Wikipedia:Village pump (idea lab)

 Policy Technical Proposals Idea lab WMF Miscellaneous 

• Table of contents
• First discussion
• End of page
• New post

The idea lab section of the village pump is a place where new ideas or suggestions on general Wikipedia issues can be incubated, for later submission for consensus discussion at Village pump (proposals). Try to be creative and positive when commenting on ideas.
Before creating a new section, please note:

• Discussions of technical issues belong at Village pump (technical).
• Discussions of policy belong at Village pump (policy).
• If you're ready to make a concrete proposal and determine whether it has consensus, go to the Village pump (proposals). Proposals worked out here can be brought there.

Before commenting, note:

• This page is not for consensus polling. Stalwart "Oppose" and "Support" comments generally have no place here. Instead, discuss ideas and suggest variations on them.
• Wondering whether someone already had this idea? Search the archives below, and look through Wikipedia:Perennial proposals.

Discussions are automatically archived after remaining inactive for two weeks.

« Archives, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58

Google Doodle advance notice

Whenever there's a Google Doodle honoring a person, it always drives a ton of traffic to their article. Sometimes we luck out and it's an FA, other times it's only start-class. Idk if Google would be willing to give us advance notice or who to ping at the WMF to get in touch with Google to ask, but it'd help to have some time to prepare the article. {{u|Sdkb}} ^talk 18:13, 8 November 2021 (UTC)

How much traffic are we talking? theleekycauldron (talk • contribs) (they/them) 18:17, 8 November 2021 (UTC)

@Theleekycauldron As a recent example, Charles K. Kao went from <400 page views per day to over 400,000 pageviews when featured in a Google Doodle earlier this week: pageviews. --Ahecht (TALK
PAGE) 18:31, 8 November 2021 (UTC)

Google states that the doodles are "surprising". Probably, they want to keep them a secret. Bada Kaji (talk • श्रीमान् गम्भीर) 18:45, 8 November 2021 (UTC)

Do New Zealand etc. get it early (when it's a global one)? If so they can alert the rest of the world. Nardog (talk) 19:21, 8 November 2021 (UTC)

holy crap. Yeah, we should figure out if there's a way we can get some advance notice. theleekycauldron (talk • contribs) (they/them) 19:00, 8 November 2021 (UTC)

Historical task force: Wikipedia:Articles for improvement/Google Doodle task force. {{u|Sdkb}} ^talk 19:12, 8 November 2021 (UTC)

Sdkb How often would this occour per years? DO you know the rough split by quality for the last yeat?

Looking at the root cause, Maybe we should ask Google to prioritize high quality Article? Of if thy still want a low quality item arrange for them to contact an editor group offline?

Nardog I am fascinated to know as well. I have asked the Kiribati Islands and Google Doodle on Quora Wakelamp d[@-@]b (talk) 03:38, 23 November 2021 (UTC)

Redoing ping of Sdkb and Nardog. Also @Wakelamp:, in your edit that added these pings, you used closing parentheses instead of braces. This caused an unclosed template which meant that all the section links after this one stopped working. I've gone and fixed it. Graham87 11:38, 23 November 2021 (UTC)

Are we too focused on number of edits?

I can't find any proposal to add another measure that recognizes the people who create lots of non-reverted content, or for the people that fix up issues, or people who do AfD /AfC, or discuss on talk, etc.

"What gets measured gets done." may have been said by W. Edwards Deming who would have made a very good editor, Wakelamp d[@-@]b (talk) 12:59, 15 November 2021 (UTC)

Wakelamp d[@-@]b (talk) 12:59, 15 November 2021 (UTC)

Wakelamp, This has been a topic of numerous academic studies. Some have proposed various better measures, but they did not get much traction. In the end, most of these require database analysis that is either difficult or simply nobody bothered with implementing them, so we are still stuck with a simple edit count. _{Piotr Konieczny aka Prokonsul Piotrus| reply here} 11:23, 22 November 2021 (UTC)

The database analysis issue might not be that bad - we already access every single record in the user file and every single record in the edit history. If we can use fields on those files it might be easier. . My thoughts were number of edits would be split into tool/non tool and the same for number of characters

• Ok - Number of characters
• ??? - Tool Assisted
• OK - BOT edit
• Ok - Roll back / reverts
• ??? tags removed - tags added  ???Good faith New Editor genuine Interactions -
• ?? - Roll back / reverts
• ?? - Number of AfD or speedy template
• ?? - Number who have not edited since you AfD/Roll Back/Revert Wakelamp d[@-@]b (talk) 15:57, 22 November 2021 (UTC)

Registered editors by edit count (all registered accounts)

If you have made...	you are about 1 in	then you rank in the...	or the...	That's more than...
1 edit	3	top 30% of all users	top 14,200,000 of all users	70% of all users
2 edits	5	top 20% of all users	top 9,500,000 of all users	80% of all users
5 edits	10	top 10% of all users	top 4,700,000 of all users	90% of all users
10 edits	20	top 5% of all users (the autoconfirmed)	top 2,370,000 of all users	95% of all users
100 edits	100	top 1% of all users	top 475,000 of all users	99% of all users
500 edits	400	top 0.25% of all users (the extended confirmed)	top 118,000 of all users	99.75% of all users
1,000 edits	1,000	top 0.1% of all users	top 47,000 of all users	99.9% of all users
10,000 edits	4,000	top 0.025% of all users	top 11,800 of all users	99.975% of all users
25,000 edits	10,000	top 0.01% of all users	top 4,700 of all users	99.99% of all users
50,000 edits	20,000	top 0.005% of all users	top 2,300 of all users	99.995% of all users
100,000 edits	50,000	top 0.002% of all users	top 900 of all users	99.998% of all users
250,000 edits	200,000	top 0.0005% of all users	top 200 of all users	99.9995% of all users
500,000 edits	1,000,000	top 0.0001% of all users	top 50 of all users	99.9999% of all users
1,000,000 edits	3,300,000	top 0.000031% of users	top 13 of all users	99.99997% of all users
For the purposes of this table, a "user" is a person who has a registered account on the English Wikipedia.

Registered editors by edit count (only successful contributors)

If you have made...	you are about 1 in	then you rank in the...	or the...	That's more than...
1 edit	1	–	one of 14,200,000 contributors	–
2 edits	1-2	top 65% of contributors	top 9,200,000 of all contributors	35% of all contributors
5 edits	3	top 30% of contributors	top 4,200,000 of all contributors	70% of all contributors
10 edits	5	top 20% of contributors (the autoconfirmed)	top 2,855,000 of all contributors	80% of all contributors
100 edits	40	top 2.5% of contributors	top 356,000 of all contributors	97.5% of all contributors
500 edits	133	top 0.75% of contributors (the extended confirmed)	top 107,000 of all contributors	99.25% of all contributors
1,000 edits	200	top 0.5% of contributors	top 71,000 of all contributors	99.5% of all contributors
10,000 edits	1,000	top 0.1% of contributors	top 14,200 of all contributors	99.9% of all contributors
25,000 edits	3,333	top 0.03% of contributors	top 4,200 of all contributors	99.97% of all contributors
50,000 edits	6,666	top 0.015% of contributors	top 2,100 of all contributors	99.985% of all contributors
100,000 edits	14,000	top 0.007% of contributors	top 900 of all contributors	99.993% of all contributors
250,000 edits	66,666	top 0.0015% of contributors	top 200 of all contributors	99.9985% of all contributors
500,000 edits	250,000	top 0.0004% of contributors	top 50 of all contributors	99.9996% of all contributors
For the purposes of this table, a "contributor" is an account with at least one published edit on the English Wikipedia.

Of course we should have a measure which is more meaningful than number of edits... but it's hard to imagine an automated measure that could not easily be gamed. Encouraging pointless edits (as we currently do) is certainly counter-productive. Fabrickator (talk) 13:40, 15 November 2021 (UTC)

See WP:EDITCOUNT.--♦IanMacM♦ ^{(talk to me)} 13:43, 15 November 2021 (UTC)

Wakelamp, I made this table for you. WhatamIdoing (talk) 17:32, 16 November 2021 (UTC)

@WhatamIdoing: Just curious, are those edit counts global or just for enwiki? 192.76.8.91 (talk) 03:33, 17 November 2021 (UTC)

Thank-you for the table. I think it does gives a good indication of experience up to a point, but do you think it encourages peoples to use automated tools? And to encourages people to move from content creation to tool use? And maybe the word top is not the best? And it also has the disadvantage that it makes it impossible for Junior editors to catch up.

There are a number of statistics already in place and measured

X tools gives you 1/ edits in main, talk, wiki 2/ manual or bot/tool assisted 3/ % of edits classified as small, medium, large. 4/ You can also see that in the last 30 days there have only been 100 active admins

http://en.wikichecker.com/

I just found a research project that also measured whether words that you added stayed roughly in the same place versus the the current version. It also measured the number of hours worked. It's a few years old, but the charts are great but it makes the following points

1. Registered editors in English Wikipedia are not getting less productive despite a dramatic reduction in the active population of editors. {BUT it also discusses that this is due to tool use}
2. Anonymous editors contribute substantially to overall productivity; however, their proportion of overall contribution has been steadily declining since the beginning of 2006.

Wakelamp d[@-@]b (talk) 11:18, 17 November 2021 (UTC)

I think that focusing on any metric can encourage people to "win" by that metric. Having multiple metrics can balance these effects.

OTOH, I think that if more of the people on this page knew that we were (almost) all in the 99.9th percentile, we might be less inclined to think that we were "typical" editors and that what works for us is what works for all editors. WhatamIdoing (talk) 23:02, 17 November 2021 (UTC)

These numbers are enwiki only. WhatamIdoing (talk) 19:19, 17 November 2021 (UTC)

@WhatamIdoing: Won't the results be rather skewed then by potentially millions of editors who signed up to edit other languages and other projects, who had accounts created here automatically by central auth without the user asking for one (or even necessarily wanting one)? 192.76.8.95 (talk) 10:28, 18 November 2021 (UTC)

Yes. It also includes people who tried to edit but couldn't figure out our software. WhatamIdoing (talk) 17:24, 18 November 2021 (UTC)

...and users who just created an account to set up their display preferences or keep a watchlist. Cabayi (talk) 14:33, 20 November 2021 (UTC)

I have created a table just as a discussion piece for some major areas, and how I think their work relates to the number of edits. I have added in experience level and automation, to show my understanding that experienced editors use the automated tools more for instance (NPP Editor requirements). Part of this was inspired by reading the very interesting essay User:Cullen328/Smartphone_editing by User:Cullen328. It made me understand the work that dedicated "artisanal" editors do. He uses no tools, but has done 80 K edits; so I created a separate row for editors like him. :-) Wakelamp d[@-@]b (talk) 11:34, 17 November 2021 (UTC)

Article Stage	Work	Edit?	Edits/Hour	Experienced or Junior	Stress Cause	Part of a team	Visible measure	Backlog
New Articles	Content	Manual	1 o 2	Both	Inexperience and Rejection Article Wizard does not have the same checks as NPP Large amount of time to create first article	No	Deleted Articles	No
Redirects	links	Scripts	Immense	Exp	None	No	????	No
AfC	Content Support	Manual	Few	Exp	New User Aggression from not understanding process High Backlog Lack of Resources	Yes	See discussion on AfC page	Medium
Support , Help , Teahouse	Support	Manual	Few	Exp	Agression	Yes	Thanks	No
NPP /Page curator/ Copyrights	Defence	Tools	Large	Exp	Low - high sense of satisfaction	Yes	NPP measures	No
AfD, images, copyright	Defence	Manual	Few	Exp	Aggression - especially based on country etc Scammers Systematic issues Fraud	Yes	Backlog ????	????
CfD	Judge	Manual	Few	Exp	Very low resources Highest Backlog 650 K + categories- nearly all not accessed Over-classifiers No category approval process Very easy to create new categories on the fly No process to ensure category-article process	Yes	Age of categories	Increasing and Huge
Stub, Maintenance tags,templates	Error ID	Tools	Large	Exp	None	Yes	????	????
Small fixes	Error Fix Small Content	Some tools	Few	IP Junior Gnomes	Insufficient Resources Errors are mostly obvious - so why tag, as more work Sense of satisfaction varies Low automation / multi screen process Talk is a waste of time on low importance	Many projects are dead	No measures of tags removed No automatic rating change (for low importance)	Very large and increasing - increase in various errors - movements in quality ratings - templates are in place for years
Animators, Artists	Images	Manual	None		??	???	????	???
Talk	Content	Manua	Low	Exp	Conflict seeking, Ad Hom, OWN,, Admin shortage What is important ? What has already been fixed, but no one closed the talk? Poor user of talk (splitting threads, no consensus building)	No	None	Decades
Content Creators	Content	Manual	Lower	Junior	Overcategorisers Lack of resources	Tech Areas Yes Other Areas vary	Self Direected	Decades
Quality Ugr	Content	Manual	1 o 2	Both	Either high interest areas or Artisanal solo editors working through their interest areas	Tech areas - yes other areas vary	lone	Centuries
Featured Article	Contents	Manual	1 o2	Exp	Resourced Artisanal	Yes	Featured Articles	No
Projects	Content Improvement	Manual But Some tools	???	Exp	Many dead. No recruitment prompts Unclear purpose sometimes Community is criticised by some Some work well, but they need a clear purpose	Varies	????	????
Proposals	Proposals	Manual	Low	Exp	????	????	Qualitative measures only	????
Policies and Procedures	Talk	Manual	Low	EXP	Conflict Resistance versus Reform	????	No measure of effectiveness No readability Extra work is not of concern	????
Bot writers	Tech	Manual	Low	Exp	Maintaining data used by scripts Unclear procedures Distrust by some editors of IT and Wikimedia	????	????	????
Admin	Prosecutor Investigator Judge	Tools Manual	Low High	Exp	Aggression Resources Stress	????	????	????

Wakelamp d[@-@]b (talk) 11:34, 17 November 2021 (UTC)

@Wakelamp, I think that if you are trying to find metrics, then you should look at the edits by Alexbrn. He spends a lot of time removing bad content. His average edit "contribution size" is a negative number. WhatamIdoing (talk) 20:20, 17 November 2021 (UTC)

It's true I do. I'd be fascinated to know what my net article-space byte change number was! Alexbrn (talk) 20:34, 17 November 2021 (UTC)

@Alexbrn I would be fascinated too - I had not realised that the Ecclesiastian Editor("a time to tear down and a time to build") editor existed (There is also a Scottish word meaning Terse, but I can't remember it) In the opposite case, at the extreme , a metric that praised words, could lead to verbosity, or readability issues. Or on Terseness could lead to single word Laconic ("If") articles. ~~~ Wakelamp d[@-@]b (talk) 22:18, 17 November 2021 (UTC)

@WhatamIdoing I think you misunderstand my intent. I am not after metrics for their own sake. I have a few concerns

• That a focus a single metric, (and the proliferation of tool use that has happened over the last 5ish year), has distorted editor behaviour, and
• That large number of tag edits is not increasing the quality of wikipedia especially on low importance items,
• That all Editor types (that have the same Goals as WIkipedia - so most Edit Wars should not be a thing), should be able to visibly see the value of their contribution to Wikipedia.
• So, if metrics should be line with Goals, what are the most important measurable current, and long terms goals of Wikipedia? Wakelamp d[@-@]b (talk) 22:37, 17 November 2021 (UTC)

I'm not sure that most active editors are all that motivated by edit counts. Probably a few people are, at least for brief periods of time, but I'd guess that most people make edits out of a belief that their edits help Wikipedia. I might disagree that some of these edits are actually helpful (see, e.g., people who add Template:Uncat to articles, even though you could just use the automatic Special:UncategorizedPages if you wanted to find pages that needed categorization...), but I don't think people add those tags because it's a quick way to increase their edit counts. WhatamIdoing (talk) 04:08, 18 November 2021 (UTC)

Is there a current report on OAUTH /Bot/Tool assisted edits.. I saw a chart somewhere showing that I think 6000 users were using tools create half the edits NPP has a metric, but I am not sure what it measures

Wikipedia:New pages patrol/Backlog drives/November 2021 Wakelamp d[@-@]b (talk) 14:56, 19 November 2021 (UTC)

"most active editors are all that motivated by edit counts". Primary motivation varies as per the table, but there are lots of studies about people's behaviors changes based on how they are measured.. A few thought experiments

• A proposal is put forward that the number of edits is done is hidden for all users
• A proposal is put forward that the number of edits is no longer added to once it reaches a 1000
• A proposal is put forward that no data will be made available that ranks users by number of edits

