Wikipedia:Village pump (proposals)

 Policy Technical Proposals Idea lab WMF Miscellaneous 

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

New proposals are discussed here. Before submitting:

• Check to see whether your proposal is already described at Perennial proposals. You may also wish to search the FAQ.
• This page is for concrete, actionable proposals. Consider developing earlier-stage proposals at Village pump (idea lab).
• Proposed policy changes belong at Village pump (policy).
• Proposed WikiProjects or task forces may be submitted at Wikipedia:WikiProject Council/Proposals.
• Proposed new wikis belong at meta:Proposals for new projects.
• Proposed new articles belong at Wikipedia:Requested articles.
• Discussions or proposals which warrant the attention or involvement of the Wikimedia Foundation belong at Wikipedia:Village pump (WMF).
• Software changes which have consensus should be filed at Phabricator.

Discussions are automatically archived after remaining inactive for nine days.

« Archives, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212

Pending-changes protection of Today's featured article

The idea of automatically applying WP:Semiprotection to WP:Today's featured article (TFA) has been thrashed to death; see WP:PERENNIAL#Protecting Today's Featured Article on the Main Page. I think the most recent discussion was in July 2020, at WT:Today's featured article/Archive 14#Question about protection. The key argument (for me at any rate) is that the last thing we want to do is discourage new editors.

Nevertheless, any time an article with a hint of controversy is TFA, it turns up at WP:ANI. For some recent examples, see Wikipedia:Administrators' noticeboard/IncidentArchive1057#The Holocaust in Slovakia—TFA subject to ongoing vandalism (27 January 2021), Wikipedia:Administrators' noticeboard/IncidentArchive1057#Guadeloupe amazon—Today's TFA subject to ongoing vandalism (28 January 2021), Wikipedia:Administrators' noticeboard/IncidentArchive1057#Pyramid of Nyuserre —Today's TFA subject to persistent vandalism (30 January 2021) and WP:ANI#Hitler's prophecy—TFA undergoing ongoing vandalism (30 January 2021). They may also get posted at WP:AIV, which is less easy to search. I haven't seen reports of problems with WP:DYK articles.

During that last discussion, I suggested that TFA might be WP:Pending changes-protected for only so long as it is on the Main Page; and this idea seems to be new. IPs and unconfirmed editors would be able to post, even if their contributions didn't display immediately; and vandalism could be speedily dispatched where it belongs. A TFA's godaprents could be encouraged to help the regular pending changes patrollers. It would also solve the problem of working out when the vandalism occurred and who did it (a perennial problem with the small amount of vandalism I deal with - mostly involving links to DAB pages - which can be buried behind several recent good edits and require copy&paste from the last good version). It shouldn't be technically difficult to implement; it could be part of the script which adds articles to and removes them from the Main Page. Some other editors seem to like the idea, and it was suggested that I open a discussion here. (For the record - I'm in favour.) Narky Blert (talk) 10:36, 2 February 2021 (UTC)

I'll note that one of the primary reasons for rejections of auto semi on TFA in the past is giving the impression that Wikipedia isn't so free to edit by having our most visible page be uneditable to the majority of the audience of TFA. Pending changes protection avoids this by still allowing users to make the edit, even if there is a slight delay in publishing it live, so it would be a decent compromise between unfettered access and maximum accessibility for our most visible page, and avoiding wasting of the community's time and potential risk of bad content being put up for all to see. Regards, User:TheDragonFire300. (Contact me | Contributions). 11:01, 2 February 2021 (UTC)

As a further thought - the protection template text should be tailormade for TFA, and welcoming. Narky Blert (talk) 13:35, 2 February 2021 (UTC)

And another - WP:ANI#Pacific swift - Today's TFA receive high level of IP Vandalism (4 February 2021). Narky Blert (talk) 08:59, 5 February 2021 (UTC)

Another one, for which protection was declined - WP:ANI#Cheadle Hulme - Today's TFA receive high level of IP Vandalism (5 February 2021). Narky Blert (talk) 11:42, 7 February 2021 (UTC)

This is every edit made to Cheadle Hulme during its day at TFA. I see a grand total of two IP vandals, one of whom made two edits and one of whom made one, while every other IP edit is constructive. If any admin had protected it under those circumstances, then unless there's something I've missed they'd have been instantly hauled off to Arbcom for admin abuse. ‑ Iridescent 12:37, 7 February 2021 (UTC)

I am not an admin, and have no comment on whether any particular page should or should not have been protected. Nevertheless, I have felt it right to post here every relevant ANI post since I opened this discussion, no matter whether they help or harm my proposal. All evidence is important towards reaching a consensus. (I have not monitored WP:AIV or WP:RFPP, which would be much more difficult.) Narky Blert (talk) 20:10, 19 February 2021 (UTC)

Another one - WP:ANI#Apollo 14 - Today's TFA receive high level of IP vandalism and unsourced content (8 February 2021). Narky Blert (talk) 12:53, 9 February 2021 (UTC)

And another - WP:ANI#Bernard A. Maguire - Today's TFA suffered persistent vandalism (11 February 2021). Narky Blert (talk) 00:23, 12 February 2021 (UTC)

WP:ANI#Vandalism on TFA Grant Memorial coinage (12 February 2021). Narky Blert (talk) 05:42, 13 February 2021 (UTC)

Another - WP:ANI#Vandalism on Saturn (magazine) (14 February 2021). Narky Blert (talk) 07:30, 14 February 2021 (UTC)

Today's installment - WP:ANI#Vandalism on Silesian Wars (15 February 2021). When this one got reported to ANI, it had only a couple of hours left on the main page. Cluebot and something like six registered editors had already dealt with edits to it; add in the protecting admin, and that's a lot of work.

I spotted something I hadn't noticed before - User:TFA Protector Bot had stuck {{pp-move}} on the article on 26 January, with automatic expiry at midnight 15 February. Narky Blert (talk) 17:34, 16 February 2021 (UTC)

@Narky Blert: For several years now, it has been normal for all upcoming TFAs (which don't already have move protection) to be given a short-term move prot, set to expire the moment the article stops being TFA. Until November 2013 it was a manual process; since then it has been tasked to TFA Protector Bot, see the BRFA. This prot is usually applied some time in advance (I believe soon after the calendar slot has been approved), see the bot's log; and the use of {{pp-move}} is supplementary to that. --Redrose64 🌹 (talk) 22:33, 16 February 2021 (UTC)

I didn't know of that procedure (which was why I mentioned it), but it strikes me as an excellent one. Even a good-faith move of a reviewed and queued article would be disruptive, in the greater scheme of things. We don't want redirects on the main page.

Also - WP:ANI#Vandalism on Meteorological history of Hurricane Dorian (19 February 2021). Narky Blert (talk) 20:00, 19 February 2021 (UTC)

And another two - WP:ANI#Vandalism on SS Mauna Loa (19 February 2021) and WP:ANI#Multiple IPs adding porn or offensive image in supposedly TFA article (20 February 2021). On the second one, I count (among other reversions) seven WP:REVDELs while the article was on the main page. Narky Blert (talk) 00:53, 21 February 2021 (UTC)

Another - WP:ANI#Vandalism on Margaret (singer) (23 February 2021; also mentioned by Levivich, below), which illustrates a problem. An admin WP:SILVERLOCKed the page for 3 days (which IMO is too much in both duration and level); a non-admin closer thought that the problem should have been posted at WP:RFPP not ANI. Narky Blert (talk) 21:49, 24 February 2021 (UTC)

@Narky Blert: It was I who closed that thread; in that case, there was already a duplicate at WP:RFPP, so having an ANI thread seemed redundant. Regards, User:TheDragonFire300. (Contact me | Contributions). 04:57, 2 March 2021 (UTC)

PC is widely agreed not to work/be practical on highly edited pages. That's why no one suggests it, probably. --Izno (talk) 15:44, 2 February 2021 (UTC)

This is something that was proven during the PC backdoor attempt, by the Barack Obama and George W. Bush articles. The volume of edits was so high that the queue on those articles were perennially backlogged, and so it was and still is agreed that PC is not suitable for pages that would see high volumes of edits, something which would happen with TFA as a matter of course. —A little blue Bori v^_^v ^{Takes a strong man to deny...} 16:43, 2 February 2021 (UTC)

At least the TFAs as considered in this discussion. I have seen some fairly quiet TFAs of late. --Izno (talk) 17:17, 2 February 2021 (UTC)

Unfortunately I suspect the set of TFAs that would be suitably quiet for PC to work is almost identical to the set of TFAs where PC is not needed. Thryduulf (talk) 01:30, 3 February 2021 (UTC)

As a supporter of the proposal, and an active PCR, I don't think this would be insurmountable for most TFAs. I note that the two given examples are highly political topics, one of whom was at the time President of the United States and the other of whom had been less than a full term ago. Vaticidalprophet (talk) 06:04, 3 February 2021 (UTC)

I'm against preemptive protection of any kind, especially pending changes which makes more work for volunteers and is rarely useful. — Wug·a·po·des​ 01:53, 3 February 2021 (UTC)

• Support some sort of protection it's unacceptable to have editors (very predictably!) vandalizing articles like The Holocaust in Slovakia while they're on the main page. This diminishes our reputation much more than any protection would do. (t · c) buidhe 04:16, 3 February 2021 (UTC)
• I should note that "highly active" is a fairly low boundary - PCR starts having real issues well before, say, editors would be frequently edit-conflicting. I'm also not sure how accurate a prediction of a TFA's likely edit rate we can do. Obviously we can predict the "very active" and "not likely to be edited much" buckets, but there'd be a large middle category that is tough to order. As such, I continue to believe PCR remains distinctly problematic for any TFA use that would actually warrant PCR. Nosebagbear (talk) 07:35, 3 February 2021 (UTC)
• To be honest, I'm not totally against this. The utility of pending changes is different from semi-protection: the goal is not to reduce the workload for administrators and recent change patrollers (as others have correctly pointed out above, pending changes effectively does the opposite), but rather, to prevent the vandalism from being seen by readers and thus possibly bringing the project into disrepute. As a matter of fact, if my memory serves me right, for a brief period a few years ago, we actually did start preemptively putting PC protection on the TFA after an WP:AN discussion because of a particularly nasty LTA that would replace images on TFAs with extremely shocking ones. The traditional problems with PC on highly edited pages don't seem to be a big concern here because the protection would only last for one day, and most TFAs don't seem to be edited so frequently that large backlogs might become a problem. At the very least, I would probably support a trial period for this idea. Mz7 (talk) 07:53, 3 February 2021 (UTC)

  Thinking about this a little further, I think the relevant question would be how frequently vandalism remains undetected for longer than, say, a minute while on the TFA. Pending changes works best on articles where vandalism has a hard time getting reverted quickly, and now that I think about it, because the TFA already has a lot of eyes on it in general, it may be the case that most vandalism is already reverted within seconds, rendering the need for pending changes moot. Mz7 (talk) 08:02, 3 February 2021 (UTC)

So far as I can tell, TFA vandalism is rarely if ever reverted within seconds. I've heard averages around 10-15 minutes, which is enough to be problematic for those articles, especially with heavy or grotesque vandalism. Vaticidalprophet (talk) 11:39, 3 February 2021 (UTC)

Looking at the edit histories of the articles I mentioned in the opening paragraph (a very small statistical sample): if ClueBot spots it, within seconds; if a human, 1-40 minutes (median 8). Narky Blert (talk) 14:24, 3 February 2021 (UTC)

I think the relevant question would be how frequently vandalism remains undetected for longer than, say, a minute while on the TFA This is the crucial question for me. If our goal is to prevent disruption to the reader, we need to know how much vandalism is currently disrupting readers. I don't think a trial would change our ability to look at that, we just need someone to crunch the numbers from the page history. It could also be crossed with the day's page hits to estimate the number of readers who actually saw the vandalism, e.g., this vandalism was up for 5 minutes, so with a total of 60 thousand page views that day---5 thousand people probably saw this instead of the article. In my experience it's as you describe with the added eyes bringing faster reversions, but knowing the average and other statistics would be useful and quite possibly change my mind. If PC protection really is a substantial benefit, I'd be willing to support its use on TFA. It at least allows users without accounts to edit, and if we craft a nice message as Narky suggests, it could minimize the potential newbie biting. So my main concern in this case is usefulness because I've rarely seen PC solve more problems than it caused. — Wug·a·po·des​ 20:13, 3 February 2021 (UTC)

(I guess the maths in your example isn't meant to be taken literally but 5 mins of 60,000 pageviews is about 209 views—5000 would be 60,000 per hour. A limitation of this approach is that vandalism lasts longer the less-viewed the page is, and the pageviews vary based on time zones in areas where most of our readers are from. But I think some API somewhere can give you hourly pageviews.) — Bilorv (talk) 20:43, 4 February 2021 (UTC)

• Support pending changes protection as a trial: TFA is different from other highly-viewed pages in that there is very likely a highly active editor who is passionate about the article (the one who just promoted it to FA), and activity is limited to 24 hours (maybe the next few days as well, while it's still linked on the main page). I can't really speak for all such editors (I've only been that person once) but perhaps some would be able to look through all the changes at the very least after the dust has settled, and collect any of the changes which are productive. A counterargument is that TFAs are, well, featured articles, and so much less likely to have issues that new and unregistered users will be able to solve. But there are likely still small prose improvements that one might expect to be made in the TFA period. An alternative is maybe pre-emptively semi-protecting only some TFAs based on topic. — Bilorv (talk) 20:43, 4 February 2021 (UTC)
• Support Few useful changes are made, vandalism is a serious problem not for the number of vandals but for the black eye we get for allowing it on the front page. Volume of changes we are talking about is not great. So this is an excellent use of PC. Hawkeye7 (discuss) 04:33, 5 February 2021 (UTC)
• Support test, in my view, this is the ideal balance between preserving both the integrity of our "anyone can edit" mantra, and quality of our featured articles. I tend to agree with Hawkeye that "few useful changes are made"—to the point where the amount of vandalism or unhelpful edits vastly outnumbers the occasional random spelling fix (and any major edits/errors would better be discussed on the talk page anyways). But either way, this protection could address both the vandalism and occasionally helpful edits. Aza24 (talk) 09:07, 5 February 2021 (UTC)
  • Would just like to reaffirm my support; when my Portrait of a Musician went to TFA a couple of days ago, there was a lot of vandalism, almost all of which could have been prevented by PC; the other proses fixes and edits were by users who would not have been restricted by PC. A test is certainly worth it. Aza24 (talk) 23:37, 1 March 2021 (UTC)
• (nom) I agree that any change should be on a trial basis.

