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:
Before commenting, note:
Discussions are automatically archived after remaining inactive for two weeks. |
« Archives, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38 |
Random Article Feature
![](https://web.archive.org/web/20211106083923im_/https://upload.wikimedia.org/wikipedia/commons/thumb/e/ea/Purple_arrow_down.svg/20px-Purple_arrow_down.svg.png)
I really like the random article feature, but it can be improved. I would love to see a way to search for random articles within certain categories.
e.g. search for a random article in science, astronomy, literature, music, etc.
I'm not sure if all articles are categorized by default, but I'm sure a select group of people including myself would help categorize articles so a feature such as this could be implemented.
Thanks. Senor hams (talk) 18:25, 13 October 2021 (UTC)
- @Senor hams: you can use Special:RandomInCategory to get a random article from a specified category (example: Special:RandomInCategory/Apples). — xaosflux Talk 18:43, 13 October 2021 (UTC)
- @Senor hams: - I've moved this to Idea Lab, as it isn't really a ready to go proposal for VPR. — xaosflux Talk 18:46, 13 October 2021 (UTC)
- It is difficult to randomly explore a wide theme using categories, because most of the related articles have been split into subcategories so fall outside the range of the search. For example, a random for "Yorkshire" would only result in one of 4 articles or its 33 subcategory pages, but not to one of the 20,000+ articles in those subcategories. To improve the RandominCategory it would be helpful to have the option to include all of the subcategory pages too. An alternative is to search using the larger Wikiproject related categories (e.g. "WikiProject Yorkshire articles") as these can give more possible outcomes, but annoyingly these are on the talk pages so link there rather than directly to articles. EdwardUK (talk) 20:58, 13 October 2021 (UTC)
- @Senor hams: - I've moved this to Idea Lab, as it isn't really a ready to go proposal for VPR. — xaosflux Talk 18:46, 13 October 2021 (UTC)
- This is a perennial proposal that would work well in theory but is so hard to actually put into practice. I'm not saying I wouldn't support this, but the subcategories, as EdwardUK pointed out, would make it nearly impossible. 🐔 Chicdat Bawk to me! 10:15, 14 October 2021 (UTC)
- @Xaosflux, do you think it would be easy to set up a script that does Special:RandomInCategory based on WikiProject categories, but then redirects to the content space? Special:RandomInCategory/All WikiProject Medicine articles takes you to the talk page. Whatamidoing (WMF) (talk) 02:28, 20 October 2021 (UTC)
- @Whatamidoing (WMF): a script solution to it didn't immediately come to mind, but maybe? Strategy for that would probably depend on who the audience for it would be, and how it needs to be presented. — xaosflux Talk 12:34, 20 October 2021 (UTC)
- @Xaosflux, do you think it would be easy to set up a script that does Special:RandomInCategory based on WikiProject categories, but then redirects to the content space? Special:RandomInCategory/All WikiProject Medicine articles takes you to the talk page. Whatamidoing (WMF) (talk) 02:28, 20 October 2021 (UTC)
- @Senor hams: One clunky way to do it is using lesser known features of Special:Search. You can filter to articles in certain topics as determined by ORES machine learning with
articletopic:TOPIC
. And you can sort search results randomly by adding&sort=random
to the URL. So this link will give a randomly sorted list of "television" articles. the wub "?!" 13:55, 20 October 2021 (UTC) - Is https://randomincategory.toolforge.org/apples "just" an external version of Special:RandomInCategory/Apples? — GhostInTheMachine talk to me 20:41, 20 October 2021 (UTC)
- If you don't care how random the result is. The short short version is that Special:RandomInCategory isn't really random; it gives pages in the category in an ostensibly unpredictable fashion, but - especially in small categories - the pages aren't equally likely to be selected. The toolforge tool claims that it is really random, and I have no reason to doubt it. Also, it can filter for specific namespaces, or filter out files or subcategory pages.If you want a random article from within a category tree to a given depth, Petscan can do that (select "randomly" from Output > Sort). Harder to use, though. —Cryptic 21:47, 20 October 2021 (UTC)
- @Ahecht:'s tool (randomincategory.toolforge) is nice, I wonder if they would be open to adding a parameter to support this use case? Ahecht, specifically that would be: request a page in a category, but when providing the result, strip the TALK prefix out? Perhaps a
?returnmain=true
or the like parameter (understanding that this COULD return a redlink)? — xaosflux Talk 11:55, 21 October 2021 (UTC)- @Cryptic @Xaosflux I discovered this lack of randomness when I was working AfC. It's really quite non-random. See T230700. -- RoySmith (talk) 12:31, 21 October 2021 (UTC)
- @RoySmith: I suspect for most of the use case of this, true randomness isn't really that important? This is mostly to let readers browse around and discover new articles to read. — xaosflux Talk 13:19, 21 October 2021 (UTC)
- For the simplest use case of a idly browsing user (which, as I understand it, was the original use case for that feature), it's probably good enough. -- RoySmith (talk) 14:02, 21 October 2021 (UTC)
- I believe the pseudo-random mechanism is something like this: Pick a number between 0 and 1 (let's say that we picked 0.32), get the smallest and largest page ids for the pages in the category, and find the article whose id number is closest to 32% of the way through the range.