Wakelamp d[@-@]b (talk) 14:45, 19 November 2021 (UTC)

There is qualitative and quantitative, computers are great the second and terrible at the former. Meanwhile humans prefer qualitative and find quantitative to be a bad way to measure human performance. This is not unique to Wikipedia! Think of algos that rate people or schools. See Weapons of Math Destruction -- GreenC 21:00, 17 November 2021 (UTC)

That book looks excellent and I have now added it to my reading list. In the Other World, i have watched managers game metrics to achieve their bonus, at the expense of increasing costs/work/stress for others, and reducing the organisation overall. I found this immoral and abhorrent Based on your reading and experience, do you think that computers can measure qualitative issues that aren't about human performance (for instance Article Quality for low importance items, readability) and human performance that is not high stake (Well done on your third month of being an Artisanal editor OR achieving your personal goal OR you may be working too many hours OR are involved in many conflicts -Wikipedia can wait. ~~~ Wakelamp d[@-@]b (talk) 22:59, 17 November 2021 (UTC)

With Google resources :) Hard to say maybe it's a matter of building blocks not today but with semantic web data in place things more possible in the future. -- GreenC GreenC 05:27, 18 November 2021 (UTC)

I think we need at least a split between bot/tool, and manual edits. There has been a massive migration to tools. And tools don't create content Wakelamp d[@-@]b (talk) 09:00, 18 November 2021 (UTC)

Honestly I feel like 1 edit should not be top 50%. This means that with just a simple edit, you're already on of the 50% people with the most edits? ???????????????? I think there should be more, like 15 edits or 25 edits, something like that, to make it look serious. Because that doesn't make much sense to me. WaterflameIsAwesome (talk) 02:34, 18 November 2021 (UTC)

"Top 50%" is a generous rounding on my part. 71.3% of registered editors have never made an edit here. The next 10.4% have made one edit, but only one. The next 4.9% have made exactly two edits. If you have managed to make five edits here, then you are in the top 10% of contributors to the English Wikipedia (by number of edits). WhatamIdoing (talk) 04:04, 18 November 2021 (UTC)

71.3 % is rather large ..... anywhere else I would say it was a bot farm..... Do we know the distribution of account creation??? This doesn't include IP editors does it?? Wakelamp d[@-@]b (talk) 06:32, 18 November 2021 (UTC)

It does not include IP editors. It includes people who didn't know that accounts are unnecessary, people who edit other Wikipedias, and people who tried to edit but couldn't figure out how. WhatamIdoing (talk) 17:25, 18 November 2021 (UTC)

@WhatamIdoing "people who tried to edit but couldn't figure out how." That's just depressing. I am analyzing all the New Article pathways, at the moment and I didn't have that one :-(. (Although I did have stuff about not understanding procedures), Can wiki detect new users who start an article and give up, or are user pathways tracked? It would be interesting to know if that improves with Visual.

@WaterflameIsAwesome The only way to do that is to get more editors to edit. :

Wakelamp d[@-@]b (talk) 00:30, 20 November 2021 (UTC)

The devs can detect failed edit attempts, at least inside a Javascript-based editing environment. A "failed" attempt may be a good thing; it includes opening a page for the purpose of copying the wikitext, or opening the editing window and deciding that your planned edit (or comment) is a bad idea before you save it.

I suspect that this is in the category of sensitive data that is only kept for 90 days. The log item about the failed edit attempt probably (but someone would have to check) associates an IP address (or account id#) with the edit attempt. Once you know the editor, it should be possible to check Special:Contributions t see whether anyone using that IP had recently made a successful edit. I expect that this would be very painful to do manually, so you'd need someone with suitable privileges and automation skills to do it for you. WhatamIdoing (talk) 04:58, 20 November 2021 (UTC)

Seems like the first step in establishing percentages should be to cut off the long tail. "Editors" (people who register an account) who have never made an edit shouldn't be included in the computations. Schazjmd (talk) 15:25, 18 November 2021 (UTC)

@Schazjmd, in round numbers, that would give us:

• 30% have made one edit
• 15% have made two edits
• 10% have made three edits (this is the median editor, 45th to 55th percentile)

and you would be number #2622 out of 12.2 million ever-successfully-edited-here editors, rather than #2622 out of 42.5 million registered editors. WhatamIdoing (talk) 17:33, 18 November 2021 (UTC)

• Edit count is obviously a poor metric because it doesn't measure productivity – an edit can be good, bad or indifferent. And they are a form of cost or input, rather than being a measure of value-added output and achievement.

For an example of a better metric, consider the number of citations added which would be a better measure of quality content. Editors boost their edit count by gaming, griefing, gnoming, gossiping and grinding but these activities don't tend to result in quality content with citations. So, counting citations added might be a better proxy for measuring useful work.

For example, see performance indicator and note that it is banner tagged as needing more citations. The person who added that tag boosted their edit count but didn't add any citations.

Andrew🐉(talk) 14:27, 18 November 2021 (UTC)

How do we count citations? Citations may range over unformated urls or names of books, <ref>...</ref>, or {{sfn*}}, inline or at the bottom of the article. Some may need work to make them appear correctly, but shouldn't an ill-formed attempt to provide a citation count? I like the idea, but I do not want to be the one trying to build a bot to count citations. - Donald Albury 20:39, 18 November 2021 (UTC)

Apparently ORES data quality model counts ref tags, example use in educational dashboard here: https://outreachdashboard.wmflabs.org/courses/The_University_of_Hong_Kong/EASC4407_-_Regional_Geology_Fall_2021_(Fall_2021)/students/overview . Graeme Bartlett (talk) 23:18, 18 November 2021 (UTC)

Neat! I don't see how to access the counter. I am curious to see what it shows for my own edits. The note says it counts ref tags. I wonder if it also counts SFN* templates (which I use as often as I can, these days). I guess improperly formatted citations would have to be fixed before they were counted. - Donald Albury 23:46, 18 November 2021 (UTC)

• ORES is a machine-intelligence mechanism which has to be trained with examples of good and bad edits. This has been done for particular Wikis in particular languages and the details depend on the way that the sample data corpus is labelled as good, bad or whatever. This is a good way of building a metric because, if you have a simple rule like counting edits or citations, then people will then abuse and game it per the cobra effect, Campbell's law, Goodhart's law, Parkinson's law, &c. Ultimately, we ought to be able to run articles through such a tool and decide whether they are good or not. And then attribute this goodness to the editors who wrote it, in proportion to their contributions to the final form. And the final stage will be when the machine intelligence can also write the articles and so cut out the middle men. This is coming too... "A robot wrote this entire article. Are you scared yet, human?" Andrew🐉(talk) 09:49, 19 November 2021 (UTC)

Here's my personal opinion:

1 edit: Top 90%

5 edits: Top 80%

10 edits: Top 75%

50 edits: Top 60%

100 edits: Top 55%

500 edits: Top 50%

1,000 edits: Top 35%

10,000 edits: Top 25%

100,000 edits: Top 15%

500,000 edits: Top 10%

750,000 edits: Top 5%

1,000,000 edits: Top 1% (13 people have 1,000,000 edits)

Obviously, the list would change depending on how many users reach a certain point, how many users DON'T reach a certain point, and how many registered users there are on Wikipedia in general.

I added 1,000,000 edits because of the fact that multiple people have passed it, and it's over 10. However, this is how the list would go on (:O):

10,000,000 edits: Top 0.1%

100,000,000 edits: Top 0.01%

1,000,000,000 edits: Top 0.001%

10,000,000,000 edits: Top 0.0001%

50,000,000,000 edits: Top 0.00001%

100,000,000,000 edits: Top 0.000001%

1,000,000,000,000 edits: Top 0.0000001%

I went all the way to 1 Trillion edits :O

I know this is a far-fetched opinion but please respect it lol

WaterflameIsAwesome (talk) 01:05, 20 November 2021 (UTC)

Number of edits as a measure

As per previous comments, tools make doing a reverts a single click, gives a strong sense of community, and there leader boards showing the number of reverts

BUT there are

Consequences of Reverts using tools - good faith newcomers edit less, are more hesitant for up to a month and leave WP faster., editors active after 2 years has decreased *was 40 now 12%).

• A 2011 WMF by @EpochFail showed that sometimes what is reverted as vandalism, could be good faith newcomers, These good faith newcomers are far,far. more likely to have their first edit rejected, than the more experienced tool using editors were then they started editing
• The WMF paper found that reverting tools are increasingly and far more likely to revert the work of good-faith newcomers. (s there a quality check on false positives??)
• The WMF paper found these automated first edit reverts predict the observed decline in New Editor Retention - nearly 40% of new editors remained active for a year pre-2005, that number dropped to only 12-15% post-2007[
• The WMF paper found that Tool users often do not engage in best practice for discussing reverts or in their interactions .[Maybe a way of reducing this is to have canned comments)
• The WMF paper found that new users are being pushed out of policy articulation. Policies and guidelines are opaque and calcified, but Essays are being created to fix gaps. .
• Two papers by the same author showed an 80 % reduction of edits by new editors who have been reverted compared to new editors that weren't. It wasn't the difference between Vandals and good faith; Both groups had an equal chance of being reverted in the next 5 weeks.
• The same paper also mentions difficulty in "understanding the vast history of prior contributions, decisions, policies, and standards that the community has evolved over time.
• Another paper mentions that a study on a 1000 University students found that editors who had an edit "unfairly" reverted where more likely to vandalize or feel personal animosity towards that edit. This seems at odds with the WMF paper, but this was a qualitative survey
• This 2020 paper discusses these issues The following editors and others were recruited for a research project , were mentioned in the paper and may like to comment @Epochfail @jrmorgan, @Krinkle, @Chicocvenancio, @Rosiestep (Sorry, my second link to you today) , @Barkeep49 , @Nick Moyes @Timtempleton @SkyGazer 512 and @Ohanwe Emmanuel .I.. There are a large number of recommendations, but I would like to point to

"Users should be  incentivized  by  algorithmic  systems  to  behave  in  ways that create enduring value for the community."

• The same issue of keeping new and different editors is brought up, along with a suggestion to explain that it is the AI making that decision. I am not sure whether that is best. Maybe a scheduled edit for far longer than 5 minutes for marginal cases (to make them feel like it took time to a review), and the editor ability to choose a canned comment and a link to teahouse might be better; The paper also makes the assumption that experienced editors are best at managing conflict, But the article points out that conflict is a major reason that prolific editors leave.

"Wikipedia  has  become  like  an  ecosystem,  in  which  certain  kinds  of  people are  quite  well-adapted.  However,  “that  limits  the  diversity of  the  contributors.  So  the  ecosystem  needs  to  change  in  order  to  be  more  welcoming  to  certain  kinds  of  people.” 

“Identifying  stable  edit could  feed  back  into  a  model  that  provides  points  in  some  way that  does  encourage  good  behavior.”  @krinkle

• this paper extends it to show even experienced prolific editors finding negative communication about a revert as a major reason they leave. The authors intend to expand this metric to all Wikipedia
• Lastly, we should consider changing the publish function to run parts of ORES, so that the Editor can make the decisions to fix. Yes, Vandals will work out work arounds, but we can stop that Editor or Anon having access to the tool if they abuse it. Currently they can do the same thing. but it takes 5 minutes for NPP to give them an answer [User:Wakelamp|Wakelamp d[@-@]b]] (talk) 12:51, 21 November 2021 (UTC)

• Yes, many people are too focussed on number of edits, but the solution is not to look for another flawed metric to replace it. We should simply get on with doing the work, rather than evaluating people. Phil Bridger (talk) 12:55, 29 November 2021 (UTC)

• Couldn't agree more with Phil Bridger. Santacruz ⁂ Please ping me! 13:36, 29 November 2021 (UTC)

People are too focused on metrics

Don't get me wrong, it's nice to have metrics (# of edits, # of articles created, # of redirects created, # of GAs, # of FAs, # of DYKs, # of successful AFD nominations, # of files uploaded, etc...), especially if you like numbers. But numbers are just that, numbers. The goal isn't to create the encyclopedia whose contributors' productivity can be measured, but to create the best possible encyclopedia out there, freely available to all of humanity. That I'm 79th in 12 million, or 79th in 44 million in number of edits is really besides the point. Or if we suddenly excluded semi-automated edits from the count, and I dropped to 148th in 12 million in the rankings, the value of my contributions wouldn't suddenly be half of what it used to be. My goal isn't to climb the ranks, my goal is to make Wikipedia better.

What's worth more? 3 FAs or 4 file uploaded to commons? What if it's an FA on a topic no one cares about? What if the image is a simple arrow?

Numbers describe quantities well, not so much subjective value. Headbomb {t · c · p · b} 19:53, 29 November 2021 (UTC)

For me, WP is not just content and structure, but is also an equal WP community of readers and editors. NPOV and 'best possible encyclopedia" need an active renewing community. Productivity is not that important (although wasting other editors time and energy is), but the current measures (Edit measures, number of articles deleted, reverts..) can lead to be behaviour that causes editors to leave. A few more measures won't fix this, but it would be fairer to editors that create content. Let them decide the relative worth ("What's worth more? 3 FAs or 4 files..").

1. Thinking about it, if we just had visibility of monthly figures on the rollovers it could be good
2. "The goal isn't ... contributors' productivity"" AND "But numbers are just that, numbers." I agree. IF we want " the best possible encyclopedia", then we need to concentrate on creation, respecting each other's time, and on editor's staying. IF an action causes a good faith editor to leave, then a one-minute decision has lost hundreds of hours of improvement. Creating thousands of citation missing on articles, disrespects another editor's time and ability; There are undoubtedly a few editors whose behavior is not extreme, but who cause a lot of people to leave by being the second last straw
3. Lastly, "Numbers describe quantities well, not so much subjective value." Subjective value can also cause issue - I get frustrated reading RfCs /reputable sources, because there are no numbers, no references to readers/new editors, and just anecdote. Wakelamp d[@-@]b (talk) 11:17, 30 November 2021 (UTC)

Bot collation of questions on low-watched talk pages

In the discussion at WT:CSD#File talk pages which consist only of boilerplates, wikiproject tags, and/or text which has been copied to Commons Jo-Jo Eumerus commented that File talk page discussions are pretty common but they usually don't get an answer. to which Redrose64 replied Mainly because they only have one watcher - the person who created the discussion. This got me wondering about the feasibility of (probably) a bot that looked for new posts to pages with fewer than N (actively engaged?) watchers and produced a list (or lists?) of such pages so that editors know that the posts exist and can go and respond if required. I don't know what a sensible value of N would be.

To avoid false positives the bot should ignore posts that consist of solely adding things like WikiProject tags, old deletion discussion notices, {{Talk page header}}, {{Talk page of redirect}} and probably others (as well as redirects to these templates) - I guess experience will show others as well, so the list should be easily configurable. Maybe also excluding things like {{help me}} which already generate notifications elsewhere - indeed maybe we would want to restrict it to certain namespaces only (Talk: and File talk: definitely; maybe Wikipedia talk:, Help talk:, Template talk: and Category talk: ?).

Some of what it will find will possibly be spam or junk, but then we can just remove this sooner than we otherwise would. It would expose that these pages have few watchers, but this bot will effectively make them more watched than average negating any benefit to knowing that. Thryduulf (talk) 17:03, 20 November 2021 (UTC)

I think this is a really good idea! CapitalSasha ~ talk 17:06, 20 November 2021 (UTC)

The big issue is file talk posts on enwiki that are about Commons files - usually they can't be actioned here. A differentiated solution may be needed for those discussion page posts that are about Commons files, because they need to be actioned on Commons if anywhere. Perhaps we could ask the Commons folks if they are interested in a dedicated collation of all enwiki file talk page posts that concern Commons files? Jo-Jo Eumerus (talk) 17:16, 20 November 2021 (UTC)

Some will be able to be dealt with here, and some by editors here making edits at Commons, but even though some will require action from Commons admins I don't think that means a list here would be without value. I can certainly see the benefit of putting such comments on a separate list to comments about locally hosted files though. Thryduulf (talk) 17:37, 20 November 2021 (UTC)

I like this idea. Another benefit could be that people post in the appropriate place, not posting where "there's more traffic" (but not necessarily more—or any—interest). That leaves discussions where they are relevant, and potentially, where others in the future will see them. — JohnFromPinckney (talk / edits) 20:43, 21 November 2021 (UTC)

@Thryduulf, I wonder what you think of phab:T295392. That would let you get a list of the discussions on pages you are interested in. Whatamidoing (WMF) (talk) 20:24, 22 November 2021 (UTC)

@Whatamidoing (WMF) while that's interesting and likely very useful, it's not the same as what I have in mind here. That seems to be:

• Tell me when there are new discussions related to this list of things that I am interesting in

Whereas this is

• Tell me when there are new discussions on pages that not very many people are actively watching

The latter will include pages that nobody has expressed an interest in (files and redirects will often only have a single watcher - the creator/uploader, who may not have edited Wikipedia for years), no WikiProjects have indicated belongs to their topic area, are too new for most people to know exist, etc. Thryduulf (talk) 20:58, 22 November 2021 (UTC)

What is the size of the issue? Are there any way to get the size of the problem? (Maybe by first project they belong to, Article importance and whether it is a new user (< x edits say)??