My idea for the template text is along the lines of: "Welcome to Today's featured article. It is an unfortunate fact that these articles attract vandals. Therefore, changes by new editors are reviewed by experienced editors before they show up for everyone to see. This usually takes only a few minutes. If you are here to be part of the community whose goal is to improve Wikipedia - Thank you, and happy editing! You might find Help:Editing a useful beginners' guide. If you are here only to disrupt - Be off with you!" Narky Blert (talk) 18:04, 5 February 2021 (UTC)

• Support Would prevent instant publication of toxic or copyrighted material on such highly visited articles. ~ HAL333 02:58, 6 February 2021 (UTC)
• Upon notice I haven't explicitly said it here yet, support as one of the parties in the original conversation. Vaticidalprophet (talk) 13:53, 6 February 2021 (UTC)
• Oppose. PC protection generates significant amount of work for minimal benefit, and doesn't work on pages with a high volume of editing; those pages which attract enough vandalism to make protection worthwhile, are also going to be those pages on which PC won't work. PC also comes with significant additional drawbacks in addition to the maintenance backlog it generates; there's no way to add a summary when rejecting an edit, so when disapproving a good-faith but inappropriate edit one can't explain why in the edit summary; for BLP issues PC has little impact since it doesn't affect what Google scrapes (they pick up the most recent revision even if it's been unapproved); to approve/disapprove changes to an article at FA level often requires specialist knowledge of the topic, which the handful of people who work the Special:PendingChanges queue are unlikely to have; the whole proposal appears to be predicated on the idea that every, or at least most, FAs are going to be on peoples' watchlists, which just isn't the case. ‑ Iridescent 08:00, 7 February 2021 (UTC)

(Interesting to see you here -- I was literally just wondering what you would think of the proposition.) For what it's worth, "there's no way to add a summary when rejecting an edit, so when disapproving a good-faith but inappropriate edit one can't explain why in the edit summary" is incorrect -- the edit summary box is...huh, PC queue is empty as we speak, so my plan to take a screenshot for "right here" will have to wait, but it's prominently placed complete with giant "ADD AN EDIT SUMMARY IF WHAT YOU'RE REVERTING ISN'T BLATANT VANDALISM" bold text. Vaticidalprophet (talk) 06:37, 8 February 2021 (UTC)

In lieu of screenshotting the edit summary box itself, here's a screenshot from my recent PCR contributions showing them. Vaticidalprophet (talk) 06:42, 8 February 2021 (UTC)

It's alarming as a standalone thing to learn that Google is scraping the most recent revision on PCP pages—albeit I understood it to have a bit of a delay for general vandalism reasons (possibly it's just that it takes a few minutes for Google to scrape the article again). However, this is Google's problem, not ours. I don't want them having a single say in any of our decisions in any way. They need to change to suit us, not the other way around. Other search engines are available. — Bilorv (talk) 00:56, 13 February 2021 (UTC)

In around 2014, I was told that Google scraped a well-known social website every 20 minutes. Our goal as volunteer moderators was to take down heavy commercial spam within that time. IDK if that's a typical number, but our leader tried and failed to get them to increase it to 30. Narky Blert (talk) 06:05, 13 February 2021 (UTC)

• Support The only backlog this would generate in terms of PC is one page which would need to be checked regularly for the duration (and, well, PC isn't that large of a backlog, compared to some other places). And, well, while it might not stop everything, at least any vandalistic edit (which would need to be reverted anyway, PC or not) won't be prominently displayed to every person who gets there... RandomCanadian (talk / contribs) 01:21, 12 February 2021 (UTC)
• Support as a pending-changes reviewer, I think the added workload would be minimal compared to the benefits. With the number of eyes on the page, good changes will be accepted quickly, while bad ones won't be prominently linked from the main page. Elliot321 (talk | contribs) 05:57, 14 February 2021 (UTC)
• Support - This strikes me as a reasonable compromise that both allows IPs to edit TFAs and combats vandalism. Put another way, it's the least restrictive means of effectively protecting our most prominent articles. (Protecting one article a day certainly won't overwhelm PC, and the article will be widely watchlisted anyway.) Let's at least give this scheme a try: I think it would be effective. Extraordinary Writ (talk) 07:18, 14 February 2021 (UTC)
• (nom) To repeat a point I made earlier, in case it might get lost. This idea has both carrot and stick. It would provide a means to speedily welcome new editors and to point them in the right directions. Experienced editors know where to look or to ask for help; newbies, by definition, don't. Narky Blert (talk) 07:50, 14 February 2021 (UTC)
• Support. This is about balancing Wikipedia's positive reputation as the encyclopedia 'anyone can edit', agains the potential negative reputation for vandalism and inaccuracies being introduced and not being caught. I think pending-changes protection is an excellent compromise, allowing people to edit, but preventing stupid nonsense from showing up on our most prominent article of the day. ƒirefly ( t · c ) 14:02, 14 February 2021 (UTC)
  • Changing to strong oppose after learning that FlaggedRevs (i.e. the module behind pending changes) is effectively abandoned, with nobody on the MediaWiki dev team having responsibility for it, and that it has various odd bugs (phab:T275322, phab:T275017) that seemingly aren't getting resolved. The last thing we want is weird MediaWiki bugs on display on TFA. ƒirefly ( t · c ) 09:42, 2 March 2021 (UTC)
• Support as trial. I agree with RandomCanadian, a short period of time having pending changes protection isn't going to create much of a backlog and with so many eyes on TFA already.
  I've been wondering though (which is also why I'm voting here, for the first time at Village Pump), where would one go to suggest an entirely-new protection type? Because what if, instead of putting pending changes on TFA, there was instead a "delayed changes" protection? It would basically be pending changes, but with automatic approval after some set amount of time. Set the time to something a bit longer than the average length it's taken for vandalism to be reverted, and I'd bet it would drastically reduce the workload on heavy-vandalism days without contributing to a pending changes backlog. This would be protection for just TFA (a case of a page with high visibility/traffic for a short period of time). But since this would need to be created and isn't a matter of assigning an existing protection, it's more of a future-implementation kind of suggestion. --Pinchme123 (talk) 01:27, 18 February 2021 (UTC)

  @Pinchme123: Feature requests should be submitted at Phabricator. --Redrose64 🌹 (talk) 14:26, 18 February 2021 (UTC)

@Redrose64: Thank you for pointing me to Phabricator, as I did not know about it beforehand. But my question isn't about "submitting" for a feature request, but suggesting one, to be discussed before requesting it. I see little point in requesting a brand-spankin' new feature without having a discussion about its merits, especially when feature requests appear to be made outside of WP (since Phabulator is merely an affiliated Wikimedia entity, and I had to create a new account just to see the feature request creation page). So where might I go to suggest the feature, for discussion? --Pinchme123 (talk) 16:16, 18 February 2021 (UTC)

• Oppose, per Iridescent, and also because of the general principle of using TFA to highlight the "anyone can edit" principle. Editing with a delay is not the same thing as editing with an immediate impact, and using PC on such a prominent page would harm recruitment. --Yair rand (talk) 06:25, 18 February 2021 (UTC)
• Comment. I'm in the knee-jerk oppose camp per Iridescent & Yair rand, but I can see the problem of adapting our processes as we evolve from "the encyclopedia that anyone can edit" to "the encyclopedia that anyone can edit". If we want to use the TFA for recruitment then the edit must be visible (almost) immediately; more than 30s would, I think, take away that little burst of joy that addicted me to editing here. I believe one of the major problems that Wikipedia faces as it matures is erosion of the editor base as it becomes harder and harder to get a foot in the door as a new editor. Is there any evidence that newbies take up editing via the TFA? I've seen multiple new editors converted via ITN, but rarely other sections of the main page. My DYKs have occasionally had substantive or typo-fixing edits by redlinks or IPs who are clearly interested in the topic, and I've even very occasionally had comments of thanks or opprobrium from IPs. Personally I hate pending changes and hardly ever accept significant revisions because it puts the onus on me to put my weight behind them, and I know others have the same problem. Is there some bot-mediated mechanism that could immediately auto-accept anything that looked good faith and not obviously problematic? Could Cluebot be set to run on every TFA edit immediately? Espresso Addict (talk) 12:19, 18 February 2021 (UTC)
• Strongly support trying something after seeing today's TFA, Margaret (singer), be ceaselessly vandalized, spawning yet another ANI thread. I think the work combatting vandalism at TFAs is greater than the work involved in protecting the article, and the detriment from vandalism at TFAs is greater than the detriment of not allowing everyone to instantly edit it. I'm not sure if it's PC or semi or something else, but Something Must Be Done™. I'd support a trial of this or any type of protection. Or even a trial of both PC and semi, like an A/B test. What I strongly oppose is resigning ourselves to "well that's just how it is" and trying nothing. Strongly oppose trying nothing. Levivich ^harass/_hound 22:44, 23 February 2021 (UTC)
• Support As it's a featured article, the cost-benefit calculation here is very different than most. Newcomers vandalize many articles but also make good edits to make up for that. For featured articles, the likelihood of a newcomer making a good, constructive edit is greatly reduced (based on low potential for improvement of the article as a featured one), while the fact that it is on the front page means that the amount of vandalism increases substantially. Personally, I support pending changes protection on all featured articles because of the diminished benefit a newcomer editing them can bring. Plenty of featured articles have significantly declined in quality over time as for a pretty much completed article it is much more likely for any edit by newcomers to be unhelpful. I think people are overestimating that satisfaction people get to see their change immediately; as long as they are aware of the pending changes situation, what matters most is the fact that your work is published at all, not the exact time. Zoozaz1 talk 22:00, 24 February 2021 (UTC)

• Support Most recent TFAs follow a pattern similar to Oryzomys gorgasi, the TFA on 26 February. There were 62 edits on that day. Of those, 27 (roughly 44%) were from IP addresses. With one exception that I saw, all were vandalism. The ‘legitimate’ edits were almost all from established users, and most of those edits involved repairing the damage caused by vandals. The 'anyone-can-edit' principle is, at least from my perspective, the means and should be subordinate to the end of creating an encyclopedia by drawing upon collective knowledge.Venicescapes (talk) 08:07, 28 February 2021 (UTC)
• Support; generally works fine on dewiki on highly-viewed pages. All articles are PC-protected on dewiki, leading to a long backlog,^{(7500 pending changes)} but this doesn't seem to be a concern when the protection is limited to 24 hours and done selectively on pages that are very likely to be reviewed. ~ ToBeFree (talk) 19:52, 1 March 2021 (UTC)

• Support per proposal; yes anyone should be able to edit, which they will be in this proposal, but Wikipedia should not once again fall victim to being called the website with inaccuracies anyone can add. waddie96 ★ (talk) 20:44, 1 March 2021 (UTC)
• Support I seem to remember suggesting this a few years ago in one of the periodic conversations about semi-protecting TFA. It would prevent vandalism from being shown live while still allowing IP editing. It can get complicated with many edits, but it's worth a trial at any rate. ~ ONUnicorn^{(Talk|Contribs)}problem solving 20:57, 1 March 2021 (UTC)
• I don't like protection much, and I loathe pending changes. But perhaps it is time to move away from TFA as our poster child for "anyone can edit" and look for other ways to draw new editors in (TFA really isn't the best place for editing tests). I'd prefer semi to PC, but I won't oppose the proposal. —Kusma (t·c) 21:25, 1 March 2021 (UTC)
• Support. Prefer semi over PC. My first concern is with the reader, THEN the editor. Dennis Brown - 2¢ 22:26, 1 March 2021 (UTC)
• Support because it will best serve readers. TFA gets a lot of traffic, and it's a disservice to readers if they encounter the article while vandalism is visible. Schazjmd (talk) 22:34, 1 March 2021 (UTC)
• Begrudging support I think PC is bad, broken, and not hardly useful. I think semi is superior in effectively all cases. However, since we have decided against semi protection for FA's for years, I will settle with PC. I think it is ludicrous that we don't want to use semi on FA's. We have become stuck in a rules trap. It is written that protection should only be used after disruption is happening, so we wait until we see disruption. But we now hew to the letter of the law, not the spirit! The spirit is to prevent disruption. Sure, for most articles, we can't know when disruption will occur. But featured articles get vandalized like clockwork. Suddenly showing a page to 10 million extra people a day (including our LTAs!) makes it almost certain to get vandalized. A little dash of WP:IAR would save a great deal of time and frustration. TLDR: semi-protection would be superior, but I will settle for PC. CaptainEek ^{Edits Ho Cap'n!}⚓ 22:55, 1 March 2021 (UTC)

• Comment and Oppose. The ability to instant create edits lies at the heart of this project. And I think many editor's very first foray into Wikipedia is through attempts at a main edit. It's our welcome mat. Seeing your change live instantly provides a thrill and confirms to newcomers instantly that their suspicions of how things work ("anybody can edit") were valid. Yes, this includes lots of immature vandal edits. But those tends to be stupid things like profanity inserted very visibly. But even this helps prompt good new edits to try to fix that. In other words, the "wild west" nature of the front page is an important way to engage new users which we need. And even some of the vandals turn into good editors as they mature. Using PC makes your first edits boring and people tend not to stick around for boring things. Jason Quinn (talk) 02:22, 2 March 2021 (UTC)

• Oppose using PC for TFA, but support just about any other form of protection (semi, extended confirmed, etc). IMO, Pending Changes is an abomination that shouldn't be used at all, on any article, ever. It increases the amount of work needlessly and makes editors responsible for someone else' edits. It is also extremely confusing and difficult to understand as a feature, which for prospective TFA use makes it unacceptable. TFA will be viewed and edited by many inexperienced editors. We should not confuse the hell out of them with a monstrosity like Pending Changes. Just semi-protect TFA instead. Nsk92 (talk) 02:47, 2 March 2021 (UTC)

• Oppose. It looks as if the wind is blowing strongly in one direction on this one, but my take is that this seems like an overkill solution in search of a problem--which proposed problem I think is substantially minimal by the very nature of the context. If TFAs somewhat gain additional traffice from IPs and neophyte editors, they are also gain increased scrutiny from a broad chunk of the community over the same 24 hour period that they are up on the main page. Additionally, these articles tend to be slightly more robust and reflective of engaged editing than the average article. In short, I can't see that adding pending changes will make all that substantial a difference to protecting the article against erroneous, bad-faith, or otherwise problematic edits over the course of its day of featured status. Meanwhile, it seems absolutely certain to have an impact on what has traditionally been one of the Featured Article's perceived benefits: editor recruitment.

In fact, I would argue that having this process by which new editors are corralled into a different article each day (which article represents decent editorial standards) and that space becoming the first place in which they can appreciate the pleasure of contributing to the encyclopedia and enjoying the immediate feeling of seeing those changes go live, while simultaneously focusing community oversight on that article space in an organic fashion all sounds like precisely the balance we want to establish and why this is a "it's not broke, don't fix it" type of situation. Additionally, TFAs frequently benefit from the super-crowd-sourcing feature of the attention they get: the vast majority of even IP edits are beneficial, not deleterious, so why gum up the works by having a queue of (potentially overlapping) pending edits. And those queues don't always get addressed with alacrity: I'm a pending changes reviewer myself and have logged a fair amount of work in the area over recent years, and I can tell you that the queue being addressed is quite a variable thing with regard to both the subject matter of the article and the time of day. Lastly, this is just not the typical scenario (a n inherently problematic article) in which this form of protection is typically seen as most justified.

In short, and with respect to the proposers and supporters above, the cost-benefit analysis runs strongly counter to the proposal, as I see it. Snow ^{let's rap} 04:28, 2 March 2021 (UTC)

It's equally likely that a new editor attempting to edit a well established article (the TFA) will mess something up and get a notice or worse a stern scary looking warning on the their talk page. This also ignores that PC still allows the new editors to edit the article, while conveniently removing the incentive for at least some vandalism/LTAs. Ex. (had to dig for it a bit): TFA from May last year - the revdel'd edits are a well known LTA vandal (the same as for most of the articles on Wikipedia:Today's featured article/May 2020... - I know, this discussion is not a new idea); after the PC protection the LTA stopped, while still allowing other non-AC edits (a fair bit were common vandalism, some were good faith but misguided; but anyway PC does not create the same issues as semi-protecting would). While PC might not be effective on some articles, or maybe even on most, on the TFA it would probably be the best of both worlds: not showing vandalism to our readers, while still allowing new editors to intervene. Cheers, RandomCanadian (talk / contribs) 05:06, 2 March 2021 (UTC)

• Support. Using PC on TFA is not actually a new idea. Myself and other admins have preemptively applied it many times to guard against long-term abuse, or in reaction to vandalism. These were emergency situations that necessitated stepping outside policy and lack of consensus, but each time to my knowledge we didn't see any of the problems others frequently cite when questing use of PC on TFA. Specifically, there are enough eyes and activity on this article that it doesn't have much detriment on the backlog. Shielding one of the most visible articles from defacement, while still allowing new users to contribute, is a win-win and frankly long overdue. — MusikAnimal ^talk 02:09, 3 March 2021 (UTC)

Pending changes bugs

(Moving discussion out of archive.) Unfortunately, there are a ton of bugs affecting pending changes right now: autoconfirmed users having their edits held back for review erroneously, administrators not being able to set pending changes protection, and apparently the pending changes codebase is completely abandoned (no active developers who understand the code). See also the discussion at Wikipedia:Village pump (technical)#Pending Changes again. I wrote a comment in the discussion above expressing some support for this idea, but if we can't ensure the software reliability of pending changes, I'm afraid I'm now quite hesitant to support expanding it. Mz7 (talk) 20:06, 12 March 2021 (UTC)

As one of the more active PCRs and one of the original formulators of the proposal, I've been following the PC bug discourse with some interest. I haven't actually observed a ton of it in my work, so I wonder quite how big the problem is -- that is, what proportion of people who get false-positived go complain somewhere. (It's not just on VPs, I've seen it come up a few times at PERM.) If the majority of people who get it bring it up, it's pretty small; if this is just the tip of the iceberg, it might be a serious issue.