- The problem is that id numbers aren't random, so you're pretty much sorting by date of creation, and when the process (e.g., AFC) is rather date-dependent, then you end up with ids that look like this:
- 1, 2500, 5609, 6233, 7239, 8305, 8359, 8904, 9134, 9234, 9345, 9456, 9519, 9714, and 10000
- and about 40% of the "random" pages will be one of the first two pages, because 40% of 10,000 is 4000, which is closer to the second page id number than to the third one. Whatamidoing (WMF) (talk) 19:50, 21 October 2021 (UTC)
- For the simplest use case of a idly browsing user (which, as I understand it, was the original use case for that feature), it's probably good enough. -- RoySmith (talk) 14:02, 21 October 2021 (UTC)
- @RoySmith: I suspect for most of the use case of this, true randomness isn't really that important? This is mostly to let readers browse around and discover new articles to read. — xaosflux Talk 13:19, 21 October 2021 (UTC)
- @Ahecht:'s tool (randomincategory.toolforge) is nice, I wonder if they would be open to adding a parameter to support this use case? Ahecht, specifically that would be: request a page in a category, but when providing the result, strip the TALK prefix out? Perhaps a
- @GhostInTheMachine: Special:RandomInCategory picks articles by randomly choosing a timestamp and then finding the article added to the category closest to that timestamp. This means that a cluster of articles added around the same time are less likely to be chosen than an article added by itself. The impetus for creating the tool was when I noticed that accessing Special:RandomInCategory/Pending AfC submissions 25 times only returned about 15 unique pages out of the 1,350 pages in that category at the time (the rest were doubled or tripled). The toolforge tool literally lists out all the pages in the category and chooses one at random, and should be as "really random" as the PHP array_rand() function allows.
- @Xaosflux: I added a
returntype=
parameter that can take the valuessubject
,article
, ortalk
. The first two always return the associated article (if it exists), the last one returns the associated talk page (if it exists). For example, toolforge:randomincategory/All WikiProject Medicine articles?returntype=article works as expected, although with very large categories like that it will take a while to load the first time you use it (it caches categories for 10 minutes). It fetches the associated page using the API, so there is a bit of a performance hit compared to just stripping "talk:" from the title, but that way it will never return a redlink and it will works across all language wikis, not just those that use the "talk" prefix. --Ahecht (TALK
PAGE) 20:39, 21 October 2021 (UTC)- @Ahecht: thank you for the quick action! — xaosflux Talk 03:09, 22 October 2021 (UTC)
- @Whatamidoing (WMF): see above, this is off-wikiish, but is perhaps something we could integrate for curious readers somewhere perhaps? — xaosflux Talk 03:09, 22 October 2021 (UTC)
- In the sidebar? (I'm not sure non-editors really use it.) Special:Random takes you straight to a page, so that limits the opportunity for showing a message. Whatamidoing (WMF) (talk) 18:34, 26 October 2021 (UTC)
- If you don't care how random the result is. The short short version is that Special:RandomInCategory isn't really random; it gives pages in the category in an ostensibly unpredictable fashion, but - especially in small categories - the pages aren't equally likely to be selected. The toolforge tool claims that it is really random, and I have no reason to doubt it. Also, it can filter for specific namespaces, or filter out files or subcategory pages.If you want a random article from within a category tree to a given depth, Petscan can do that (select "randomly" from Output > Sort). Harder to use, though. —Cryptic 21:47, 20 October 2021 (UTC)
Help on Foreign Wikipedias
There are frequently questions asked at various forums on the English Wikipedia about dealing with foreign Wikipedias, typically about block appeals. (Of course, the user may or may not be blocked. Questions about blocks from editors who are editing from their own account and are not blocked do happen. That is a detail.) They have also asked about blocks on Commons, or about problems on Meta. They are always told, of course, that each Wikipedia is a separate project. However, what I am suggesting is a global Help page giving information about seeking various sorts of help on other Wikipedias, and on Commons, and on Meta, and on Wiktionary, and so on. The page, possibly with subpages, should be one that we invite cross-wiki editors to maintain as a service to other cross-wiki editors. It should have a disclaimer that the information is the best that we have here but may not be accurate, and to go to the page on the foreign Wikipedia and ask further, and please come back and update the help here if it needs updating. Robert McClenon (talk) 18:18, 16 October 2021 (UTC)
- Why would this be better than just telling people to go ask on the appropriate project? (Genuine question!) Calliopejen1 (talk) 19:44, 20 October 2021 (UTC)