What do people think are the root causes for Editors creating discussions on no-watched/Low importance pages?

{{Thryduulf}} You mentioned interests - The biggest interest groups are projects. Would an addition to the new project dashboard showed the number of outstanding unwatched discussions encourage them being answered?

{{Jo-Jo Eumerus}} Can you explain about why the Comments can't be actioned here? I don't know anything about commons, Wakelamp d[@-@]b (talk) 02:37, 23 November 2021 (UTC)

Forget about "Commons", this is an issue that can arise on any little-watched talk page in any namespace. Second, a talk page question can be asked by any editor who ends up on a particular talk page, ranging from a genuinely curious reader who seeks clarification of something on the page or suggests a change to a WP:POINTy troll leaving an inappropriate remark. Sometimes articles on relatively obscure topics are created by editors who do little else in the encyclopedia and leave before a comment is added there. It's a good general cleanup project to address. BD2412 T 02:46, 23 November 2021 (UTC)

@Thryduulf I think your suggested report is of crucial important for Editor retainment because it makes them feel listened to, not isolated, and their needs are being met. In general, I am against many reports that are not run JIT on Articles. Many reports are unvisited except by search engines, have no or few using them (stubs, categories), are sometimes have completeness issues (Category-Article lists), but in this case a report is needed.

Some other things you may wish to consider to minimize the amount of Talks to review -

• Prioritize new or returning editors
• Exclude Topics that are in action or marked as done (There are done and in progress templates and it could work like the topic subscribe)
• Excluding Topics that need no action (So we still need a done flag)
• Can we get the discussion editor to make a choice? I believe strongly that editors should have the same information. Why not tell them there are no active watcher? And why not and remove inactive editors from article watch lists"
• BOT created Topics - You would need to exclude, but do we need them at all?
• I love the idea of identifying interests, but how??
• Automation - Because of the size, maybe we need a tool with canned responses, a way of viewing a all the discussions and marking them done from Publish?
• Responsibility - Many monthly reports seem to have gigantic backlogs (categories missing, stubs ..) How do we ensure that it is actioned? Wakelamp d[@-@]b (talk) 03:12, 23 November 2021 (UTC)

• Sounds like a good idea to me. Enterprisey (talk!) 03:20, 23 November 2021 (UTC)

  @Wakelamp Responding to your comments in both messages:

  • If a page has WikiProject tags then it makes sense to flag the discussion to those projects (maybe as part of article alerts?) but not every page has them (e.g. random article took me to St. Louis School of Fine Arts where I have just created the talk page with project banners). So "interests" is not really relevant for this idea, my mention of them was distinguishing this project from the one @Whatamidoing (WMF) mentioned (where they are relevant).
  • I don't know how prioritisation would work, but if it can be done it might be useful. We shouldn't exclude established users though, as I know that I have posted comments on low-watched talk pages of articles I have come across but lack the subject knowledge to fix.
  • I'm anticipating this bot waits a while (a few hours? a day?) before flagging comments in this way, and so it would likely make sense to exclude any that have responses (by registered editors?) as that would suggest the issue is being/has been dealt with. Similarly if the comment has been removed, it shouldn't be reported here (regardless of why it was removed).
  • If the list is long (I honestly have no idea how long it will be) then some way of highlighting when someone has looked at a comment to avoid duplicating effort would be good. Maybe just remove an entry from the list when you've dealt with it? Dealing with it could just be flagging the problem somewhere someone who knows what to do should see it. For example I wouldn't have a clue how to verify whether a comment saying that $article misinterprets South African law is correct or not, but I could flag it to the South African and Law Wikiprojects.
  • Posts by bots should not be reported here, but a comment followed by an edit by Signbot should be reported.
  • I don't know the size of the problem, so I can't say anything about what level of automation is worthwhile, nor what form it should take. I think this is something that might be best developed with experience after we get a feel for how it works in practice.
  • Getting things actioned is a realistic concern. I don't know how to ensure that, but if it doesn't happen we aren't actually any worse off than we are now.
  • Letting people know that a page has few followers is a double-edged sword. We don't do it now because it would make the page a magnet for vandalism, and if we aren't reporting things instantly that would still be a potential issue. Maybe just an edit notice for talk namespaces that note not all pages have active watchers? Thryduulf (talk) 13:18, 23 November 2021 (UTC)

  For the last bullet, the "Page information" feature's "Number of page watchers" entry shows "Fewer than 30 watchers" for non-admins. This was a deliberate decision not to show the exact figure, and has been discussed at both Help talk:Watchlist and at WP:VPT. Re the last sentence, we already have Template:Editnotices/Namespace/File talk, Template:Editnotices/Namespace/Category talk and several other similar notices. --Redrose64 🌹 (talk) 15:00, 23 November 2021 (UTC)

Prevent basic errors getting into articles

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.

---

Wikipedia is based on the idea that it doesn't matter how poor your edit is; someone will fix it. In general, obvious vandalism is indeed fixed quite quickly. But edits that are well-intended but poor persist for years. Some of the most glaring issues that I frequently see include:

1. First sentences which are a pointless restatement of the article title
2. Bold text that does not correspond to the article title
3. Use of contractions and ampersands
4. Incorrectly capitalised section headings
5. Links within section headings
6. Links within bold-face reiterations of the article title
7. Misuse of /

There are of course many others. But it would be trivial to prevent or inhibit any of these basic errors from ever getting into articles. A simple edit filter could apply simple quality checks, and warn the user if their edit fails them.

I can think of any number of advantages to basic quality control, and not a single disadvantage. Interested to know what other people think. 51.6.138.90 (talk) 10:19, 22 November 2021 (UTC)

The style guide Wikipedia:Manual_of_Style has discussions on MOS:AMP MOS:&, and Use of contractions and ampersands. They also give many exceptions such as AT&T and quotes containing contractions, Exceptions are what makes programming complicated. The disadvantage is that if you disallow such changes then an editor may not finish their edit, or worse quit Wikipedia Wakelamp d[@-@]b (talk) 12:25, 22 November 2021 (UTC)

I'm not suggesting disallowing anything. Rather, if one makes an edit that contains, for example, isn't, and the text is not within quotation marks. you would simply be warned that contractions should not be used, with a link given to the relevant part of the MOS. If there is some valid reason to use the contraction, you would just click save anyway. I can't remember how but I've certainly encountered edit filters with that behaviour before, where you can save the edit after a warning. 51.6.138.90 (talk) 13:05, 22 November 2021 (UTC)

I think it would result in user experience issues. It certainly makes me mildly annoyed every time I trip an edit filter. Kleinpecan (talk) 13:20, 22 November 2021 (UTC)

I would anticipate the user experience, if the filter were tripped, to be something like the following:

1. user who did not read the manual of style makes an edit including text like "It should be noted"
2. user is prevented from saving the edit immediately, with the reason displayed including a link to the MOS
3. user revises their edit, and is less likely to make a similar mistake in the future

How would you see it playing out? 51.6.138.90 (talk) 13:43, 22 November 2021 (UTC)

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

Article wizard - Check reference reliability and do some notability checks

The aim is to avoid new editors rage quitting after article deletion, by telling them as early as possible that the article will be deleted at AfD, The idea is to make the article stay in wizard and in draft space until some notability requirements are met,

Steps In the Wizard

1. Requires a portal/category, and the notability requirements would be advised
2. A notability quiz is done for GNG
3. Proposed References are added in
4. NPP page curator checks the reference reliability
5. A new development would check the reference for relevance (# of mentions in google books)
6. The editor can then delete the failed references and find a better one
7. Once these issues are fixed, then the article leaves Wizard
8. If the article is still in article wizard in 30 days it is deleted

—Wakelamp d[@-@]b (talk)

• I'm a bit lost on exactly what this idea wants to do? It seems like it is proposing that a new article workflow gets "google" to evaluate the article, and some "fact check" provider to provide some sort of feedback to the editor? Am I missing something? — xaosflux ^Talk 23:43, 30 November 2021 (UTC)

  "The Wizard" is just a guided form that helps people set up a new page before the first revision of it is published (even as a draft or sandbox page) - I don't think it would be a good idea to prevent saving (and therefore collaborative editing) of drafts until they meet new standards, or require that a page go from idea to this point in only one editing session. I do think that giving people additional tools to help determine if a current version of a draft or sandbox is in good order prior to moving it to an article space could be helpful though. — xaosflux ^Talk 14:37, 1 December 2021 (UTC)

  1. I agree it's currently guided, but I was hoping that the wizard could be extended to have a link chooser like complicated.
  2. With collaboration on drafts, AfC has a 2 month wait and an 80 % fail because the article has no hope of notability.
  3. Ir we wait until the article is saved, before we do a rough notability check then there may be a great chance they have wasted their work
  4. I also want more checks on articles to be available for the editor in the article.I don't want the article marked up with tags though, but highlighted with the possible error. But I do want Rater to use
  5. Saving an article causes NPP to review it within 5 minutes, which 20 % of AfD articles survive (as this page by {{User:JPxG}} shows. Wakelamp d[@-@]b (talk) 16:02, 1 December 2021 (UTC)

  • Note, I wasn't talking about publishing an "article" - I was talking about publishing to Draft space, or to a sandbox. — xaosflux ^Talk 16:52, 1 December 2021 (UTC)

New User Experience and emotional states

New User Experience and emotional states (talk) 14:18, 23 November 2021 (UTC) New Editor interaction Exp Editor Opinion Good Article Good Artticle AfD Good Article AfC Vandal Vandal AfD Creates Login NA Confusion Anticipation Starts an article style="background-color:#32cb00;";" Excitement Joyful Uses Wizards Confusion Edits Worry Saves - Draft Article NA Worry NA Saves - Draft Article to AfC Uncertainty Waits for AfC Backlog to fix Indifference Frustration AfC - Disccussion Satisfaction Frustration Pleased Confusion Upset AfC - Fail Uease Anger AfC - Pass Satisfaction Relief Publish Article Backlog Pride Amusement Screen dump of article NA Pride Status Publishes Article accidently and NPP/Other Edits immediately Satisfaction Upset Amusement Tools - tags Satisfaction Frustration NPP - AFD Pride NA Shame Anger Fear Further Edits NA Ok Upset Pride the longer it lasts Amusement Edit Reverts Satifaction Guilt Anger Amusement Amusement AfD -Pass NA Relief Annoyed Amusement Amusement AfD - More work Frustration Frustration Amusement Amusement AfD - Fail Anger Anger Amusement Amusement User page for Deletion Indifference Anger Hatred Amusement Amusement User Page edited (Personal Page) Anger Hatred Amusement Amusement User Talk (personal space) Safety Thankful Friendly Relief Confusion Rejection Frustration Anger Troll Talk Satisfaction Tea house Satisfaction Unease Frustration Thankful Relief Frustration Help Desk Satisfaction Stress Frustratio — Preceding unsigned comment added by Wakelamp (talk • contribs) 14:23, 23 November 2021 (UTC) User:Wakelamp, I especially appreciate the analysis of emotional states, which is often so overlooked. As the rest, can you provide an executive summary of what you're saying here, and thinking as to possible actions? I'm not super smart and am having trouble interpreting the chart. I think you are thinking along lines of changes to some software rather than just rules? Herostratus (talk) 16:31, 24 November 2021 (UTC)

Notability Standards reduction for underrepresented topics

NO Notability Standards reduction for underrepresented topics If nothing else, en-wiki has made it very clear that it does not support reducing notability standards for under-represented topics, even if the WMF would like it. The notability standard holds two purposes: it acts as a filter from everything being included, but it also works to ensure a certain minimum standard of sourcing and thus content and accuracy for that article. Changing it (and also changing our definition of reliable/independent source) would inherently have a negative impact on every article, not just those within the category, as readers would note a decline in quality but not be aware of the split-structure (just as they aren't aware of, say, Geoland's lower hurdle or NCORP's higher hurdle). There is a semi-automated tool for rating, but we deliberately require a manual look, because automating what makes a good wikipedia article isn't the same as a regular piece of text. — Preceding unsigned comment added by Nosebagbear (talk • contribs) 14:59, 23 November 2021 (UTC) @NosebagbearI can see your point, but I find it unfair that New (non-sockpuppets)Editors can no longer create a new article and pass AfD; Even experienced editors disagree about what is notable, so how can we expect them to? And after wasting 4 hours on their first one I can see why they leave. https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Log/2021_November_26. # Notability standards :Can you point me to the last RfC? "en-wiki has made it very clear that it does not support reducing notability standards" BUT WP:CONEXCEPT WMF is responsible for legal issues. Equal opportunity is a legal issue, a charity issue, and a reputation issue. Women who starred in a major film - no. But a minor cartoon character - yes. IMDB reputability is a particular problem. We use their information without acknowledgement. Either with no references. TV shows have all the non-main cast/crew in the order as IMDB (number of episodes), even many of those people do not appear on credits. Movies have them same order. or we have used them to inspire our own searches for often incredibly obscure links (posters, blogs, and very low rep. blogs) Wikipedia:WikiProject_Film/Resources#Questionable_resources IMDb content is mostly user-submitted and often subject to speculation, rumor, hoaxes, and inaccuracies. The use of the IMDb on Wikipedia as a sole reference is usually considered unacceptable and is discouraged. Its romanization of Chinese titles does not follow the standard. Reliable sourcing from established publications cannot be stressed enough. Anonymous or pseudonymous sources from online fansites are generally unacceptable. So, while itself discouraged as a source, IMDB might provide information leading editors to the preferable reliable sites. Wakelamp d[@-@]b (talk) 14:53, 26 November 2021 (UTC) The WMF didn't do the easiest job at archiving, so the best way to see how hostile the Community was to changing the notability rules is to go to any page in the strategy recommendation process on it and take a look at the talk pages. For example this was the first draft of one, and then there are multiple talk pages of the later iterations. It was also made clear on the village pump, and in innumerable zoom calls. As the WG was responsible for assessing commentary on its own work, it still persisted into the final form, despite a unanimous opposal on the final version. Take from that what you will. While the WMF has responsibility for legal functions, equal opportunity a) applies to things like employment, not encyclopedic content, b) would require a level playing field. Were we to say "female biographies need 8 good sources, male need 3", that would be uneven (though still not a legal matter). In futher response to what you say, in a sense, we don't expect new editors to know how to handle the notability guidelines. We discourage writing a new article as the first thing to do, and also heading straight to AfD. However new editors can create new articles (but through AfC) and pass AfD - it's not like we auto-fail any argument backed by a new editor in AfD, they just may not be aware of what those are (as are some experienced editors). Nosebagbear (talk) 17:16, 26 November 2021 (UTC) A question. If you created an Article what do you think your chances would be to get no NPP templates, or fail at AFd? ( and 30 % or the Articles that end up at AfD have no notability issues) 1/I am sure I read about it on an EU website applying to written work...There is also something about allowing equal access to work and not discriminating,,, In Australia, volunteering and work have the same safety rules. Our current process removes far more more women 2/ With discouraging - not so much with the 'Be Bold' everywhere. And after a SIMPLE :-) spelling error on search, you get "Create the page "SdifjoIWEHF" on this wiki!" 3/ AfC is not forced, but they told me that 80 % fail for lack of references 4/ I would love the split for New Article into speedy, AfD (pass, fail, quit, rename)m AfC 5/ And I just found out that ORES does rated articles up to C CLass using Rater Wakelamp d[@-@]b (talk) Wakelamp d[@-@]b (talk) 19:55, 26 November 2021 (UTC) In Australia, volunteering and work have the same safety rules. Our current process removes far more more women. Those "safety rules" would apply to the volunteers themselves, not to the articles they write. Do we have statistics that new editor retention is far worse for women? --Ahecht (TALK PAGE) 20:12, 26 November 2021 (UTC) I will post in women in red. I think Our male/female split has got worse, which means that it must be so Anyway I have seen 4 articles Women and men are far nearer to 50/50 to edit WP, but are far more find it unwelcoming, And are far likely to leave. Educated women are more likely to leave workplaces that are aggressive Women are more likely to respond negatively and personally to critical computer warning messages Women in power imbalance situations are more likely to be passive or avoidant. There only restorative is to exit Wakelamp d[@-@]b (talk) 13:50, 27 November 2021 (UTC)

