But what we need now are reviewers who are genuinely interested in patrolling new pages, and among them, many more who have a thorough knowledge of Asian media. We are all very interested to hear the results of the Zoom meeting. Kudpung กุดผึ้ง (talk) 22:21, 14 July 2022 (UTC)
- @Kudz, they recorded the meeting, so you can watch it as soon as they get it uploaded. As you may have anticipated, it was more of a trustee
vanitymeeting where they provided updates of their accomplishments and introduced some of the programs they're working on, including the new Enterprise. I wasn't even given the opportunity to speak. The panel harvested and responded to a few text questions originating from the chat (YouTube?) portion of the meeting. The meeting chair (Elena Happen) harvested the questions and the respective trustees, or project leaders responded on Zoom. I thought it odd because a few days before the meeting, WMF sent us a request via the e-list to submit our questions in advance of the meeting so they could prepare, and get them worked into the agenda. One by one the agenda items were presented and discussed only by the trustees/project leaders. An hour+ later, I wasn't getting even a hint that my reason for attending was going to be discussed so I sent Elena a private text asking when the pre-submitted questions would be addressed. A short while later she responded and publicly referenced the presubmitted questions, but started and ended with mine, saying it was not an agenda item but went ahead and asked Shani Evenstein to respond. Long story short, Shani admitted that she knew about the NPP backlog, yada yada,but it appeared to me that she neither had a clue as to what I wanted to discuss nor did she seem interested in doing so. She ended by basically saying the issues at NPP are not inWMF'sBoT's purview, and that it was something the community had to figure out - discussion over. Elena moved to another topic, so I sent Shani a private text asking if I could discuss it with her via email, and she texted back with an email address. Atsme 💬 📧 01:28, 15 July 2022 (UTC) strikes and edits made on 16:07, 17 July 2022 (UTC). See Shani's response directly following Kudz response below, which is also followed by my apology for misunderstanding the BoT's position. Stay tuned - there is more to come, hopefully with the help & guidance of our BoT's. We are already starting to get support from WMF techs - EXCITING!! Atsme 💬 📧
- Atsme, Shani Evenstein is Vice-Chair of the Board of Trustees and her home Wiki is Hebrew Wikipedia. Needless to say therefore that I am not surprised that she is not familiar with NPP and gave you a throwaway answer. NPP is very much in the WMF's purview - they developed it and nobody else can address the issues, nor can the local en.Wiki NPP community be expected to. The WMF is currently looking into some of the NPP issues at Phab but there are no signs of an ETA. User:Rosiestep is also a board member and it may be worth contacting her. Kudpung กุดผึ้ง (talk) 03:44, 15 July 2022 (UTC)
- Hello, @Atsme: and @Kudpung:. Thank you for all you do, for attending the last Open Conversation with the Board of Trustees (BoT), and for sharing your view in these posts. There's a lot to consider. Note, we might have different points of view, but I will do my best to answer in good faith and with the facts so people can have a full picture. A few points to consider:
- The talk was live streamed automatically on YouTube. Here is a link. You can find the exact question mentioned at 1:09:45. Check what I said and if you share quotes, please ensure accuracy. I did not say “NPP are not in WMF's purview”, but rather “…not the BoT purview”. There is a big difference. I also stated the BoT does not deal with operations, but with strategy, and that such matters, such as NPP, are not related to Board work. I added that there is a true desire among WMF staff to help. I believe the Product department has already replied, as they are the ones with the expertise to do so, not the BoT.
- In the past, the BoT was criticized by the community for being non-transparent, at times referred to as “Black Box”, the community not knowing what it does, and at times why it makes the decisions it makes. The fairly new Community Affairs Committee (CAC), which I serve as founding Chair of, has several goals (see Charter), one of which is as a platform to help the BoT be more transparent and allows dialogue between the community, WMF, staff and the BoT. Part of our efforts is conducting Open Conversations with Trustees; and opening askcac{{@}}wikimedia.org email, available for anyone. The BoT has never been as transparent and accessible as it is now. This is not a hoax, nor a publicity stunt, and certainly not coming from a place of vanity. We take our work very seriously. Various people, including Trustees (who are volunteers as well, BTW), spend many hours making these meetings as meaningful and inclusive as possible, with true commitment by the whole BoT, and our CEO and staff, to be as transparent as possible, to have honest and meaningful conversations (even if tough at times). We do our best to listen, address issues that have not been addressed (at times for years), and hopefully, also rebuild trust.
- For the Open Conversations to be efficient, effective and inclusive, they have to be well organized in advance. We prepare for these meetings and send an agenda of main topics, and post it on Meta, a week in advance, and make sure to go through these agenda items. Not only Trustees and staff speak. The main agenda items are followed by a Q & A on the main topics (as opposed to general updates at the beginning). We also have 20-30 minutes in each meeting for open questions in the room (and on YouTube), or pre-submitted questions - not part of the agenda topics. We view it as the best way to run such meetings, but listen to feedback and make changes to make the format more effective almost every time.
- If a community member does not like the format - they can let us know. There are proper channels to do that, like the CAC email, and the survey at the end of each meeting. We are on record for already changing multiple things to the request of the community. We’ll seriously consider a community member’s request to change the format, and answer questions in a timely manner.
- Attendees were told they could write questions/comments in the chat or raise a hand and request to speak. Perhaps you hadn't arrived at the Zoom meeting when that was stated, so we'll remember to mention it more than once in future meetings. If the point is that it was important for you to ask the question live in your own words, then point taken. In future meetings an effort will be made to allow people to ask themselves, if they are in the room.
- On a regular basis, the BoT receives questions from people in the movement on topics that have nothing to do with Board responsibilities. One of the goals of these Open Conversations is better informing the community what the BoT does and what it does not. Having a mechanism for pre-submitted questions allows the BoT to forward questions not within its remits to staff, to give more informed answers. Sending your question in advance, we were able to give an answer that included our Product department perspective. When any Trustee answers a question, these are official answers. They are usually checked and researched beforehand. I happen to know EnWiki quite well, including this issue and others, but even if I didn’t, I would not speak without checking first.
- The BoT has limits to what it can do. “We understand, we care, but it’s not related to our work and here’s where you need to go for answers” may not have been a desired answer, but it is a legitimate answer nonetheless. It’s how organizations work - each part has a different role. This question did not touch a topic under the purview of the BoT. Does that mean the Board doesn’t care? Absolutely not. We care deeply, certainly enough to make it heard, ask staff to give more context and think strategically about longer term solutions. But the BoT does not have ‘magic’ solutions. Not always, anyway.
- Trustees serve the WMF and the Movement as a whole, across all projects and all stakeholders. Any Trustee would have given the same answer, no matter what their home wiki is as the answer that was given was prepared in advance, after checking it with relevant staff. It may not have been the answer you wanted, but it was an honest answer. The BoT understands and cares about anything to do with our most important stakeholder - the volunteers who run the projects. We will continue to look for solutions and support the communities in our Movement in any way we can, even if we can’t fix everything, for everyone, right away.
- Discussions do not have to be over when an Open Conversation meeting is finished. I ended my reply with an open invitation to share more details. I shared the askcac@wikimedia.org email (the proper channel to get in touch with us) multiple times. Atsme, you have also reached out privately, and I immediately responded that I am happy to continue this via email. And obviously, discussion can continue here. Hope this helps.
- Best, Shani (WMF) (talk) 14:01, 16 July 2022 (UTC)
- Hello, @Atsme: and @Kudpung:. Thank you for all you do, for attending the last Open Conversation with the Board of Trustees (BoT), and for sharing your view in these posts. There's a lot to consider. Note, we might have different points of view, but I will do my best to answer in good faith and with the facts so people can have a full picture. A few points to consider:
- Atsme, Shani Evenstein is Vice-Chair of the Board of Trustees and her home Wiki is Hebrew Wikipedia. Needless to say therefore that I am not surprised that she is not familiar with NPP and gave you a throwaway answer. NPP is very much in the WMF's purview - they developed it and nobody else can address the issues, nor can the local en.Wiki NPP community be expected to. The WMF is currently looking into some of the NPP issues at Phab but there are no signs of an ETA. User:Rosiestep is also a board member and it may be worth contacting her. Kudpung กุดผึ้ง (talk) 03:44, 15 July 2022 (UTC)
- @Kudz, they recorded the meeting, so you can watch it as soon as they get it uploaded. As you may have anticipated, it was more of a trustee
- Hi, Shani (WMF), thank you for stopping by; it was a welcome surprise and very much appreciated. First, my apologies for any misunderstandings, and for my poor choice of words in the use of vanity, a word I considered as innocuous as vanity publishing (self-published). I'll try to be more thoughtful in the future. My initial expectations for the meeting were probably more wishful thinking than reality based, hoping there would be active discussions among attendees & trustees resulting in either suggestions or solutions. I take full responsibility for any misunderstandings I may have had in that regard. Perhaps after you receive my proposal via email, it will bring more clarity while providing the best approach for accomplishing what NPP needs in the way of tools, maintenance and improvements. I am happy to say that we have received some encouraging input from Johan (WMF) and a few other techs at Phab, but there is still a great need for a coordinated/designated effort to protect the efficiency of NPP/AfC and the quality of en.WP overall. I'm looking forward to further discussions with you! Atsme 💬 📧 03:00, 17 July 2022 (UTC)
- Hey folks, a couple of things in the horizon which might be helpful here (and for others dealing with the same issues).
- We hope that phab:T273220, StopForumSpam, should be helpful for combatting automated insertion of content, and might be relevant here. For the next step, at some point this quarter, it will be tested on a few wikis in report mode, so we can see that it works well in real-world setting.
- Atsme is already involved in this ticket, but for the rest of you, one interesting way of detecting hoaxes might be phab:T312841#8075735 – see the conversation there. That's promising, but still fairly theoretical – difficult to say what will work and how yet.