- User:Calliopejen1 - It is not always obvious on a given project where or how to ask for help. I recently encountered a page on Meta that I thought was a very bad attempt at humor that had no real purpose, and so should be deleted, but it was not obvious on Meta either how to ask for help or how to nominate a page for deletion. I did answer my own question, and the page was deleted. A less experienced user might have been baffled. How to look for help is different on different projects, and the differences can be confusing to a user, especially to a user who lacks experience with online systems. Some of the questions that we see about foreign Wikipedias, or about other projects (e.g., Commons, Wiktionary), seem to indicate that the user either has tried and failed to find help, or is exasperated and has stopped looking on the other system. Robert McClenon (talk) 17:00, 23 October 2021 (UTC)
- I don't see anything wrong with maintaining a page that tells an editor how to do that. That would give people something to link to if editors ask here, which many people seem to do in the mistaken belief that the English Wikipedia has some sort of primacy over the other languages. Phil Bridger (talk) 19:57, 20 October 2021 (UTC)
- An example of where this could be useful can be seen here. I see a few of these every month, and it would be good to be able to direct the editors to the correct place to ask for help on the other project. Phil Bridger (talk) 09:27, 24 October 2021 (UTC)
Preparation for IP masking
I read few weeks ago here and on meta wiki that IPs are going to be masked. Then some editors proposed that IP editing be banned but many opposed it. So banning will be controversial. But, are there some other ideas which can be implemented to counter vandalism? I am relatively new at wikipedia so I don't know what will that be exactly. I propose that now a userscript be made which mimics IP masking. Interested vandal fighters can add in there js file so as to get a feel on how it will look after update, so that community will get better idea about it and vandal fighters will be more prepared for actual masking. I have no idea how to make such script but just thought came in mind so shared it. -- Parnaval (talk) 11:27, 22 October 2021 (UTC)
- @Parnaval: while an interesting preparation idea, I would think one of the bigger issues here is that it's not the regular vandal fighting workflow that is impaired by masking but the socking and range-hunting. A lot of that is done by cross-editor efforts (rather than one person handling everything themselves), so you'd have to make sure everyone was on it. Otherwise you'd be getting "can you block 5%£u343ur/64" and others would be "who is 5%£u343ur/64 - I'm seeing 4.343.121.12/64" etc Nosebagbear (talk) 16:06, 22 October 2021 (UTC)
- @Nosebagbear: Then just aggressively add this to every editor who is rollbacker/admin/checkuser. And post a message to talkpage of everyone in these usergroups. And if this is not feasible, there should be something else which enwiki should do before masking goes live. Some sort of brainstorming needs to be done like one going on for RfA review. -- Parnaval (talk) 06:45, 23 October 2021 (UTC)
- Parnaval, those of us who have been around a long time and have been instrumental in getting important policies brought into line with today's reality know that change on Wikipedia happens slowly. One medium sized Wikipedia has recently banned IP editing with resounding success, while the WMF is currently supporting a 6-month trial on another. That said, consensus can change, and sometimes dramatically so, hence banning IP editing might not be quite so controversial after all. These issues are discussed in depth on Meta. Kudpung กุดผึ้ง (talk) 05:59, 24 October 2021 (UTC)
- @Nosebagbear: Then just aggressively add this to every editor who is rollbacker/admin/checkuser. And post a message to talkpage of everyone in these usergroups. And if this is not feasible, there should be something else which enwiki should do before masking goes live. Some sort of brainstorming needs to be done like one going on for RfA review. -- Parnaval (talk) 06:45, 23 October 2021 (UTC)
- Why are you even doing IP masking. I ask: why? I see no benefits and many disadvantages of hiding IP numbers from all but the privileged few. 🐔 Chicdat Bawk to me! 12:11, 25 October 2021 (UTC)
- @Chicdat: -- Well, we aren't doing it. The WMF is. We don't want it, but alas -- Rockstone[Send me a message!] 00:22, 28 October 2021 (UTC)
Roud Folk Song Index and Child Ballads nr in the WP:LEAD of articles about songs
It's common [1] that a WP-article on a folk song (?) will mention something called a Roud Folk Song Index number in the WP:LEAD, sometimes also a Child Ballads number. See also List of folk songs by Roud number. Sometimes in the first sentence:
Sometimes a little further down:
IMO, this looks like strange WP-writing, like having "The imdb number is tt0068646" in the lead of The Godfather, and not "a summary of [the article's] most important contents". I'm guessing Roud and Child are more academic, though. A previous discussion Wikipedia:Village_pump_(policy)/Archive_156#Stop_putting_Roud_index_numbers_in_the_lead_of_folk_songs mentioned external links, infobox and wikidata (it can be added there apparently [2]) as alternatives to lead, but nothing really happened afaik.