Redesigning and simplifying the English Wikipedia website

I would like to start a discussion on redesigning the English Wikipedia website. I'm not sure if the WMF would need to get involved with this. Here are some design changes I would like to see and would like to know your thoughts on them. To start, I would like to see the sidebar's contents be moved to the top of the page as menus and be replaced with the table of contents of a particular article. Britannica does it best there with its articles (example). I would also like to see the main page to be redesigned. Since the 2006 design, the design has changed very little and I think the main page is due for a redesign. Britannica also does good here with this factor with its modern look. The current main page is dated and I was hoping we could make an RFC on what we could do to make it more user friendly. I would appreciate any comments about where to start with how we can simplify and redesign the Wikipedia website and the main page. Interstellarity (talk) 23:04, 25 November 2021 (UTC)

@Interstellarity: have you tried the "new vector" interface? — xaosflux ^Talk 00:36, 26 November 2021 (UTC)

@Xaosflux: No, how do I access it? Interstellarity (talk) 01:29, 26 November 2021 (UTC)

@Interstellarity: in Special:Preferences#mw-prefsection-rendering select the skin "Vector", uncheck "Use legacy vector", check "enable responsive mode". You should be able to tell you are in it if you can collapse the left sidebar. — xaosflux ^Talk 02:26, 26 November 2021 (UTC)

Hi Xaosflux,

While I think this is an improved version of the old vector design, I think it could be improved by moving the sidebar's controls into tabs like most websites do. I think for articles, instead of placing the contents in the article, they should go where the sidebar's current controls are. I think if you want to navigate from one section of an article to another, you would have to scroll up to the contents to go where you want to go. If the contents were on the sidebar, it would be a lot of easier to navigate from one part of an article to another. These are my ideas of how to make the user interface of Wikipedia more user friendly. Interstellarity (talk) 16:09, 26 November 2021 (UTC)

@SGrabarczuk (WMF): Do you have any mockups or wireframes related to the upcoming proposals? Whatamidoing (WMF) (talk) 19:51, 6 December 2021 (UTC)

Workspaces

What about if editors could rearrange the window as workspace. If you edit a page the page appears on the left say, the talk on the right, and the history at the bottom. Wakelamp d[@-@]b (talk) 08:56, 26 November 2021 (UTC)

Adding "AfD closer" status

So, AfD sometimes works non-excellently, help from non-admins is needed, non-admins can't really process AfDs well since they can't make "delete" closes because they can't delete articles, so let's address that. We're not ever going to (and shouldn't) give non-admins the bit to delete articles, so forget that, on to Plan B.

More exposition on this (optional)

So, my basic starting point is the admin corps is understaffed (IMO it's worse than people think, because while few admin tasks aren't taken care of, some seem rushed. And, IMO AfD doesn't work as well as it might. So, one way of dealing with that has been to offload tasks onto editors -- flag to allow non-admins to edit templates etc etc etc.

So, we have Wikipedia:Non-admin closure. And non-admins close various things, and WP:NACAFD addresses closes of AfDs specifically. Any editor can close an AfD but really only speedy-keeps, or expired ones, and those have to be closed as "keep".

That's not very useful. Only being able to do "keep" closes is inherently unbalancing, like being a judge who can only acquit. And also, I'm not going to spend an hour or hours looking at a discussion (old ones are often the thorniest), only to determine that "delete" is in order so I have to throw away my work. Who would.

And I mean, I personally am not convinced that admins overall are really any better at closing AfD than any other good editor. Admins are good at many other things, they are carefully vetted because they do important and difficult things and handle sensitive info and situations -- edit wars and ANI complaints and vandals and reviewing deleted articles etc etc etc. and I'm confident they're good at it. But they're not really extensively vetted, usually, on their qualifications for AfD closes specifically. I'm not saying admins are bad at that but I'm not convinced they're any better than other good editors.

Another point is that AfD closer would be a helpful step for people who do want to be admins. You know, like "100 AfD closes with no complaints" would be another data point for someone wanting to undergo RfA, in addition to the useful experience.

This suggestion is that we make a new page, "Wikipedia:Requests for permissions/AfD closer", which looks and works a lot like say Wikipedia:Requests for permissions/Template editor. An editors applies there and, just as for template editors etc., an admin grants or denies the status after consideration. (Presumably the editor would have previously participated thoughtfully in AfD discussions, gaining experience and also WP:AfD stats to aid consideration).

If "AfD closer" status is granted, there's no bit added, the editor just gets put in Category:AfD closers. A new criterion is added to Wikipedia:Criteria for speedy deletion and in Twinkle, something like "A12, deletion requested by a certified AfD closer following an AfD 'delete" close".

If an AfD closer closes an AfD as "keep", the procedure is the same as before. But if she closes it as "delete", she slaps an A12 CSD on it. The next admin clearing out CSD deletes it. But as you see there's an an admin backstopping the process and catching anything that looks wrong.

Criticism and advice requested. Herostratus (talk) 17:27, 27 November 2021 (UTC)

I realize this is contradicted a bit by the existence of non-admin closures but closing contentious AfD discussions has always seemed to me like an area requiring enough wisdom and discretion that there is value in having community input into who is given that responsibility. It seems different than things like page mover or template editor permissions where the main issue is that the user can be trusted to not break things too badly with the tools. CapitalSasha ~ talk 18:38, 27 November 2021 (UTC)

OK, but as I said a person who can only close "keep" is like a judge who can only acquit -- not helpful. Yes and let's say all admins are good AfD closers, there are also many non-admins who would also be good closers, and there aren't enough admins. We need some sort of solution.

And remember, an admin cannot be removed except in rare exceptional cases, so if they're a mediocre AfD closer and can't be talked out of doing it they could still be doing it in 2040. An "AfD closer" can be instantly removed from the category by a single admin (and allowed to re-apply). Herostratus (talk) 00:58, 28 November 2021 (UTC)

• I would oppose. It would be a hassle for admins to regulate and monitor these closers, AfD is incredibly subjective and conflict-ridden. The thing about admins, they have a lot more at stake then someone with a special-purpose bit, admins are more judicious in deciding a close. A special power is a hammer looking for nails, once you have it you will want to exercise it. Such a power is a mandate of one's supposed superior abilities in deciding to delete. It would also form a club of sorts, where fellow bit holders will find it useful to support one another, making it difficult to oppose a group of deletion specialists and their supporters. -- GreenC 18:43, 27 November 2021 (UTC)

OK. I don't agree with some of your points, but I've been wrong before. But do you agree that AfD closes sometimes don't get the involved consideration they require, and if so what should we do instead? Herostratus (talk) 00:58, 28 November 2021 (UTC)

• The thing about admins, they have a lot more at stake then someone with a special-purpose bit - is this actually true? It is very rare for an admin to be de-adminned, and I can't recall it happening over bad AfD closures in recent memory. --Aquillion (talk) 22:31, 30 November 2021 (UTC)

• I'm intrigued. I have no intention of ever becoming an admin, but I could imagine requesting this permission. I find it interesting to keep an eye on the book-related AfDs and I think I have gotten to know the process well. I'd close uncontroversial deletes periodically. Some threshold requirements -- e.g., account at least a year old, !votes in at least X AfDs, whatever seems likely to increase AfD competency -- could go a long way in screening for good candidates. But none of those metrics will do a good job of screening for the interpersonal skill and judgment, not just "experience", which makes for a good AfD close. Having admins do the actual delete provides some "safety net" but a bad AfD closer could sow a lot of discord. I think it's worth considering what processes might give this permission to the most suitable candidates. ~ L 🌸 (talk) 06:05, 30 November 2021 (UTC)

Yes of course, votes in X number of AfDs would be necessary. I don't know if X should be a hard number. Ten is not enough, 100 is, in between it depends; obviously nobody is going to consider someone with ten, we hardly need to say that. And OK, Interpersonal skill is fine, but I don't have it and I've closed hundreds of AfDs with no complaint. Patience, analytic skill, very good knowledge of the relevant rules and usual practices, and diligence are more important IMO.

Since this the idea lab, I'll keep idea'ing. I like simple, cos that give best chance for adoption, but you could add all kinds of stuff:

1. Rrequire the signoff of X admins... 3 or 5 or whatever, instead of just 1.
2. Or, instead of just an admin(s) given the status (as in Wikipedia:Requests for permissions/Template editor etc, You could have a full-blown "Wikipedia:Requests for closership" -- like WP:RFA but very much simpler and shorter (see hatted material).
3. Or, other. Let's hear it.

Some important points which I'd recommend reading, but skip if you wish

The editor is not asking for the ability to block persons, see deleted material, edit protected pages, adjudicate WP:ANI cases, and so on. Just for the one ability (which isn't even a bit, and is backstopped). All questions could revolve just around that. Extensive knowledge of rules outside those useful at AfD, being good at vandalism patrol, at handling edit wars, etc etc aren't in play. (Demonstration of good character and intelligence would be of course.)

And, to that end, we have Wikipedia:AfD stats. here's mine. Easy to read charts showing the editor's level of engagement. How often the editor did or didn't vote in accordance with the eventual outcome. How "deletiony" or "keepy" the person is, compared to the general population. Links right there to check her contributions, a random sample or all. You can also see just her AfD nominations, read them, and see what percentage were agreed with. Your vote practically writes itself. (Of course this info is also available to admins if we go the "Requests for permission/xxx" route.)

Also, good grief, I just saw Wikipedia:Successful requests for adminship. We have had seven new admins this year. This isn't sustainable and sooner rather than later we are likely to end up with some AfDs being closed too hastily. If that hasn't happened already. And remember, a big effect isn't just to help AfD but to free admins to do other tasks better.

Come, on, people, we have got a non-excellent situation here. It is not going to get better. We need a fix. IMO we don't need to hear "Well, but then there's this downside" (EV-ER-Y thing has downsides) or "Well, but maybe this bad thing might happen" (In which case, we just stop doing it or fix it) or "Well but maybe a troll or deep-cover vandal might get in" (trolls and vandals get in everywhere, a few; we unddo their work and deal with them) or "But what if the AfD closer turns out suck at it after all" (a single admin can kick him out of the category (and he can re-apply), and so on. Don't take counsel of your fears, don't fear that our experienced editors who want to help out here are dolts. If they are, let's find out. Individual dolts are removed -- no trial (unless people want that). If the new AfD closers turn out to be dolts generally, the admin corps will stop appointing (or the community electing) new AfD closers, and it just dies and is eventually deleted. There's only one way to find out. We need to keep adjusting to circumstances -- 7 new admins this year, under 20 in a typical year, and that's not accounting for retirements -- or face a decline.

Seven new admins this year. We have to take some burden off the admin corps. If you've got a better idea, let's hear it. Don't just be a naysayer, please. Herostratus (talk) 19:04, 30 November 2021 (UTC)

Herostratus, I've always thought of you as a good editor, but I guess I'd like a persuasive statement from you as to exactly why you think this new permission is needed. Sometimes, AfDs are not closed after exactly seven days, but usually there's a discernible reason (low participation, the possibility of further comments' leading to a clearer result, particularly problematic discussions that would probably be beyond the competence of your proposed AfD closers, etc.), but I've not noticed that there's often a severe backlog of AfDs that need closing. What prompted you to propose this? Deor (talk) 21:53, 30 November 2021 (UTC)

Hi. Well, I explained in detail above, such detail that that I hatted a lot of it to avoid TL;DR syndrome. It's been in the back of my mind for some months. So anyway, I'm not so much concerned about lack of backlog, as quality. Don't get me wrong; quality is good! Most closes are fairly easy anyway. Most are deletes, and properly so, or they wouldn't be at AfD. Occasionally we lose an OK article. I personally hate to see that, but it's not a crisis. Occasionally I see (IMO) a close, keep or delete, that seems rushed. It's not a crisis. I guess... maybe 19 out 20 closes are OK. That's good. It's not humanly possible to get that last 5%, probably. I don't know... A part of what I'm trying to do here is think of ways to allow the admin corps to have more time and energy to focus on the many things that only they can do.

As to "beyond the competence", mnmh. I don't see a clear admin/editor line such that the editor corps is mostly blockheads, and I mean after all you're talking about people who have participated significantly in AfD, competently, have not shown poor character, and want to be closers, and meet any other requirements the people granting the status want to personally impose, such as length of service or whatever. But that's my opinion. Herostratus (talk) 00:46, 1 December 2021 (UTC)

• TBH while there are areas that lack for admins, I'm not sure AfD tops the list. The really big problem is WP:AE, which is largely handled by just a few admins - until recently El_C was doing a huge chunk of it, and right now it's just a handful of people. Fortunately we've generally had evenhanded admins handling it, but no matter how good they are it's still not ideal to have so few people making those decisions, not in the least because of the load it puts on them. And AE isn't really something we can farm out to non-admins. --Aquillion (talk) 22:31, 30 November 2021 (UTC)

Well... I mean if El_C needs help, what about the admins who are closing AfDs right now? I don't know how fungible admin time is, but maybe some. And I mean you cannot have non-admins doing WP:AE, so where can we find a way to support the admin corps?

Let's see... at AfD, 24 November, 95 articles. 25 November, 68. 26 November, 87. That's a lot. Lot of those are quick slam dunks, a fair amount aren't. Some of those, you need to check that BEFORE was done well enough, weigh other matters. It can take time. How many man hours to close 95 articles? Say 15 minutes average, that's 24 person-hours right there. 15-20 admin-hours every single day that could be spent on WP:AE or elsewhere. I don't know if we average 15 minutes per close, but if we don't that may just mean we can't, not that we shouldn't. Herostratus (talk) 00:46, 1 December 2021 (UTC)

I don't know that they would. Closing AFDs (even the controversial ones!) is usually a lot easier than handling AE. And another concern I have is that if we start slicing off the "easier" parts of adminship and handing them out as individual roles, we're going to both see fewer people applying as admins, and even higher standards at RFA (because people will invariably say "oh, if all you're going to do with the tools is close AFDs, why not just become a closer?" or "I'd like to see you get closer permission and close a bunch of AFDs first, since that's a big part of what admins do.") --Aquillion (talk) 19:20, 1 December 2021 (UTC)

• Meh, if we want to empower some group to delete pages, we can give them access to delete pages. It doesn't require the ability to also undelete or viewdelete - if they screw up, they can just ask for help at AN. — xaosflux ^Talk 23:39, 30 November 2021 (UTC)

  • That's a reasonable and logical point, but unfortunately it hasn't been the consensus whenever unbundling of the delete function has been brought up. --Ahecht (TALK
    PAGE) 21:09, 2 December 2021 (UTC)

• I'm in the "AFD is basically never backlogged, if they're so great at closing discussions they can run an RFA" crowd. As a more productive suggestion -- it shouldn't be a category easily added to user pages, it should be a full-protected project-space page like WP:AFC uses. User:力 (powera, π, ν) 03:15, 1 December 2021 (UTC)
• I don't get where this is coming from. AfD is not significantly backlogged, I haven't heard any of the other admins who close them asking for help, and I haven't seen any signs of declining quality of closes (e.g. an uptick in DRVs). If you want to be an "AfD closer", you can ask at WP:RFA. And I guarantee you will be 'vetted' on your knowledge of the deletion policy and conduct in deletion discussions. – Joe (talk) 08:29, 1 December 2021 (UTC)
• The main issue with AfD is lack of participants, rather than lack of closers. Large backlogs of discussions needing closure aren't exactly common, so adding more AfD closers won't necessarily improve things. Hut 8.5 08:48, 1 December 2021 (UTC)

That's an excellent point. After all, other considerations aside, lack or participants offering data and cogent arguments leaves more burden on the closer to do research rather than just adjudicating what's presented. I'll open a section below to address this. Herostratus (talk) 21:04, 4 December 2021 (UTC)

• What problem is this solving? We haven't had a spate of inaccurate or ill-founded non-admin closures lately. Typically those few NACs that are challenged end up being upheld. --WaltCip-(talk) 13:04, 1 December 2021 (UTC)
• As someone who works AfD when I have time and when there is a need (i.e. a large number of open AfDs and/or AfDs which are eligible to be closed but have not been for more than a day or so) I have not recently seen a need beyond a few short periods. On the whole, and in comparison to other processes, such as AE, I think AfD is appropriately staffed by admin and already overstaffed by non-admin. Best, Barkeep49 (talk) 18:05, 2 December 2021 (UTC)
• I rather like the idea but have trouble with the logic presenting it. It seems to me the closure part of deletion discussions, by admins, works remarkably well and doesn't really need tinkering from a quality point of view. That being said, unbundling the admin toolkit is a Good Thing, and I could see a plausible argument that weighing consensus needs a different skillset than using the block button well. Finally, I could see that it would be more efficient to allow wholly noncontroversial Delete NAC just as we now allow Keep ones; which could presumably be implemented by a nonadmin Delete closure slapping a (new) CSD tag on that an admin would act on, implicitly endorsing the NAC close but having to do less work. However, controversial closes would continue to be left to Admins. So, the idea maybe has legs, just I feel it's currently being approached from the least-strong angle. Martinp (talk) 20:19, 2 December 2021 (UTC)
• I'm of the mindset this could be accomplished much more effectively by adjusting the WP:NACD policy and the WP:CSD#G6 to include a non controversial/contentious AFD NAC close. At this point I don't think we are in need of any different approach here for AFD to be more effectively handled. This is defiantly a tool looking for a purpose at this point. McMatter ^(talk)/_(contrib) 23:19, 2 December 2021 (UTC)

• OK OK I get that editors think AfD closing is basically OK as it is and no change is needed, and fine. It's the idea lab here and no way to find out these things without running it up the flagpole and see who salutes, and thank you all for your time.