- Technical solutions to specific workflows is pretty far from what the Board of Trustees work on. Community Tech did a lot of New Page Patrol work in 2019. For specific tools beyond what's linked above, I think the Community Wishlist is probably the best realistic venue. /Johan (WMF) (talk) 17:38, 15 July 2022 (UTC)
- Thank you, Johan (WMF). Your reply is much appreciated...and encouraging. Atsme 💬 📧 18:03, 15 July 2022 (UTC)
- Hello Atsme; it's been a while since we interacted last. Hope you're doing well! Thanks, too, for being so involved with NPP, including your mentee program. I imagine that's a lot of work. And, Kudz, thanks for the ping (Rosiestep). I won't try to paraphrase or readdress what Shani (WMF) has already said above in her thorough response. I see that Johan (WMF) responded yesterday in this section, and that is great (thanks, Johan!) as Johan is one of the WMF staff involved in this work so I think we're on the right track. --RStephenson (WMF) (talk) 19:53, 16 July 2022 (UTC)
- @RStephenson (WMF), Johan (WMF), and Shani (WMF): thank you all so much for chiming in here, especially Shani for her long and detailed response. Personally I am not a fan of live video meetings, or even Internet chat channels - which might all be temporarily necessary - and I don't participate in them. I look forward to the resumption of face-to-face meet ups, and conferences such as Wkimania, which is how I and two others were able to convince the WMF's VP and C-level devs to develop the new Page Triage system some 10 years ago. Better and more willing WMF support for the volunteer attendees at such conferences should now also be possible.
- NPP is so important it definitely should be part of short, medium, and long term strategy if Wikipedia is to survive. We feel therefore that there is much the BoT could do to persuade the WMF to channel its focus. Unfortunately one of the greatest hurdles for the volunteer community to get its head around is understanding the (former) lack of transparency in the way both the WMF and the BoT operate, and the way the wealth of funds is used. This can and does cause misinterpretation of comments and jumping to conclusions by a volunteer community whose enthusiasm has been abused in the past by having to repair costly WMF mismanagement without thanks or compensation. The NPP volunteers are now nervous and jumpy and more than a few are jumping ship.
- Also unfortunately, the rift between the encyclopedia communities and parts of the WMF's 550-strong staff seems to be still growing, especially since around 2013-2014 when with the mass exodus of senior staff and WMF leadership changed. Since then, and after the initial and obviously expected exponential growth of traditional content slowed, the Foundation's policy has come to be perceived as the growth in the number of articles is more important than the quality, and many volunteers feel that the WMF has lost touch with the grassroots of the organisation. As Atsme says:
It's better for us to have 5,500,000 quality articles than 6,500,000 that include a million garbage articles. Funders donate because they are expecting some level of quality in what we publish.
- The en.Wiki is the WMF's flagship project and New Page Patrol is the only mechanism that both prevents inappropriate articles from entering the corpus, and encouraging, as a field ambulance triage station, creators of articles with potential to improve their work. NPP is therefore crucial to upholding the quality of the encyclopedia and repairing its oft criticised accuracy and neutrality. This is why the encyclopedia needs a system that looks good, performs well, and encourages users of the right calibre to use it and to use it responsibly. NPP bugs and backlogs have become too much of a challenge for those who signed up for it. There are 700+ authorised New Page Reviewers, nearly half of them have never made a patrol, only a tiny fraction are doing all the work and some of these have burned out - it is unreasonable to subject enthusiastic volunteers to such stress, and guilt for not doing enough.
- The 2018 ACPERM driven by NPP volunteers was an immediate success but 4 years later is now already negated by the mutation in new content (see my comment below: in Asian Media). A very significant and growing percentage of new articles are either inappropriate or not in a fit state for immediate inclusion, causing spikes in the backlog which our truly active patrollers are now unable to contain. The current backlog drive is having some impact but the backlog will rise again to untenable proportions when it is over.
- The bugs and features listed at Phabricator (see our local list here) list are important and and despite successful campaigns with the Wishlist in previous years are now too serious to have to compete with the cosmetic and convenience gadgets that are traditionally requested at the Wishlist. We fully understand that the code base used by the WMF for Page Triage is not fully compatible with current iterations of MediaWiki; NPP is a WMF responsibility and its needs override the Wishlist project. The community therefore hopes that a team of developers can be assigned to addressing in a timely manner the technical requests that are either unfulfilled WMF promises or community consensus before the system collapses completely. Kudpung กุดผึ้ง (talk) 01:30, 17 July 2022 (UTC)
- Kudpung & Atsme, thank you both for your replies and additional historic context. I acknowledge reading it. I cannot promise an immediate solution, but I can promise it will be properly discussed. To make sure expectations are realistic, I will add that this topic / issue / problem statement requires further discussions internally, both with our CEO and our new CPTO. As you may know, we have just hired a new CPTO, who will be starting in August. Whatever operational solutions WMF will come up with for this stated problem, it will have to include her. Till she settles in, and till we are able to strategize around this topic (and other related technologically-related topics) further in collaboration with staff, our Product department continues to be aware and continue to work on it to the extent they can; but I hope it is clear that talking about longer-term solutions, a bit more time will be needed to make sure this is properly discussed. Atsme, we look forward to further details you may want to share via email. Best, Shani (WMF) (talk) 11:02, 17 July 2022 (UTC)
Asian media
- Perhaps a NPP School template should be advertised on the main page for the project (WP:NPP). It could help alert users that they should come by if they want to join the patrol, or advise newer patrollers who need a refresher on the process. In regards to Asian media, maybe an essay can be written on foreign media to help native English speakers correctly patrol these pages. Not necessarily a policy, but some advice from an experienced patroller that has reviewed a lot of Asian media. CollectiveSolidarity (talk) 22:48, 14 July 2022 (UTC)
- CollectiveSolidarity, essays get read far less often than you would imagine. The problem here is that the Southasian continent has a greater population than all the traditional English L1 speaking regions together. To write such an essay that would be helpful to the majority of NPPers, every one of the thousands of possibly reliable Asian sources would need to be tabulated. At least in the native English speaking world we are fairly clear about what constitutes a reliable source and even one of the best known mainstream national daily newspapers is banned as a source. Kudpung กุดผึ้ง (talk) 02:12, 15 July 2022 (UTC)
- @Kudpung. So that’s it then? More work expanding this massive list? It would undoubtedly just create an issue at WP:RSN. I wish that considering all the L2 speakers that have joined the project in recent years, we could have at least had a multilingual task force assisting with source assessments. I would’ve suggested translating source assessments from the Southasian language wikis, but they may have different processes for determining reliable sources. CollectiveSolidarity (talk) 03:58, 15 July 2022 (UTC)
- CollectiveSolidarity, essays get read far less often than you would imagine. The problem here is that the Southasian continent has a greater population than all the traditional English L1 speaking regions together. To write such an essay that would be helpful to the majority of NPPers, every one of the thousands of possibly reliable Asian sources would need to be tabulated. At least in the native English speaking world we are fairly clear about what constitutes a reliable source and even one of the best known mainstream national daily newspapers is banned as a source. Kudpung กุดผึ้ง (talk) 02:12, 15 July 2022 (UTC)
- I'm beginning to wonder if we're slowing the growth of other language encyclopedias by accepting so much foreign language material and sources into this English encyclopedia? It's becomes even more concerning when it's UPE and they're marketing a person, place or thing citing foreign language sources that we have no access to, much less are able to read. I realize there's a push for globalism, but we simply don't have enough resources to cover different languages coming at us from countries with populations in the billions. I've got a headache just thinking about it. Atsme 💬 📧 04:10, 15 July 2022 (UTC)
- CollectiveSolidarity, you may find some of these resources helpful: WP:NPPSG, Wikipedia:WikiProject Korea/Reliable sources, User:Novem Linguae/Scripts/CiteHighlighter –Novem Linguae (talk) 04:25, 15 July 2022 (UTC)
- @CollectiveSolidarity and Atsme: That's it, folks, we just don't have enough resources. All the Wikipedias are owned by the WMF but being responsible for their own content they operate very different criteria for notability and sourcing. The French Wiki, for example, is notoriously lax. The en.Wiki is extremely strict. I have often said (and I hope it won't get me into trouble) that en.Wiki is fast becoming the South Asian Wikipedia in English... Kudpung กุดผึ้ง (talk)
- I doubt many have given much thought to what a global democracy would look like, much less a global WP in English. It will become so large and unwieldy readers will lose interest. Right now, globalism seems like a great deal because most are focused on the low hanging fruit of a global economy and the $$ it will bring. What they probably haven't thought about is how things will operate once states and then entire nations lose their autonomy, and are forced into a global currency, global laws, and the mass majority will be governed by some form of global democracy wherein the majority wins. The important question then becomes: who will comprise the majority? It's hard enough to get consensus on WP, but just thinking about what could go wrong in a globalist democracy gives me pause. It appears to me that foresight is seriously lacking on all fronts. Atsme 💬 📧 18:39, 15 July 2022 (UTC)