In terms of the proposal -- the thing that stands out to me is most of the oppose votes are saying "We shouldn't do this, we should semiprotect instead", not "We shouldn't do this, TFA should be status quo". I wonder how the vote distributions would stand out for the following two proposals:

"TFA goes under either pending changes or no protection, only two options, semiprotection is completely off the table and will never happen"

"TFA either goes under semi, PC, or none; rank the options by preference"

It'd be an interesting RfC either way. Vaticidalprophet (talk) 03:11, 13 March 2021 (UTC)

I go through PC occasionally (when I'm not otherwise busy) and I don't encounter the bug too often (non-improvements and vandalism are, alas, more frequent), or when I do it's usually not too much of a hassle to just accept the change - if I have a particularly large amount of spare time I'll leave a notice about this to the affected editor. The fact that the code is unmaintained and apparently some form of a mess isn't a good sign, however, and well we'd have to weight whether the risk of further issues developping is less of an annoyance than potentially doing something about TFA LTAs... Cheers, RandomCanadian (talk / contribs) 05:42, 13 March 2021 (UTC)

Update: I'm aware there's now a BRFA intended to address this issue. If that goes through, then I think I'll withdraw my hesitation. Mz7 (talk) 19:06, 15 March 2021 (UTC)

I think hotfixes via a bot are untenable long-term. As more and more bugs crop up that can't be fixed in the extension, we can't keep using bots and other hacks to get around them locally. I'd personally like to see FlaggedRevs get picked up in the code stewardship review before expanding its use. ProcrastinatingReader (talk) 23:26, 15 March 2021 (UTC)

• Why don't we scrap pending changes reviewer? and grant the rights to extended-confirmed? I mean, the point of PC is a basic flag to stop vandalism/obvious DE, right? It's akin to a less extreme semi-protection. So it doesn't really make sense, imo, to put it to a separate right. The right should be bundled into extendedconfirmed and the separate right removed I think (+ EC is a reasonable level of 'trust' against pure nonsense vandalism + they can already edit areas like Israel-Palestine). That would also fix this particular bug. ProcrastinatingReader (talk) 17:17, 16 March 2021 (UTC)

  ProcrastinatingReader, honestly I'm not opposed. Back in 2018 I floated this idea on the idea lab village pump, and there was some mixed reception, and my interest in pursuing the change kind of just fizzled out. (Looking back I really wrote too much, heh. Concision was not/is not my strong suit.) Mz7 (talk) 19:42, 21 March 2021 (UTC)

  @Mz7: afaics there's two main objections there:

  1. That people would get a permission enabled without knowing what it does / directly asking for it. But this is true with autoconfirmed too: you get a bunch of buttons automatically and probably won't use many of them. And for adminship, too. You don't need to use a button just because you have access to it.

  2. That editors would game the system / increase edit count to gain it. I don't think this is true for a few reasons. 1) same could be said about the 'privilege' to edit in ARBPIA. 2) same could be said for pages currently autoconfirmed/ECP protected (gaming edits up means that you can now accept {{edit protected}} requests onto the page, and PCR is practically equivalent to {{edit protected}} except that it's directly supported by the software interface).

  So neither objection really makes sense imo. ProcrastinatingReader (talk) 19:55, 21 March 2021 (UTC)

  While we're at it why not scrap pending changes altogether? Is there any reason other than sunken costs to keep it, rather than use other forms of protection? I've often noticed that experiments on Wikipedia are hardly ever deemed to be failed, often because people are following that fallacy. Phil Bridger (talk) 19:53, 21 March 2021 (UTC)

  The main advantage of Pending Changes from my point of view, is that it acts as a form of protection against vandalism on pages without blocking their ability of IP's to actually edit. Asartea ^{Talk | Contribs} 07:54, 25 March 2021 (UTC)

RFC: Should certain succession box content, be deleted

Should we delete succession box content such as the following? Examples: "Oldest-living British prime minister". What say all of you? GoodDay (talk) 21:57, 4 February 2021 (UTC)

Survey (succession boxes)

• Yes. We ought not to have any of that trivia in succession boxes. There are often many boxes, dozens even, and additional clutter is unhelpful. One would be very hard-pressed to find reliable sources discussing how James Callaghan "succeeded" Alex Douglas-Home as the oldest British prime minister, and I dare say that no biography of either of them mentions it. Surtsicna (talk) 22:15, 4 February 2021 (UTC)
• Yes. Only posts with transitions (or successions) would need succession boxes. -- Kautilya3 (talk) 22:56, 4 February 2021 (UTC)
• Not here. Trivia should be deleted, non-trivia should not be. Whether any specific succession is or is not trivia needs discussing individually at an appropriate forum - that forum is very much not here. Thryduulf (talk) 02:21, 5 February 2021 (UTC)
• No As Thryduulf says trivial ones should be, non-trivial ones should not be. That needs to be done on a case by case basis. -DJSasso (talk) 18:03, 5 February 2021 (UTC)
• Yes I find succession boxes unnecessary in general since they usually duplicate the infobox, a navbox, and/or the text. When it's not duplicative like this example, it's often trivial that doesn't need a box to itself. Reywas92^Talk 19:36, 5 February 2021 (UTC)
  • You dislike them, whereas I find the clear, consistent presentation extremely useful. Succession boxes, infoboxes, navboxes and prose all serve different functions and so the same link appearing in more than one place is a Good Thing. Some succession boxes should be removed, but most should not be. Which is the case for any given succession box can only be determined by a consensus discussion about the succession box in question. Thryduulf (talk) 01:04, 6 February 2021 (UTC)
• Yes and would support getting rid of them all. Redundant and pointless clutter. Why not have separate boxes for marriages, spouses, and works dispersed througout the article while we're at it? ~ HAL333 02:50, 6 February 2021 (UTC)
  • What are they redundant to? Links in prose and links in succession boxes have different purposes so they are not redundant to each other. Boxes (for anything) scattered throughout the prose would be completely disruptive to the prose, which is why succession boxes are not placed in the middle of the prose? Thryduulf (talk) 11:58, 6 February 2021 (UTC)
• Yes - Not sure we even need them for positions, but definitely not for superlatives. Levivich ^harass/_hound 17:22, 6 February 2021 (UTC)
  • All superlatives? What about tallest buildings and longest bridges? Those seem extremely useful succession boxes to me. Thryduulf (talk) 18:24, 6 February 2021 (UTC)
• Comment: I feel the validity or triviality of succession boxes should be discussed on a case by case basis on their respective talk page. Not sure what a global decision one way or another on "trivial" succession boxes could achieve, when it does nothing to establish what is trivial and what is not. PraiseVivec (talk) 14:01, 7 February 2021 (UTC)
• Yes for purely trivia succession box such as "oldest living X" - unless said role is notable in its own right (not sure if this actually applies anywhere). Elliot321 (talk | contribs) 05:54, 14 February 2021 (UTC)
• Yes certain ones should be deleted. At a minimum, the spirit of WP:NAVBOX #4 seems applicable: There should be a Wikipedia article on the subject of the succession box.—Bagumba (talk) 10:11, 22 February 2021 (UTC)
• Depends - Yes for "Oldest-living British prime minister" "successions" as trivia, but without a discussion on others is will depend on the case, this should not be used as consensus to delete non-cited examples. KylieTastic (talk) 10:15, 23 February 2021 (UTC)
• Yes Succession boxes are not needed, in most cases it's a duplicate.Sea Ane (talk) 21:43, 26 February 2021 (UTC)
• Oppose deciding this specific one here, support a general rethink of succession/navboxes and what they are good for (if anything). Most of the succession boxes at John Major are less trivial, but there are so many of them that they are not a useful way to navigate anything, and are hidden away by default (just like the ungodly amount of navboxes). Once an article has more than three succession boxes, they tend to stop being useful. —Kusma (t·c) 21:50, 1 March 2021 (UTC)
• Yes - A lot of succession boxes seem to be pure trivia. If there isn't an article for the topic it shouldn't exist as a succession box. Nosferattus (talk) 04:05, 9 March 2021 (UTC)
• ? Oppose Confused, how can this RFC decide anything? The RFC is vague and the answers equally ambiguous; everyone saying "Yes" is supporting something different. Who decides which ones are trivial? Aza24 (talk) 23:34, 15 March 2021 (UTC)

Discussion (succession boxes)

• Why? Ivanvector's squirrel (^trees/_nuts) 22:17, 4 February 2021 (UTC)

  Because, those are not political or party offices. They're merely trivial. GoodDay (talk) 22:19, 4 February 2021 (UTC)

  Why do we need a global policy or guideline or anything about this? If you think a given succession box is trivia, discuss that specific succession box somewhere (talk page, WikiProject, TfD, there are plenty of venues). I genuinely don't understand what you are trying to achieve here - there is no chance that there will be a consensus that we should have succession boxes for everything (10th place qualifier in a British Touring Car Championship race as an unquestionable example), but with the wording of the discussion you cannot get consensus (for or against) regarding any individual example. Thryduulf (talk) 02:18, 5 February 2021 (UTC)

  You expect there to be a separate discussion on each individual bio article? Anyways, I don't exactly know what you're posting about. GoodDay (talk) 17:29, 5 February 2021 (UTC)

  I expect you to get consensus for each individual succession. That could be on an article talk page or a centralised discussion. e.g. if you want to get rid of oldest living British Prime Minister, you need to have a discussion specifically about removing the "oldest living British Prime Minister" succession box from every article it appears on, with notifications to (at least) the talk page of all the affected articles. In other words you need to do exactly the same as you would for any other bulk change. Thryduulf (talk) 22:26, 5 February 2021 (UTC)

  That's too time consuming. There's too many 'useless' topics in these succession boxes, to go through one-by-one. GoodDay (talk) 22:30, 5 February 2021 (UTC)

  A centralised discussion about the succ boxes for oldest living British Prime Minister could be held at Wikipedia talk:WikiProject Politics of the United Kingdom. --Redrose64 🌹 (talk) 23:40, 5 February 2021 (UTC)

  But the British prime ministers example, is just one example of these trivial topics in suc-boxes. GoodDay (talk) 23:41, 5 February 2021 (UTC)

  The issue is that we have no idea what other succession boxes you regard as trivial so we (and most importantly editors of the articles they appear on) have no way of knowing whether we agree or disagree. While oldest living British Prime Minister doesn't seem very useful, others such as Prime Minister of the United Kingdom is very much not trivial, where to draw the line between them needs to be determined by consensus. Thryduulf (talk) 00:59, 6 February 2021 (UTC)

  Subjects like "Tallest President of country", "Fattest Prime Minister of country", "Oldest Stock Car racer", "Youngest 100 meter dasher", would be trivial topics for suc-boxes. GoodDay (talk) 02:18, 6 February 2021 (UTC)

  The first two undoubtedly would be trivial, the latter two maybe - it would depend if there was any significance given to this in reliable sources. However a short list of topics (which a cursory search suggests are not in use) that you consider trivial do not go any way towards addressing the issue in my comment. Thryduulf (talk) 18:20, 6 February 2021 (UTC)

  I don't know what your complaint is. Perhaps if you would put down examples of suc-box topics that you believe should & shouldn't be deleted, would help. GoodDay (talk) 18:26, 6 February 2021 (UTC)

  My point is that they need to be discussed individually or in small, closely related groups (e.g. perhaps succession boxes related to British Prime Ministers). No matter how many examples are listed here you cannot get consensus for anything not listed here, and the more examples you list here the greater the liklihood of a trainwreck. Thryduulf (talk) 00:10, 7 February 2021 (UTC)

  RFC is already in full swing. We'll see how it ends up. GoodDay (talk) 00:47, 7 February 2021 (UTC)

  Consensus that some but not all succession boxes should be removed, but no consensus for the removal of any one in particular - that will need further discussion. It's pretty much the only way it can end. Thryduulf (talk) 02:20, 7 February 2021 (UTC)

  Besides, I find Wiki-Projects tend to garner less attention. Was considering moving another RFC to Village Pump, for the same reason. GoodDay (talk) 23:55, 5 February 2021 (UTC)

  Your subjective opinion that something is "Too time consuming" does not grant you the right to ignore WP:CONSENSUS. If you propose a course of action on a WikiProject and advertise the discussion to the relevant talk pages but nobody opines after a reasonable time (~week) then you can go ahead and remove the succession boxes listed in the discussion. Thryduulf (talk) 00:59, 6 February 2021 (UTC)

  If you want to advertise this RFC on the related WikiProjects? then by all means, do so. GoodDay (talk) 02:18, 6 February 2021 (UTC)

  If this discussion was actually about specific succession boxes that would be appropriate, but as the discussion is actually a vaugie and unfocused attempt to get rid of an unspecified list of succession boxes you (or presumably any other editor) personally dislikes then it is impossible to know which projects and articles are relevant. On the other hand, if you did deign to list those boxes you deem trivia, then it would I suspect rapidly become a trainwreck due to the large number of disparate boxes editors will have different opinions about. Thryduulf (talk) 11:55, 6 February 2021 (UTC)

  Since my concerns grew out of the discussion at James Callaghan, one would link to Political WikiProjects. GoodDay (talk) 18:41, 6 February 2021 (UTC)

All should be delete as links spam due to duplication of links and undue because of overwhelming size. Never understood why we have giant boxes with very few links in them overwhelming the sections. It's definitely a point of contention for content editor that these undue boxes are spamed automatically without consideration of over linking or unwarranted linking.--Moxy 🍁 00:04, 6 February 2021 (UTC)

Very simply because many (not all) of the succession boxes are very useful for readers. Thryduulf (talk) 00:59, 6 February 2021 (UTC)

Not sure how duplication of lnks and overwhelm sections is good for readers. We have protocols for these 2 points....just ignored by template spammers.— Preceding unsigned comment added by Moxy (talk • contribs) 01:19, 6 February 2021 (UTC)

See my reply in the section above for why duplication is not a problem, and I disagree that the presentation is overwhelming. That some succession boxes are trivia does not mean every succession box is spam. Thryduulf (talk) 01:41, 6 February 2021 (UTC)

We will simply have to disagree. By placement practice alone indicates there very low value to the community. See also links dumped at the bottom of articles because they duplicate existing links and the format is not responsible anywhere else in the articles. It's horrible that these boxes are more prominent than the actual topic-specific navigation boxes. Odd they were not omitted from mobile view as load junk.--Moxy 🍁 02:01, 6 February 2021 (UTC)

How does placement indicate they are low value? Of course the format is not appropriate anywhere else in the article - see also links and links in succession boxes have a completely different purpose to links in the prose. You need to explain why the duplication is problematic. Thryduulf (talk) 11:55, 6 February 2021 (UTC)

It's why we have a guideline on this..... distracts readers from the links that are actually important Wikipedia:Overlink crisis. They are so overwhelming that editors hide them in collapsible templates Abraham Lincoln#External links. In many other cases the amount of them is simply overwhelming...mass link spam John A. Macdonald#External links.--Moxy 🍁 17:59, 6 February 2021 (UTC)

You've explained why you think having too many boxes is an issue, and reiterated your opinion that some boxes are low value (which basically nobody is disagreeing with). However you have not explained why having any boxes is problematic or why their position in the article indicates this. Thryduulf (talk) 20:24, 7 March 2021 (UTC)

As I said in my oppose above, who ever closes this is going to find themselves in an impossible situation. There is no uniformity in any of the "Yes" comments. Additionally what does "certain succession box content" even mean?—who decides which ones are being discussed? Aza24 (talk) 23:08, 24 March 2021 (UTC)

RFC: Citation Style 1 parameter naming convention

Should non-hyphenated parameters be fully removed from the CS1/2 family of templates? Primefac (talk) 02:14, 10 February 2021 (UTC)

Background (CS1)

In 2014 an RFC determined that all parameter names in Citation Style 1 templates should have an alias which is in lowercase that has separations with hyphens between English words, between (not within) acronyms, or between English words and acronyms. The documentation is to show this lowercase, hyphenated version as the one for "normal use". This meant, for example, that |access-date= would be shown as the preferred parameter while |accessdate= was shown as acceptable but discouraged from use. In the following years there have been various trends and discussions formally deprecating many of the non-hyphenated parameters, from a small handful (2019 example) to the current list which contains over 70 entries. Many of these are the non-hyphenated variants of the preferred/hyphenated versions, which are being removed to decrease the maintenance burden and increase the uniformity between templates (i.e. "ease of use"). Other changes have included having RefToolbar give the hyphenated params and setting AWB's genfixes to replace some parameters.