(I personally continue to question whether the closing process is entirely excellent, as for instance right now I'm kind of rolling my eyes over a four-word close on a contended (not slam dunk) AfD, and I can't figure why we're (even if only occasionally) seeing that. It's not time constraint, it's not ability constraint, so....?? Color me puzzled, but that's just me I guess.)

Participants

It's a point well taken that participation is a sticking point. It's often fine but sometimes sparser that you'd like to ssee. I don't know how to address that. But since it's the idea lab, I'll throw these against the wall.

1) "Did you know" has a mechanism to ensure that each entry gets someone to look at it, similarly we could have "If you want to nominate an article, you have to meaningfully participate on one (or X number of) AfD first". There's an obvious downside tho: articles that should be nominated and normally would be, wouldn't be. Maybe there's a way to adjust for that, dunno.

2) You could have a robo-invite of random editors to random AfD as there is now for RfC. This is opt-in so not that likely to bring in many new participants, and there's the chance of getting lower-quality input (altho input from editors who are not familiar with subject is actually helpful often). And I personally do got to RfC where I wouldn't without the invite.

3) Other? Herostratus (talk) 21:04, 4 December 2021 (UTC)

URFA addition to article milestones

Hi! I was recently made aware of WP:URFA by SandyGeorgia. I was not aware of this before and I noticed there are many articles there that have notes such as "satisfactory", etc. to indicate they still satisfy FA criteria. However, this is not shown in article talk pages in the milestones thing but could be useful information (e.g. seeing an article passed FAN in 2004 but its FA status was last discussed/checked positively in 2019), both to make the great work volunteer editors do in maintaining the curatorial standard in Wikipedia be more transparent to newcomers and lurkers, and also to encourage more participation in that process. Articles can change immensely in the years (sometimes decade+) since they were listed as FA, but that does not necessarily mean their quality has dropped in relation to the encyclopedia's other articles. I'm not saying there is a dire need for more reviews (nor saying there is not), but just a small idea. I thought of posting this in URFA's talk page but was also wondering about the perspective of less involved editors on this. My main concerns have to do with increasing the visibility of these reviews while not making them too bureaucratic to perform (e.g. many links, templates, etc. to perform), as well as the fact I myself am not experienced with this process so might be unaware of previous consensus in regards to something like this.Santacruz ⁂ Please ping me! 22:33, 27 November 2021 (UTC)

Hi, AC! So glad you have taken an interest. Actually, the current link is WP:URFA/2020 (URFA is the original, but old, effort).

Yes, we need more reviewers; WP:FAR was moribund for years, and now we have quite a backlog to get through. And we sure do need a way to draw more attention to the effort to review older FAs.

As far as Article milestones go (I was involved in helping develop that), we really shouldn't include the informal "satisfactory" marks from URFA as a milestone, because those are only intended to indicate FAs that are good enough to not need to be submitted to FAR, whereas those that are actually submitted to FAR, are recorded as an entry at article milestones. With so many needing review, we had to have a process that allowed us to focus on those that are truly outdated or neglected, because we can't actually review that many thousands! @Hog Farm, Buidhe, and Z1720: (the other main participants at URFA/2020) and @WP:FAR coordinators: . Open to ideas, though Best, SandyGeorgia (Talk) 22:42, 27 November 2021 (UTC)

I missed some others who have worked at URFA/2020: @HJ Mitchell, David Fuchs, and Femkemilene: (probably missed more, but didn't mean to :) SandyGeorgia (Talk) 22:47, 27 November 2021 (UTC)

Not opposed to this idea. A concern I've had with URFA is that, at our current pace, it will take years to check every article. An article marked satisfactory today may deteriorate in five years, but will not be rechecked at URFA/2020 because it is already deemed "satisfactory". Having this added to article history might allow us to keep track of the satisfactory FAs and know to re-check them after a certain number of years. Something like the above idea might not be necessary if URFA/2020 were to complete at a quicker pace, which is why I hope more editors will get involved in this process. I welcome more brainstorming and ideas on this topic. Z1720 (talk) 22:51, 27 November 2021 (UTC)

(ec, essentially duplicating Z1720) I like the idea of including it in the article milestones, or noting somewhere that the article has been looked at and deemed satisfactory. A record might come in handy in another few years when we come to review our oldest FAs again. HJ Mitchell | Penny for your thoughts? 22:54, 27 November 2021 (UTC)

Z and HJ, possibly one way to add those marked Satisfactory by at least three reviewers to Template:Article history would be to name a new process to be added to the template (something akin to FAR-light), but then someone would have to program that in to the template, and we'd have to either add those deemed satisfactory manually, or get a bot to do it ... and we'd need a page to record them (like a FAC or FAR page, unless we just used a diff to the URFA page) ... so quite some work is involved in adding a whole new thing to article milestones, just food for thought, because some of those template editors are always coming up with good stuff. I'm not sure we're ready for it yet ... we need more FA-experienced reviewers first! SandyGeorgia (Talk) 22:56, 27 November 2021 (UTC)

I would rather add URFA to the template in our early process. This will advertise the project on talk pages, causing editors to discover or become reminded about the process. Also, adding it early will mean there are less articles to update when this is first implemented. And I fully agree with Sandy that URFA/2020 needs more reviewers. Z1720 (talk) 23:21, 27 November 2021 (UTC)

Maybe someone with the technical expertise will come along and suggest how we might do that via only using diffs, rather than having to institute new archive pages ... or something ... creative ... SandyGeorgia (Talk) 23:29, 27 November 2021 (UTC)

HJ Mitchell we do have a record (see here) but as AC has pointed out, it’s not very visible, in fact, is quite transparent to most people, who would never know where to look for it. But when/if we have to do this again, we do know where to get the data. SandyGeorgia (Talk) 23:44, 27 November 2021 (UTC)

The milestones idea was the first thing to come to mind, and I do see your point above not including informal marks as a milestone. I don't have enough experience here I think to have any great ideas to propose, but perhaps I can be of help as a representative for the new editor/lurker perspective :P. Backlogs are truly the wildest horses to tame. Santacruz ⁂ Please ping me! 22:53, 27 November 2021 (UTC)

Yep, but not my first rodeo :) The original WP:URFA looked insurmountable when we first tackled it, but that we did. Keep the ideas coming; I see that both Z and HJ like your idea, so maybe someone can put all the pieces in place to make it work, as the list is becoming unwieldy. SandyGeorgia (Talk) 22:59, 27 November 2021 (UTC)

I don't necessarily have too many ideas here for the actual article history thing, but I wonder what y'all's stance on finding some way to implement quid-pro-quo into the GA/FA process would be. I have found being made to do a QPQ when nominating for DYK a great way to make sure too many of those aren't ignored, so maybe some way of "include qpq link of urfa review when nominating for FA" might go some way to slow the backlog creep Z1720 accurately identified above. Santacruz ⁂ Please ping me! 23:04, 27 November 2021 (UTC)

That has been talked about quite a few times at WT:FAC, and it opens the door to some problems so that it has been repeatedly rejected. It would be better to ask that at WT:FAC, so we can keep your ideas going here on how to advance the WP:URFA/2020 effort. SandyGeorgia (Talk) 23:07, 27 November 2021 (UTC)

I'm a little bit hesitant to put an informal assessment such as URFA/2020 into article history, but I think this is a good idea in background and there's going to be a way to do this right. I'm pretty sure the GA sweeps from years ago was in article history, but I'm not sure if that's comparable, because GA is a one-reviewer process, which makes it quite different from FA. One way to solve the linking problem would be to make it a standard practice to create a URFA section on the talk page and note that you're marking it as satisfactory there, which could then be linked to in the article history section. The one issue would be if old items are auto-archived, but I reckon this could be worked around by fixing the thread so that it does not get archived. Hog Farm _Talk 00:55, 28 November 2021 (UTC)

As long as the URFA log line is clear enough I don’t think there’s much difference between URFA and peer review, which does get an article history notice. People can either ignore it or hit the link and get an explanation of the process like everything else in article history. Der Wohltemperierte Fuchs ^talk 14:20, 28 November 2021 (UTC)

That is now three in favor (and some against). If we go this direction, I suspect we need to go to subpages for the process, which I fear will deter involvement, and we need involvement! On the other hand, perhaps it will encourage involvement from those who want the line in article milestones … SandyGeorgia (Talk) 15:06, 28 November 2021 (UTC)

Could you explain a bit more the issue with subpages? I agree they're not an ideal solution but want to understand your personal perspective here a bit better, SandyGeorgia. Santacruz ⁂ Please ping me! 15:10, 28 November 2021 (UTC)

Most events that are recorded in article milestones have an event link entry to a subpage where the evaluation occurred (DYK, Peer review, GA, FA process, etc). There are a few that don’t (ITN, OTD, main page date as TFA— all involving only a date). Those are typically recording only a date, rather than an assessment. (Study Template:Article history). I’m not sure a one-line-link assessment like ITN or OTD or maindate will work for this case, as we expect at least three reviewers to assess an article as satisfactory before marking it at WP:URFA/2020, so that seems to be calling for a full assessment page. A whole ‘nother process, which would be akin to WP:FAR lite. As an example, look at Talk:Malcolm X in edit mode, where you will see the difference between events and dates, and then look at it not in editing mode. My concern is that, to go this way, we would have to essentially repeat a FAR-like process so that we could record an event page. Easier said than done, and will there be unintended consequences? SandyGeorgia (Talk) 15:19, 28 November 2021 (UTC)

So, to summarize my points above, the short question for those who like this idea is, would it suffice to add only a one-line link (like this with a date) to articlehistory, or would we be forced to a full subpage event? @Z1720, HJ Mitchell, and David Fuchs: If we are forced to a full subpage new event process, I would be seriously opposed, as we would essentially be creating a FAR-lite new process. If we can add a one-line event, the template coding to that might be simple, but not my skillset, could be wrong. Also ping Hog Farm to revisit this idea based on this new feedback. SandyGeorgia (Talk) 15:29, 28 November 2021 (UTC)

If we could add a one-link item to articlelhistory, it would essentially be the same way OTD is done— where we add a date and a diff, the diff in this case being the third editor to mark satisfactory and move the article to the Kept or FAR not needed section at WP:URFA/2020. SandyGeorgia (Talk) 15:31, 28 November 2021 (UTC)

DYKs also have subpages, albeit for their nominations, and are edited by at least 3 people (nom, reviewer, and promoter). I haven't found the subpage issue to be as impactful there. One possibility that could maybe work here is when an article promotes to FA a bot automatically creates a subpage (either of the article, URFA or elsewhere). If there is a minimum time for when it will be reviewed by URFA, that could be added to article milestones with that link (e.g. 18 February, 2023 - Next scheduled URFA (the accurate wording here is hard for me to find)). This could greatly increase the visibility of URFA throughout the project, as editors wouldn't learn about it only through articles who have gone through urfa (which seems like not enough). The bot-generated subpage could also include pre-made text/templates to make editing there less cumbersome, which could also help make it easier for new editors to be involved. As a new editor, noticeboards and pages like URFA that have many sections, tables, and stats can be quite scary, so maybe subpages could help from that point of view. DYK subpages felt like a very accessible way to get involved in that. Santacruz ⁂ Please ping me! 15:38, 28 November 2021 (UTC)

I'm not necessarily for or against the idea but I think the way to do it without involving a subpage is to use the specific revision current at the time the third person "signs off" on the URFA listing. It shouldn't need a devoted subpage but a link to the specific revision that was ruled satisfactory is useful, especially as it means if the article does deteriorate in future there's a "last good version" that can be referred to. ᵹʀᴀᴘᴘʟᴇ ꭗ 15:42, 28 November 2021 (UTC)

Yes, that agrees with what I see as the easiest way to do it, should we decide to do it. I will later ping some others in, but want to first see if this idea truly has support, and then would also want an RFC (since I don’t like adding things to milestones without broad consensus). SandyGeorgia (Talk) 15:45, 28 November 2021 (UTC)

The FA process already has automatically generated subpages (for FAC, FAR), and those are maintained in articlehistory by a bot, as are DYK entries. (TFA is a manual addition by TFA Coords). I’m not entirely following what you are suggesting. If we hear back from HJ, Fuchs and Z, I will try to ping in others familiar with how to make this happen … later today. SandyGeorgia (Talk) 15:43, 28 November 2021 (UTC)

Agree with Grapple X. I don't see a need for individual subpages. That would add a whole new layer of bureaucracy to the process. A one-line link is fine as a record that the article was looked at and deemed satisfactory. As was mentioned above, the GA sweeps were recorded in article history and some of those reviews were only very slightly more verbose than "satisfactory". HJ Mitchell | Penny for your thoughts? 15:44, 28 November 2021 (UTC)

Great, so far we agree on this part then. SandyGeorgia (Talk) 15:45, 28 November 2021 (UTC)

I would not like to use a new subpage. I think linking the revision version that was deemed "Satisfactory" by the third editor will suffice, as noted by Grapple above. Z1720 (talk) 16:07, 28 November 2021 (UTC)

I also don't think a new subpage would do anything good here, and would just bog down the process. I know article history can handle an item with no link - that was done for some of the old GA sweeps ones, and I've seen it for GA reviews from before they created dedicated GA subpages. Would be certainly doable to just have a link in there. Hog Farm _Talk 16:20, 28 November 2021 (UTC)

If we were to ask

... template and bot editors how to make this happen, I suggest it would look something like this. (I don't want to actually ask them until/unless we are in agreement, but it does appear that all agree that a new process/subpage should not be created, and we should just link to the third "Satisfactory" diff.)

Sample diff: This is what an edit that marks the third "Satisfactory" at WP:URFA/2020 looks like, as it also moves that entry from the list of articles that need review to the list that has been marked as "FAR not needed". (Those that are "Kept at FAR" are already automatically processed by FACbot.) That would be the diff upon which the Template:Article history entry would be based.