- @Atsme: you've been reading too many dystopian novels. We all did as teens in the 1950s and 60s - indeed some of it was set reading for our exams, but I believe they are largely out of fashion now. If NPP isn't brought up to snuff very soon, within the next 20 years (after I am long gone) Wikipedia will either be gone, be sold to Elon Musk, or have become a government instrument of propaganda and Montgomery Street a prison for dissidents. Kudpung กุดผึ้ง (talk) 03:40, 17 July 2022 (UTC)
Kudz, I don't read novels – I wait for the movie to be released. I cut my teeth reading the Encyclopedia Britannica, and it still amazes me how many WP articles I can read/absorb in one sitting whereas attempting to read a novel puts me right to sleep. Perhaps my living 50 mi north of Venezuela adds to my wokeness about what can/does happen in a global context. I've also noticed how most people rarely think about dying, even though it's inevitable. I guess it's because we're too busy living. We're reminded how fragile life is when we lose a loved one, or barely escape what could have been a fatality. I've experienced both in the past 4 months. It's human nature to discount things we don't give much thought to...until we're personally affected by them. Atsme 💬 📧 12:23, 17 July 2022 (UTC)
- @Atsme: you've been reading too many dystopian novels. We all did as teens in the 1950s and 60s - indeed some of it was set reading for our exams, but I believe they are largely out of fashion now. If NPP isn't brought up to snuff very soon, within the next 20 years (after I am long gone) Wikipedia will either be gone, be sold to Elon Musk, or have become a government instrument of propaganda and Montgomery Street a prison for dissidents. Kudpung กุดผึ้ง (talk) 03:40, 17 July 2022 (UTC)
- I doubt many have given much thought to what a global democracy would look like, much less a global WP in English. It will become so large and unwieldy readers will lose interest. Right now, globalism seems like a great deal because most are focused on the low hanging fruit of a global economy and the $$ it will bring. What they probably haven't thought about is how things will operate once states and then entire nations lose their autonomy, and are forced into a global currency, global laws, and the mass majority will be governed by some form of global democracy wherein the majority wins. The important question then becomes: who will comprise the majority? It's hard enough to get consensus on WP, but just thinking about what could go wrong in a globalist democracy gives me pause. It appears to me that foresight is seriously lacking on all fronts. Atsme 💬 📧 18:39, 15 July 2022 (UTC)
- @CollectiveSolidarity and Atsme: That's it, folks, we just don't have enough resources. All the Wikipedias are owned by the WMF but being responsible for their own content they operate very different criteria for notability and sourcing. The French Wiki, for example, is notoriously lax. The en.Wiki is extremely strict. I have often said (and I hope it won't get me into trouble) that en.Wiki is fast becoming the South Asian Wikipedia in English... Kudpung กุดผึ้ง (talk)
NPP at ANI
There is a discussion at ANI that is relevant to NPP. Wikipedia:Administrators noticeboard/Incidents#NPP Polyamorph (talk) 16:16, 9 July 2022 (UTC)
I do recommend that all reviewers read that thread, especially newer ones and ones with high patrolling counts, because there are lessons in it for all of us. Kudpung กุดผึ้ง (talk) 02:21, 10 July 2022 (UTC)
Now archived at Wikipedia:Administrators' noticeboard/IncidentArchive1103 § NPP. MarioGom (talk) 14:00, 16 July 2022 (UTC)
AP editor at RfD
An editor has been taken to RfD for very many redirects yesterday (and today's log too). I started to look at their creations, as they are on WP:RAL, so not previously patrolled by us. I found a few more to nominate and another few that I bundled to a previous nom - I'm going to bed but if anyone wants to take a quick look to catch what I missed, I'd appreciate the second set of eyes, as mine are quite tired. Happy Editing--IAmChaos 05:31, 11 July 2022 (UTC)
- Looking at your contribs, I assume you're referring to articles such as Filipinos in c –Novem Linguae (talk) 14:11, 11 July 2022 (UTC)
- The editor who created Filipinos in c (and the large number of similar redirects) appears to have begun realising that these redirects aren't deemed helpful by the rest of the community, but their future creations will definitely need to start getting patrolled again. – Uanfala (talk) 10:40, 15 July 2022 (UTC)
- An WP:RAL (redirect whitelist) user with a large number of RFDs is a bit concerning. I suppose we should ping the editor Castncoot so they have a chance to see this discussion and chime in if they wish. Also pinging admin Rosguill. –Novem Linguae (talk) 11:03, 15 July 2022 (UTC)
- I've removed them from the autolist. signed, Rosguill talk 14:45, 15 July 2022 (UTC)
- Thanks for the ping User:Novem Linguae and User:Rosguill, I just saw this now. I had misunderstood the intention of redirects as being time-saving shortcuts, as with Google-search autopopulating remaining characters or even phrasing, but now I understand that in fact they are meant to hold more weight as a standalone moniker. It would have been nice if someone had mentioned this to me all these years instead of pointing out years of apparent misdirects years later all at once in a tsunami. But now I stand corrected and will be careful to make these fit this tighter criterion. Best, Castncoot (talk) 00:05, 16 July 2022 (UTC)
- Hi User:Rosguill, I’m not sure what list is being referenced above. Whatever it is, may I please be instated on it..or alternatively what are the requirements for instatement..? As I mentioned above, I believe this was a misunderstanding that stands corrected now and going forward. Thank you and best, Castncoot (talk) 01:22, 17 July 2022 (UTC)
- @Castncoot the list is the one linked by both myself and Novem Linguae, WP:RAL. The criteria is listed on the talk page:
The criteria for this pseudoright is an established track record creating uncontroversial redirects
. To apply, you would use the talk page, but it would require 3 patrollers to agree with your request, which, with all due respect, may not be likely while all of your RfDs are so fresh they haven't been closed yet. Happy Editing--IAmChaos 01:33, 17 July 2022 (UTC)- OK, thank you User:IAmChaos, I will let some time wash this wave over first. :) Best, Castncoot (talk) 02:23, 17 July 2022 (UTC)
- @Castncoot the list is the one linked by both myself and Novem Linguae, WP:RAL. The criteria is listed on the talk page:
- Hi User:Rosguill, I’m not sure what list is being referenced above. Whatever it is, may I please be instated on it..or alternatively what are the requirements for instatement..? As I mentioned above, I believe this was a misunderstanding that stands corrected now and going forward. Thank you and best, Castncoot (talk) 01:22, 17 July 2022 (UTC)
- Thanks for the ping User:Novem Linguae and User:Rosguill, I just saw this now. I had misunderstood the intention of redirects as being time-saving shortcuts, as with Google-search autopopulating remaining characters or even phrasing, but now I understand that in fact they are meant to hold more weight as a standalone moniker. It would have been nice if someone had mentioned this to me all these years instead of pointing out years of apparent misdirects years later all at once in a tsunami. But now I stand corrected and will be careful to make these fit this tighter criterion. Best, Castncoot (talk) 00:05, 16 July 2022 (UTC)
- I've removed them from the autolist. signed, Rosguill talk 14:45, 15 July 2022 (UTC)
What PageTriage bugs and features would you like us to prioritize?
Howdy folks. I was recently asked to be the NPP technical issues coordinator and I happily accepted. MB and I are going through the old PageTriage bug reports and feature requests. PageTriage is the software that pops up the NPP toolbar on the right of your screen. There's a ton of old bug reports and feature requests, and some are outdated. Anyway, are there any bug reports and feature requests that jump to mind that you think we should prioritize? Feel free to name one or two issues that are important to you, to get them on our radar. Thanks. –Novem Linguae (talk) 08:42, 14 July 2022 (UTC)
- As a relatively new NPP reviewer, I have fallen foul of a bug in the AfD option three times. The AfD process is started but the tool doesn't complete it properly. I have had to re-edit the article to manually add the deletion tag on the article (plus edit summary) and edit the article's deletion discussion page to add the recommended code ("{{subst:afd2 ..." etc). This occurred most recently (yesterday: 13 July 2022) in relation to Susan Ruth Kamunya. Thanks Paul W (talk) 09:33, 14 July 2022 (UTC)
- Do you get any popups when you experience the AFD bug? If so what do they say? –Novem Linguae (talk) 21:03, 14 July 2022 (UTC)
- Sadly, no. The initial outcome is a deletion tag that includes a lot of italicised text. When it first happened to me, I didn't know it was faulty (but wondered why it didn't immediately appear in that days's AfD listing) - another editor (User:JPxG) kindly fixed the problem and alerted me to it. Paul W (talk) 09:26, 15 July 2022 (UTC)
- Do you get any popups when you experience the AFD bug? If so what do they say? –Novem Linguae (talk) 21:03, 14 July 2022 (UTC)
- Why can’t the NPP tool use Twinkle (or the Twinkle code) for everything Twinkle does. That way, we’re reducing the maintenance effort for NPP and using a code base that’s heavily used and well-maintained. Then place the NPP effort on all the other components’ bugs. — rsjaffe 🗣️ 11:16, 14 July 2022 (UTC)
- Unfortunately this one is harder than it looks and would likely require dozens of hours of dev time on both codebases. The problem has to do with tightly coupled code. –Novem Linguae (talk) 20:23, 14 July 2022 (UTC)
- I guess it’s too late to state that tight coupling is considered a “code smell” in many programming environments. — rsjaffe 🗣️ 20:29, 14 July 2022 (UTC)
- Hmm. One simple implementation of this would be to just have clicking certain PageTriage buttons (such as the "tag" button) pop up the Twinkle tag window. Now that I'm talking this out, that might be an idea worth exploring. Eliminating code duplication would be a win for maintainability, and this would also fix the AFD bug. One complication would be that the user would have to have the Twinkle gadget turned on and fully loaded. Another complication would be that we'd lose Special:Log page curation logging of deletion and tagging actions, and notification tray notifications to the article creator. –Novem Linguae (talk) 20:43, 14 July 2022 (UTC)