So, is there some sort of consensus that having these in the lead should be dealt with somehow (how?), or is it ok to have them there? At the time of the previous discussion I edited Cotton-Eyed Joe like so [3], noone has changed it. Gråbergs Gråa Sång (talk) 20:08, 22 October 2021 (UTC)
- I know they're an important identifier for those songs. It would be useful to have some standard method of incorporating them in those articles, so readers and researchers can rely on where to find them. Maybe as fields added to Template:Infobox song? Ah, I see that was generally the consensus in that discussion (minus the lead/lede/introduction tangent). But that would also imply an expectation that all of those types of articles include an infobox. It could be a standard section after the lead? Schazjmd (talk) 20:45, 22 October 2021 (UTC)
- How about a "Folk Song Index" section above See also/References? Gråbergs Gråa Sång (talk) 21:11, 22 October 2021 (UTC)
- I like that idea! (Properly sentence-cased, of course) Schazjmd (talk) 21:11, 22 October 2021 (UTC)
- But of course. Gråbergs Gråa Sång (talk) 21:37, 22 October 2021 (UTC)
- I like that idea! (Properly sentence-cased, of course) Schazjmd (talk) 21:11, 22 October 2021 (UTC)
- How about a "Folk Song Index" section above See also/References? Gråbergs Gråa Sång (talk) 21:11, 22 October 2021 (UTC)
- Roud numbers at least (not sure about the Childs nos.) are akin to the opus numbers (or other recognized cataloging numbers like BMV or Köchel numbers) we see in the lede -- usually the introductory phrase -- of most classical music articles, e.g. The Symphony No. 6 in B minor, Op. 74...; or The Requiem in D minor, K. 626... or Fugue in G minor, BWV 578.... They have recognized academic status as identifiers to unambiguously identify a particular work. They ought to be in the lede.
- This is nothing like IMDB numbers, which are just the database numbers used by a particular vendor to navigate its database, and whose stability Wikipedia takes advantage of for linking. Roud numbers, like opus, BMV and Köchel numbers, are stable, academically recognized and not tied to a vendor who might arbitrarily change them in the future (as Turner Classic Movies did some time ago, breaking Wikipedia templating in the process). TJRC (talk) 23:44, 22 October 2021 (UTC)
- Interesting. Opus numbers has been around longer though, and Bach-Werke-Verzeichnis/Köchel catalogue is more limited in scope, like opus numbers for one musician at a time. Roud seems more like a limited ISBN or ASIN, catching them all, or trying to. Gråbergs Gråa Sång (talk) 08:02, 23 October 2021 (UTC)
- Identifying Roud and similar catalogue numbers is important, but the way they are presented in many WP articles is awkward and not encyclopedic. A while back, I remember several articles that were little more than "'Ballad X'" is Roud 12345" followed by lyrics that were italicized and formatted with a series of <br /> and : (rather than something like {{Poemquote}}), with a link to the Roud index as the sole ref. It reminded me of the mass creations of superstubs in the early days that seemed to ignore WP:NSONGS "Notability aside, a standalone article is appropriate only when there is enough material to warrant a reasonably detailed article; articles unlikely ever to grow beyond stubs should be merged to articles about an artist or album."
- Interesting. Opus numbers has been around longer though, and Bach-Werke-Verzeichnis/Köchel catalogue is more limited in scope, like opus numbers for one musician at a time. Roud seems more like a limited ISBN or ASIN, catching them all, or trying to. Gråbergs Gråa Sång (talk) 08:02, 23 October 2021 (UTC)
- The catalogue number should be worked in to articles with some explanation or context; a reader who is only interested in one song may not know what a Roud number is or why it is significant. Also, Roud and other ID numbers can be added to Template:Infobox musical composition using the
|catalogue=
parameter. This template also has several other useful fields more applicable to traditional songs than Template:Infobox song, which is more geared towards commericial recordings. An example follows.
- The catalogue number should be worked in to articles with some explanation or context; a reader who is only interested in one song may not know what a Roud number is or why it is significant. Also, Roud and other ID numbers can be added to Template:Infobox musical composition using the
- Maybe compare it to how the classification scheme for folk tales is done. My thought is that the Roud song index should be in the info box with a link to the official Roud site in the external links as there are 60,000 codes and we only have 3 or 400 hundred.