1. Who is programming/maintaining Template:Article history these days? (Gimmetrow, who designed and used to run a bot to add every content review process to the milestones is basically gone.) That is, who can code this new entry in to the template?
2. If the editor who made the final "Satisfactory" move, then placed that diff (sample) at a subpage (WP:URFA/2020/Bot archive or something similar), could that be the trigger to call a bot to process the item in to article history, just as the move to FAC and FAR archives now triggers FACbot to do the rest of the bookkeeping?
3. Can a bot pull the article name from a diff like that ?
4. Just as FACbot then pulls a diff to the corresponding version of the article at the time it was marked satisfactory, could that be done here as well ?
5. Can FACbot do this ? Or is this type of task more suited to the bot that processes DYK, or the bot that processes ITN?
6. Then, could FACbot (or another) process that archived diff in to article history with an event that looks like this (where x represents the next open event number and the sample diff listed is used)?
7. And then, could that result in a line in articlemilestones that reads: Date (linking to article oldid) ... Unreviewed featured articles (linking to sample diff) ... Marked satisfactory ?

| actionx       = URFA
| actionxdate   = date from diff as in sample above
| actionxlink   = diff as in sample above
| actionxresult = Satisfactory (this would not technically be needed, as they would all be "satisfactory" by definition of how we are doing this, but leaving it off may confuse article editors)
| actionxoldid  = oldid that the bot pulls from the article that corresponds with the date of the sample diff

| currentstatus = (this would not change ... the article remains an FA)
}}

If I have this right, I will next ping in those who can opine. I would also recommend a formal RFC before we do this, if the bot and template people sign on. SandyGeorgia (Talk) 16:59, 28 November 2021 (UTC)

This makes sense (to me anyway). However, to play devil's advocate—anyone monitoring URFA is likely already an experienced editor familiar with article history templates; could we theoretically begin doing this manually at first to see how it pans out before engaging a bot to automate it once we see how a few instances of it will pan out? There are still some article history entries that are entirely handled manually (FTC/FTRC for example) so this wouldn't be unheard of and I for one don't mind doing the bookkeeping of updating these things if someone wants to pawn it off, but it might identify the best method to automate if we do it in practice for a bit first. ᵹʀᴀᴘᴘʟᴇ ꭗ 17:00, 28 November 2021 (UTC)

ugh ... I do TONS of manual maintenance of article milestones, and It's a Bitch. The worst part of doing it manually is that DrPda's (or Gimme's, can't remember) script to return an oldid based on a date is no longer working, so finding the oldid is a bunch of work, that is (somehow) already programmed in to FACbot. One of the main things you find is the number of times unfamiliar editors mess them up! Also, most of the code for this should be already present in the (at least three) bots that are already doing this.

But I do agree with you in principle ... that we would a) first ask it to be coded in to the template, b) second add all (or most) of those already done manually to see how it goes (I could do that), and only after that, then c) ask for a bot to continue. Great thoughts ! SandyGeorgia (Talk) 17:08, 28 November 2021 (UTC)

If this were something I anticipated a lot more newer or intermediately-experienced editors doing then sure, the ability to mess up or miss a step would certainly be magnified but I think it's a safe bet we're likely to only see editors much more familiar with intricate operations working on this. If we want to identify some initial articles to test it out on after any RFC then I'd be happy to take a list of a few dozen and grunt through it just to see what works. ᵹʀᴀᴘᴘʟᴇ ꭗ 17:15, 28 November 2021 (UTC)

OK, when/if we get to that stage, you and I can divide the list to be initially processed. Thanks! SandyGeorgia (Talk) 17:20, 28 November 2021 (UTC)

PS, we already have a perfect division of labor at WP:URFA/2020, as we have two different sets of "FAR not needed" by date. You could take one, and I could take the other ... when/if the time comes, we're on! SandyGeorgia (Talk) 17:38, 28 November 2021 (UTC)

Hawkeye's MilHist Bot and FAC Bot already add to the Article History template, so if he has time perhaps a bot task would be something Hawkeye7 could help with. HJ Mitchell | Penny for your thoughts? 17:42, 28 November 2021 (UTC)

I was hoping to wait until I was sure we were all on the same page, and only then ping in all those technical people at once. So, since you've already pinged Hawkeye, I will go ahead and ping the rest of them, via a post at the talk page of the articlehistory template. SandyGeorgia (Talk) 18:07, 28 November 2021 (UTC)

Queried Template talk:Article history (and I also placed crossposts at WT:FAC and WT:FAR). SandyGeorgia (Talk) 18:20, 28 November 2021 (UTC)

Getting more freely licensed media by allowing fair use images of living persons

So hear me out: many living people have no image on their Wikipedia article, and the subjects in question generally just don't care. They only start caring once they do have a picture and they don't like it.

I've seen this on multiple occasions as I've uploaded many crops from group photos (where flaws that weren't too noticeable in the full picture become more apparent) and screenshots from freely licensed videos (some people literally never close their mouth while keeping their eyes open when they're in frame) and added them to articles. Nobody cared for years, but after I've added a less than perfect picture (because that's the best we got), the subject wakes up.

An example of this that I remember is Jaap Smit. His article was created in 2011 on nlwiki. On 27 June 2018 I added a photo, File:Jaap Smit CNV 2012.jpg. That's a crop and his glasses magnifying his left eye combined with the angle make the photo not the most flattering there could be. (but the only freely licensed media we had) On 20 August 2018 the photo was replaced by Rob van de Webredactie provincie Zuid-Holland (Rob from the web editorial staff South Holland) who was instructed by Smit himself (talk page: "Het is de wens van de heer Smit om zijn foto te (laten) wijzigen.", translation: "It is the wish of Mr. Smit to have his photo changed") This eventually resulted in a CC BY 4.0 license being added to https://www.zuid-holland.nl/overons/bestuur-zh/gedeputeerde-staten/cdk-jaap-smit/beelden-publicatie/ and we now use File:Jaap Smit 2 (cropped).jpg.

So if we allowed fair use media for living persons but somehow made sure the media isn't overly flattering (the 0.1 megapixel restriction for fair use in general clearly isn't enough for that), we just might get more free media. This is pretty much a brain fart, but it may be worth brainstorming on how to exploit people's vanity. — Alexis Jazz (talk or ping me) 10:35, 28 November 2021 (UTC)

@Buidhe, Nikkimaria, and Elcobbola: I doubt this is doable, but they know better than I do. SandyGeorgia (Talk) 15:08, 28 November 2021 (UTC)

Even without fair use we could perhaps do more. For example, we could have the infobox automatically insert a generic silhouette when there's no image that links to instructions on how to license a photo properly. Probably not as effective, but indicates more clearly that something is missing. — Alexis Jazz (talk or ping me) 15:21, 28 November 2021 (UTC)

We used to do that, but the practice was deprecated after a community discussion. See WP:UPPI and Wikipedia:Centralized discussion/Image placeholders for the background. Nikkimaria (talk) 15:27, 28 November 2021 (UTC)

We could re-introduce them so we don't have to rely on the French Wikipedia (or other Wikipedias that are more pro free content than we are) for asking for free images. The discussion linked by Nikkimaria isn't exactly clear, and there might well be consensus to bring back begging-for-image placeholders for some classes of people (politicians, actors, sports personalities) where this might result in a free image being donated. Of course not every biography needs a picture, just like not every article needs an infobox. Do we have any data on how effective the placeholders were in soliciting free images? —Kusma (talk) 10:53, 2 December 2021 (UTC)

I admit I am a little uneasy at manipulating people in this manner. It feels a little bit like extortion - give us a free image or this unflattering one will remain up instead. Jo-Jo Eumerus (talk) 15:33, 28 November 2021 (UTC)

I don't know that it will necessary work like that—we essentially have that in place at present, free images are often taken from poor angles, at seated events or behind microphones, that sort of thing, so donating an image to improve the aesthetics of an article is already a possibility. That said, I'm leery of this for the simple fact that it will undoubtedly lead to tens of thousands of non-free images being plucked from the internet and added to articles which don't need them, and I don't believe that's something to encourage, especially as the knock-on effect will be "well, we have an image, no need to try to obtain a free one now", and the search for/creation of freely-available images will be curtailed. ᵹʀᴀᴘᴘʟᴇ ꭗ 15:47, 28 November 2021 (UTC)

It certainly does work that way, especially when they find that news sources often use our images. I get a lot of complaints about unflattering images, and know of many cases where the athletes were sufficiently aggrieved to upload a better image. Most would prefer that we use a flattering PR image. Hawkeye7 (discuss) 19:26, 29 November 2021 (UTC)

But that's the point I was making--it already works that way, it wouldn't be a byproduct of a change to fair use, it just is how it is now. ᵹʀᴀᴘᴘʟᴇ ꭗ 19:41, 29 November 2021 (UTC)

I just read about Personality_rights. It looks like celebrities have more rights depending on the country Wakelamp d[@-@]b (talk) 07:06, 2 December 2021 (UTC)

I question the idea that adding pictures is always of encyclopedic value. Let's take a subject like Hans-Lukas Kieser, who currently has no photo. Kieser is a historian known for his writings, not what he looks like. Even leaving aside the copyright issues, what benefit to the reader is advanced by letting them know he looks like this? Sure, if there was a free image I would have no objection to adding it but I'm skeptical that it does much to raise the value of the Wikipedia article. (t · c) buidhe 21:34, 28 November 2021 (UTC)

Humans use facial recognition as a primary tool for recognizing other humans. So anyone hoping to recognize the person (perhaps to meet them a conference, or to identify them when seeing them in a documentary or on television if there is no graphic indicating their name, ...) would benefit a photo of their face. CapitalSasha ~ talk 22:39, 28 November 2021 (UTC)

That may well be the case but Wikipedia is not LinkedIn, we're here to educate and not to network—if there is an informative need for a nonfree image that's one thing (perhaps a snippet of music that is being discussed, or a piece of visual art that cannot be conveyed in words alone) but in the case of a living individual we might actually be better served using free files that illustrate their work or the context around their lives than by a simple non-free portrait. In Buidhe's example, for instance, Kieser is an historian of the late Ottoman Empire and his bibliography lists several works on genocide; perhaps a contextual illustration like File:Assyrian_population_1914.svg conveys more about the subject than claiming "fair use" on a picture of him alone could. I'm also a fan of illustrating articles with quotes where possible as well; has this writer said something succinct which conveys who they are or what their work means which could fill a quote box instead of an image? ᵹʀᴀᴘᴘʟᴇ ꭗ 22:54, 28 November 2021 (UTC)

I am inclined to wrap this thread in {{archive top}}/{{archive bottom}}, in order that no more time should be wasted. Please see my closing summary at Wikipedia:Village pump (proposals)/Archive 170#Allow fair use non-freely licensed photos of politicians. If we can't allow it for politicians, we can't allow it for anybody else who is still alive. --Redrose64 🌹 (talk) 23:12, 28 November 2021 (UTC)

Redrose64, this is VP idea lab, not VP proposals. Your proposal has the goal of just allowing more fair use because you've given up on a high quality free alternative. This thread is about the idea of exploiting people's vanity to get more freely licensed content. I grant you that actually doing this with fair use content would require a change to foundation:Resolution:Licensing policy, but if the community would feel that it's worth it such a change wouldn't be impossible. It's unlikely to happen in the form I described (never say never, but, unlikely), but perhaps there's another way to achieve something similar. — Alexis Jazz (talk or ping me) 14:20, 29 November 2021 (UTC)

As a BLPN regular I'd very strongly oppose this. I know the headers tells me to be creative and positive but sorry I can't be. This proposal needs to be killed with fire. I was thinking the title made no sense, how would relaxing NFCC result in more free media? Then I read carefully and realised the proposal is effectively to punish living persons by using images they might not like to force them to release a freely licenced image. Although there's no suggestion of intentionally choosing unflattering images, the fact remains this is what is being proposed. Sorry but this is completely unacceptable and should never ever be acceptable on Wikipedia. Yes because of our free image requirements sometimes the images we chose is something the subject doesn't like and sometimes a subject may release a free image because of it. But there's a difference here in that we aren't choosing such images because we know the subjects may not like them. In fact one of the things we struggle with is how bad is too bad i.e. when is a free image so bad that we shouldn't use if even if it's the only image we have especially when a subject complains. I'd also note that in so much as other sources using our images which causes additional concern among subjects, it's likely one reason they do is because they're free. So even this part of what currently happens will probably be different when the images aren't free. Nil Einne (talk) 13:17, 1 December 2021 (UTC)

Nil Einne, this isn't a proposal (yet), it would be better hashed out if it was. As you've also noticed, unflattering images sometimes result in free images. Reading the comments here and thinking some more about this, we could perhaps research this effect: we select 100 people (actually 200 so we'd have a control group) from any English-speaking country who are alive, have no photo on their non-stub article and (to alleviate buidhe's concerns) have made a public appearance in the past 2 months. We select a non-free photo for them using the same criteria we use to select a non-free photo for people who passed away. (so we wouldn't select unflattering photos to start with) We make all 100 photos less attractive in some to-be-determined but consistent way. Maybe add some overlay, scale them down (0.01 megapixel?), blur, pixelize, black-and-white, contrast, brightness, newsprint filter, noise or some combination. We put them up for a year (will need to obtain an exemption from the foundation) and afterwards we see if this group has more free media than the control group. It's a fairly best-case scenario as it would require subjects from English-speaking countries (as this is English Wikipedia) with non-stub articles who make public appearances, but the goal would be to determine how strong the effect is. If it would turn out to be barely measurable it clearly wouldn't be worth pursuing this any further. If the effect is stronger than expected, there's perhaps some possible balance where the benefits outweigh the cons. Again, idea lab, no actual proposal at this point. — Alexis Jazz (talk or ping me) 17:37, 1 December 2021 (UTC)

I disagree with the original premise: I don't feel it is desirable to try to exploit people's vanity. Using a lower resolution fair-use image is of course already a standard practice, but I don't agree with deliberately degrading photo quality in a way to make them less attractive. This runs counter to the goal of presenting the best quality information available on a subject in a neutral manner. isaacl (talk) 21:28, 1 December 2021 (UTC)

I can already see the headlines; "Wikipedia badgers celebrities for better photos." - Donald Albury 00:06, 2 December 2021 (UTC)

Oppose forcing pictures into articles. Wikipedia is not a yearbook and does not need a pic in every article. Readers who want to see what a person looks like can easily google them and be lead to scads of pics - some authentic and some photo-shopped. MarnetteD|Talk 00:28, 2 December 2021 (UTC)

@Redrose64 I think the conversation is worth continuing, if we ask is there another way for the BLP to give us the image, but without our requesting it or lowering the photo quality.

1. does the Open Social login or Open id store a photo?
2. Do we need a better photo required list, that we share?
3. Is there a way for a BLP to upload an image to Commons as CC? Maybe holding a WP sign :-) The current process means that someone has to have a celebrities email Wikipedia:A_picture_of_you
4. Do the Article authority control link to somewhere with an image? I know the open scholar id (or whatever is) does and that does have an image
5. [[User:Isaacl|isaacl] Nil Einne Vanity is a good point, except {{[[1]]}} mages? And I agree with {{Alexis Jazz}} point that there is no point in not choosing a better photo.

and

Photos are encyclopaedic by precedent (Britannica),by readers like of them, and by need (they help comprehension for ESL speakers, children, and those with dementia. (As long as we provide other option

Anyway This may all become moot in a few years (with verified ids and pictures) Wakelamp d[@-@]b (talk) 03:44, 2 December 2021 (UTC)

I'm uncomfortable with fair use images in an otherwise freely reusable Encyclopaedia. I'm not convinced that putting up low quality images will bring forth more good quality images under open licences than we already get by requiring images of living people be freely licensed. I see a huge problem in such a change because of people who have previously released openly licensed images now wanting to move to a "fair use" model. But most of all I agree with those above that we should not be putting unflattering images in biographies. ϢereSpielChequers 07:22, 2 December 2021 (UTC)

If our intent is not purely educational in picking the images, then our fair use claim seems to weaken; fair use accounts for the purpose and character of the use. I'm somewhat skeptical that coercing people to agree to license copyrighted content under a free license by using a bad non-free image of them would survive a serious challenge—especially if we are digitally altering them towards the end of portraying their appearance in a negative light as in the case of the floated A-B testing idea. I'd imagine WMF legal would need to give a (conditional) sign-off before this goes to a proposal, if nothing else. I'm not necessarily opposed to using fair-use images of living individuals who it is going to be extremely difficult to get a free picture of (for example, people in solitary confinement or those in prison for the rest of their life) or for people for whom it is unclear as to whether or not they are alive (verifiable missing people). Maybe something more narrowly tailored could get a consensus, but I don't think that something so broad is going to receive sufficient support for adoption. — Mhawk10 (talk) 07:34, 2 December 2021 (UTC)

I think we all need to remember that most people are not nearly as obsessed with Wikipedia as we regular editors are, so most of the images suggested to be included would not be replaced with better freely-licenced ones. I cannot support any idea that makes the quality of BLPs depend on their subjects' willingness to spend any time (even just a few minutes) thinking about Wikipedia. Phil Bridger (talk) 09:29, 2 December 2021 (UTC)

Merging flag template system with the ISO 3166 system

I was looking for a template similar to {{Country name}} that would take a country code and output a wikilink, such as {{Country name|USA}} → United States, but I sure can't find one. I could simply put Country name in [[ ]] like I did just now, but that doesn't account for the two Georgias, among other disambiguates. I kept finding myself looking through the guts of the WP:WikiProject Flag Template system and the Module:ISO 3166 system and I wondered why they weren't operating off of the same internal data. I feel like Template:Country data United States and Module:ISO 3166/data/US should be using the same code base. –Fredddie™ 00:02, 29 November 2021 (UTC)