- Twinkle Is already aware of NPP and has options pertaining to marking pages reviewed. Perhaps that other functionality could be added to it. — rsjaffe 🗣️ 05:33, 15 July 2022 (UTC)
- Removing tagging and deletion icons from the tool and just having reviewers use Twinkle could also help with the icon spacing issue. It would also remove those “dangerous” buttons from the interface which could improve the Ux (I.e., if I needed to do something about the article other than approval or wiki love, I’d move to Twinkle—conceptually, that seems clearer to me, and would be great for people who already use Twinkle for non NPP purposes). — rsjaffe 🗣️ 07:49, 15 July 2022 (UTC)
- Twinkle Is already aware of NPP and has options pertaining to marking pages reviewed. Perhaps that other functionality could be added to it. — rsjaffe 🗣️ 05:33, 15 July 2022 (UTC)
- Hmm. One simple implementation of this would be to just have clicking certain PageTriage buttons (such as the "tag" button) pop up the Twinkle tag window. Now that I'm talking this out, that might be an idea worth exploring. Eliminating code duplication would be a win for maintainability, and this would also fix the AFD bug. One complication would be that the user would have to have the Twinkle gadget turned on and fully loaded. Another complication would be that we'd lose Special:Log page curation logging of deletion and tagging actions, and notification tray notifications to the article creator. –Novem Linguae (talk) 20:43, 14 July 2022 (UTC)
- I guess it’s too late to state that tight coupling is considered a “code smell” in many programming environments. — rsjaffe 🗣️ 20:29, 14 July 2022 (UTC)
- Unfortunately this one is harder than it looks and would likely require dozens of hours of dev time on both codebases. The problem has to do with tightly coupled code. –Novem Linguae (talk) 20:23, 14 July 2022 (UTC)
- The message box under "Mark as Reviewed" should have the language adjusted to say " Add a message to Talk page for the creator and/or other Reviewers. This message will also be posted to the creator's talk page" This would better match the message left on Article Talk page when using and also hopefully encourage reviewers to use the feature on difficult articles, you know the ones we all skip over repeatedly. Slywriter (talk) 11:58, 14 July 2022 (UTC)
- I have two requests. 1. Move the “mark as reviewed” tickbox from just above the “add tags” button to the right of it. It is easy to hit both together on a touch screen interface and separating them would prevent this. 2. Add “orphaned” to the common tags list. I’ve never got the curation toolbar to do AfDs properly but I think we should either fix it or strip that functionality out. Mccapra (talk) 12:43, 14 July 2022 (UTC)
- #2 would be easy to implement by putting in an edit request at MediaWiki:PageTriageExternalTagsOptions.js. Making this comment as a note to myself to do this, after testing the adjusted code. –Novem Linguae (talk) 20:25, 14 July 2022 (UTC)
- I have problems with the touch interface too. Consider increasing the vertical spacing. — rsjaffe 🗣️ 20:33, 14 July 2022 (UTC)
- The list of unfulfilled bugs and features is at Open Phab requests replete with a handy table made by MB and Novem Linguae that includes the Phab tracks. It might be best to start by listing the items in the table in order of reviewers' priority rather than start requesting new features. That's why this thread has been opened in order to obtain feedback. New patrollers might not be familiar with the history of NPP and its long road to development and further upgrades. No one here would want to go back to the ancient page patrolling system which some of us have been around long enough to remember. On a more personal note, I would be against reverting to Twinkle for all or any of the features of the Curation tool. In my opinion a compact system for patrolling and controlling the quality of new articles is paramount to the ultimate quality of the encyclopedia. Therefore, as the post-2012 system was a salaried WF development (albeit with a great deal of direct collaboration with us), the effort should be to persuade them to continue to maintain and improve it rather than burden the volunteers with it, even if it means rewriting the MediaWiki extension from the ground up. One of the days the system will want to be adopted by other language Wikipedias that don't use Twinkle. Kudpung กุดผึ้ง (talk) 21:36, 14 July 2022 (UTC)
My own experience is having AFD on it malfunction many times. Usually the AFD page not getting created, or getting created without the contents. I started to try to ask about it here and the experts just said use Twinkle. North8000 (talk) 22:56, 14 July 2022 (UTC)
This is a long list of things that are quirks/hurdles for newcomers to overcome but then are fine after one gains experience. Let me know if you want me to list those. North8000 (talk) 23:01, 14 July 2022 (UTC)
- North8000. Sure. Feel free to list them if they are software issues you'd like us to prioritize. –Novem Linguae (talk) 21:32, 15 July 2022 (UTC)
- OK here goes....."list of things that are quirks/hurdles for newcomers to overcome but then are fine after one gains experience":
- Lack of nuts-and-bolts documentation on how it works / what it does. I don't mean coding-level stuff, I mean user level stuff...like "this is specifically what happens when you click this". In other words, the missing real manual at the curation tool info.
- Some way to turn it on. Right now the only way I know is to go to the page feed via page curation and look at articles that it lists. For example, to use it to un-mark a reviewed article.
- OK here goes....."list of things that are quirks/hurdles for newcomers to overcome but then are fine after one gains experience":
- Sometimes the "message bar" default is at "message the reviewer" and if you don't watch for that you end up messaging yourself instead of the editor. I don't know why that even exists, but it should never switch to that automati9cally.
- Switch off or bring a warning screen up for "message the creator" when the "creation" was more than a year ago. Inevitably on those the "creator" is like somebody who created a redirect 10 years ago and not the actual creator which converted the redirect into an article 3 months ago. Or have it go to the person whose action turned the "reviewed" flag off.
- Have the creator in the page feed be the person whose action turned the reviewed flag off, not the one who originally created the article. So if "person A" created it as a redirect 10 years ago, and "Person B" changed the redirect into an article 3 months ago, we see the info on "Person B" not the info on "Person A"
- Sincerely, North8000 (talk) 13:13, 16 July 2022 (UTC)
- The list is not really long but it obviously contains some requests that are more important and/or urgent than others. The ultimate aim is to convince the WMF rather than the self-appointed 'policy controllers' at Phab that they really should be completed as soon as possible, and that the work should be done by paid devs and not expected to be carried out by community volunteers who are already at breaking point. Theoretically the WMF works for us and not vice versa. Some of the requests go back 10 years and are not complex to implement. It should not be too difficult for reviewers to make a top 10 list of priorities, based not only on their personal preferences, but on what they believe is most important for NPP. Twinkle is not an option, Triage/Curation was based explicitly on the premise that it should be compact and not force patrollers to dash around all over the site to find a function they need, and we should not be expected to go back to the pre-2012 NPP system..Kudpung กุดผึ้ง (talk) 23:48, 14 July 2022 (UTC)
- As noted by Slywriter above, and previously by li'l ole me, it would be nice to expand the 'add message' functionality to include another contributor (maybe a drop down of contrbibutors? Shouldn't be too long with newish articles - and it's already there in the 'Wikilove' option...) other than the creator, for instance where redirects have been turned into articles. I'd also like that functionality to be open if you aren't tagging the article - so that a note can be shared with the creator/major contributor to improve the page without dropping it on the article talk and where a tag is not necessary. For instance, I find many editors neglect to put a project tag on the article talk page... Best Alexandermcnabb (talk) 05:10, 15 July 2022 (UTC)
- Alexandermcnabb, most editors neglect to put a project tag on the article talk page. The newbies don't know about projects, and others are in the firm belief that it's up to the community to finish the articles. Naked URLs and missing cats are more things that annoy me and put me off wanting to patrol. Nobody can really blame them though, under the mantra Wikipedia is the encyclopedia anyone can edit and the WMF's policy that the number of articles is more important than the quality, not enough is done to catch new users and say "Woa! Have you completed all the essential bits of your article before you click 'publish'? Here's a list: ..." Easy enough to do, never been done. Kudpung กุดผึ้ง (talk) 09:44, 17 July 2022 (UTC)
- I agree and I agree - and generally find that when I pop 'em a reminder, editors do add project tags... Best Alexandermcnabb (talk) 11:31, 17 July 2022 (UTC)
- Alexandermcnabb, most editors neglect to put a project tag on the article talk page. The newbies don't know about projects, and others are in the firm belief that it's up to the community to finish the articles. Naked URLs and missing cats are more things that annoy me and put me off wanting to patrol. Nobody can really blame them though, under the mantra Wikipedia is the encyclopedia anyone can edit and the WMF's policy that the number of articles is more important than the quality, not enough is done to catch new users and say "Woa! Have you completed all the essential bits of your article before you click 'publish'? Here's a list: ..." Easy enough to do, never been done. Kudpung กุดผึ้ง (talk) 09:44, 17 July 2022 (UTC)
AI wishes
- Auto tag & draftify - stubs & articles that have -0- sources, which would include no sections/sub-sections titled References or Sources, and no citations (raw or otherwise). The auto header tag would be citations needed. Once checked and tagged, draftify with a friendly note to the author's UTP advising them that citing RS is required, and suggest they watch this tutorial, or read Wikipedia:Citing_sources, or they can seek mentorship help at WP:TeaHouse.
- Auto tag stubs/articles that have sources but need citations only.
- See How AI could help make Wikipedia entries more accurate. There may be something in that article relative to finding/determining hoaxes & accuracy.