- Aarne–Thompson–Uther Index is a global classification scheme for fairy tales and covers most of the globe. Introduced in 1900, with last update printed in 2004, it has 2100 codes, and has been used in 8000 ish academic papers. The wiki article Cinderella for instance has ATU of 510A "persecuted heroine" and this is mentioned in the lede and the infobox. The ATU wiki page is active this year, and the corresponding List of fairy tales is quite active and uses ATU, but only covers a very small number of the stories. A casual reader would be able to find similar stories and realises the connectedness of folk tales
- Roud_Folk_Song_Index is a pre-WW2 folk English language (UK and US) song identification scheme, which references to variants and recordings.Introduced in 1900, with updates still occurring online, it has 60,000 possible indices and 250,000 references, and has been used in 1800ish academic papers. The Roud wiki page is not active the year, but the list has had a lot of work on this, but there are only 300 to 600 folk songs . A casual reader would find it of limited use - it's just an id. Copying SMeeds (talk) and AlexanderHovanec (talk)) who seem to be the people who have done lots of work and may be able to explain things better Wakelamp d[@-@]b (talk) 07:57, 1 November 2021 (UTC)
Extended content
| ||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
{{Infobox musical composition | name = | type = | composer = | image = | image_size = | alt = | border = | caption = | translation = | native_name = | native_name_lang = | other_name = | catalogue = | genre = | written = | text = | language = | composed = | published = | publisher = | first_recording = | misc = }} Subset parameters
|
- One of the problems with folk songs and ballads is that they often exist in many variant versions, no one of which is definitive. Even the titles may vary; or, conversely, two or more songs may have identical titles, but entirely different content. The purpose and great strength of the Child and Roud lists is that they bring together all variants under a single authoritative identifier, and perhaps save the student a lot of wasted time and effort in chasing "ghosts". So to my mind this is important information that should be included both in the lead, and in the infobox if there is one. On the other hand, it's a technical reference which won't immediately mean much to many general readers, so shouldn't be given undue prominence, or made the basis of the definition. Putting the reference in brackets after the title (e.g. Whiskey in the Jar), or briefly stating it at the end of either the lead paragraph or lead section (e.g. Ring a Ring o' Roses, Holmfirth Anthem), are all entirely acceptable ways of achieving what we want – getting core information out there, succinctly, without overcomplicating the issue. But making it the sole content of the definition (e.g. Robin Hood and the Bishop of Hereford, admittedly a stub) is bad practice, and to be deprecated. GrindtXX (talk) 13:13, 24 October 2021 (UTC)
- It seems to me that Roud/Child nr is not "core information," that's stuff like year/country of origin, composer and what Donald Trump tweeted about it. If "Whiskey in the Jar" was a chemical element like lead, Roud is not atomic number, it's electron configuration. However, I think it would be helpful that included Roud numbers has a cite with link included, like.[1] Not necessary if this is under External links, of course. I'm not pushing infoboxes beacause I don't want to "force" an article to have an infobox just to have a place to put the Roud nr. Gråbergs Gråa Sång (talk) 08:41, 25 October 2021 (UTC)
- One of the problems with folk songs and ballads is that they often exist in many variant versions, no one of which is definitive. Even the titles may vary; or, conversely, two or more songs may have identical titles, but entirely different content. The purpose and great strength of the Child and Roud lists is that they bring together all variants under a single authoritative identifier, and perhaps save the student a lot of wasted time and effort in chasing "ghosts". So to my mind this is important information that should be included both in the lead, and in the infobox if there is one. On the other hand, it's a technical reference which won't immediately mean much to many general readers, so shouldn't be given undue prominence, or made the basis of the definition. Putting the reference in brackets after the title (e.g. Whiskey in the Jar), or briefly stating it at the end of either the lead paragraph or lead section (e.g. Ring a Ring o' Roses, Holmfirth Anthem), are all entirely acceptable ways of achieving what we want – getting core information out there, succinctly, without overcomplicating the issue. But making it the sole content of the definition (e.g. Robin Hood and the Bishop of Hereford, admittedly a stub) is bad practice, and to be deprecated. GrindtXX (talk) 13:13, 24 October 2021 (UTC)
References
- ^ "Whiskey in the Jar, Roud No 533". Roud Folk Song Index. Retrieved 25 October 2021.
Refining the rollback right
At present, the rollback user right is predominantly given by administrators to trusted editors (typically counter vandalism patrollers) and allows access to more complex and powerful tools, the most notable tools being rollback links (the namesake of the right), Huggle, and some features in RedWarn.
As time has progressed, the rollback feature and right has been used for more than just rollback links (i.e. a link that's clicked and immediately reverts an edit without summary) and as such the current rollback policy is outdated, confusing and conflicts with what many people use it for (i.e. through Huggle or other tools with edit summaries). Because of the lack of clarity, many times WP:ROLLBACKUSE has been brought up in discussions, even when it's usually not applicable to use of rollback with edit summaries (i.e. by using the rollback link).
We should clarify this - as such I think renaming the right on the English Wikipedia to "Tool Confirmed" or similar (thinking along the lines of auto-confirmed, extended-confirmed, tool confirmed, but I welcome proposals for any better name) that clearly describes what the role does, why the person has it and how they can be held accountable for their use.
The rollback link policy should then be merged into the wider reversion policy as no matter where, tool use without edit summaries should only be appropriate in certain situations and autoconfirmed users can replicate the use of rollback links by using a tool such as Twinkle or RedWarn. This means there won't be fragmentation between what essentially in both cases is just reverting an edit.