In October 2020, all non-hyphenated parameters were added to the "current list" linked above. In November 2020, a bot request was submitted ("Monkbot 18") to remove or replace these deprecated parameters so that they could be removed from the base module and simplify the template code further. While acknowledging that this task was largely cosmetic in nature, BAG and other venues (primarily templates for discussion) have historically sided with the idea of a "maintenance burden" as a valid reason for edits such as these; see for example Monkbot 17, which replaced (cosmetically) one parameter for a better-named value for ease of use.

The issue for Monkbot 18 arises from the number of edits it is/was required to make; a conservative estimate would put the number of edits it has made for this task over a two month period (Nov 2020-Jan 2021) at around 1 million edits; as discussed on the task's talk page, this has essentially removed all but five of the non-hyphenated parameters, but another 2-3 million edits taking up to four additional months will still be required to fully "clear out" these parameters. Additionally, those opposed to the bot also expressed concern that the relatively small-scale discussions to deprecate these parameters had not reached a wide enough audience to merit what they felt were sweeping changes.

Proposal (CS1)

There are three main options with regard to the CS1/2 family of templates, and by extension Monkbot 18's task.

• Option A: Non-hyphenated parameters should be deprecated and removed; the bot is free to continue its work.
• Option B ("status quo"): Non-hyphenated parameters are formally deprecated, but should not be immediately removed. Deprecation can be bundled into genfixes or performed along with other non-cosmetic changes, but (ignoring a possible Cosmetic Bot Day) should not be done on its own by a bot.
• Option C: Non-hyphenated parameters should not be deprecated; deprecation should not be continued and bot approval should be revoked. This will also mean that the deprecated parameter list will need to be updated to remove the non-hyphenated parameters.

Please note that this discussion is primarily about the CS1/2 template parameters and whether two full sets of parameters should be kept/maintained. It is not the place to re-litigate the various discussions about Monkbot 18's task; Option B is provided for those who feel the task should not proceed.

Survey (CS1)

• First choice A, second choice B. I'd be happy to see AWB's genfixes take on some of this burden, and I'd be happy to see this happen a little more slowly, but it should happen, even though it's occasionally inconvenient for me. Also, when any individual parameter reaches a sufficiently small state (e.g., potentially still thousands of uses, but not hundreds of thousands of articles), the template should be updated to disallow that particular parameter (not merely advise against it in the documentation), so that they won't keep creeping back in, because, hey, it still works, so why should I bother? WhatamIdoing (talk) 19:12, 10 February 2021 (UTC)

  @WhatamIdoing: AWB's genfixes already handles this through WP:AWB/RTP, so manual edits and other bots can help whittle down the list while making other (non-cosmetic) edits. GoingBatty (talk) 04:12, 11 February 2021 (UTC)

  GoingBatty, are you sure that |accessdate= will be replaced by |access-date= by editors using the current version of WP:AWB/RTP? I think it was removed. – Jonesey95 (talk) 04:33, 11 February 2021 (UTC)

  @Jonesey95: You're right! While the functionality exists (and other parameters are still replaced), editors have removed some hyphenation replacement rules. GoingBatty (talk) 04:59, 11 February 2021 (UTC)

• Option A. I support completing the nearly finished move away from unhyphenated multi-word parameters. See below for more details about this process, which is being questioned now by a very few editors after seven years of work, and when it is more than 90% complete. With any other template, it would have been the work of a few days to standardize on one style of parameter name and convert all of the transclusions. The only reason that the process for CS1/CS2 templates is different is that there are millions of transclusions instead of hundreds. – Jonesey95 (talk) 04:33, 11 February 2021 (UTC)
• Option C. This is pointless make work; see extended comment in discussion section. Espresso Addict (talk) 07:08, 11 February 2021 (UTC)
• Option C If it was only confined to a few thousand pages, it wouldn't be a big deal. But when it's upwards of 3 million pages, maybe it should be the other way round - IE no hyphen. Lots of pointless bot cloggage in going through millions of pages for a trivial change. Lugnuts ^{Fire Walk with Me} 08:20, 11 February 2021 (UTC)
• B if not C As per comments above; many editors are clearly quite happy with the unhyphenated forms, so why not allow both? Changing is pointless. Lots of other templates allow aliases as parameter names. I fail to see the problem. But we are where we are, so I favour B rather than C. Peter coxhead (talk) 09:04, 11 February 2021 (UTC)
• Option A (second choice B), per Jonesey95. Let's just get everything simplified, as it should be. Get on and finish the job. --NSH001 (talk) 10:18, 11 February 2021 (UTC)

  Update Add a qualification: I still very much hope that the bot will be allowed to resume, but subject to this condition: only if Monkbot 18 drops its removal of entirely blank lines within citation templates. For the reasoning behind this, see my conversation with Floydian in the discussion section (this is quite an easy change to make to the bot). While I'm here, I'll add a second qualification, also arising from the discussion: the list of articles on which the bot runs should be filtered to remove articles that have already been visited by the bot. This is in order to reduce the alleged problem of "bot spam" objected to by some editors (who says I don't listen to objections?). --NSH001 (talk) 06:52, 12 March 2021 (UTC)

• Option C. It is a totally pointless exercise. The unhyphenated ones are better in my eyes as they do not wrap in the edit window and can be underlined as a typo so are clearly visible in the edit window. Keith D (talk) 13:01, 11 February 2021 (UTC)
• C (first choice) or B (second choice). I've read many of the discussions about this issue and I've never yet seen anything the convinces me that deprecation actually benefits the encyclopaedia in any way. Even if we assume for the same of argument that it somehow does, the real and evident disruption caused by the bot so far and the extent of the changes the bot operator notes will be required to complete the task, very clearly and very significantly outweigh that benefit. Thryduulf (talk) 15:20, 11 February 2021 (UTC)
• Option A this change is a bit painful, but better for editing in the long-run. It's better to make this change than to not make it. The best time would've been fifteen years ago - but the second-best is now. Allow the bot to continue its operation, get rid of all the parameters, and once all are removed, start generating cite errors. Elliot321 (talk | contribs) 17:09, 11 February 2021 (UTC)
• Option A. Unfortunately, the visual editor exists... but isn't useful for large editing projects and thus many people still use wikitext editing - condensing to one form (specifically the hyphenated one as it is easier to read for editors) will help ensure consistency between articles at least in the CS1 templates. Obviously this won't make every article easier to edit as there are articles with non CS1 style citations, but it'll help the millions that do use CS1 look the same to editors instead of having a hodge-podge of hyphenated names. I further agree with the bot continuing to run now, and then running maybe once per week or so after this initial run to fix any CS1 non-hyphenated parameter names. -bɜ:ʳkənhɪmez (User/say hi!) 17:13, 11 February 2021 (UTC)
• Option C. The point of templates and bots is that they should work to make editors' lives easier, not that editors should change the way they do things to make template and bot creators' lives easier. This bot has it completely topsy-turvy, and if the bot-approvals group has approved this then that is a problem with that group, not with a few unhyphenated parameters. I can't help feeling that that group is looking at the interests of a few bot operators rather than of the many more editors and still many more readers. There's no great complexity in having a few synonyms for template parameters, and there's no problem at all with exporting data - if the synonymic parameters point to the same place in code then they can be exported in the same format with no extra effort. I can't believe that forty years after I went into IT we still expect users to be slaves to the systems that are supposed to help them. Phil Bridger (talk) 18:21, 11 February 2021 (UTC)

  The point of templates and bots is that they should work to make editors' lives easier. Exactly so. That's what this bot is for. It makes the parameter names easier to read (taking them as a whole, not just access-date on its own), reduces the size of the template documentation, makes the parameter names all consistent. All these combine to make learning how to use the cite templates easier. It also makes maintaining the templates easier. a win-win situation. The only reason we're having this discussion is that Mediawiki's watchlist software is so bad at handling bot edits. Otherwise it would be a no-brainer. Some people here seem to be under the mistaken apprehension that this is just to advance the interests of "a few bot operators". Well, I'm quite sure Trappist (the operator of this bot) could do without the stress of planning, writing and running this bot. He does a bloody fantastic job, and deserves a huge amount of credit and appreciation for his work. I can't believe that forty years after I went into IT we still expect users to be slaves to the systems that are supposed to help them. The whole reason for this bot is to make users' life easier. FWIW, I also have experience of IT work more than 40 years ago (seconded for 2 years to work on a (very successful) project - mathematical programming on big data for a life insurance company) and frequently had to liaise with IT people since then. I'm well aware of the problems that can arise when you just allow systems to get more and more complex, so I appreciate Trappist's (and many other people's) efforts to simplify matters. --NSH001 (talk) 23:37, 11 February 2021 (UTC)

  Clarifying: Re-reading this again this morning, it might appear to readers who don't read carefully that I'm agreeing with Phil Bridger. Nothing could be further from the truth, I still support Option A. Option C makes no sense, for all the reasons set out by SMcCandlish. --NSH001 (talk) 08:35, 12 February 2021 (UTC)

  And Phil Bridger's hyperbole stance is fallacious. Option B imposes nothing on anyone. "I don't like option A" does not equate to "only option C can work". Option B is the status quo, and it has broken nothing. I'm rather shocked at how stark obvious this is, yet at least 10 editors don't seem to have noticed. I know that we have a lot of populism running around in the world – a lot of "I would feel very strongly about this, if it were true, and it feels good to pretend its true, so I'm going to pretend it's true" behavior. But that stuff needs to be left at the door.  — SMcCandlish ☏ ¢ 😼  04:21, 12 February 2021 (UTC)

  No, my stance is not hyperbole or due to populism. The reason I reject option B is the word "immediately". It still leaves current consensus that unhyphenated versions of parameters will be removed, just not immediately, and it is that consensus that should change, however many years it has been stable for. I am happy with simplifying the documentation, but not with removal of the option. Phil Bridger (talk) 09:19, 12 February 2021 (UTC)

  As noted above: Deprecation is not synonymous with disabling. You're confusing replacement of the parameters as written in a template transcluded in a page, with removal of the runtogether parameter variant's ability to function in the template. Only option A would do the latter. We have many, many templates that support variant-spelling parameters but do not "advertise" them in the documentation, and it breaks nothing whatsoever to bot-replace them with the canonical version, just as the same bot will replaces calls to a redirect name for the template with the actual page name of the template. I.e., you're having a strong negative reaction to an argument that option B is not actually making.  — SMcCandlish ☏ ¢ 😼  17:50, 13 February 2021 (UTC)

  Option B very clearly says "but should not be immediately removed". Either the word "immediately" has some meaning, and this option would lead to removal, but just not immediately, or it has no meaning and has no business being there. Phil Bridger (talk) 18:10, 13 February 2021 (UTC)

  SMcCandlish, this was explained in the previous discussion at CS1: ""Deprecation", by the definition used in the context of CS1/CS2, means that if the parameter is used, a red maintenance message will be shown and it will appear in a tracking category. It is a phase before removal of support for a parameter (in which case only an "unsupported parameter" message and optionally a hint on the new parameter will be shown). It is possible to stay in this state for extended periods of time, but the idea is that eventually the functionality will go". So the intention is to error and then remove the functionality of these parameters, not simply not advertise them in the documentation. If you want to propose a variant of option C that removes the parameters from the documentation but still allows them to function, go ahead, but option B is not that. Nikkimaria (talk) 20:09, 13 February 2021 (UTC)

  What I want to propose is a variant of Option B that removes the parameters from the documentation before letting some bot go around changing the articles on my watchlist as if the parameters are somehow bad. If the params are really bad, and if there's a consensus to that effect, then Step One is to take them out of the documentation (either by dragging them away in the dark of night, silently, like ninjas, or by transparently mentioning them as being deprecated, so everybody knows what's going on). The first phase of real deprecation is telling people to stop using something. Then you can start throwing up red messages and it's not so damned rude or surprising. — JohnFromPinckney (talk) 21:40, 13 February 2021 (UTC)

  Sure, I would support that variant, too. Or a variant that kept mention of those versions of the parameters but deprecated them. Whatever gets us closer to having consistency in the actual deployed templates being used in citations.  — SMcCandlish ☏ ¢ 😼  04:15, 14 February 2021 (UTC)

  This stuff should be in a tracking category, if that's what bots and other tools are going to work with. If we don't like the error messages, then we can disable them. This is just template and module code, it's not etched forver into a mountain face like Mt. Rushmore.  — SMcCandlish ☏ ¢ 😼  04:15, 14 February 2021 (UTC)

• Option C I agree strongly with Phil Bridger here. In other words, I fail to see what exactly is accomplished by making millions of cosmetic edits and then deliberately breaking things that would otherwise have worked. * Pppery * _{it has begun...} 20:21, 11 February 2021 (UTC)

  Option B breaks nothing. You, et al., are providing an argument against option A, not an actual argument for C, and just ignoring B. Also, a !vote for C is an !vote for overturning a status quo that has been stable for years, in favor of chaos, yet without an actual rationale to do so. The closer should take that into account, per WP:NOTAVOTE. In the event of a "no consensus" result we still end up with the status quo. Things like this are why I keep telling people they need to write RfCs better. "Anything that can be misinterpreted will be."  — SMcCandlish ☏ ¢ 😼  04:21, 12 February 2021 (UTC)

  SMcCandlish, option B supports disabling of the non-hyphenated parameters - that is a breaking change. The difference between A and B is speed, not outcome. Your !vote below suggests you want the non-hyphenated variants to remain supported, but of the options provided only option C accomplishes that. Nikkimaria (talk) 16:18, 13 February 2021 (UTC)

  Nope. See my response to same claim by Phil Bridger, above.  — SMcCandlish ☏ ¢ 😼  17:50, 13 February 2021 (UTC)

  Yep - see above. You can also look at what has happened with previously deprecated CS1 parameters such as |authorfirst= - they don't work. Nikkimaria (talk) 20:09, 13 February 2021 (UTC)

  Which is perfectly fine, given enough time that people stop actually using them. Thus remove them from the template documentation and replace them in deployed template translcusions. Eventually the "monkey see, monkey do" effect of being exposed to deprecated parameter variants goes away, through decreased and eventually zero exposure. I'm mean, come on, that's what the point of deprecation is. It seems to me that you [plural] are coming from a "give me C or give me death" perspective, artificially conflating A and B because you just will not tolerate the idea of any parameter name variants ever going away for any reason. If I'm wrong about that perception, then please explain.  — SMcCandlish ☏ ¢ 😼  04:18, 14 February 2021 (UTC)