My focus is more on automation, and lessening the work load for patrollers, and what they have to do after landing on a new article. I may return to expand my list if anything more comes to mind. Atsme 💬 📧 14:56, 17 July 2022 (UTC)
Bangla, Thai, Indian and similar television
I notice a significant part of backlog is Asian television shows. Does anyone have a good indicator of reliable sources? or a general rule to follow - Channels that imply notability? I've approved some that at least listed ratings (though also discovered at least one "news site" regularly inflates, so even those require a deeper dive), prodded a few for only sourcing from promotional materials but been skipping the same ones for weeks now and wondering how we can cut down that part of the queue. Slywriter (talk) 21:27, 14 July 2022 (UTC)
- I have noticed that the majority of the articles in the New Pages Feed are nowadays on Asian topics and that a great many of them either fail GNG or are in a state where they can't sensibly be put in mainspace. It would cramp the style of the native English speaking and/or non Asian reviewers to have to decide on the reliability of Asian sources of which there is often a plethora in some even short articles. Article creators who scrape the Internet's barrel for every website that has a fleeting mention are not being helpful, especially where around half the 750 holders of the New Page Reviewer right have never patrolled a page, and fewer than 100 reviewers are truly active. Kudpung กุดผึ้ง (talk) 21:49, 14 July 2022 (UTC)
- This is one of the reasons we need to stick closer to WP:NOT - WP should not be turned into the TV Guide, a tourist brochure, a schedule for sports broadcasting, a list for game scores, a venue for promoting new movies, etc. We have become the antithesis of NOT, and now that we've got new articles coming at us in uncontrollable numbers as translations from different languages and what not, we will either have to (a) develop AI to control it, (b) put a monthly limit on translated submissions or better yet, (c) have each language encyclopedia provide for their own English translations in their respective encyclopedias, most of which are governed differently from en.WP. For example, their articles can having various language flags at the top of their original article's translation, and keep those translated articles in their project pages. We either organize properly or who knows what will happen, but it doesn't appear to me that we're headed in a good direction if we don't make some preventative changes quickly. We simply don't have the workforce that can handle billions. ~ Eternal Optimist but fading rapidly 12:31, 16 July 2022 (UTC)
- @Slywriter and Atsme: See the latest developments in the thread above at Concerns over AI are warranted, and don't hesitate to state your preferences above at What PageTriage bugs and features would you like us to prioritize? Kudpung กุดผึ้ง (talk) 03:57, 17 July 2022 (UTC)
- Kudpung, until there's a shift in mentality, I doubt any of my ideas will get much traction. I'm an editor, not a content creator. I spend my time in counter-vandalism, AfC, NPP and patrolling contentious pages. I see the mission as curation and improvement at this point. Breaking news (really all new articles) should start as drafts, notability for celebrities/sportspeople should be ratched way up to cover those who actually make an impact in their profession, article should have locked sections for confirmed data that will never change, pending changes should be the default for new users. We should act like the go-to place for the world to get their information that we are, instead of hiding behind "Wikipedia is unreliable" and a "work in progress". Slywriter (talk) 17:10, 17 July 2022 (UTC)
- @Slywriter: if you read everything on this talk page before some of it gets archived again (it won't take you long), you'll get a good overview of the challenges we are permanently faced with and what we are currently trying to do to get some of these issues resolved. You might also find some areas where you can help. Kudpung กุดผึ้ง (talk) 18:01, 17 July 2022 (UTC)
Question about granting the NPP flag
What are the chances of making it a requirement for editors with under 5k edits to attend & graduate from NPPSCHOOL before they can apply for or be granted the right? Atsme 💬 📧 23:03, 17 July 2022 (UTC)
- I don't know so let me argue in both directions. Pro: We get a lot of people who are just hat collectors, and for folks serious about wanting to do NPP this is a great way to get them started and more likely to succeed. Con: NPP is 90% about thoroughly knowing the relevant areas of Wikipedia and only about 10% about learning NPP. Current NPPSchool isn't running 100% and lacks understanding / focus / training on what that 10% is. So it would be just another hurdle to getting NPP'ers and thus reducing the number that we can graduate into real NPP'ers and and then fluent/ big number NPP'ers<> North8000 (talk) 02:15, 18 July 2022 (UTC)
- I don't see a problem such a policy change would solve. Elli (talk | contribs) 07:46, 18 July 2022 (UTC)
- This looks like a good idea as otherwise the gate is wide to any comer, however clueless, but is there not a bottleneck in terms of the restricted numbers of sufficiently experienced NPPers willing to undertake the commitment of leading such courses? I remember when I did it that it was quite intensive and time-consuming, for the trainer as well as the mentee, and not everyone from the in any case short list will want to put aside that much time. But this not a suitable area for enthusiastic beginners. I can't agree that it's better to have lots of untrained NPPers rather than a smaller number of effectively trained and experienced ones - in the long run it creates more work in fixing the reviewing errors, and if the errors are left unfixed it detracts significantly from NPP's standing. Ingratis (talk) 08:23, 18 July 2022 (UTC)
- It seems like a decent idea, but we should probably get more trainers before we try to action this or else we will have a trainee backlog too. | Zippybonzo | Talk | 08:48, 18 July 2022 (UTC)
- agree w/ Zippybonzo--Ozzie10aaaa (talk) 14:11, 18 July 2022 (UTC)
- I like the idea of NPP school for newish editors. It certainly helped me. Notability isn't something that can be cracked by reading the PAGs. The PAGs aren't kept 100% accurate compared to results at AFD, GNG is kept vague on purpose, and many of the SNGs are too detailed and cluttered with noise. Notability has to be painfully reverse engineered, or taught by someone experienced. I'm undecided on making NPPSCHOOL required, but I do like the idea of nudging more newish editors towards it. In the long run, NPPSCHOOL is an area where we may want to appoint a coordinator, and also standardize it. For example I notice there are two common curriculums used by trainers, and one curriculum takes a week, and the other curriculum takes 3 months. –Novem Linguae (talk) 18:59, 18 July 2022 (UTC)
- The NPP School was created in 2012 after the 2011 WP:ACTRIAL was disallowed by the WMF. The school was still underused and the quality of patrolling was low, so in 2016 a user right was created for NPP and the better performing patrollers were grandfathered into the new user group.
- With a new official hat for collection the number of reviewers has grown to 731 . Almost half of them have never made a single patrol, and most of the rest have very low activity. In a last desperate attempt to stem the tide of junk WP:ACPERM was finally rolled in 2018. The quality of patrolling, while much better than pre 2016, still left much to be desired so a probationary feature was added to Requests for permissions.
- The NPP software has been improved to make the task more streamlined and attractive to patrollers but it's still not possible to force them to be good at it or to prevent more hat collectors from signing up. Backlog drives have temporary impact but some prolific patrollers might be reviewing too fast (in some cases a review every 10 seconds). The deliberate misuse of the right is either on the increase or editors are getting better at detecting it, and the amount of it is causing concern.
- Requiring all NPP candidates to attend the school is not a new an idea, but while it would put an end to the hat collecting and much of the deliberate abuse, it would discourage many editors of the right calibre from applying. Apart from the possibility of compulsory training and raising the bar for the NPP right, there are not many solutions. One compromise would be to put all new, accepted applicants on 3 months probation during which they must graduate from the school. Some editors might contend that such a measure would introduce too much added bureaucracy. OTOH if the candidates have already done their homework, they should sail through it. Kudpung กุดผึ้ง (talk) 20:49, 18 July 2022 (UTC)
- I support this idea. I graduated from WP:CVUA and later instructed a couple students there. Just having the aspirant talk to an instructor is enough to suss-out who's just looking for a hat. Similarly, I graduated from WP:GARC to properly do GA Reviews. The tough part is having enough instructors to fulfill student need but I think such a process is invaluable. Beyond preventing new patrollers from making massive mistakes, the cadre of instructors help create a better sense of community shared standards as each student graduates in the mold of their instructor. This might help reduce the practical amount of variance in the project as (I suspect) many NPP'ers don't stick strictly to the flow chart. Chris Troutman (talk) 20:57, 18 July 2022 (UTC)
- I also naturally support the idea, but as I hinted above, it's going to need some workshopping before a solution can be adopted. Kudpung กุดผึ้ง (talk) 21:17, 18 July 2022 (UTC)
- I support your take on this, Chris Troutman. I am only a probationary reviewer, and have made several mistakes in the process even with six months tenure and a GA under my belt. Probably the only solution is to raise the bar with NPP, even if more than half the reviewers are kicked to the curb. It’ll be worth it if less substandard articles end up slipping through the cracks.CollectiveSolidarity (talk) 22:05, 18 July 2022 (UTC)
- @CollectiveSolidarity: I've been here since 2018 and even now I feel like I've only recently started understanding notability as an overall concept, beyond GNG, within the last year or so. That had a lot to do with my (still ongoing) training at NPP school. I still don't feel like I'm quite there in terms of understanding everything, but my self-confidence wavers at times and I have minimal experience with AfDs so I feel like that makes sense. My point is that notability is nuanced and I think that more people could benefit from NPP school if they want to participate in it for that reason. At the same time, I'm not sure how I'd feel about making it mandatory for someone with less than x amount of experience or y amount of edits. Seems kind of like editcountitis that could lead to people trying to game the system. I think I'd prefer something along the lines of "it's strongly encouraged to participate in NPP school if you're an intermediate editor" with some description of what is meant by intermediate editor? Clovermoss (talk) 23:02, 18 July 2022 (UTC)
- @Clovermoss, I would think that intermediate in this sense would be more than 90 days and 500 edits to main space (the bare minimum for the right), but less than a year and 2000 non-automated edits to main space. Again, this proposal is a work in progress, so further suggestions to help refine it would be appreciated. If I was to propose something, I would add a template on WP:NPPR suggesting that prospective reviewers should sign up for NPP school if they want experience for the right. CollectiveSolidarity (talk) 23:26, 18 July 2022 (UTC)
- @CollectiveSolidarity: Yeah, that sounds about right for a description of an intermediate editor. Where it gets complicated is that editcounts can mean very different things. Fixing thousands of typos manually (so technically not an automated edit) vs writing multiple GAs could lead to the former having a much higher edit count despite the latter potentially having a better understanding of overall policy. Also someone who's been a very active editor right from the start might have a better understanding of policy compared to someone who's been more casually participating for years. What I'm saying is that there shouldn't just be an arbitrary cutoff that prevents people that could be good new page patrollers from participating (at least beyond the minimum criteria that already exists, at least imo). There should be some amount of leeway even if there's general guidelines for when someone's ready for the userright, is what I'm trying to get at. Clovermoss (talk) 23:35, 18 July 2022 (UTC), edited 23:41, 18 July 2022 (UTC)
- @Clovermoss, I would think that intermediate in this sense would be more than 90 days and 500 edits to main space (the bare minimum for the right), but less than a year and 2000 non-automated edits to main space. Again, this proposal is a work in progress, so further suggestions to help refine it would be appreciated. If I was to propose something, I would add a template on WP:NPPR suggesting that prospective reviewers should sign up for NPP school if they want experience for the right. CollectiveSolidarity (talk) 23:26, 18 July 2022 (UTC)
- @CollectiveSolidarity: I've been here since 2018 and even now I feel like I've only recently started understanding notability as an overall concept, beyond GNG, within the last year or so. That had a lot to do with my (still ongoing) training at NPP school. I still don't feel like I'm quite there in terms of understanding everything, but my self-confidence wavers at times and I have minimal experience with AfDs so I feel like that makes sense. My point is that notability is nuanced and I think that more people could benefit from NPP school if they want to participate in it for that reason. At the same time, I'm not sure how I'd feel about making it mandatory for someone with less than x amount of experience or y amount of edits. Seems kind of like editcountitis that could lead to people trying to game the system. I think I'd prefer something along the lines of "it's strongly encouraged to participate in NPP school if you're an intermediate editor" with some description of what is meant by intermediate editor? Clovermoss (talk) 23:02, 18 July 2022 (UTC)
Handing out user rights, as everyone here knows, is done through WP:PERM by admins. The minimum required activity levels are the basis and usually much more experience is required to be demonstrated. However, a lot of frivolous requests still get made at PERM, many perhaps in good faith by editors who 'think' they are qualified, many from non native speakers of English who just don't understand the rules, many who believe the 90 days and 500 are an automatic pass for the right, and of course there are the inevitable hat collectors. High edit counts are a poor metric for judging a user's experience, but most of the admins who view the requests for New Page Reviewer are familiar with the challenges of the NPP system and investigate further. Kudpung กุดผึ้ง (talk) 01:31, 19 July 2022 (UTC)
Just as a minor start, the wording of the info banner at the top of the New pages patrol page has been modified. Kudpung กุดผึ้ง (talk) 01:53, 19 July 2022 (UTC)
Does anybody use Special:Log -> Deletion Tag Log?