A sample of what my idea of a summary would look like:
Tool confirmed editors are editors approved by an administrator to have access to more powerful semi-automated tools. A tool confirmed user has access to features in tools such as X, Y and Z, also access to rollback links.
This could also supersede the AutoWikiBrowser request page, if technically possible.
We should also consider creating a tag named something along the lines of "Tool edit" that's required for all automated tools, userscripts etc on the English Wikipedia where appropriate and possible to increase accountability and other benefits such as analytics.
What do people think? ✨ Ed talk! ✨ 23:49, 24 October 2021 (UTC)
- It makes sense to me. I requested Rollback because I'd heard intriguing stuff about Huggle, then had Rollback removed when I discovered that I didn't like using Huggle. I couldn't see any other use for having Rollback, because Twinkle provides rollback/vandalism links in both article history view and diff view, and enables rollback of multiple edits (sequential by a single editor) in one click. So the Rollback permission isn't really associated to rollback functionality. Schazjmd (talk) 00:07, 25 October 2021 (UTC)
- I do not think of the rollback permission as mainly functional or technical—giving access to tools—rather as a matter of trust, in effect a licence to revert vandalism without an edit summary. IMO it doesn’t really matter how the revert is done, be it with a helper script, the rollback button, or simply null-editing an old version of the page (which last any user can do).—Odysseus1479 01:45, 25 October 2021 (UTC)
- What does the rollback right grant, other than the "[rollback]" link? --Redrose64 🌹 (talk) 07:01, 25 October 2021 (UTC)
- Permission to use Huggle, and nothing else. 🐔 Chicdat Bawk to me! 10:22, 25 October 2021 (UTC)
- That's incorrect. In fact the rollback right is required by quite a few tools for both technical and other reasons. "Tool confirmed" could also extend to providing trusted users with more powerful features and merge already complex permission systems together. ✨ Ed talk! ✨ 17:47, 25 October 2021 (UTC)
- Permission to use Huggle, and nothing else. 🐔 Chicdat Bawk to me! 10:22, 25 October 2021 (UTC)
- What does the rollback right grant, other than the "[rollback]" link? --Redrose64 🌹 (talk) 07:01, 25 October 2021 (UTC)
- I do not think of the rollback permission as mainly functional or technical—giving access to tools—rather as a matter of trust, in effect a licence to revert vandalism without an edit summary. IMO it doesn’t really matter how the revert is done, be it with a helper script, the rollback button, or simply null-editing an old version of the page (which last any user can do).—Odysseus1479 01:45, 25 October 2021 (UTC)
- Literally the only thing that membership in the
rollbackers
group does is confer access to therollback
permission. The rollback permission allows an user to request a server-side reversion. None of these other scripts or clients are official and are subject to change on the whims of their maintainers, or in some cases by any user that wants to fork or recompile them. — xaosflux Talk 17:56, 25 October 2021 (UTC)
- I agree that the current situation doesn't make as much sense as it should and that it should be improved. I specifically agree it's silly how we draw such a distinction between rollback and undo. Putting all of countervandalism behind a role might not be such a bad idea. Hear me out! Perhaps without the role, you could only improve existing edits by adding a citation, or rewording bad edits, but you wouldn't have an easy way in the software to undo them. Enterprisey (talk!) 08:30, 5 November 2021 (UTC)
Spoken wikipedia on the main page
This is probably more likely to apply to Today's featured article, Today's featured list, Today's featured picture, and possibly Did you know than it is to apply to In the news and On this day, but I think there should be a way for people to volunteer to add a narration to the blurbs that appear on the main page. This probably wouldn't work for ITN and OTD, because those are dynamically updated lists, but it might work for DYK, because the entire list is updated at the same time, and it could really work for TFA, TFP, and TFL, which have thousand-word blurbs. Maybe it's just the icon, instead of the whole playback box, but I think there should be a way to add spoken tracks to the main page. They'd have to be reviewed, of course, but I think it could be done. theleekycauldron (talk • contribs) (they/them) 01:44, 29 October 2021 (UTC)
- Is a spoken-word reading of material which, like a mayfly, lives only one day (or even less) really a good investment of editor time and effort? And think of the pronunciation disputes! EEng 18:31, 29 October 2021 (UTC)
- For sure. I don't think it'd be a requirement for every appearance, but if someone wants to make a recording, there should be a way to display it—and we have MOS:VAR for pronunciation disputes. The main page is seen by millions of people every day—it does seem worth the investment if someone wants to make it. theleekycauldron (talk • contribs) (they/them) 18:32, 29 October 2021 (UTC)
- Where will you find readers who can pronounce all those foreign names? —Kusma (talk) 18:44, 29 October 2021 (UTC)