{{flagg|xx|{{Country name|GEO}}}} → Georgia is one solution. Agree there should be a {{Country link}} template, and unified database if possible. Sod25 (talk) 12:33, 30 November 2021 (UTC)

It's a good start but {{flagg|xx|{{Country name|USA|GA}}}}Georgia is the wrong Georgia. –Fredddie™ 01:24, 1 December 2021 (UTC)

Same with {{flagg|xx|{{Country name|USA-GA}}}}Georgia –Fredddie™ 01:50, 1 December 2021 (UTC)

@Fredddie: The way to do it correctly is {{flag|Georgia (U.S. state)|name=Georgia}} which produces Georgia. ― Blaze The Wolf^Talk_{Blaze Wolf#6545} 14:30, 2 December 2021 (UTC)

Also, |name=Georgia is optional, however without it the name displays as "Georgia (U.S. state)" instead of just "Georgia". I found the code for it at Georgia (U.S. state)#External Links. ― Blaze The Wolf^Talk_{Blaze Wolf#6545} 14:33, 2 December 2021 (UTC)

However I do agree that, in situations like this, being able to use something like that would be helpful. ― Blaze The Wolf^Talk_{Blaze Wolf#6545} 14:43, 2 December 2021 (UTC)

Yeah I'm trying to do it with location codes so inputting "Georgia (state)" is not really tenable. –Fredddie™ 02:52, 3 December 2021 (UTC)

Add a "Funding & Donors" tab under large corporations/organizations

In the heart of democracy, transparency, and informative research, there should definitely be a more standardized category under most large corporations, organizations, non-profits, etc. that discloses said org's donors and funders. Open for discussion! — Preceding unsigned comment added by Tadism (talk • contribs) 00:15, 30 November 2021 (UTC)

@Tadism: our articles don't have "tabs". As far as categories go, not sure how you would want to integrate these to the category system? If the funding for an organization is noteworthy and referenced, you can add edit the article to add it already - either somewhere it fits, or in its own section. I don't think it would be encyclopedic to simply include a list of names of everyone that ever donated to an organization in their article, - only ones of special relevance. — xaosflux ^Talk 11:54, 30 November 2021 (UTC)

If you're using the mobile skin, then the ==Section headings== look like a bit like "tabs". Whatamidoing (WMF) (talk) 20:05, 6 December 2021 (UTC)

I imagine that the funding for particular organizations could be incorporated into the body of the article if it’s reliably sourced and constitutes due weight. I imagine that there could be space for “key people” to be listed in nonprofit infoboxes if there is a person or two who bankrolls an entire operation, but that has to be a case-by-case thing. — Mhawk10 (talk) 21:27, 30 November 2021 (UTC)

Showing special relevance makes sense, butI don't think WP should do things in detail, that NPOV groups do better. They are experts in collecting primary and secondary, sources and weighting it, Maybe, we could link to an article that discusses all the different transparency sites or financial declarations? Or we could use a tool to create an acknowledge current list. There are categories for this, but they are incomplete. Wakelamp d[@-@]b (talk) 22:37, 2 December 2021 (UTC)

Do editors quit because of changes to their user page and aggression on user talk? Options?

User pages are hidden from google and main search. So, does it matter what legal things people do on their user page? Or if there is a  ? (As an aside many MfD discussions seem to be look like payback, are petty, or are people imposing their views.)

Experienced editors (especially admins) see talk as a collaborative space, a place to show appreciation, a task list, and a place to show achievements. New editors and overwhelmed prolific editors see their user page similarly, but also as a private safe space. If they do see it like, then changes to user, anger on talk, or complaints on MfD would be seem as deeply personal. They might perceive these chnages as stalking, harassment, as deeply critical and disrespetcful, or as vandalism. Any of these perceptions might make them quit. if another editor is going to be angry, they should do it the full light of day on article or their own talk.

So, what do people think, and what should we do>? Wakelamp d[@-@]b (talk) 23:51, 2 December 2021 (UTC)

First, what exactly is it that you're asking? You made a lot of semi-related claims (not all of them seem to be substantiated/valid), alongside some that are hard to understand (what 'legal things' are you talking about), and talk in vague generalities. Once you have a clearer question in mind, we can opine more substantially. There are lots of "someone might do this" in there. The realm of possibilities in the human condition is vast. I can post a link to a song on youtube, and someone might get offended by it. But that doesn't mean we need to do anything about the perceived offense.

In general MfDs are for dealing with miscellaneous pages that violate one or more policies or guidelines. We are, amongst other things, not a webhost, and user pages have guidelines to which they must adhere. While in general we tend to let people put more or less what they want on their user page concerning their own interests, background, etc..., there are limits to it because this is specifically not a social networking website. If User:Example is interested in wargaming or baking, they can say so, maybe even put a picture or two up. But userpages dedicated to painting techniques they've personally used, or their grandmother's roast beef sandwich recipe belong on other sites.

Hope that answers some of your question. Headbomb {t · c · p · b} 03:38, 3 December 2021 (UTC)

Thank-you for your reply. I do have some ideas (referenced even). But I would like to understand your viewpoint as an experienced editor first

• Do you really think that Wikipedia has no features of a social network? And are social networks important for retention? (See the Discord channel with avatars and off topic areas)
• Now that user space is hidden from search, what do you see as the benefit to Wikipedia of the MfD policies? And is the cost of editors leaving worth it?
• Can you see how a new editor might find aggression on talk, more upsetting than on article?
• Lastly, do you think younger editors should conform to our policies, or we should acknowledge they may be different? Wakelamp d[@-@]b (talk) 22:29, 3 December 2021 (UTC)

First you need to understand that Wikipedia is an existing community, with its own dynamics, and your questions seem to be mostly an academic/corporate HR exercise concerning what ifs and other hypotheticals, rather than concrete situations.

1. "Do you really think that Wikipedia has no features of a social network?" It doesn't matter if Wikipedia has some features of a social networking site, it is not our goal to be one. We tolerate/encourage social network-like things to the extent that they align with our goals of building an encyclopedia. Letting others know you have a general interest in wargaming lets others know, if they stumble upon a wargaming article and they are unsure what to do with it, that you might be a person with an informed opinion on the subject. So yes, networking between people is a thing, but we are still not a social networking site like MySpace or Facebook, or Livejournal (see WP:NOTWEBHOST).
2. "Now that userspace is hidden from search" It was always hidden, or at least has been hidden for longer than I can remember. As for MFD, see WP:MFD for what it deals with. It ranges from misuse of userpages (from attack pages, hosting copyright violations, violations of WP:WEBHOST like a memorial page to deceased members of your family), to various issues found in the Wikipedia:, Help:, Draft:, Mediawiki:, etc... namespaces.
3. "Can you see how a new editor might find aggression on talk, more upsetting than on article?" Sure, but again, unless you have specifics, this is really a discussion about generalities. I really doubt telling something upsetting (Anywhere from "I disagree with you for these reasons" to "This violates our policies" to "PISS OFF U IDIOT") on a talk page would upset them less if they were told the same thing at a different location (for example, in the edit summary of an article).
4. "Do you think younger editors should conform to our policies" Everyone who wants to edit Wikipedia should conform to our policies. It doesn't matter if they're 82, 28, or 12. Anyone that's here to build an encyclopedia is welcomed to join and help. Anyone that's WP:NOTHERE is welcomed to stay away. Age only matters the extent that competence is needed.

Headbomb {t · c · p · b} 22:54, 3 December 2021 (UTC)

@Wakelamp and Headbomb: Regarding Now that userspace is hidden from search, this has been the case since phab:T104797 was actioned on 23 November 2015. Before that, you needed either an explicit __NOINDEX__ or an option in a header template such as {{userpage|noindex=yes}}, something that I did here but never got around to removing again. --Redrose64 🌹 (talk) 12:18, 4 December 2021 (UTC)

Starting off with point 3.

1. You stated "not all of them seem to be substantiated/valid), see references below.
2. Can you advise your proof that"I see no evidence that harassment on user pages is more damaging than harassment on user talk pages."
3. "It is not that "someone might get offended by it", but it is whether WP:5P4 causes them to leave. Maybe think of it from the perspective of a new editor, or someone who avoids conflict

• They don't have your power to protect your page https://en.wikipedia.org/wiki/Wikipedia:Requests_for_comment/Protect_user_pages_by_default
• Women are the far more likely to use user space to collaborate and to thank, they create more content edits, and avoid conflict [[2]] [[3]] [[4]]. [[5]]

"We revisited their talk pages and found that in many such cases the editors had to face bullying from other editors. Those bullying instances are usually mentioned as a part of the article talk pages. Sometimes the user has also been found to express his/her grievances on their personal user page. Figure 5 shows few example cases of toxic arguments and explicit reasons for leaving the platform that the editor had to face ...When expertise gone missing: Uncovering the loss of prolific contributors in Wikipedia

• New users don't have your experience and are judged far harder than in the past; they react to user page templates negatively. A WMF paper by Epochfai 50 % of good faith new editors have their first edit reverted and a template added to their user page automatically, They were then less likely to use user talk, create content, and were more likely to leave within 2 months. Don’t Bite the Newbies: How Reverts Affect the Quantity and Quality of Wikipedia Work Don’t Bite the Newbies
• A new editor would see the current policies as offering little support; they cane be summarised as toughen up Wikipedia:Harassment#User_space_harassment, Wikipedia:KEEPOFF "it is not enforceable to tell somoene to stay of your talk page", Wikipedia:RPA "t "Reving unquestionable personal attacks from your own user talk page is RARELY a matter of concern.", "Similarly, discussion of a user's conduct or history is not in itself a personal attack when done in the appropriate forum for such discussion (for example, the other editor's talk page, ..." , but this disagrees with [Universal code], WP:AVOIDYOU, WP:HOUNDING and [Wikimedia terms of use] Wakelamp d[@-@]b (talk) 14:00, 5 December 2021 (UTC)

What am I missing: Requiring email for admins

I was having a little look at November's inactive admins and noticed that GermanJoe was unable to be email notified of their inactivity - this got me thinking, could we make having a verified email a requirement for administrators?

I can't think of any real negatives, and the security benefits are fairly significant:

• Password resets
• Notification of failed logins
• Contactable by bureaucrats

Similar sorts of discussions (i.e. "requiring 2FA for admins") tends to fall down on implementation and availability of 2FA devices as not everyone has access to a mobile phone - there are a number of email providers, such as Gmail, which are entirely free and are no more difficult to use than MediaWiki...

For what its worth, 989 of our current 1067 admins (~93%) do have email enabled already.

So, save me typing out a doomed RfC, what am I missing here? -- TNT (talk • she/they) 04:00, 4 December 2021 (UTC)

I think it's a reasonable proposal and I would support it. It's also possible to use email as a 2FA method, but that's tangential.

One question I have though, is the requirement to just have an email set, or would it be to have a working email set? Email addresses do regularly stop working or people lose access to them. Would we have regular pings and confirmations to make sure the email keeps working? If emails bounce enough and the address becomes unconfirmed, do they automatically lose their admin privileges? Legoktm (talk) 05:28, 4 December 2021 (UTC)

I think it's reasonable to discuss this, and I'd tentatively support the tenets premised in the original posting. I will, however, say: I do not think we could (for reasons of privacy) or even should (for multiple reasons) require an affirmation of receipt or reply in any form. Best regards.--John Cline (talk) 15:37, 4 December 2021 (UTC)

@John Cline: why would there be a privacy issue? Legoktm (talk) 12:00, 5 December 2021 (UTC)

Thank you for asking this of me Legoktm; I'll offer this clarification:
Of the many possible ways a negative issue with privacy could affect our ability to mandate email access and, in particular: email correspondence, the one I referred to (in my comment) which almost certainly would, involves the apparent conflict (and inherent direct contravention) that exists with formalizing such a mandate in light of the painstakingly deliberate implementation efforts associated with Wikipedia's system of email delivery.
That system thoroughly describes:
1.) the efforts emplaced to prevent disclosure of your email address (as the recipient)
2.) an allusion to reasons for emplacing this protective measure enveloped in cautioning against arbitrarily disclosing your email address without due consideration, and
3.) the belabored instruction that a reply is not required coupled with the belabored explanation that choosing to reply would, thereby, effect and include the very disclosure ill-advised therein.
Consider, for example, the statements given to the recipient in advance of the email message itself; which says:
"The sender has not been given the recipient's email address, nor any information about the recipient's email account; and the recipient has no obligation to reply to this email or take any other action that might disclose their identity. If you respond, the sender will know your email address." emphasis mine.
I apologise for the lengthy reply but know not of another way when efforts at concision have failed, having been tried. Please follow-on, if questions remain, and (Lord forbid) I haven't been clear regarding the answer to the question asked. Best regards.--John Cline (talk) 19:08, 6 December 2021 (UTC)

It's difficult to see what problem this is addressing. An inability to do password resets is a risk anyone is entitled to take. Failed logins are a non-issue - we rarely notify anyone about them. Contact by bureaucrats can be done on the talk page. I have certainly been through periods where the only wiki emails I get are harassment, or not interesting, or I have simply had no need for private communications. Adding to that, email accounts may be free, but they usually require regular logins. And if someone is logging in only every few months to keep an email account alive, what is the point of email? What you're probably wanting is an email address which will reach a user promptly. At times I offer no such thing, and I don't see why anyone else should. -- zzuuzz ^(talk) 16:04, 4 December 2021 (UTC)

I haven't thought about this enough to decide if I'd support it, but one thing to be clear about is that there's a difference between having an email for your account and having the "let other users email me" option checked. Some admins may prefer not to have that option checked, since it opens up vectors for harassment, and some prefer on principle to handle everything in the open for transparency. {{u|Sdkb}} ^talk 04:30, 5 December 2021 (UTC)

How does adding an additional attack surface (a possibly poorly secured email account) help to increase security? Also, we need more admins, and desysopping people because they do not have an email address set seems to go in the wrong direction. —Kusma (talk) 13:33, 5 December 2021 (UTC)

I have an email attached to my account, just not enabled for "let other users email me" option. I get emails all the time from editors I know well, notifications from the Wikipedia Library, or various others. Before I became an admin, I had to change my email address and remove the option to let others email me - because an editor added me to receive their daily blog. I am not interested in what users do in their personal lives. And then there were a few others who wanted to dish dirt on other editors, or trying to get me to take an admin action without discussing it on Wikipedia. Not enabling the "let other users email me" option allows me to remain above board on admin actions. — Maile (talk) 13:58, 5 December 2021 (UTC)

I have a gmail attached, and an emailnotice which discourages frivolous or harassing emails. I see no problem requiring email access to admins, just as I see no reason it should be an IRL email address. We can only assume GermanJoe made a conscious decision about email and was aware of the possible consequences. Cabayi (talk) 15:08, 6 December 2021 (UTC)

Project banners should be mostly determined by categories, rather than applied to pages individually

I'm guessing this has been considered before at some point, but it's quite inefficient that we try to remember to tag each biography talk page with {{WikiProject Biography}} rather than just automatically having every article in a person category (or defined as a person on Wikidata) fall into that project's scope automatically. Ditto with most other project banners—rather than having to remember to tag every page for Iranian topics with {{WikiProject Iran}}, just have the project cover everything within Category:Iran. This would require us to solve leaky categories, but we need to do that anyways (and if we can't, maybe we should use Wikidata instead). {{u|Sdkb}} ^talk 04:36, 5 December 2021 (UTC)

Would this be something a bot is capable of doing? We could manually approve a few core categories for each WikiProject and then have a bot assign the WikiProject tags for those. I don’t think that this is something that should go into every category , but something like this could work in a limited scope. — Mhawk10 (talk) 04:42, 5 December 2021 (UTC)

I'm thinking about this mostly in the less practical, more abstract sense of how we might totally overhaul our category system. But yes, in the practical realm, there's a lot of work to be done in this area. Anomie's WP:WikiProjectTagger could be put to much wider use than it has been so far. {{u|Sdkb}} ^talk 06:51, 5 December 2021 (UTC)

Before any large scale overhaul happens, we should be clear on what project tagging is supposed to be good for. The only thing I really use is that they feed into article alerts so I get informed about what is going on at some projects I watch. That wikiprojects operate with a "tag" system instead of a "tree" system based on categories might actually be beneficial. —Kusma (talk) 11:42, 5 December 2021 (UTC)