Does anybody use Special:Log -> Deletion Tag Log (pagetriage-deletion)? I'm thinking about deleting it from the code. The code currently writes all deletion tagging to both that log and the main Page Curation Log (pagetriage-curation). Seems unnecessary to have two logs logging the same thing. @Tol, which log does your backlog drive bot use to check these (hopefully pagetriage-curation)? –Novem Linguae (talk) 23:33, 17 July 2022 (UTC)
- @Novem Linguae, it currently uses
pagetriage-deletion/delete
. I can take a look at usingpagetriage-curation
instead. Tol (talk | contribs) @ 21:32, 18 July 2022 (UTC)- @Tol. Thanks for the reply. Yes would be great if you could change it to pagetriage-curation/delete. Quarry query. I think it's exactly the same log entries. If you don't set log_action = delete, would just need to make sure to ignore some other types of logs in there such as "USER tagged ARTICLE with orphan tag" –Novem Linguae (talk) 23:20, 18 July 2022 (UTC)
The Heat Is On
- 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.
See the AfD @ Wikipedia:Articles for deletion/The Heat Is On (TV series) and you decide. Atsme 💬 📧 02:49, 19 July 2022 (UTC)
- Interesting bit of canvassing here, Atsme, especially on an article that dates back to 2005. It’s also interesting how easily you’ve been fooled by a hoax with absolutely no corroborating evidence except for a few very obvious circular references, but haven’t even tried to verify it with the TV network in question, which has a website with every programme they’ve ever made listed in it… except this one. — TREY MATURIN has spoken 04:25, 19 July 2022 (UTC)
- I just voted there - it is NOT a hoax. Very old British people like me will have a vague recollection of an awful, marginal, silly piece of TV filler with Bobby Davro. Game shows like this were padding for better (more expensive) programming. It's not notable for sure, but it ain't no hoax. Best Alexandermcnabb (talk) 04:54, 19 July 2022 (UTC)
- Then you'll surely also remember the tabloid expose of Davro's "cocaine-fuelled prostitute orgies", how the show was nominated for 6 BAFTAs but only won the award for "Best Female Production Assistant", or when Jamie Oliver hosted the show at the age of 12? Read the first iteration of this article in 2005, it was obviously a joke. gobonobo + c 05:22, 19 July 2022 (UTC)
- Surely this discussion belongs on the AfD? I can't be responsible for the content people posted on WP in 2005, I can just tell you that the show existed and save you the embarrassment of posting it in the Hoax Museum only to find that the show itself was a thing - even if editors wrote gibberish about it. Editors write gibberish all the time... Best Alexandermcnabb (talk) 05:25, 19 July 2022 (UTC)
- I struck my comments at the AfD. Hoax or not (and the evidence for its existence is certainly now clouded, if not obscured, by great billowing gusts of WP:CITOGENESIS), it should be deleted. The issue is that few WP:RS would cover Davro - but there's certainly evidence the creating editor used WP as an outlet for his attempts at satirical output. Best Alexandermcnabb (talk) 05:50, 19 July 2022 (UTC)
- Surely this discussion belongs on the AfD? I can't be responsible for the content people posted on WP in 2005, I can just tell you that the show existed and save you the embarrassment of posting it in the Hoax Museum only to find that the show itself was a thing - even if editors wrote gibberish about it. Editors write gibberish all the time... Best Alexandermcnabb (talk) 05:25, 19 July 2022 (UTC)
Trey Maturin, what I did here is NOT canvassing. You might want to strike your aspersion and read the guideline which clearly states:
An editor who may wish to draw a wider range of informed, but uninvolved, editors to a discussion can place a message at any of the following:
- The talk page or noticeboard of one or more WikiProjects or other Wikipedia collaborations which may have interest in the topic under discussion.
- A central location (such as the Village pump or other relevant noticeboards) for discussions that have a wider impact such as policy or guideline discussions.
If you don't think NPP reviewers have an interest in this topic, think again. Atsme 💬 📧 13:17, 19 July 2022 (UTC)
- No, I don't have any interest in seeing AfDs listed at the NPP talk page unless they have some direct bearing on NPP. An AfD for a page created in 2005 has no interest at all for NPP, and only pollutes this talk page. Whether it is canvassing or just a misguided posting is a different discussion, but please don't do this in the future. Fram (talk) 13:32, 19 July 2022 (UTC)
- Then don't respond to it. It's that simple. I think it is important for NPP reviewers to see these types of things, especially new patrollers. Atsme 💬 📧 15:50, 19 July 2022 (UTC)
- I’m not striking it because it’s not an aspersion: I’m directly accusing you of canvassing by posting a link to an AFD on a hoax article created in 2005 to a forum specifically for brand new articles — moreover an AFD where you are basically a lone voice demanding a link be maintained because of circular references to the hoax article, inviting its recreation. — TREY MATURIN has spoken 16:50, 19 July 2022 (UTC)
- You are incorrect, you have skin in the game as the AfD nominee whose speedy delete was rejected by EurekaLott. The following further validates that you cast an aspersion against me for canvassing and have refused to strike it:
...for discussions that have a wider impact such as policy or guideline discussions.
We are currently working with our tech team to create an algorithm that can sniff-out hoaxes to prevent something like this from happening again. Hoaxes are a big part of NPP - it's something that should have been caught in the beginning and wasn't. It now serves as an excellent example for my trainees, as well new & veteran patrollers. I have better things to do - good day. Atsme 💬 📧 17:06, 19 July 2022 (UTC)
- You are incorrect, you have skin in the game as the AfD nominee whose speedy delete was rejected by EurekaLott. The following further validates that you cast an aspersion against me for canvassing and have refused to strike it:
- I’m not striking it because it’s not an aspersion: I’m directly accusing you of canvassing by posting a link to an AFD on a hoax article created in 2005 to a forum specifically for brand new articles — moreover an AFD where you are basically a lone voice demanding a link be maintained because of circular references to the hoax article, inviting its recreation. — TREY MATURIN has spoken 16:50, 19 July 2022 (UTC)
- Then don't respond to it. It's that simple. I think it is important for NPP reviewers to see these types of things, especially new patrollers. Atsme 💬 📧 15:50, 19 July 2022 (UTC)
Interesting ANI discussion
This is interesting reading about inclusion vs exclusion in Wikipedia. The conversation is winding down, so it's better to not participate unless truly important. Wikipedia:Administrators'_noticeboard/Incidents#BilledMammal_nominations_of_Danish_international_footballers. — rsjaffe 🗣️ 16:13, 19 July 2022 (UTC)
Example draft help
Often I'll see page creations in draft space that are identical to this, or that started like that and then have been gradually changed. It seems like they have found an example article or template somewhere, load it in the edit window, save the page, then go back and replace the placeholder text with text about the subject they want to write about. What I want to know is - where is that coming from? It's not the placeholder text that appears when you use the article wizard. There are approximately 3,000 pages that use that, or portions of it, mostly in userspace or draft space, but I can't find the gadget or tool they are using to load that. ~ ONUnicorn(Talk|Contribs)problem solving 20:42, 21 July 2022 (UTC)
- It comes from {{Biography}}. DanCherek (talk) 20:52, 21 July 2022 (UTC)
- Beat me to it! Yes, I've seen folks paste in the template and not realize they have to subst it. Chris Troutman (talk) 20:54, 21 July 2022 (UTC)
- Thank you both! ~ ONUnicorn(Talk|Contribs)problem solving 21:23, 21 July 2022 (UTC)
- Beat me to it! Yes, I've seen folks paste in the template and not realize they have to subst it. Chris Troutman (talk) 20:54, 21 July 2022 (UTC)
Measuring the backlog
Hi folks - I'm a statistician and fellow NPP member and I'm curious to learn more about how we are tracking the size of the backlog. In particular I'd like to research the volume of articles flowing into and out of the backlog each day in hopes of coming up with an estimate of what it would take to permanently reduce the size of the backlog and keep it hovering as close as possible to zero. Yes this may sound crazy but I think it's a good exercise at the very least if nobody has done it yet. (Clearly the answer is more reviews and/or reviewers, but how do we calculate how many?)
So my specific question is - does anyone know how to pull a list of timestamps for articles both when they are **entering** and **exiting** the queue? Paradoxsociety 00:07, 23 July 2022 (UTC)
- I have had it in the back of my mind to ask for better statistics, such as the number of in and out per day (both article and redirects separately). I don't believe this is readily available at present. All we currently "track" is the total article backlog. That is recorded twice a day I believe (there is a link on the reports section). You can get a count of the redirect backlog from the pages feed. Improving this would be nice, but there are probably higher priorities right now. MB 03:47, 23 July 2022 (UTC)
- I used to use Quarry to take 'snapshots' of the queue back when we were trying to bring it down from the high of 20-25k in 2017. It gave a good visual description of the queue. Did require a bit of work but was pretty useful IMO. For example [1] shows the backlog as it existed on 28 Jan 2018. I don't think there is any easy way of retroactively doing this, I had to run queries on Quarry to get snapshots on the day. As far as your specific question, I feel like it would be possible to query 'articles under 6 months old', 'date created', and 'date reviewed' in some meaningful manner to get the info that you want, since all of these metrics are in public logs. I'd ask on WP:QUARRY if you have an idea of what you want and someone can help with the code. — Insertcleverphrasehere(or here)(or here)(or here) 18:47, 23 July 2022 (UTC)
If you looked at it that way, over time, with 1% fewer reviews, our backlog would be ~40,000 and with 1% more reviews our backlog would be at ~0. In reality, the rate of reviews changes in response to the backlog (and other things) To keep the backlog low, we need happy reviewers who stick around and respond as needed. Right now the reviewing process is painful for most people and they just grin and bear it and many just pitch in when the barn is on fire. North8000 (talk) 11:32, 25 July 2022 (UTC)
- Maybe the key is to figure out how to make it less painful/more fun.