- The IPA can help there, no? If not, it'll probably be weaker on that side, but again, we don't need to have one for every blurb. There should be the option, though. theleekycauldron (talk • contribs) (they/them) 18:46, 29 October 2021 (UTC)
- Personally I find IPA an eh-NIG-muh (or eh-nig-MUH if you're over 35 or went to private school) [5]. EEng 19:23, 29 October 2021 (UTC)
- I have a PhD in Linguistics and don't speak or read IPA. - Donald Albury 21:32, 29 October 2021 (UTC)
- Personally I find IPA an eh-NIG-muh (or eh-nig-MUH if you're over 35 or went to private school) [5]. EEng 19:23, 29 October 2021 (UTC)
- The IPA can help there, no? If not, it'll probably be weaker on that side, but again, we don't need to have one for every blurb. There should be the option, though. theleekycauldron (talk • contribs) (they/them) 18:46, 29 October 2021 (UTC)
@Theleekycauldron: it's a good idea, if the main page projects start with a few to test the waters, and see how it goes. It doesn't have to be an everyday thing, but selected by the individual projects. Let them decide on some trial runs. — Maile (talk) 23:52, 29 October 2021 (UTC)
- Maybe we start with TFA, then—see how it goes. I'll ping some of the coordinators over there. theleekycauldron (talk • contribs) (they/them) 23:55, 29 October 2021 (UTC)
- I consider this a great idea, especially for TFA and TFL. I doubt whether it would be easy enough to do this at DYK, as sometimes, we tweak a hook while it is on the main page. In rare conditions, the we swap a hook. It isn't as stable as TFA and TFL. I wonder, would we be able to change the recording simultaneously? – Kavyansh.Singh (talk) 17:17, 30 October 2021 (UTC)
- No, at that point we'd have to pull the recording—you're right, DYK is pretty unstable. theleekycauldron (talk • contribs) (they/them) 17:20, 30 October 2021 (UTC)
Truth seeking team
There's a lot of cloudy information in the news that can make its way onto Wikipedia. Perhaps someone is pronounced dead in absentia on CNN, but people in the real world see the man walking around and get clear tape and pictures of him. Maybe there are little white lies, unverified claims, and other such commodities on a page. Enter the "Truth Team."
Much like the Typo Team (who seek to find Typos and correct them), the Truth Team looks to find lies and false info in Wikipedia pages and correct them or even remove them. For instance, someone may say that "John Doe was assassinated at 3:30 AM" but in reality, John Doe was assassinated at 4:30 AM, as found by eyewitness testimony. If the exact time can't be found, say at "approximately 3:30 AM." Stuff like that.
It would clear up a lot of false evidence, unite Wikipedia users to find the truth and to make Wikipedia a better source of information in general. xdude (talk) 16:54, 29 October 2021 (UTC)
- In my role as the party pooper, i'd direct you to WP:Wikipedia is wrong and WP:Primary sourcing. theleekycauldron (talk • contribs) (they/them) 16:56, 29 October 2021 (UTC)
- May I also direct your attention to Wikipedia:Verifiability, not truth - Donald Albury 18:21, 29 October 2021 (UTC)
- Well, maybe as a sort of close approximation he could join the tooth team? EEng 18:33, 29 October 2021 (UTC)
- seems more like a candidate for the troupe team. theleekycauldron (talk • contribs) (they/them) 18:35, 29 October 2021 (UTC)
- I would say that we strive for “Accuracy” rather than “Truth”. The latter has connotations that come too close to “belief”. Blueboar (talk) 18:54, 29 October 2021 (UTC)
Listen, Xdude gamer, I made a little joke above, but I want you to know that I (and everyone else) recognize that you're trying to improve things. But if you read the links above you'll see that actually, trying to fix articles by direct evidence isn't a good idea. EEng 22:02, 29 October 2021 (UTC)
Module code TOC
![](https://web.archive.org/web/20211106083923im_/https://upload.wikimedia.org/wikipedia/commons/thumb/e/ea/Purple_arrow_down.svg/20px-Purple_arrow_down.svg.png)
I could use a "Module code TOC" to navigate a module page. Module code can be too long to keep overview.
For example, Module:String does have a ../doc TOC, all fine. But in development stage, or in code-research situation, I'd want to go to line "function p._main
". The /doc could have like "==code== ===_main===". -DePiep (talk) 20:34, 30 October 2021 (UTC)
- @DePiep: are you proposing that we adopt a certain coding style (can you show an example of what it would look like when complete) - or is this a feature request for pages in the Scrubinto content model? — xaosflux Talk 11:25, 1 November 2021 (UTC)
- I think it is a feature request. For example, Module:String has function name code
function str._match(...)
. - The TOC I propose then could find this, and create a TOC line equivalent to (the known effect of)
===str._match===
- Bottleneck IMO is the "find" part in this. I'm not sure this can be done by Lua module only.
- -Piep (talk) 11:34, 1 November 2021 (UTC)
- You are asking for anchors such as Module:String#L-140? Gonnym (talk) 11:49, 1 November 2021 (UTC)
- @Gonnym: great! line number could be part of the solution indeed. Especially when code is stable. Then TOC to create manually.