• Option C per Phil Bridger. Ealdgyth (talk) 21:28, 11 February 2021 (UTC)
• Option C - Commenting here probably falls under my doctors' lists of "things HF shouldn't do while he has a concussion", but I don't want to miss this discussion. While I understand that maintaining the citation templates is not a particularly easy job, in the end, rigidness in the citation template is not desirable. The citation templates should be easy to use, and having a couple aliases for the most common parameters makes it easier to use. And a trout to the bot guild for approving a bot task that was designed to deprecate template parameters with no consensus for that deprecation. Hog Farm _Talk 22:15, 11 February 2021 (UTC)
• First choice A, second choice B. Wikipedia source text is becoming increasingly complex and difficult to manage making it less accessible, except for the type of experienced editors participating here. Anything we can do to reduce complexity is a win, fewer options the better when plainly redundant. -- GreenC 22:41, 11 February 2021 (UTC)
• Option C per Phil Bridger, personally I prefer non-hyphenated parameters, and I find deprecating them to be an absolutely pointless exercise that breaks things for absolutely no reason other than to satisfy the cosmetic preferences of a few editors. Devonian Wombat (talk) 00:14, 12 February 2021 (UTC)
• Option C. Largely per Hog Farm. Keeping extremely commonly used aliases is helpful for editors using these templates. Nikkimaria (talk) 01:17, 12 February 2021 (UTC)
• Support option C, neutral on option B, strongly oppose option A. My fingers are used to typing many of these parameter names without hyphens. Deprecating them and the accompanying gnomework of replacing them with the un-deprecated versions is already causing me significant hassle, both in trying to remember that really now they should be hyphenated, and in trying to pick out the important changes on my watchlist among all the pointless gnomery. Removing the unhyphenated forms altogether would be worse. —David Eppstein (talk) 01:41, 12 February 2021 (UTC)
• Option C. I'm with Phil Bridger and Hog Farm. I don't know if this is bike shedding or yak shaving, but it's just not productive and a bad look. Alternatives exist because it's simpler than remembering which of two common possibilities are acceptable, rather than forcing one or the other (as the other options do). Discussing efficiency because of a off-row character on the other hand (oh, so we're making this choice based on everyone using QWERTY?) is the kind of ignore-practical-facts reasoning that yield platypus-shaped end results. -- Mikeblas (talk) 01:52, 12 February 2021 (UTC)
• A is better in the long run, but B for now, and have the conversion covered by AWB genfixes and similar. Headbomb {t · c · p · b} 02:29, 12 February 2021 (UTC)

  Comment Unfortunately, in this case, leaving it to AWB genfixes alone would be neither effective nor desirable. Because there are likely to be so many of these parms on any given page, the genfixes are likely to swamp the main, intended change, causing understandable annoyance. Much better to leave it to this excellent bot, which is clear and open about what it is doing – no nasty surprises. That's why I prefer option A to option B, but either of these is vastly preferable to option C. --NSH001 (talk) 08:31, 7 March 2021 (UTC)

• Option C per Phil Bridger. SarahSV ^(talk) 02:51, 12 February 2021 (UTC)
• Option C. This "everything must be hyphenated" approach doesn't work well with the text editor. For some common parameters like |accessdate=, it is simply better being unhyphenated because the source text is quicker and easier for a human to parse because the parameter doesn't word wrap in the editor's textbox. Jason Quinn (talk) 03:09, 12 February 2021 (UTC)
• Option B. We should stop listing the nonhyphenated ones in the documentation at very least, so that between editorial shift and AWB/bot genfixes cleanup, we get more consistent over time. It's too soon for option A, if ever, because the templates serve us, we don't serve the templates. It's perfectly fine for templates to quietly support non-hyphenated variants so they don't just break if people try them. But we should not continue listing those variants in the docs. It's antithetical to the purpose of templating, for us to perpetuate inconsistency (without good reason, like an ENGVAR color vs. colour distinction). And it pointlessly makes the documentation longer and more complex for no gain at all; no one looking for how to specify the Archive.org URL needs to know anything but |archive-date=, and telling them |archivedate= also works is just stuffing pointless trivia into their head. Yes, do continue converting to hyphenated versions in genfixes and other automated edits (when doing something more substantive at the same time). Finally, option C is rather pointless. We've regularly been (gradually) deprecating various old parameter names, and it has worked just fine. Option B will not break anything, and will have (already is having) positive results.  — SMcCandlish ☏ ¢ 😼  04:21, 12 February 2021 (UTC)
• Option B Option C. These are all valid aliases, as there is zero confusion between say |access-date= and |accessdate=. The status quo works fine: hyphenated is preferred because it's easier to read, but unhyphenated is acceptable because there is no ambiguity and evidently plenty of people type it that way. Cosmetic changes should adhere to policy. – Finnusertop (talk ⋅ contribs) 05:53, 12 February 2021 (UTC) Changed from B to C as I am opposed to the implications of the "formally deprecated" part of these valid aliases. – Finnusertop (talk ⋅ contribs) 09:26, 12 February 2021 (UTC)
• Option C first choice (B second). Editors should be allowed to use either hyphenated or non-hyphenated versions. Consistency is not better than flexibility here: the only people reading the parameter are editors, so let the editors decide whether or not they want to use a hyphen in their template parameters. I share the general concern about the disconnect between code writers and content writers, and the frustration with some template and bot maintainers imposing decisions on everyone without consensus. Fait accompli editing across millions of edits or transclusions is kind of a big deal. Levivich ^harass/_hound 07:20, 12 February 2021 (UTC)
• I Like C: I especially echo Levivich, Thryduulf|, and Phil Bridger's comments regarding these perennial, periodic, new surprises where editing articles is concerned. That is disruptive and we should just stop it. C, therefore, will bring the sanity back. GenQuest ^"scribble" 07:42, 12 February 2021 (UTC)
• Option C. These are templates for use by editors, they have no impact on readers, and it's a complete waste of time making rules and regulations about them and then writing bots to enforce those pointless rules. Not to mention that when AWB goes through "fixing" all these in an article, it can drown out the genuine edits and make it harder for people to track what's going on. Get rid of this ridiculous rule, delete it from AWB, and then maybe we can get on with actually building an encyclopedia.  — Amakuru (talk) 09:13, 12 February 2021 (UTC)
• Option C (which is the status quo ante). I just don't see what problem people are trying to fix, so follow WP:NBDF. The hyphenated parameters are useful and improve wikitext legibility, I personally prefer them, but allowing both forms makes things easier for editors. Deprecating them appears pointless, and removing them entirely seems actively harmful. It's created millions of pointless busywork edits clogging up watchlists for no good reason. We don't have a problem with template aliases, so why the concern about parameter aliases? None has been convincingly articulated. Modest Genius ^talk 12:10, 12 February 2021 (UTC)
• Since I have been pinged, neutral all round. I'm have nostalgia for the status quo for some reasons already presented (muscle memory, line breaks), but I recognize that once we pass through the valley of the shadow and emerge into the bright uplands yonder of a cleaner implementation, we'll have forgotten the pain. Mind you, I'll probably be 90 years old and beyond caring. But I have some implementation concerns in the Discussion. David Brooks (talk) 16:54, 12 February 2021 (UTC)
• Option C - Standardization of this sort may be useful to researchers or developers, but not to regular users. Adding citations should be as easy as possible. To that end, I want to minimize the chances that the interface will not know what I'm trying to do, or that I'll get an error because I entered an underscore instead of a hyphen. The rest of the internet is trying to increase flexibility to make user experience easy and intuitive. We do that too in many ways. But here we seem to be doing the opposite: removing flexibility and requiring users to know how we've worded/arranged things.
  I don't care if there's a bot that goes in an standardizes them afterwards or if someone runs AWB behind me. Again, I get the appeal of standardization. We should just be doing everything we can to make the experience intuitive for regular users. — Rhododendrites ^talk \\ 23:04, 12 February 2021 (UTC)
  • Alternative: Option D - let the bot and/or AWB standardize, but never disable the parameters. Standardization is good; degrading usability is bad. — Rhododendrites ^talk \\ 23:12, 12 February 2021 (UTC)

    I would certainly support your option D if it was on the table, but cannot support anything that says that this functionality should be removed, as option B (despite the illiterate claims above that it doesn't) clearly mandates. Phil Bridger (talk) 18:10, 13 February 2021 (UTC)

• Option A with B as my second choice. Maintaining our complex citation templates is not an easy task, and if the people who are putting in the work want to do this to make their jobs easier, I'm in favor of it. Legoktm (talk) 23:48, 12 February 2021 (UTC)
• Option C with Option B as second choice: Modest Genius makes an excellent comment, as do several others. The community needs to stop wasting its time on this citation formatting nonsense and do the hard labor of introducing citations in articles instead, the place we actually have a front-facing desperate crisis which damages our reputation. I use |accessdate= and I have no intention to stop. I remember very clearly the process I went through of learning citation templates in 2014—they were confusing at first but I never had a single confusion in parsing "accessdate" as the two word phrase "access date", very clearly meaning the date you last accessed a reference. I can see that in theory this would be marginally better for new users, and I can see that in practice some people who don't like the look of "accessdate" are escalating it ("my opinion" becomes "the correct view" becomes "a moral imperative to enforce"), but this just doesn't outweigh the actual genuine pain it will cause editors like me to retrain a years-long developed muscle memory; almost every mainspace edit I make for months will have a disrupted flow (interrupts my thought process, wastes my time re-typing) if this is to change.
  As for the bot that has been wasting several minutes of my time per day, I strongly opposed it in the first place and have had more than enough of it. (No, I can't ignore it, because if I turn bot edits off I will miss acts of vandalism, unconstructive changes or cleanup tagging that are hidden by a subsequent bot edit.) BRFAs need wider advertisement when their scope is so enormous and their violation of WP:COSMETICBOT so overt. I don't accept that option B reflects past consensus in practice in that I can't remember anyone ever approaching me about my use of |accessdate= or changing it to |access-date= manually. Past local consensus among technical groups who don't work in article space, sure. But enwiki community consensus? No. — Bilorv (talk) 01:07, 13 February 2021 (UTC)
  • @Bilorv: if I turn bot edits off I will miss acts of vandalism, unconstructive changes or cleanup tagging that are hidden by a subsequent bot edit -- this always takes me aback. Just curious, but why not enable the "Expand watchlist to show all changes, not just the most recent" + "Group changes by page in recent changes and watchlist" settings? I see no bot edits, and they don't cover anything up. — Rhododendrites ^talk \\ 02:50, 13 February 2021 (UTC)
    • @Rhododendrites: appreciate the suggestion! It's never occurred to me that combining those would produce that functionality (there are so many complex combinations of Watchlist filters and Preferences options), but I've enabled those two and kept "Latest revision" on and enabled "Human (not bot)" as the Preference option "Hide bot edits from the watchlist" didn't seem to do that and now I think (small sample size in my watchlist atm) I see the latest change only, including if it's a bot, but only if at least one non-bot edit has been made (which is exactly desirable). — Bilorv (talk) 11:34, 13 February 2021 (UTC)
• Option A with B as my second choice. Maintaining the complex citation templates is not an easy task. AManWithNoPlan (talk) 02:52, 13 February 2021 (UTC)
• Option A on the basis that standardization is good. I don't actually care whether the hyphens are there or not, but it's better one way or the other rather than having a mix. In general I'd rather see us migrate away from using wikitext for reference information, though, it's like doing your finances in a text document. Thanks. Mike Peel (talk) 18:26, 13 February 2021 (UTC)
• Option A I have always preferred the hyphenated form personally because it allowed the spell checker to verify that there was no typo, whereas the unhyphenated form is always flagged as a typo, although the preview now informs me of errors. I disagree with the contention that humans can parse |accessdate= more quickly than |access-date=; spaces were invented for this reason. I also know the overhead that permitting multiple parameters that mean the same thing in templates entails. I'm aware that this has been cluttering my watchlist with Monkbot edits, but my !vote is to continue. Hawkeye7 (discuss) 22:12, 13 February 2021 (UTC)

  Note that this is mostly an WP:ILIKEIT rationale and dismisses the views of those who indicate that they can parse "accessdate" without issue as wrong without any evidence. It also dismisses the very real disruption caused to others as necessary to reduce overheads, whereas this overhead, even if it is a significant as some claim (for which no convincing evidence has ever been presented) it's still trivial compared to the disruption caused by Monkbot's edits and by the unnecessary disruption to editors using the templates. Thryduulf (talk) 15:29, 14 February 2021 (UTC)

• B or C. I've no strong preference between accessdate or access-date, both seem useable, if one is formally preferred then fine. However, the massive ongoing bot spam for something that has literally no effect on readers, and barely any on editors, is unwarranted. In addition to being everywhere on the watchlist, Monkbot makes it much harder to disentangle various series of diffs in the edit history for little benefit. A user making an edit inserting "accessdate" isn't an egregious issue that should cause a bot to come running, and such a bot action then obscures the edit in question from watchlists. CMD (talk) 18:18, 14 February 2021 (UTC)
• Option C. Removing a common way to type parameters in templates reduces ease of use for the end users. Having a bot going around and "fixing" these has a negative impact on the readability of page history and watchlists . No one in this conversation has demonstrated that the maintenance burden on the template is so significant that it would justify all of these downsides. It also seems that the maintenance burden is not something that would be difficult to track like accidental blue links to primary topic articles when non-primary topic articles were intended, since the template code are on the template pages.---- Patar knight - ^chat/_{contributions} 18:53, 14 February 2021 (UTC)
• Option C. The previous discussions linked above (a six-year-old RFC with seven people participating in it, which specifically promised that nothing would be depreciated, followed by a handwavy argument about maintenance burden), are not remotely sufficient to justify such a sweeping change. Yes, maintenance burden is a pain, but it affects a relatively small handful of bot authors; removing the most widely-used version of a template parameter affects a huge swath of editors, who need to be given deference here on account of being a larger group. And obviously there is not a standing consensus that maintenance burden justifies such changes (at least not one on a scale necessary to justify this), or this discussion wouldn't be so clearly split. Therefore, the unhyphenated version should never have been depreciated, which makes the bot's edits pointless clutter at best and an attempt to push through a controversial change without sufficient consensus via fiat accompli at worst. Furthermore, if maintenance burden is the only concern, obviously the solution is to reverse the 2016 RFC (which, again, had only seven people participating it and agreed merely to create the hyphenated versions as alternatives) and remove the hyphenated version, which currently sees little use and would therefore be far easier and less disruptive to discard. --Aquillion (talk) 22:12, 15 February 2021 (UTC)
• Option B-ish. I think it's reasonable to have the hyphenated forms be the canonical version of the parameter, but I see no reason to make mass-edits to change from one form to another or to change the usual rules about cosmetic bots here. I see no harm in Option C, but implementing Option A will trade current problems for new ones. --AntiCompositeNumber (talk) 02:15, 16 February 2021 (UTC)
• Strong Support A: standardization for template parameters is important & useful.

Mild Support B: the # of |accessdate= per page is too damn high (much of the time), so much so as to interfere with checking regular WP:GenFixes (i.e. many single-screen diffs become many-screen diffs) — I would Strong Support B IIF Monkbot was allowed to continue & hyphenate at least this parameter.

Strong Oppose C: antithesis of A.