- I think the fun part of NPP is finding an article that isn't very good but has potential and working to improve it, but that is time consuming and if the focus is the backlog then people feel the need to tag the issues and move on quickly rather than doing the fun part.
- I also think there is an unhealthy focus right now on deletion/spam. Yes - the point of NPP is to keep that stuff out and it does need to be kept out - but the idea of "fighting a demoralizing losing battle against a never ending flood of literal garbage" is rather unappealing for most people.
- This needs to be about improving the encyclopedia - that's how you get people involved. ~ ONUnicorn(Talk|Contribs)problem solving 16:54, 25 July 2022 (UTC)
Watch out for draftifying old articles
I've seen this scenario play out three times:
- An old article somehow gets into the new pages feed (in two cases it was because of a history merge operation and in one case it was due to a vandal moving a page out of mainspace and the user reverting the vandalism not having autopatrolled)
- A new page reviewer draftifies the article
- I revert the draftification since we're not supposed to draftify old articles
So, a reminder to everyone here: make sure to check the history properly before draftifying articles found at the back of the queue. * Pppery * it has begun... 13:04, 24 July 2022 (UTC)
- With the backlog, we're also seeing quite a few articles now that are over 90 days old. Which is annoying, 'cos on a few occasions now I've reached for Twinkle's nice red Draftify button just to check and find they're older than three months. Grr! Best Alexandermcnabb (talk) 14:15, 24 July 2022 (UTC)
Prevention really is worth a pound of cure
WHAT IF...our techs could create a script that will prevent an editor from executing publish changes if the material contains a copyvio? Perhaps a copyvio prevention tool would work similar to the way the blacklist tool works; i.e., preventing material from being added. I'd appreciate your thoughts before I add this to our wishlist. Atsme 💬 📧 10:46, 24 July 2022 (UTC) Adding - I just read a comment by SandyGeorgia at WP:ANI Tangential discussion on possibility of a Copyvio algorithm, and she mentioned a bot possibly called Corenbot, that checked every new article. I wonder if she's thinking EranBot, which is a tool that DanCherek brought to my attention in this comment? 11:21, 24 July 2022 (UTC)
- I don't think AI technology is good enough to use a strategy of blocking the edit completely. Let's say you're writing a stub about a professor and you include a bibliography of their works. An article that is mostly the names of their papers will set off Earwig with a very high percentage, but is not a copyright violation. Also, any edit to an old article will also set off the copyright detector, because people will have copied from Wikipedia over the years. If we did something like this, we'd need to two two things:
- 1) Only do this check for new article creation.
- 2) Make the edit a warning that they could click through rather than a full stop.
- Another practical limitation is that copyright detection is A) slow and B) expensive. Earwig takes like 30-60 seconds to load, and is sometimes down (and it is down because it ran out of credits for the day). The reason for both A and B is that it uses a search engine API to grab random snippets of the article and search for those chunks. Search engine APIs are a performance and cost bottleneck.
- It may be possible to run another AI on top of Earwig and train it to have a very low false positive rate. Training an AI to have a very low false positive rate was done with User:ClueBot NG, the most successful AI on Wikipedia. In general, AI works by "training it" with humans giving a yes/no answer, and then it adds data to a "yes" dataset and a "no" dataset, and then it invents algorithms to pass all the "yes" testcases and fail all the "no" testcases to the best of its ability.
- Just my two cents. Hope it helps. –Novem Linguae (talk) 13:13, 24 July 2022 (UTC)
- Thanks, Novem - I think the AI technology actually is advanced enough. We already know machines learn fast, depending on what humans teach them; therefore, it is more likely that limitations are HI-based, not AI. The latter is my AIP (Anything Is Possible) way of thinking, but rest assured, it is only dangerous in early morning, and wanes substantially as the day progresses. It is gone by Happy Hour, which can occur at any time of the day.[FBDB] But seriously,
eitherboth of your suggestions are great!! Perhaps the authors of prior scripts/code (EranBot + Earwig) could put their heads together with you, Danny & the other brilliant techs working for WMF (or possibly our own dedicated team of super-duty techs once our proposal is granted), and can actually make this happen. I know Rome wasn't built in a day, but they didn't have super computers back then, or the need to develop faster, more capable and user friendly software. FB is dabbling in it, and we know WP has their big toe in the deep end, so why not give it a try? Just think about the resources such a tool will free-up, not to mention all the other suggestions on the wish list! Atsme 💬 📧 13:54, 24 July 2022 (UTC)
- Thanks, Novem - I think the AI technology actually is advanced enough. We already know machines learn fast, depending on what humans teach them; therefore, it is more likely that limitations are HI-based, not AI. The latter is my AIP (Anything Is Possible) way of thinking, but rest assured, it is only dangerous in early morning, and wanes substantially as the day progresses. It is gone by Happy Hour, which can occur at any time of the day.[FBDB] But seriously,
- SandyGeorgia was referring to User:CorenSearchBot, which last edited six years ago (contributions). "The source code can be perused here, but it's not pretty." Yup, not at all pretty :( – wbm1058 (talk) 14:59, 24 July 2022 (UTC)
- Yes, thanks; carry on :) SandyGeorgia (Talk) 17:01, 24 July 2022 (UTC)
- From User talk:Coren/Archive/2016/May:
Hey Coren, as you may have noticed, CorenSearchBot has been down since April 2nd. Is there anything that the Community Tech team can do to help CorenSearchBot migrate to the new API? I put some notes for you at phabricator:T131169. Kaldari (talk) 19:18, 11 April 2016 (UTC)
- The bot lost access to Yahoo's index, so they made the new Google Search API available for use from Tool Labs.
- ...but I'm not sure whether Coren ever transitioned his code to use the new Google Search API :-(
- So I'm not sure how helpful the source code would be, even if it could be found somewhere... wbm1058 (talk) 15:30, 24 July 2022 (UTC)
- Wikipedia:Bots/Requests for approval/CorenSearchBot (approved August 2007). Part of the infrastructure that supported this is still hanging around:
- Templates were moved from Coren's user space
- Someone was concerned about this bot's editing.
CorenSearchBot seems to have become over-elaborate and may be causing problems.
- Aha! Source code from August 2007 — wbm1058 (talk) 16:11, 24 July 2022 (UTC)
- I just noticed that the CorenSearchBot BRFA mentioned another copyvio bot, User:Wherebot, which was blocked in November 2007 because it was malfunctioning (contributions) – wbm1058 (talk) 16:34, 24 July 2022 (UTC)
- Aha! Source code from August 2007 — wbm1058 (talk) 16:11, 24 July 2022 (UTC)
- CorenSearchBot used to compare newly created pages to a web search, and reported the results at Wikipedia:Suspected copyright violations for copyvio patrollers to go through. As wbm1058 points out, it stopped being active in 2016. That same year, CopyPatrol was launched with EranBot, which compares new pages and large additions of text to existing pages in mainspace and draftspace using Turnitin's plagiarism detector. So CorenSearchBot has since been completely superseded by this new system. The Turnitin detection works pretty well (though, like Earwig, each plagiarism check uses a Turnitin credit, and we don't have an infinite number of those), but it does require the human review at CopyPatrol just because there's so many possible scenarios that need different responses – copyvio that should be removed, un/attributed quote, un/attributed copying within Wikipedia, un/attributed translation, false positive, simple reverts, restoring previously removed content, etc. DanCherek (talk) 17:05, 24 July 2022 (UTC)
- Ah hah!! Speaking of bots, see User:Eubot. Correct me if I'm wrong, but User:Eubot#Spring 2006 suggests to me that articles were being created by bots as far back as 2006. Perhaps Warofdreams would be kind enough to fill us in on some of the details that possibly may help WMF techs create some of the suggestions I made relative to copyvio, copyediting, corroborating/adding RS, rooting out hoaxes, etc. Atsme 💬 📧 01:13, 27 July 2022 (UTC)
- @Atsme: See Wikipedia:History of Wikipedia bots. Bots were creating articles as far back as 2001–2002. They were a significant factor in bootstrapping Wikipedia into becoming what it is today. – wbm1058 (talk) 01:27, 27 July 2022 (UTC)
Substantial AFD discussion about edge-case election article
The discussion was at Wikipedia:Articles for deletion/1996 Chorley Borough Council election. Besides the normal reasons, I took this one there requesting substantial discussions which could provide useful guidance. This was a typical "all stats" article (meaning it had an intro sentence or 2) for a medium sized area (~107,000 people) that is not a state/province or a country, with no shown GNG-type sources (albiet where it could be argued that they might exist) The discussion was substantial and was open for over 5 weeks. Closed by a highly experienced admin. The result was delete. Sincerely, North8000 (talk) 13:30, 26 July 2022 (UTC)
- One interesting thing to do is first read the closing statement, then read the "save" rationales. I believe that none of the "save" !votes had policy backing. — rsjaffe 🗣️ 15:27, 26 July 2022 (UTC)
- If there's one thing for us here to learn, I think that it's to always keep policy and guidelines in mind before these things get out of hand. I'm not advocating being overly tough, but to keep them in mind. This makes one of several blocks of articles which grew big while clearly not complying with policies and guidelines. People start doing large blocks of these as an activity. then a few get kept at AFD by "because there's lots of articles like that" arguments. And then maybe get added at the wp:outcomes essay and then people quote the essay as evidence of a consensus which overrides policy and guidelines, despite the essay itself clearly saying to not use it in that way. This happened at train stations. I don't know what's next on elections, this was just one where I requested a very thorough discussion. North8000 (talk) 17:58, 26 July 2022 (UTC)
- The mind boggles. There are gazillions of “East Cupcake Council election” articles sourced to stats. I’m going to need a longer retirement. Mccapra (talk) 20:12, 26 July 2022 (UTC)
- I think that a gradual approach is to start with just dealing with new articles. And the convince folks that are doing these to redirect their energies. For example to build one nice "East Cupcake Elections" article instead of an article on each of East Cupcake's 50 elections. North8000 (talk) 20:39, 26 July 2022 (UTC)
- The mind boggles. There are gazillions of “East Cupcake Council election” articles sourced to stats. I’m going to need a longer retirement. Mccapra (talk) 20:12, 26 July 2022 (UTC)
- If there's one thing for us here to learn, I think that it's to always keep policy and guidelines in mind before these things get out of hand. I'm not advocating being overly tough, but to keep them in mind. This makes one of several blocks of articles which grew big while clearly not complying with policies and guidelines. People start doing large blocks of these as an activity. then a few get kept at AFD by "because there's lots of articles like that" arguments. And then maybe get added at the wp:outcomes essay and then people quote the essay as evidence of a consensus which overrides policy and guidelines, despite the essay itself clearly saying to not use it in that way. This happened at train stations. I don't know what's next on elections, this was just one where I requested a very thorough discussion. North8000 (talk) 17:58, 26 July 2022 (UTC)
New Dratification discussion
See Wikipedia:Village_pump_(policy)#Message_to_Author_When_New_Article_Is_Draftified. MB 03:06, 27 July 2022 (UTC)
An idea - would patrollers use this?