- More thoughts: 1. more automation; 2. read & show function name in the TOC; 3. adjust linenumber when code changes; 4. keep distinct from possible /documentation function name (=like when regular "===_match===" exists)'; 5. 'TOC-depth' options applied to public/local functions, top code section. -DePiep (talk) 13:27, 1 November 2021 (UTC)
- I imagine something like the documentation generated by Javadoc for Java code. isaacl (talk) 21:55, 1 November 2021 (UTC)
- You are asking for anchors such as Module:String#L-140? Gonnym (talk) 11:49, 1 November 2021 (UTC)
- I think it is a feature request. For example, Module:String has function name code
- @DePiep: moved here from VPR, since (a) not a ready to go proposal (b) doesn't seem like something we can directly do here? — xaosflux Talk 11:39, 1 November 2021 (UTC)
What links here
![](https://web.archive.org/web/20211106083923im_/https://upload.wikimedia.org/wikipedia/commons/thumb/e/ea/Purple_arrow_down.svg/20px-Purple_arrow_down.svg.png)
I would like to propose that "what links here" feature to sort articles alphabetically. I think that would be more useful. Bada Kaji (talk • श्रीमान् गम्भीर) 19:41, 2 November 2021 (UTC)
- So do I, though ideally by namespace then title. Meanwhile, you may be interested in User:GhostInTheMachine/SortWhatLinksHere. Certes (talk) 20:22, 2 November 2021 (UTC)
- Bless you, Certes, for pointing me to this. I have been craving something like this (be it the OP's dream version or Ghost's slight-daly-while it sorts version). Thanks to GhostInTheMachine, too, of course. — JohnFromPinckney (talk / edits) 23:44, 2 November 2021 (UTC)
- Yes, add this as a sorting option. Not the default, though. BD2412 T 20:25, 2 November 2021 (UTC)
- @Bada Kaji: I moved this from VPR, as it isn't a ready to go proposal. This isn't an option we can set here on the English Wikipedia. You could perhaps maybe build a very inefficient javascript for it, but that would be in userscript land for at least a while. If you would like this added as a new software feature for the software please see: Wikipedia:Bug reports and feature requests for how to submit a software feature request on the mediawiki software that runs Wikipedia. — xaosflux Talk 23:25, 2 November 2021 (UTC)
- My biggest ask for What links here is to filter out "things that link here only because of a transcluded template". For example, every park in NYC links to every other park in NYC because they all include {{Protected areas of New York City}}, but that's almost never what I'm looking for. I vaguely remember there was a phab ticket for this which was basically closed as "infeasible" due to the way things are parsed. But I still want it. -- RoySmith (talk) 23:32, 2 November 2021 (UTC)
- Exactly. Trawling through an unsorted (very random-looking) list of articles which, when you go there, don't seem to have the link anyway, until you expand all the navboxes at the bottom... well, it can be quite maddening. — JohnFromPinckney (talk / edits) 23:41, 2 November 2021 (UTC)
- Maybe you might find User:V111P/js/What Links Here link filter useful. But see also Wikipedia:Village pump (technical)/Archive 132#What Links Here generating too many listings due to navboxes. Thincat (talk) 22:26, 5 November 2021 (UTC)
- My biggest ask for What links here is to filter out "things that link here only because of a transcluded template". For example, every park in NYC links to every other park in NYC because they all include {{Protected areas of New York City}}, but that's almost never what I'm looking for. I vaguely remember there was a phab ticket for this which was basically closed as "infeasible" due to the way things are parsed. But I still want it. -- RoySmith (talk) 23:32, 2 November 2021 (UTC)
Report :Do proposal discussion Editors match Active Editor Demographics
The Wikipedia:Wikipedianscommunity proposal consensus is only true consensus if it does not have involvement of all demographics in terms of years experience,and diversity. Not as a percentage, but least as a voice to explain different views. The Wikipedia foundation board is thinking of expanding to reflect diversity, but we don't know whether our consensus is being dominated by one group
A way of doing this would be to have reports that
- Compare between proposal editors and Active Editors. the years of active editing, diversity, and editor type. (Diversity could be based on user page templates or Wikipedia:WikiProject_Accessibility membership).
- Compare the voting percentage (Abstain, Support, Oppose) by years of active editing and editor type
- Confirm whether a few voices dominate proposals. Consensus does not means the majority always wins, but you would expect that the majority would be in line with consensus most of the time
Why should we care? There is more editing to be done than has already been done; there are huge numbers of maintenance tags, article tags, open Talk topics especially non replied, various error reports, missing content, procedure readability improvements, missing or questionable references, missing or incorrect categories,mentors needed,stubs, low project membership, article quality etc,... Wakelamp d[@-@]b (talk) 07:47, 6 November 2021 (UTC)