~ Tom.Reding (talk ⋅dgaf)  19:23, 16 February 2021 (UTC)

• To whom is standardization for template parameters important & useful? Phil Bridger (talk) 20:39, 16 February 2021 (UTC)
• How and why is standardisation more important and useful than editors being able to improve the encyclopaedia without needing to know the exact format of parameter names and deal with watchlists and page histories full of cosmetic edits? Thryduulf (talk) 20:45, 16 February 2021 (UTC)

• Are y'all suggesting we bring back all deprecated parameters, and adding more so that every user may choose to use the parameter names that are most comfortable/understandable/intuitive to them? I'm ok with soft-deprecation - allowing both but discouraging/converting one, in bulk once via bot, then gradually/passively via WP:GenFixes & other tools, but that is not one of the options.
  Re: "watchlists": may be configured to ignore bots until it's done.
  Re: "histories full of cosmetic edits": the bot only requires 1 edit; hardly "full of"; regardless, there is community consensus for WP:CBD.   ~ Tom.Reding (talk ⋅dgaf)  12:16, 17 February 2021 (UTC)

  Yes, I am suggesting that all the two-word parameters should accept hyphenated and non-hyphenated varieties. It's fine for one to be preferred over the other in documentation but both should work and continue to work as this is by far the least disruptive to editors and allows them to spend their time producing/maintaining content rather than worrying about finicky syntax. I don't have massively strong objections to general fixes substituting one for the other when making non-cosmetic changes to the page, but I wouldn't actively encourage it as it will clutter diffs for no benefit. I strongly oppose bulk bot runs and making one option non-functional in the future. Thryduulf (talk) 15:29, 17 February 2021 (UTC)

  Re "Are y'all suggesting we bring back all deprecated parameters" - yep, that's spot on. They should never have been deprecated in the first place. This is a template, which sits in the background, and exists for editor convenience, nothing more. Deprecating parameters reduces that convenience. And it's well-established that bots and editors shouldn't be running through making cosmetic changes to wikitext that do nothing to alter the page, so those need to stop as well.  — Amakuru (talk) 12:48, 18 February 2021 (UTC)

• Option C, I guess, as I've not been persuaded as to the marginal utility of the hyphenated versions. Without that clarity, we shouldn't be doing these kinds of changes. (And if we had consensus, then B, but with documentation changes as the first step.) — JohnFromPinckney (talk) 01:48, 17 February 2021 (UTC)
• Option C If it works don't fix it. Andrew🐉(talk) 13:01, 17 February 2021 (UTC)
• Option A, strongly oppose option C: anybody working regularly on templates or modules will appreciate the value of settling on a single style for issues like parameter names. I don't care what the agreed style is, as long as it's consistent, and the argument about whether it should be hyphenated, underscored, camel-case or run-on has been settled with hyphenated as the preferred style. It is then nonsensical to fail to implement that style, and I'd prefer it was done as soon as possible. This whole debate is reminiscent of the date-linking wars where strong objections were made to unlinking dates, yet within a few months of a binding decision being reached to delink dates, everyone had moved on, and nobody today would even consider linking dates. Once we have standardised on hyphenated parameters, future editors will look back and think how lame and time-wasting this sort of debate is. --RexxS (talk) 19:46, 19 February 2021 (UTC)
• A or B but not C per Scott, Hawkeye, and RexxS. Usingahyphenismoreuserfriendly, andwhilethereissomeupfrontworkonourparttomaketheswitch, itisbetterinthelongruntohavesomedelimiterbetweenwordsinsteadofrunningthemtogether. Just like we stopped linking dates and no longer SmashWordsTogether, this is worth the temporary inconvenience. — Wug·a·po·des​ 00:46, 20 February 2021 (UTC)

  Notwhen it'sjust twowords :-P Levivich ^harass/_hound 08:10, 20 February 2021 (UTC)

  ...is what your comment would look like if we allowed people to use whichever method worked for them and still made it work. If we turn off those parameters, your comment wouldn't have appeared. If we allow people to use whatever works and use a bot to clean it up afterwards, your comment would look just like everyone else's after some period of time (during which your comment would still be functional, even if pairs of words were sometimes combined). :) — Rhododendrites ^talk \\ 20:15, 22 February 2021 (UTC)

• Option A for all the arguments already made at length in the CS1/CS2 talk pages over the years and the good arguments brought forward above. Definitely not Option C. --Matthiaspaul (talk) 02:06, 20 February 2021 (UTC)
• Option C. Such an absurd waste of time and energy and an enormous source of watchlist spam, all to achieve something that will make editing more difficult. Toohool (talk) 02:26, 22 February 2021 (UTC)
• Option A as standardizing is better in the long term. Let the bot continue its work. If people have a problem with their watchlist getting spammed, they have the option to filter out bot edits. AVSmalnad77 ^talk 05:59, 23 February 2021 (UTC)
• Option C if not B While I love consistency, I'm struggling to see what the actual issue is here. I guess they can be gradually changed by the bot along with other more useful changes, but this is just cosmetic to the code and clutters edit histories. Reywas92^Talk 06:21, 24 February 2021 (UTC)
• Option C. Benefits for bot or template builders don't outweigh the inconvenience for other editors. Fram (talk) 08:13, 24 February 2021 (UTC)
• C. I don't see the argument for mandating hyphens. Just let people use both, consistency on this does not matter. Fences&Windows 17:00, 24 February 2021 (UTC)
• Comment - my default is Option D... I refuse to use citation templates, and format my citations by hand. It means that I can ignore all the silly debates about parameters and what not. Blueboar (talk) 17:49, 24 February 2021 (UTC)

  Yeah and we can instantly clear the WP:PHAB backlog by writing articles with paper and pen. We can solve many technological problems by abandoning technology. Levivich ^harass/_hound 00:52, 25 February 2021 (UTC)

• Option C. Bots do some useful work but their code can be changed as needed. Here I am much more concerned about regular editors who use citation templates when making manual edits. These are live human beings and there are many more of them than bots. Making the template syntax too rigid will make their life much more unpleasant. Extra flexibility is both useful and helpful for regular editors. Nsk92 (talk) 02:26, 25 February 2021 (UTC)
• Option C. I've been typing accessdate for the last 15 or so years, sometimes while reaching for my drink with my other hand (Qwerty keyboard). Why wreck my muscle memory and deprive me of a sip of tea?-gadfium 04:04, 25 February 2021 (UTC)
• Option Cper Phil BridgerSea Ane (talk) 21:53, 26 February 2021 (UTC)
• Option C, stop the craziness hitting watchlists (although most of that damage is already done). SandyGeorgia (Talk) 00:58, 2 March 2021 (UTC)
• C sounds best, followed by B. Removing parameters that were widely used in the past will break old revisions of articles for very little practical advantage. —Kusma (t·c) 11:44, 2 March 2021 (UTC)
• Option C. Others have already hit on why – watchlist spam, waste of time and energy for no genuine benefit, unnecessary imposition on editors, etc. I'm increasingly warming to the idea of writing out citations manually (as someone else here mentioned), just to avoid the constant tinkering around that seems to take place with these templates. JG66 (talk) 11:51, 2 March 2021 (UTC)
• Option C per Aquillion. Monkbot should never have been approved --In actu (Guerillero) ^{Parlez Moi} 20:22, 2 March 2021 (UTC)
• A or B This seems like a choice between having unnecessarily having several different ways to write a given parameter, and between standardizing after the fact with a lot of minor edits. Many of the arguments in favour of C appear to presuppose that editors will land in trouble for writing the unhyphenated parameters, but I don't see evidence of that. Jo-Jo Eumerus (talk) 10:30, 4 March 2021 (UTC)
• Option A. From real world experience, I understand how difficult it is to maintain code without occasional deprecation. Opponents' fears of unacceptable disruption and inconvenience don't match what I encountered when the bot was running, and typing the hyphen quickly became second nature. --Worldbruce (talk) 05:07, 7 March 2021 (UTC)

  You may not personally have encountered disruption and inconvenience, but I did and there are a great many other editors in this and other discussions that have reported unacceptable disruption and explained exactly why the inconvenience is not justified. Thryduulf (talk) 20:29, 7 March 2021 (UTC)

• Option A. Long term benefit is worth the short-term disruption. There is too much inconsistency in templates that causes me to waste far more time (e.g. is it "image=" or "Image=" or "photo=" in this particular template); we should move towards more consistency. I would support keeping the parameters functional for a few years (but undocumented and with a warning) for those who really are troubled by typing the hyphen, knowing the bot will come by and change them. MB 15:23, 7 March 2021 (UTC)

  The better solution to "is it "image=" or "Image=" or "photo=" " is simply to support all of them. That way there is no need for any disruption, short-term or otherwise. Thryduulf (talk) 20:31, 7 March 2021 (UTC)

• Option A—standardization helps simplify code maintenance, and that means more time can be devoted to future improvements. Imzadi 1979 → 17:34, 7 March 2021 (UTC)
• The question before us is: Should non-hyphenated parameters be fully removed from the CS1/2 family of templates? Yes, absolutely, for most if not all of the reasons enumerated above. Consistent parameter naming should have been implemented c. 2007 when the various, independently-developed, cs1|2 templates were converted to use {{citation/core}} as the common citation rendering engine. In early 2013, en.wiki migrated to the Module:Citation/CS1 suite. Since then, approximately 180 other-language MediaWiki wikis plus some number of non-MediaWiki wikis have also migrated to the module suite. In the time since en.wiki switched to the module suite, we have added new parameter names to support new functionality while at the same time, we have pared away quite a few parameter names because of redundancy, peculiar name-style, non-use, and other reasons. This reduction includes most of the nonhyphenated multiword parameter names so that today, the only remaining nonhyphenated parameter names are |accessdate=, |archivedate=, |archiveurl=, |authorlink= (and its two enumerated forms), and |origyear= (there are 263 basic parameter names and 77 enumerated parameter names). The worldwide adoption of the cs1|2 module suite has caused us to add support for internationalization. Non-English wikis employing the cs1|2 module suite should retain all of the English parameter names because, very often, articles developed at en.wiki are exported and translated to those other languages. That means that a fully implemented module suite at a non-English wiki must support the 340 English parameter names plus 340 local parameter names. It is best, I think, to have a single consistent style for multiword parameter names so that translating editors don't have to learn about or deal with redundant parameter names (this same applies to beginner editors at en.wiki). The cs1|2 templates are complex enough, we don't need to add to that complexity by maintaining lists of synonymous parameter names that don't have semantic meaning (for example, |chapter=, |article=, |entry=, |section=, etc, are treated by cs1|2 as synonyms but, to editors, convey different meanings – the inclusion or omission of a hyphen conveys no meaningful distinction). So yeah, non-hyphenated parameters [should] be fully removed from the CS1/2 family of templates and since we can't go back to 2007 to do what we should have done then, we should do it now, we should do it quickly, and we should get it done. A is my preferred option but, if needs must, then (sigh) B; never that other option. —Trappist the monk (talk) 00:50, 8 March 2021 (UTC)

  Thanks for weighing in, Trappist; I just wish you had done it a month ago. Your explanation is persuasive, but I still think Option A is silly if we don't at least simultaneously change the documentation to tell users, "don't use that parameter anymore". Having the bots fight against the human editors is silly and inefficient and possibly even dangerous. — JohnFromPinckney (talk) 02:03, 8 March 2021 (UTC)

  So to summarise Trappist's arguments: It's much easier for developers to only have a few names so we should disrupt editors and thus the wiki to make their lives easier, regardless of the costs (it's been explained at length previously how having two names do the same thing is not at all a problem for editors). Sorry, but that is not persuasive in the slightest: templates exist to support editors not the other way around. Thryduulf (talk) 03:06, 8 March 2021 (UTC)

  No. Trappist said no such thing. He helpfully explained why this change is necessary. With the possible exception of watchlists, this change does not seriously "disrupt editors" (there's no way having to type in one extra character can be said to be a problem of that magnitude); on the contrary, in the long run it makes life easier for editors, because there is only one simple, easy-to-remember rule: all muti-word parameters use a hyphen to separate the words. On watchlists, see #Worth noting below, which seems to be a promising solution. And if the worst comes to the worst and that solution doesn't work for you, then I sympathise, but at least the "disruption" is only temporary until the bot is finished. --NSH001 (talk) 10:05, 8 March 2021 (UTC)

  No, the disruption is not temporary. Not only are people likely to continue using the "old" parameters, but removing support for these parameters from the templates means that millions of old revisions will have errors in them instead of simply displaying the references like they used to. Creating errors in the references of old versions of countless pages, so that you only have to maintain 340 instead of 349 parameters (well, not even that, 340 parameters + 9 synonyms)? That seems like a lot of disruption for little gain. Fram (talk) 11:13, 8 March 2021 (UTC)

  Doesn't really matter, it's relatively uncommon to cite old versions of pages, and the error messsages can safely be ignored – only the latest page matters. They also suffer from deleted templates (which can be much more serious for the page as actually displayed to the user) and wikilinks which have gone red because the target page has been deleted. No doubt other things I haven't thought of yet. One just accepts these imperfections. Doesn't seriously "disrupt" anything. --NSH001 (talk) 14:05, 8 March 2021 (UTC)

  A redlink because the target doesn't exist isn't an issue (it is even an improvement). The others are problems, and knowingly adding much, much more of the same is a serious issue. A version of Salvador Dalí from 5 years ago now already has 10 or so errors: the lang text ones are the most annoying. But the planned defenestration of e.g. accessdate will suddenly add 24 extra errors here. Basically, every edit that MonkBot does for these, will mean one or more errors in every history revision, or millions upon millions of old versions which will become worse, and no current of future versions which will become better. Fram (talk) 14:33, 8 March 2021 (UTC)

  "Basically, every edit that MonkBot does for these, will mean one or more errors in every history revision, or millions upon millions of old versions which will become worse". That's false. Monkbot 18's edits make no difference whatsoever to the display of error messages; in fact it will reduce (drastically) the number of error messages that are eventually displayed when and if the remaining are formally deprecated, where "formally deprecated" means changing the CS1/1 module suite to actually generate the error messages. --NSH001 (talk) 09:37, 11 March 2021 (UTC)

  Not false, just shorthand. One can see the number of errors that will exist by looking at the number of edits that Monkbot does for this. No, Monkbot doesn't cause the errors directly, the deprecation does; but the two belong together. If option C is chosen, then there will be no "when and if", the remaining will not be formally deprecated, and no error messages will be generated, not in historic versions, not in live versions. None. Fram (talk) 10:56, 11 March 2021 (UTC)

  "One can see the number of errors that will exist by looking at the number of edits that Monkbot does for this." That's very misleading. The number of edits by the bot has nothing whatsoever to do with the errors displayed in the live version (except that, as I've already explained, it reduces them). You're trying to mislead readers here by implying that every single edit by the bot causes an error message, when in fact it only does so for historic versions. Error messages in historic versions don't matter. --NSH001 (talk) 12:00, 11 March 2021 (UTC)

  No, that's not misleading if one reads the actual conversation, which is about historic versions. "removing support for these parameters from the templates means that millions of old revisions will have errors in them instead of simply displaying the references like they used to." and "Basically, every edit that MonkBot does for these, will mean one or more errors in every history revision, or millions upon millions of old versions which will become worse, and no current of future versions which will become better." That's what you replied to, and which we are still discussing. We can disagree whether errors in old revisions are a problem or not, or whether the maintenance cost of keeping a few synonyms alive is negligible or significant; but please don't start accusing people of "trying to mislead readers" (a bit rich coming from someone claiming that "access-date" is easier to use and more meaningful than "accessdate"). Fram (talk) 12:12, 11 March 2021 (UTC)

• Option C And this issue is largely why I am of the current position BAG is not suitable for purpose and needs nuking from orbit. Only in death does duty end (talk) 15:31, 8 March 2021 (UTC)

  The BAG has a few functions. It does a good job of some of them - notably it's pretty good at ensuring bots do only what their authors intend them to do before they are unleashed. The main issue is with ensuring that only tasks with consensus to be done and consensus for a bot to do the job get approved - I can't recall an instance when it failed to approve something that should be approved, and it does stop the truly egregiously bad ideas from proceeding, but there are multiple instances (including this one) where bar for consensus has been set too low and a small local consensus has allowed a bot without wide community approval to operate. If we have bots then we do need some sort of approvals process, so don't nuke the one we have without coming up with something better first. Thryduulf (talk) 16:54, 8 March 2021 (UTC)