So, I have an idea - User:ONUnicorn/Articles for cleanup. This idea was inspired in part by several discussions here in New Page Patrol, but also on the village pumps, ANI, and the current ArbCom case about conduct in deletion discussions. Please take a look. Does this look like something you as patrollers would use? Would you be inclined to use it instead of draftification? Instead of PROD? Instead of AFD? In conjunction with one or more of those? Do you think it would be workable, an improvement over what we have? Or would it be a mess? Would it get sufficient participation? Would you feel obligated to use it? Would you be confused as to when it is appropriate? Right now it's a very rough idea, and any feedback is welcome. ~ ONUnicorn(Talk|Contribs)problem solving 18:54, 27 July 2022 (UTC)
- My concern is that there is already too little participation at AfD, so this would likely not see enough editors to make a difference. ScottishFinnishRadish (talk) 19:01, 27 July 2022 (UTC)
- I see the first three problems as almost certainly occurring. — rsjaffe 🗣️ 19:11, 27 July 2022 (UTC)
- I think the larger problem is the ill-placed inclusionism of the idea. NPP triages pages already in mainspace. There is no reason to try to "save" junk content. In fact, by improving dreck in the mainspace, even if that were possible, would create a perverse incentive for drive-by randos to just drop junk on our website to clean up. I think WP:Requested articles is a better way for the non-editing readers to indicate which articles they would like to see so our experienced volunteers can do it the right way. Chris Troutman (talk) 19:13, 27 July 2022 (UTC)
- Neither for nor against at the moment, but I'll note that Wikipedia:Cleanup, while not used too often, does serve a similar function. -- Tamzin[cetacean needed] (she|they|xe) 19:15, 27 July 2022 (UTC)
Article moved to draft space
One of the reviewers User:Bruxton has moved an article I was working on to draft space giving the reason as "I unable verify any of the references." But they are all cited in the article? I logged in to work on the article and saw it was moved again even after I did everything the previous reviewers asked for (added at least 2 citations - I added 4, with inline references to a 5th study I haven't been able to find yet). The books are not all available in a publicly accessible way, is it then not allowed to use them in the article? I have read the information that was given me and I think the article is ready but it was moved again. what can I do about it? Moonspiel (talk) 09:44, 28 July 2022 (UTC)
- Could you please link to the article/draft in question? -MPGuy2824 (talk) 10:15, 28 July 2022 (UTC)
- It's currently at Draft:Perrissona Gappit case where I submitted it for additional review. Moonspiel (talk) 10:17, 28 July 2022 (UTC)
- Seems to be genuine and the sources are good. The lede is absolutly terrible and needs to be expand. The lede should always have location and dates. There is no info it it at all. scope_creepTalk 10:44, 28 July 2022 (UTC)
- It's currently at Draft:Perrissona Gappit case where I submitted it for additional review. Moonspiel (talk) 10:17, 28 July 2022 (UTC)
IMO the article complies with wp:not and wp:notability and that is sufficient to exist in mainspace. The inability to verify the given references is not a valid reason to say otherwise. That said, the article in such terrible shape that it really isn't an article. Besides the "no lead" issue, the whole article is written like somebody just making some comments rather than presenting the topic. In short, the article does not cover the topic. So, while IMO the article is technically OK to exist in mainspace, why not work on it in draft some more and then move it / get it moved? North8000 (talk) 11:47, 28 July 2022 (UTC)
- @Moonspiel: Despite being technically eligible for mainspace, IMO no reviewer is going to want to be the one to OK that non-article for mainspace. What do you think about working on it in draft space and then moving / getting it moved to mainspace? I'd be happy to get pinged for the move then if you wish. North8000 (talk) 14:01, 28 July 2022 (UTC)
It's not mandatory by any means, but I'd also consider online versions of those sources - they DO exist. There are links to Kieckhefer's work here, for instance! No harm working on it in draftspace, as per my colleagues above. Best Alexandermcnabb (talk) 12:41, 28 July 2022 (UTC)
- And page 32, though not 33, is available on Google books and verifies Perrussone Gapit. "I am unable verify any of the references" seems an inappropriate grounds on which to draftify: presumably the reviewer would not accept any article based on print sources - books or newspapers - but has to see websites? Hmm. @Moonspiel: while you are still working on an article, the {{under construction}} template is helpful as it tells reviewers just that, and the lead needs to summarise the article (particularly, to have geographical context). PamD 13:19, 28 July 2022 (UTC)
- The Ref 3 doc is available and you can verify all of it. Its in German right enough, but it is verifiable. scope_creepTalk 14:24, 28 July 2022 (UTC)
- Here is the document: page 107. It is all valid scope_creepTalk 14:30, 28 July 2022 (UTC)
- The Ref 3 doc is available and you can verify all of it. Its in German right enough, but it is verifiable. scope_creepTalk 14:24, 28 July 2022 (UTC)
- Comment. Good morning all. Thank you all for the comments. My own opinion was that there was no harm in having the editor continue to work on the article. I know that it might cause consternation on the part of the OP to have the article moved to draft - I see they were upset with another reviewer who moved a different article to draft telling them not to do it again. Any reviewer is welcome move the article back to main space at any time. Thank you all for your hard work in this section of the project, and I apologize that I have caused editors extra work with my action. Bruxton (talk) 14:41, 28 July 2022 (UTC)
PageCuration toolbar AFD and PROD bugs
Hello all. I'm happy to announce that I wrote some patches that I think fixed the AFD and PROD bugs in the Page Curation toolbar. These bugs have been around for years and are probably the most common bugs that cause people to get fed up with the toolbar and switch to Twinkle. I would encourage folks to try switching back to the Page Curation toolbar for deletion tagging, and let me know if the below bugs are gone, and let me know if I need to fix any additional bugs.
- PROD now writes the reason
- AFD template now only placed on article page one time
- AFD page now created consistently
cc @ComplexRational, Paul W, and Mccapra: –Novem Linguae (talk) 02:12, 29 July 2022 (UTC)
- That’s great thank you so much. I’ll try it today. Mccapra (talk) 05:21, 29 July 2022 (UTC)
- Tried it out on an AfD and it worked perfectly. Paul W (talk) 14:47, 29 July 2022 (UTC)
- Great work! Curbon7 (talk) 15:04, 29 July 2022 (UTC)
- Thanks for the ping. For some reason, though, my AfD page still did not get created [2]... could it be a lag/cache problem or might there be another outstanding bug? ComplexRational (talk) 15:18, 29 July 2022 (UTC)
List of retracted paleontology papers
If someone would like to comment on the notability of List of retracted paleontology papers that'd be welcome. Some discussion happening at Wikipedia talk:WikiProject Academic Journals#List of retracted paleontology papers. My current take would be that this ought to be draftified until at least two sources showing general topic notability are provided. --Elmidae (talk · contribs) 12:06, 29 July 2022 (UTC)
- Really, what we need is an article on retractions in paleontology. Just having a list gives no context. I have similar concerns with other list articles where it seems that more effort goes into making a list than goes into making the list have meaning. — rsjaffe 🗣️ 18:54, 29 July 2022 (UTC)
Backlog count display
Location
When I first created this template, I added it to the top of all the NPP sub-pages (above the navigation tabs). I could see it at a quick glance at any page I went to. It was moved without any discussion, because someone felt it "looked ugly" there (according to a edit summary). Now it is different places - before the toc, after the toc, in the graph, in a text block. This really bugs me because I have to hunt for it instead of it just being on top on all pages. Does anyone object if I move it back?
Backlog count thresholds
The template just displays the backlog count in black text. The Pending Changes counter also displays a color-coded "level":
- blue = very low
- dark green = low
- light green = moderate
- red = high
- inverted-orange = very high.
Originally, I didn't think this would be useful for NPP. At Pending Changes, very high is over 18 and a few people can review the changes and get it back low relatively quickly - the level typically varies throughout the day; the colors probably do call attention to the backlog and cause some people to review PCs. If we implemented this here, the boundaries would need to be several thousand and it wouldn't work the same way.
Now, I think it might help. After the May newsletter, the backlog dropped about 1,500 then flattened out. After the June newsletter, there was another 1,000 reduction and then flat. Then the July backlog drive led to a reduction of 4,500 in the first half of the month, and then it's been mostly flat again. This shows that just keeping the backlog in people's minds has an affect. Using these thresholds would be another way to do that. Some people might see we are rising and be motivated to do more reviews to keep from crossing into a higher category. (or do more if we are close to reaching a lower threshold). It would be a permanent goal. It may help, probably not much, but we need every little bit we can get in any way. I can't think of any real negative - it's just a line of text wherever people have put the template. We could try this:
- very low <2000
- low 2000-4000
- moderate 4000-6000
- high 6000-8000
- very high >8000
In spite of all the effort the past few months, with this scheme we would still be "very high", but getting down to "high" is not too far off. Comments? MB 05:30, 30 July 2022 (UTC)