As many projects currently are more or less inactive, I don't think doing this would be productive for every possible project. Perhaps a trial of the idea with more active projects willing to opt into it would clarify the benefits. - Donald Albury 17:44, 5 December 2021 (UTC)

@Donald Albury I think WikiProject Disability might be a good candidate for testing these ideas. It is a fairly small project and is quite neatly associated with a category tree with Category:Disability as the "root". -- Roger (Dodger67) (talk) 17:54, 5 December 2021 (UTC)

I thought the purposes of talk page banners were mostly to make empty talk pages look fancier and to advertise WikiProjects, neither of which is accomplished by doing this with article categories. User:力 (powera, π, ν) 17:46, 5 December 2021 (UTC)

@力 That's probably how it looks from the article end, but having a project banner on an article enables various functions and tools at the project page, so it's definitely not mere decoration. Roger (Dodger67) (talk) 17:58, 5 December 2021 (UTC)

• We could still display the banners in the same way, even if they're determined differently. {{u|Sdkb}} ^talk 18:16, 5 December 2021 (UTC)

I think that's a good idea if possible. However, there are WikiProjects which tag articles with things other than just importance+class. Would any of these present implementation difficulties? Dege31 (talk) 18:35, 5 December 2021 (UTC)

Dege31 I don't think that's a serious concern, once the basic tag has been added the project members can easily add their additional parameters. I sometimes do it for WP:MILHIST (which has a particularly comprehensive set of banner parameters. Roger (Dodger67) (talk) 19:26, 5 December 2021 (UTC)

• A semi-related comment: I'd like to see the whole stub category system go away. It more or less duplicates the project hierarchy and I don't see that it serves any useful purpose now that projects are an established part of how the encyclopedia is organized. -- RoySmith (talk) 18:49, 5 December 2021 (UTC)

  Definitely related, and I agree. I think the ultimate goal should be to only have to categorize stuff once, and to have everything else build off that core system. {{u|Sdkb}} ^talk 22:26, 5 December 2021 (UTC)

  @DannyH (WMF), did you ever write up the "neighborhoods" idea? A tag-once system would be useful for that. Whatamidoing (WMF) (talk) 20:31, 6 December 2021 (UTC)

• The solution here is do both. We have dozens of tagging bots. But categories need to be super specific to be bot-friendly. Just because something mentions Canada doesn't mean it's of interest to WikiProject Canada. Headbomb {t · c · p · b} 18:53, 5 December 2021 (UTC)

• Adding to Headbomb's point User:AlexNewArtBot has a far too high false-positive rate. At WP:DISAB my thumbsuck estimate is that only about a quarter of the new articles it lists are actually sufficiently relevant to receive the project's tag. So we should look at how the bot identifies relevant projects and try to do better than that. Roger (Dodger67) (talk) 19:26, 5 December 2021 (UTC)
  • The new article alerts bot just searches new articles for a list of regular expressions, giving and taking away points based on which ones are contained, and then lists all of the articles which are above the threshold. The particular rules for WP:DISAB are here – and are fairly limited. (My experience is mostly with the Classical Antiquity articles, which has a rather more substantial ruleset and seems to do relatively well at picking relevant articles). Caeciliusinhorto-public (talk) 11:01, 6 December 2021 (UTC)

Please remember the guidelines of WP:CATVER and WP:CATDEF. That means that there are some articles that might not be in a category but are eligible to be part given wikiproject. MarnetteD|Talk 15:10, 6 December 2021 (UTC)

Quality control

Please correct me if I am wrong; I believe there is no mechanism on Wikipedia to prevent bad edits from being made. I also believe that there are no mechanisms in place to effectively deal with bad edits once they have been made. By this, I mean edits which might be well-meant, but are of poor quality. I ask this because I observe that poor quality content and serious errors can be found in almost any article you might care to look at, and if ever I investigate the article history, I usually find that the poor quality material has been in the article for many years. DSMN-IHSAGT (talk) 14:50, 6 December 2021 (UTC)

DSMN-IHSAGT, WP:Sofixit. - Cabayi (talk) 14:53, 6 December 2021 (UTC)

In what way do you think that answers my question? DSMN-IHSAGT (talk) 14:57, 6 December 2021 (UTC)

You didn't ask a question DSMN-IHSAGT you made a declarative statement. Yes, that situation exists on numerous articles here at the 'pedia. Please be aware that there are edit filters that erase some bad edits. For the rest it is the nature of a website that anyone can edit that there will be problems that need fixing. Please feel free to improve articles as you see fit. MarnetteD|Talk 15:05, 6 December 2021 (UTC)

OK then, to be clear, my question is: is there really no effective mechanism to prevent bad edits from being made, or from dealing with them once they have been made? DSMN-IHSAGT (talk) 15:34, 6 December 2021 (UTC)

What 'mechanism' would you propose? AndyTheGrump (talk) 15:37, 6 December 2021 (UTC)

I can think of plenty. I just want to be sure that my interpretation is correct; that there is no mechanism in place to prevent bad edits from being made, or to deal with them once they have been made. A specific example: 14 years ago, an anonymous editor added a lot of text to Neuchâtel that was not encyclopaedic in content or tone (see Special:Contributions/132.156.198.16). Today, I found that much of that poor content was still in place. I observe that a majority of articles contain significant quality deficiencies, so I see two possibilities: a) there are mechanisms in place to deal with poor quality, but the rate at which poor material is added exceeds the rate at which it is dealt with; or b) there are no mechanisms in place. If the former, don't they need improving? If the latter, don't they need to be put in place? DSMN-IHSAGT (talk) 15:47, 6 December 2021 (UTC)

There are lots of processes and mechanisms in place to deal with "bad" edits - depending on the definition of bad. You can learn more about these at Wikipedia:Recent changes patrol. — xaosflux ^Talk 15:50, 6 December 2021 (UTC)

OK, well, then it seems to me that these mechanisms are just not up to the task of ensuring that edits are predominantly good, or are dealt with rapidly if they are bad. Shouldn't they be improved?DSMN-IHSAGT (talk) 16:58, 6 December 2021 (UTC)

It is true that many edits deteriorate the contents of Wikipedia, but the model of letting anyone edit, including making less than optimal edits (at least until an editor demonstrates that their edits have a serious detrimental effect on the encyclopedia) is what has brought Wikipedia to its present state. We must be careful that any method of preventing "bad" edits does not end up killing Wikipedia. - Donald Albury 16:41, 6 December 2021 (UTC)

I agree with you that the model brought Wikipedia to its present state. But that state is that most articles have serious deficiencies. How would a method of preventing bad edits ever kill Wikipedia? DSMN-IHSAGT (talk) 16:58, 6 December 2021 (UTC)

• Because good edits will also be prevented. If it was easy for software to determine whether an edit was good or bad, then it would be easy to have the software write Wikipedia articles.
• Because humans, including Wikipedia editors, learn from making mistakes. If you prevent all bad edits, you prevent many useful forms of learning.

WhatamIdoing (talk) 20:35, 6 December 2021 (UTC)

I can easily imagine simple mechanisms that would prevent certain types of bad edit from being made, and that could not possibly prevent any good edit from being made. I disagree entirely with your premise that "determining whether an edit is good or bad" is a problem of equivalent complexity to "writing an encyclopaedia article". And part of my point is precisely that there is no mechanism for anyone to learn from their mistakes. Some simple QC checks which warn the user about common errors would provide precisely that. Preventing bad edits provides a useful form of learning. I've commented similarly below. DSMN-IHSAGT (talk) 21:44, 6 December 2021 (UTC)

• There are many "mechanisms" to prevent poor quality edits being made, but none work perfectly, and once in place for a while, poor edits often remain for a long time. As an example, a rather high proportion of your own c. 40 edits (all in the last 2 days) have already been reverted - so the system does work, fairly well! Johnbod (talk) 17:03, 6 December 2021 (UTC)

I like your use of passive voice to describe your own actions. You provide another excellent example. You made a series of edits a month ago, in which you started with a pointless restatement of the article's title (do not introduce the list as "This is a list of X"...), misspelled "officially", wrote "This is the case in Ireland also" instead of the idiomatic "This is also the case in Ireland", used relative time references twice, and falsely claimed that the government of Ukraine is seeking "to make the rest of the world say Kyiv". There were no checks in place to stop you doing this, and your errors remained in place for a month until I fixed them. Why on Earth would you have just put them all right back? Did you not understand that they were errors? Did you simply resent someone else fixing your errors? Either way, it illustrates the issue extremely well. Had there been even so much as a spell-check in place to warn you that you'd made a mistake, it would have been a positive for the encyclopaedia. DSMN-IHSAGT (talk) 17:30, 6 December 2021 (UTC)

No, apart from the one-letter typo, which I corrected, I do not think they were "errors", which is why I reverted. It is remarkable how high a proportion of your edits were problematic. Given the very bad-tempered tone of your edits here and your edit summaries, and your fluent command of in-house jargon, one is bound to suspect this is not your first rodeo. Johnbod (talk) 17:42, 6 December 2021 (UTC)

It is remarkable how many of my edits, which corrected grammar, spelling, factual and style errors, you have a problem with. Nobody else has reverted any of my edits. And why would they? You seem to me to exemplify many problems with Wikipedia; you contributed extremely poor content and not only was there nothing to prevent you from doing so, there is nothing now to prevent you from aggressively restoring your errors. And you even argue that they are not errors! I linked to the easy-to-find (and also common sense) guidelines that I was following. Do you not understand them, or are you consciously flouting them? DSMN-IHSAGT (talk) 18:01, 6 December 2021 (UTC)

No longer true - see Canal - your change there was reverted within 2 minutes. Johnbod (talk) 18:37, 6 December 2021 (UTC)

And my, you must have been giddy with excitement when you saw that. I had inadvertently introduced an ugly repetition; someone reverted that. Fair enough. Ideally, they would not have simply restored the problem that I'd fixed but instead improved the article still further. But I then did that. Unlike you, I am capable of understanding that I made a mistake, and would not dream of reacting as inappropriately as you have done to being corrected. DSMN-IHSAGT (talk) 18:58, 6 December 2021 (UTC)

• Discussion on article talk pages is important for quality control, as well as other things. Try it. Phil Bridger (talk) 18:05, 6 December 2021 (UTC)

To take the example I highlighted above at Neuchâtel, the poor text was never discussed when added in 2007 or at any time since. So I would argue that talk pages are not an effective mechanism for quality control. DSMN-IHSAGT (talk) 18:09, 6 December 2021 (UTC)

They are if you follow WP:BRD and discuss reverts rather than get offended by them. Why haven't you started a discussion at Talk:Neuchâtel? What makes you think that you have any less of an obligation to do so than any other volunteer editor? A project like Wikipedia can't get anywhere if people refuse to use the mechanisms that we have. Phil Bridger (talk) 18:14, 6 December 2021 (UTC)

The user who is reverting my edits at Neuchâtel is doing so because they felt offended by my correction of their errors at another set of articles. I don't believe they are acting in good faith so I don't see any particular benefit to discussing anything with them there. DSMN-IHSAGT (talk) 18:18, 6 December 2021 (UTC)

I would have reverted some of your edits there (example – this word means different things in different English variants, so explaining it helps readers). WhatamIdoing (talk) 20:41, 6 December 2021 (UTC)

MOS:ENGVAR says use one English variant consistently within an article. It doesn't say provide synonyms for any word where usage might differ. DSMN-IHSAGT (talk) 21:51, 6 December 2021 (UTC)

This discussion is rapidly turning into nothing but editors griping at one another over disagreement of each others' edits. That's not what this forum is for, and those of us who monitor it for constructive discussions are having our time wasted.

Please limit your discussions here to the topic of general issues and suggestions for improvements of Wikipedia. TJRC (talk) 19:33, 6 December 2021 (UTC)

Quality control mechanism

There are many, many common sense guidelines which it would be easy to automatically check that edits comply with. A simple set of regexes applied to the diff of an edit could check, for example, that section headings were correctly capitalised, acronyms defined, list articles did not start with "This is a list", etc, etc. People take it very personally if you tell them they have made a mistake (see above) but it is exceedingly common for mistakes to persist in articles for years. So automated warning is really essential, I think, to prevent mistakes from being made. A simple warning that your edit might not comply with some guidelines - what downside could there be to that? DSMN-IHSAGT (talk) 18:18, 6 December 2021 (UTC)

Alphabetically, or in order of likelihood?

Seriously, this is partly done with the Special:AbuseFilter, user scripts, and anti-vandalism bots, but processing speed (you would have to check thousands and thousands of rules) and the damage caused by both false positives and false negatives are significant barriers. As a thought exercise, imagine what rules you would need to write to evaluate just the single article Profanity. Write it in regex, or write it in plain English, but write it down, and see how many rules it takes you just to reject edits that add words related excrement, except when it's potentially encyclopedic content (e.g., Poop humor, Fart joke, Toilet humour, Scatology, etc). WhatamIdoing (talk) 20:52, 6 December 2021 (UTC)

You're talking about vandalism I think. I'm talking about quality control checks, and specifically ones that would be simple to implement. Like, checking that section headings are correctly capitalised. My thought experiment is as follows:

• case 1, no QC mechanism. User writes ==Section Heading== and hits save. Article now uses incorrect style. Likely outcome: nobody says anything to them, and the incorrect style will remain in place for a long time. Possible outcome if someone does say something to them: user gets offended.
• case 2, QC mechanism in place: User writes ==Section Heading== and hits save. But instead of the save going through, they see a warning saying something like "section headings should be in sentence case, not title case. See Wikipedia:Manual_of_Style#Article_titles,_sections,_and_headings. Please either fix this, or just press 'save' if the guideline does not apply (eg a person's name is in the heading)." Likely outcome: user fixes the capitalisation before saving. Also likely outcome: the user, now being aware of the guideline, will correctly capitalise headings in the future.

What would the downside be? DSMN-IHSAGT (talk) 21:39, 6 December 2021 (UTC)

That might deal with a few trivial errors, but we have a much bigger problem with editors who are so convinced that they are right that they don't need to follow normal consensus-building practices, such as using article talk pages. Dealing with such editors would help us much more with quality. Phil Bridger (talk) 21:53, 6 December 2021 (UTC)

Content disputes affect a minority of articles; simple and easily avoidable errors, a majority. Are you saying that simple and easily avoidable errors are not worth worrying about? See Broken windows theory. DSMN-IHSAGT (talk) 22:20, 6 December 2021 (UTC)

I think many of the things you cite as examples are both extremely minor errors (and indeed, are actually stylistic variations that are "wrong" only by Wikipedia MOS standards) and subject to false positives. Either of those characteristics indicate things that ought not to be baked into the underlying configuration; where both both are true, that's especially so. They might be good candidates for bots, which can be both shut off in case of excessive false positives and can be expressly warded off with things like {{not a typo}}, {{text}} or the like. TJRC (talk) 23:06, 6 December 2021 (UTC)

Which things that I've cited do you regard as too minor to care about, and which do you think could not be checked without giving false positives? I think that fixing a simple error later with a bot has exactly the same advantages and disadvantages as raising a warning before the error can be saved, except that the person making the error would not be made aware of it. DSMN-IHSAGT (talk) 23:17, 6 December 2021 (UTC)

Capitalization of section headings is an example. Sometimes, title case is correct. And the reader gets the same value from the article whether section heads use title or sentence case. Schazjmd (talk) 23:20, 6 December 2021 (UTC)

In the process I outlined above, I noted that specific possibility. If title case were correct, the user could simply ignore the warning and save anyway. I don't think the reader does get the same value from a badly-presented article as they do from a well-presented one; that's really the whole reason that any publication has a manual of style, isn't it? DSMN-IHSAGT (talk) 23:28, 6 December 2021 (UTC)

Just so we're clear, I haven't said anything is "too minor to care about". I've said the things you point out are too minor to be automatically attempted to be "corrected" by error-prone technology which is as likely to make things worse than better.

Are you actually here for a serious discussion or are you just interested in having a fight, being inflammatory and burning straw men? TJRC (talk) 23:23, 6 December 2021 (UTC)

Sure, yeah, it could have potential. See Wikipedia talk:Making editing easier 2021#Text reactions. People worry about false positives but I suspect the correct balance to be drawn is not all the way on the side of "don't show hints if there's a chance they'll be wrong". Enterprisey (talk!) 00:05, 7 December 2021 (UTC)

Nobody has mentioned ClueBot NG (talk · contribs) - example edit. --Redrose64 🌹 (talk) 09:30, 7 December 2021 (UTC)