  "only what their authors intend them to do before they are unleashed." On a technical level this is correct - what BAG doesnt do is any sort of even half-decent job of confirming that bots *continue* to only do the tasks they were initially approved for, or that that approval in the first place has consensus amongst those editors who it will *affect* that the task is wanted or needed, or that there is a clear benefit to those effected by it. The problem with automated editing is rarely the technical aspect, its the business case for it. If we did a review of current active bot tasks (and editors using automated editing to perform mass edits) do you genuinely think they will all be able to point to a discussion that shows a level of consensus proportionate to their effect on the wider editing populace, rather than the walled garden (and the term I would actually prefer to use here as its more accurate is circle-jerk) of 5 or 6 data & machine readable obsessed editors who want it? Only in death does duty end (talk) 08:30, 11 March 2021 (UTC)

  On the technical vs business-case aspect, I agree, that was much of the point I was trying to make. I also agree that a regular review of bot and automated editing tasks should be undertaken. There are some that I'm sure still have community consensus (e.g. the anti-vandal bots) but I can't say that will hold true for every task. I also agree that there must be an active consensus of editors that will be affected before a task goes ahead - in the case of monkbot almost all affected editors were entirely unaware that it was even proposed. Thryduulf (talk) 14:20, 11 March 2021 (UTC)

• Option B - I can easily see both sides of this debate. As someone who works in IT, I fully agree that IT should serve the needs of users first wherever possible, rather than simply making life easier for the people who maintain the systems. I see far too many systems that make life harder than the non-digital alternative, which is just ridiculous. There are of course times when things need to be deprecated because they're either incompatible with other things, have security holes, or take up too much development time to maintain. This case is one of the latter, although I'm not sure what extra burden the deprecated (non-hyphenated) paramaters actually add. I would recommend that we formally deprecate the unhyphenated parameters and clean them up with AWB genfixes but I can't support continuing to run Monkbot 18 given the strength of feeling here against such a task. Personally I do not care about 'watchlist spam', but clearly many people do, and we cannot simply dismiss their opinion because "it'd make life easier for us techs". ƒirefly ( t · c ) 07:42, 10 March 2021 (UTC)

  As I understand it, there's a framework for supporting parameter aliases, which are configured in Module:Citation/CS1/Configuration. As there are many other aliases supported, the framework would remain in place. isaacl (talk) 15:23, 10 March 2021 (UTC)

• Option A / B per above.  Spy-cicle💥  ^Talk? 18:50, 10 March 2021 (UTC)
• Option A, strongly oppose Option C. Hyphens should be standard and we should discourage them. I do not mind continuing to support the non-hyphenated version per B, but B also implies the bot isn't free to do its job, so A it is. SportingFlyer T·C 12:57, 11 March 2021 (UTC)

  Why should one version over the other be discouraged? What benefit to the encyclopaedia does it bring that outweighs the disruption that removing support for a long-standing and well-used feature causes? Why should the bot be free to continue to do a job it didn't have consensus for? Thryduulf (talk) 14:22, 11 March 2021 (UTC)

  We're trying to gain consensus here, right? I support standardising the reference tags, and I don't think the opponents have good counter-arguments. I'd appreciate if you just let me (and others) have our opinions without the need for commenting - you're not "correct" and we're not "wrong." SportingFlyer T·C 15:01, 11 March 2021 (UTC)

  I'm not trying to stop people having opinions. I'm trying to get people to explain them and to actually counter arguments left explaining preferences different to their own. Why is standardising reference tags more important than the disruption standardisation causes? Why is a desire to avoid this disruption "not a good counter-argument"? I may or may not be "correct" but unless you can explain why forcing standardisation against the wishes of a very significant number of editors (who will not see any benefit from such standardisation but will experience significant ongoing disruption due to it) is somehow desirable then there is no hope of getting consensus for your views. Thryduulf (talk) 16:02, 11 March 2021 (UTC)

  You're chiding me for not countering your argument, even though your argument above is simply "I don't see the point." We've been working on this for awhile, it's been discussed on talk pages, it makes it easier to maintain the encyclopaedia, and it's a good idea to finish the project as opposed to having it held back by users who don't like it. I also thought I was just !voting here, I think this is obvious and I don't want to be drawn into a long argument about this, so not only did I not appreciate your response, I'd appreciate it if you left this conversation alone going forward. SportingFlyer T·C 16:49, 11 March 2021 (UTC)

  As with many of these types of questions, the conclusion each person draws depends a lot on the weighting they give on the relative priorities of different factors. We lay out our lines of reasoning with the tradeoffs, and others can use it as they wish to figure out what tradeoffs they would like to make. isaacl (talk) 17:13, 11 March 2021 (UTC)

• Option C Unless there is an issue with bots having to read accessdate and access-date as the same thing (and there doesn't appear to be a problem from what has been described, as bots can do this) then I'm not convinced that forcing users who already write accessdate to make an error when they continue to do so after its use is stopped that bots will then flag up for humans to solve is going to be a useful thing to do. Creating a situation which is going to frustrate and demotivate volunteers for no valid reason other than "conformity" doesn't sound like a good plan. Indeed, it was stressed in the original 2014 RfC that "Establishing this uniform parameter name convention does not preclude the existence of any other alias for a parameter, merely that a lowercase, hyphenated version will exist for each parameter." And some of those supporting did so because: "This will significantly reduce editor confusion. They don't have to think about: "Is this the parameter where the words are mushed together, or is it one where they are separated by an '_' ?" Hopefully this will make the templates easier to use." - User:Makyen. What we should be looking at is making bots read the varying ways that users may write a word or phrase in a template. I hate it when, for example, I use a capital letter by mistake, and the template doesn't work, and I have to work out what the problem is. I don't follow how frustrating, demotivating and alienating volunteers by changing how a template works and so creating problems for them will "decrease the maintenance burden" on those users. It is sometimes things like functionality being changed, that drives users away. SilkTork (talk) 01:09, 12 March 2021 (UTC)
• Option C. I haven't checked whether this move was done under a six-year old RFC with seven people, or under a seven-year old RFC with six people. But no one disputed that it was an at least six years old RFC involving at most seven people. Now, it seems that there are a large number of users who disagree. Perhaps telling them they "don't see the larger picture" will be enough to silence the dissenters. Maybe the "Visual Editor reception" will happen again. Who knows the future? Pldx1 (talk) 15:07, 12 March 2021 (UTC)
• Option A. I do not see any real argument for why citation templates should be a mish-mash of two different parameter names. It's an eyesore, it's a pain in the ass, and it brings zero utility. There are a couple ways of fixing it: for the new parameter names to be integrated into every tool, and every adjacent task, and every automated editing tool, and then maybe in ten years 90% of them will be gone (but the remaining 10% will be scattered willy-nilly across the entire project and impossible to hunt down). Or we could just have there be one short run and be done with it. I, for one, would be glad for the issue to be concluded. jp×g 05:33, 15 March 2021 (UTC)

  Have you actually ready any of the comments left by others? There are at least half a dozen different reasons given for the utility of multiple forms of parameteres. Even if option A is selected, disruption will not be short term but will continue for years for the reasons explained multiple times elsewhere on this page. Thryduulf (talk) 12:27, 15 March 2021 (UTC)

• Option A. Over time, we evolve different conventions. Once we have done so, we need to clean up the uses of the old conventions. Option B will be too slow and leave this issue dragging on for many more years. So, firstly ensure that all documentation is unambiguous about using only the new names. Secondly, make sure that all editing tools use only the correct parameter names and all genfix rules are up to date. Wait for the bot to complete a full pass and then treat the old parameter names as solid errors — GhostInTheMachine ^{talk to me} 15:36, 15 March 2021 (UTC)
• Option C. The encyclopedia should be run for the benefit of readers, and subject to that, for the benefit of editors. Creating busywork and random conventions for the sake of it is not useful. Stifle (talk) 10:19, 16 March 2021 (UTC)
• Comment: The argument made here and elsewhere about how much easier things would be for template- and module-writers under Option A, is a completely misguided and arrogant one. It seeks to benefit a small subgroup who are neither the consumers of the encyclopedia, nor its principal creators. Namely, it calls for optimizing ease of work and convenience for the very small number of template/module writers, rather than optimizing for (the much larger number of) article editors. The trade-off must benefit the largest number; if you make it harder for editors, then the article readers will suffer, and they are the largest group of all and the reason the encyclopedia exists. This should be axiomatic, but the encyclopedia isn't here for the convenience of template and module writers. Personally, I'm happy writing and improving templates and breaking my head against squirrely, confusing, and sometimes infuriating template problems, just to make things slightly easier for editors. If it's too inconvenient or too much work for template/module-writers, then they should abstain from "helping", and just improve article content, rather than seek to make their own lives easier. Mathglot (talk) 23:42, 20 March 2021 (UTC)

  Thryduulf said it better (and briefer) than I, here: "Those who do care about template mechanics should always be acting in ways that actively put readers first, editors second and programmers third." Precisely. Mathglot (talk) 00:04, 21 March 2021 (UTC)

• Option C (preferred) or Option B (seoncd choice) -Beyond My Ken (talk) 03:06, 23 March 2021 (UTC)
• Strongest possible option C Per WP:KISS and WP:IFITAINTBROKE. If it ain't broken, don't fix it, and don't waste our time with meaningless bot tasks. We should be making templates easy to use. Having different aliases for the same parameter is helpful. Insisting on one particular format because of some concept about what is correct and what is not is WP:CREEP, and generally to be discouraged. RandomCanadian (talk / contribs) 01:59, 24 March 2021 (UTC)
• Option C. If we were starting today, and writing brand new template to handle citations, I would strongly advocate using one consistent format for parameter names. But we aren't. Many thousands of editors are used to using the no-hyphen versions of those parameters, and I see no reason to disrupt their editing lives in order to (allegedly) make life easier for a handful of template editors. The system we have now works fine, I see no reason at all to waste anybody's time changing it. Chuntuk (talk) 20:41, 30 March 2021 (UTC)

Discussion (CS1)

Some suggestions regarding the options:

• Pedantry: "Non-hyphenated parameters" should read "Non-hyphenated multi-word parameters" in all three options. Parameters that contain only a single word or acronym do not need a hyphen.
• In Option C, there is no point in suggesting that "the deprecated parameter list will need to be updated to remove the non-hyphenated parameters"; there are no instances of those deprecated parameters left in the affected namespaces, so support for them will be removed shortly, just as support has been removed for dozens of unhyphenated multi-word parameters already.

Some history and a status update, for those here on VPP unfamiliar with the long history of updates and changes to the Citation Style 1 templates ({{cite web}} and its siblings): As far as I can tell, there are only six unhyphenated multi-word parameters left – |accessdate=, |airdate=, |archivedate=, |archiveurl=, |authorlink=, and |origyear= – out of an original population of many dozens. So far, through the work of scores of editors and bots over the seven years since we standardized on hyphenation of multi-word parameters, we have deprecated, removed from pages in affected namespaces, and then removed from the CS1 templates themselves (as WhatamIdoing suggests above), many dozens of different unhyphenated multi-word parameters in CS1/CS2 templates. All new multi-word parameters during that time have been introduced using only a hyphenated form. This RFC is essentially asking: should we finish the job, or leave it at over 90% done, in a sort of limbo state, with six parameters as exceptions to the overall pattern, just because those parameters are used in a lot of articles? – Jonesey95 (talk) 04:38, 11 February 2021 (UTC)

• Problem with "Option B": Option B is not the status quo. If it were, I'd be much happier. Since it's not, I don't know whether to !vote for A or B. The documentation at Cite web, for example, makes no mention of accessdate being deprecated. This means that users will be quite justified in blissfully adding and readding templates with accessdate, even after the bot has come through already and changed accessdate to access-date in 20 places. Each iteration tends to address fewer instances, but each run is a separate entry in my watchlist. Watchlist entries seem to be part of the complaint against this kind of work, so the first step should be turning off the faucet at the source, before spending energy to, um, bail out the boat. — JohnFromPinckney (talk) 06:37, 11 February 2021 (UTC)
  • The first sentence is true; all but six of the unhyphenated multi-word parameters have been formally deprecated and removed. As for "turning off the faucet at the source", that would be Option A, but in the past, when we have made red error messages appear in large numbers of articles as a first step in standardizing the citation templates and noting errors in them, there has been significant pushback. In order to avoid that pushback, the bot was acting to fix 90+% of the non-standard parameters before turning on the error messages, but that work has been stopped as well. Option A will allow that work to be completed. – Jonesey95 (talk) 15:09, 11 February 2021 (UTC)
    • I feel we are not understanding each other. "Deprecation" involves communication as a first step; removal is a second, later step. If we are going to have a bot changing pages (e.g., accessdate to access-date), causing some distress to the populace (and this is the status quo), then we should at least change the documentation to say that accessdate is deprecated, and is not a usable alias. I am strongly against a bot going around to change parameters to the "good" names when we don't ever tell the humans they shouldn't use the "bad" ones. That's just an endless cycle of watchlist-cluttering edits causing great irritation. If you want to avoid pushback, tell the people not to use the unhyphenated versions, then run the bot, then remove support some time later. — JohnFromPinckney (talk) 15:41, 11 February 2021 (UTC)
      • Deprecate in terms of the CS1 templates means to emit a red error message indicating deprecation. When we emit any red error messages for parameters in CS1 that are used often (or lack thereof for parameters that should be used more often), a lot of people get very irate. We do not change the documentation to indicate deprecation until after the module begins telling people inline about deprecation. So yes, you are not understanding each other, but it's a question of terms of art. --Izno (talk) 16:44, 11 February 2021 (UTC)
        • Why do people get irate? Because suddenly millions of readers go from nothing being wrong to seeing a load of red error messages all over hundreds of thousands of articles (and you couldn't seriously expect someone to debug a CS1 error as their first contribution). These citation templates aren't for us. They're for the reader and they need to be functional with low rate of error messaging 24/7 with no exception, because even small amounts of downtime have significant reputational damage. — Bilorv (talk) 02:54, 15 February 2021 (UTC)
          • I would be more sympathetic to such an argument of "significant reputational damage" if there weren't over 100,000 pages with active errors on them (and that's ignoring the hidden Category:CS1 errors: missing periodical) and another 250k in its sister (also hidden). --Izno (talk) 20:22, 15 February 2021 (UTC)
    • That "we" (who is that, some class of super-techies whose opinions count for much more than those of us "normal" editors?) get significant pushback when creating error messages is simply a demonstration that the wider community does not agree with what "we" have decided. The answer is to stop doing what you are doing, not to do it more stealthily. Phil Bridger (talk) 18:37, 11 February 2021 (UTC)
      • You are welcome to participate at Help talk:CS1, the same as any other editor. Decisions are made by consensus there, inline with how consensus is practiced everywhere else on wiki. Don't like a decision "we" made? Get involved. --Izno (talk) 18:42, 11 February 2021 (UTC)
        • I saw absolutely no discussion of the removal of the freetext editors field on the talk page. Where did that happen? I don't think it's coincidence that all the editors whose names I know well from content creation/editing are voting B/C and all the editors voting A are those I don't recognise at all. Espresso Addict (talk) 20:08, 11 February 2021 (UTC)
          • The difference is between those people who think this is an encyclopedia and those who think that it's a playground for techies. Phil Bridger (talk) 20:43, 11 February 2021 (UTC)
            • Your ad hominem has no welcome here. Move along if that's the best you can offer to the discussion. --Izno (talk) 20:47, 11 February 2021 (UTC)
              • That is just the kind of reply that people who are here to create an encyclopedia tend to get from the techies. No response to genuine concerns but just an order to "move along". I have no wish to monitor whatever pages are used to ignore end users, but I have a right to expect that any decisions taken will respect the interests of everyone, not just a self-appointed clique of people who "know" what is right. Phil Bridger (talk) 19:53, 15 February 2021 (UTC)
                • It's the kind of reply to people who needlessly create category A and category B and then line themselves up in one or the other categories. If you (specific/personal) don't want to monitor whatever pages, that's your prerogative. Do know that it is your choice and you are responsible for that choice, and choices have consequences. Changes are always announced ahead of time and consensus is sought for non-obvious changes (and even obvious changes with non-obvious implementations), so you have no excuse not to tune in at least once every couple months when the regular "Shit is Changing" post gets made. Secondarily, the scare quotes are not indicated by this discussion, nor any prior discussion that I can see, whatsoever. I have not seen any such 'clique' nor any of the users who would prospectively be in such a 'clique' claim they "know" what is right. As I said, if you have nothing to contribute but smears and attacks and divisiveness, move on. --Izno (talk) 20:22, 15 February 2021 (UTC)
          • In reverse chronological order: the patch notes indicating removal, the removal discussion, the patch notes indicating deprecation, the deprecation discussion. 4 times mentioned over a period of half a year on that talk page, a pre-existing maintenance category indicating a soft deprecation, and my removal of over 4k instances of the parameter over the year and a half preceding that deprecation discussion. Basically by hand (my hand, as it happens). --Izno (talk) 20:45, 11 February 2021 (UTC)
          • As for I don't think it's coincidence that all the editors whose names I know well from content creation/editing are voting B/C and all the editors voting A are those I don't recognise at all., I invite you the same as I invited Phil Bridger. You may watch and edit Help talk:CS1 at any time, the same as me. "I don't recognize these people" is a trash association to make and is the same kind of ad hominem that I have asked Phil so kindly to stop employing. It is certainly not sufficient cause to say "I don't like this change". --Izno (talk) 20:49, 11 February 2021 (UTC)
            • I think there's a genuine problem here that there's a disconnect between the editors maintaining the citation templates and the editors employing them to write and maintain articles. I didn't make the decision to abandon years of using citation templates (CS2 actually, but the same parameter changes are happening there too, even less announced) lightly. Espresso Addict (talk) 21:35, 11 February 2021 (UTC)

              The disconnect is real and significant. Someone whose skills and interest lie in writing or maintaining article content shouldn't need to care about what happens at pages like Help talk:CS1 (and how on earth is a new editor even meant to know that page exists?), let alone have to pay attention to discussions there. Those who do care about template mechanics should always be acting in ways that actively put readers first, editors second and programmers third. The only time there should ever be breaking changes or a need for cosmetic edits is when after detailed examination there are literally no alternatives available. We've ended up here because that hasn't been happening and a local consensus has ignored the needs of end users. Thryduulf (talk) 01:38, 12 February 2021 (UTC)

              I almost commented last night, and then put that version into a text file and went to bed. Now I have been spurred by a comment above to reply here. The editors using these templates to write and maintain articles are the same as the editors maintaining the citation templates. Get that through your head. If we prioritize different qualities versus this other supposed separate group, it's because we have the experience to do so. If you don't, let me reiterate, get involved. Come and say "I don't like this" or "this is a painful interaction" or X, Y, and Z, and provide a suggested fix. That's how consensus starts forming, like every other page or process on this website. Others will say "I don't want this to work like that suggested fix because A, B, and C". Then you discuss the tradeoffs and make a decision. If you're unwilling to step up and discuss the paper cuts, then we end up having mass RFCs or AN drama or what have you over what would otherwise be small issues or ones that could have been discussed informally with the ordinary "let's figure it out" consensus view of the world. That's a disconnect too, and blaming editors interested in one page versus those apparently disinterested is not the right way to move forward. Want something to be better? Ask. Propose. Cajole. Make the effort to put forth the minorest of social interactions required to start a discussion in the one place where everything about that thing is discussed. We're all volunteers and our heart is in just a right a place as yours, but if we should have disagreements, then we talk to each other and find out how to fix the problem or agree to disagree on the points of interest. Not have constant complaints of "they didn't listen to me". Consensus is not unanimity.

              The only time there should ever be breaking changes or a need for cosmetic edits is when after detailed examination there are literally no alternatives available. This is quite frankly an opinion lacking any community consensus whatsoever. We've recently approved an RFC allowing cosmetic edits on a regular basis and from what I could see in that discussion (and murmurings elsewhere), cosmetic edits might be closer to having consensus than not, it's just inertia that leaves us not performing them regularly. Moreover, hundreds of templates, and certainly all those which go through WP:TFD, have a process applied which causes breaking changes or which results in more-or-less cosmetic edits. Are you claiming that TFD does not have consensus as a process? That templates which are being cleaned up because of a talk page discussion on that template's talk page should be stopped? You want your cake (to not care about how things work) and eat it too (have all of your opinions and claims listened to, when they aren't even articulated in the first place). Get real. --Izno (talk) 20:22, 15 February 2021 (UTC)

              I fear this just demonstrates the truth of what Thryduulf writes. "The editors using these templates to write and maintain articles are the same as the editors maintaining the citation templates. Get that through your head." is self-evidently false, as well as impolite. Speaking only for myself, what I actually want is to be able to get on with writing & improving articles, rather than having long side discussions on matters that aren't directly relevant. Espresso Addict (talk) 00:10, 16 February 2021 (UTC)

              get on with writing & improving articles Then do that. No-one is stopping you from not caring. I am just calling you hypocritical for raising a fuss when you don't and then changes happen elsewhere that affect you. Don't like it? Change your behavior. (I won't reply to you again.) --Izno (talk) 00:34, 16 February 2021 (UTC)

              The editors using these templates to write and maintain articles are the same as the editors maintaining the citation templates. Get that through your head. Both uncivil and not even close to correct. While it is possible that some, maybe even many, of the editors maintaining templates also write and maintain articles there are literally hundreds of editors writing and maintaining articles who have never been near a discussion about the template code, let alone written any code. Writing encyclopaedic prose and writing computer code are very different skills; nobody is or should be required to do both. However given that the purpose of this project is to write an encyclopaedia everything that is not writing an encyclopaedia should be done for the benefit of (first and foremost) readers of the encyclopaedia and (secondly) those who write and maintain the encyclopaedia. The convenience of those dealing with tools is least important. Thryduulf (talk) 00:23, 16 February 2021 (UTC)

              Writing encyclopaedic prose and writing computer code is a false dichotomy and a blatant misrepresentation of what the two paragraphs I had to say. I did not ask you to do both. Nor do I serve you (general) in your supposed role as an article writer. There are no (formal) hierarchies on this encyclopedia, and even considering yourself as supposedly above me or my efforts is the actual uncivil statement. Lastly, I place myself fully in both supposed groups given the thousands of articles where I have bettered their citations. (And as I said to Espresso, I will not reply in this section again.) --Izno (talk) 00:34, 16 February 2021 (UTC)

              My role here is not as an article writer (I do very little of that), I spend the majority of my time on the project trying to ensure that readers can find the content they are looking for and those that do write and improve articles without needing to worry about nonsense like this that will needlessly make their job harder. There might not be a formal hierarchy but their should be: (1) Readers are head-and-shoulders above anyone else. (2) Those who write and maintain the encyclopaedia a short way above (3) Those who support those write and maintain the encyclopaedia are a long way above (4) Those who hinder any of the above. I place myself in the third category alongside many of those who maintain templates. Thryduulf (talk) 21:02, 16 February 2021 (UTC)

• Comment. What is the merit of citation templates? Let alone the increasing creep to make them more and more rigid and harder and harder to use? The only reason I can see is to make it easier to export and sell data. No-one ever seems to give a thought to those who type citations manually; it is far harder to type "access-date" (which requires two hands and a hand movement) than "accessdate" (one hand, no move). I don't understand what the benefit of this change is at all. In fact, broadening this discussion, I don't understand why the freetext editors field was suddenly withdrawn this January. Personally I've decided to meet the latter change by reverting to writing out references by hand, which is easier, more flexible, and seems to have no downsides. Espresso Addict (talk) 07:04, 11 February 2021 (UTC)
  • Citation templates make standardizing citation style and look much easier. They can, for example, automatically standardize date display. Machine-readability is also a good thing, imo. I suppose access-date is slightly harder to type - but editors editing wikitext manually would probably not mind. Otherwise, tools for inserting these citations exist. Elliot321 (talk | contribs) 17:12, 11 February 2021 (UTC)
    • It's not slightly harder to type, it's a great deal harder to type for a touch typist. And here's an editor editing wikitext manually who does mind, enough to bother responding here. There's no tool I know of that creates citation template code in the form I prefer for ease of wikitext editing afterwards; they all make code soup that takes longer to fix than retype. Espresso Addict (talk) 20:08, 11 February 2021 (UTC)

      A bit slower, yes, but I personally don't find it a great deal harder—touch typists generally have practised it a lot while learning. Nonetheless, I don't think ease of typing by some metric is the primary issue. isaacl (talk) 20:40, 11 February 2021 (UTC)

      Speaking as one of those touch-typists, and as a person who learned to type on an actual typewriter, I don't find the hyphenated version any harder to type, and I do recall my typing teacher saying that it was faster and easier to type words that were split evenly between hands, instead of all in one hand. WhatamIdoing (talk) 22:49, 11 February 2021 (UTC)

      Yeah, personally I imagine it has a greater effect for a hunt-and-peck typist. When I said a bit slower, I was thinking compared with a word of equal length that consisted only of letters. Of course, in this case, the hyphenated name has one more character which would comprise most of the difference for me. isaacl (talk) 23:02, 11 February 2021 (UTC)

      For me, not an issue on a regular keyboard, but a bit of a pain on a tablet. Then again, so much is a bit of a pain on a tablet... · · · Peter Southwood ^(talk): 05:39, 2 March 2021 (UTC)

    • I use citation templates myself, but I respect the wishes of those who prefer not to use them. It is precisely the fact that I use them that leads me to the opinion that obvious synonyms for parameters should not be removed, as proposed here. What on Earth is the problem with allowing both hyphenated or unhyphenated forms? All it means is a few extra bytes of storage, and no extra code if it is done properly. And the work has already been done, but this proposal is to undo it. Do we still live in the days when I started in IT working for one of the world's leading business information companies whose UK operations were all run from a computer with 1MB of memory and where all online programs were limited to 12KB? I thought we had moved on from there. Phil Bridger (talk) 20:34, 11 February 2021 (UTC)

      I am reminded of a story about a lady, who (decades ago) bought a large supply of personalized stationery and then discovered that the post office was renaming her street. She convinced the local postmaster into letting her continue to use the old address until her stationery was used up. This saved her some money and effort, but it had external costs. For many years to come, every mail carrier had to learn that "123 Main Street" wasn't on Main Street and wasn't number 123, but instead had to be delivered to the other side of town.

      Aliases are often low-cost in storage and computational terms, but they are not actually free. "Not free" can add up to a quite significant cost when it is repeated a million times. WhatamIdoing (talk) 22:56, 11 February 2021 (UTC)

• Although I sympathize with the desire for all templates to align with one standard (from a user's perspective, I would personally find it easier), I just don't think it's feasible at this point in time. In which case, I sympathize with those who think computers should make our lives easier, and just accept both formats. On the third hand, from an implementer's perspective, I appreciate that it adds a lot of noise to template syntax, if not implemented with a Lua module. Since the CS1-based templates are implemented with Lua, the overhead can be minimized, and so I think both formats should continue to be supported. I don't feel there is much advantage to converting en masse, by any mechanism. isaacl (talk) 20:40, 11 February 2021 (UTC)
• Pinging some previous participants in related conversations who may not yet be aware of this discussion: @SandyGeorgia, Jason Quinn, Oknazevad, Tom.Reding, Brainulator, DavidBrooks, SMcCandlish, David Eppstein, Matthiaspaul, Headbomb, Gracefool, Ss112, JG66, Mikeblas, Gonzo fan2007, Sariel Xilo, Modest Genius, and SlimVirgin:. Nikkimaria (talk) 01:17, 12 February 2021 (UTC)

  Thanks for the ping, I was indeed unaware of this discussion. Modest Genius ^talk 12:16, 12 February 2021 (UTC)

• Does anyone have any evidence that having alias for parameters makes things harder for end users? In my experience as a trainer and long-time user, having multiple ways to achieve the same ends allows editors to spend their time and energy working on content rather than puzzling over making sure that the template parameters are named in exactly the right way. I've never met anybody who was comfortable enough to be dealing with templates in the first place who was not completely comfortable with the idea that "you can write it as either access-date or accessdate, it doesn't make a difference." (or similar). Speaking personally, I don't want to have to learn whether it is "accessdate", "access-date", "access_date" or "access date", I just want them all to be accepted so that whichever I input the template works and I can worry about more important things. Thryduulf (talk) 01:29, 12 February 2021 (UTC)

  The inconsistency across all templates does make things harder. Users have to either remember which templates only support one form and which one, or look it up each time. I appreciate though that there is extra complexity in trying to support lots of aliases, particularly for ones that aren't implemented with a Lua module (either each use is going to become more elaborate and verbose, or the template could delegate the detailed implementation to a helper template, with the top-level one only doing parameter normalization), so personally I wouldn't want to require that every template must support aliasing. Thus I think we're kind of stuck with inconsistency. isaacl (talk) 02:12, 12 February 2021 (UTC)

  This inconsistency should be tolerated. Some, particularly high-use, templates having aliases is important for the great number of people who use these templates. They don't have to remember was it |access-date= or |accessdate=. If it doesn't work on some other template, well fine. But eliminating aliases and simply making the computer say no on all templates makes things a lot harder for the vast majority of users. That inconvenience is far greater than the fact that a few lesser-used templates don't support aliases and you might need to take a look at the documentation sometimes. – Finnusertop (talk ⋅ contribs) 06:05, 12 February 2021 (UTC)

  Yes, as I said in another comment, I feel both forms should continue to be supported for the CS1-based templates. I'm pretty sure the vast majority of templates don't support both hyphenated and non-hyphenated parameters—for example, from what I can tell from the code and a quick test, {{Infobox}} doesn't. That's just the way it goes when trying to make as many aspects of Wikipedia markup accessible to as many editors as possible: standardization won't always happen, and often simpler markup will be preferred by more editors than more complex markup, even if it would add more functionality for users. isaacl (talk) 22:22, 12 February 2021 (UTC)

• In Option C, there is no point in suggesting that "the deprecated parameter list will need to be updated to remove the non-hyphenated parameters"; there are no instances of those deprecated parameters left in the affected namespaces, so support for them will be removed shortly, just as support has been removed for dozens of unhyphenated multi-word parameters already. This argument directly goes against WP:FAITACCOMPLI. Also, stop referring to them as "depreciated" - it is clear that there was insufficient consensus to declare them depreciated in the first place. I can understand that it's frustrating to have something you thought was decided fall apart like this, but it is extremely important that sweeping changes get more discussion before being implemented; the only way we can reasonably encourage that is by refusing to allow things like Monkbot's use here to determine policy, ie. it is necessary to make the work it did moot, or people will have a constant incentive to take the easy route and avoid bothering with time-consuming, often-frustrating, and unpredictable (but necessary) discussions before a change of this magnitude. That means, yes, allowing and even encouraging people to resume using non-hyphenated parameters even though you thought you were finally done with them; if you want to avoid that happening next time, seek larger-scale discussions first before making large-scale changes to avoid unexpected blowback if it turns out you don't have the consensus you thought you did. No matter how well-intentioned this was, we absolutely cannot risk rewarding the use of a tool to make a wide-spread change without sufficient consensus - which means, yes, as painful as it is, very deliberately taking the wrench to the kneecaps of the intended improvement this change was meant to establish, and intentionally ruining it even after the "cost' has been paid. It's not ideal, but it shows why it's so important to have broad discussions involving many users before making wide-spread changes; having policy effectively set by bots performing mass edits and establishing things as fait accompli is just not acceptable. I can sympathize with the amount of work that went into this that will be wasted as a result, but reaching a clear, unequivocal consensus involving a large number of editors should have been the first step for something of this magnitude, so you wouldn't suddenly run headlong into an RFC like this and find the consensus wasn't what you thought it was. --Aquillion (talk) 10:07, 4 March 2021 (UTC)

Arbitrary break 1

• Thanks to Nikkimaria for the ping based on my earlier involvement. I'm appallingly neutral (or torn) on the main topic but I earlier did some analysis that gives me concern about impacts on Template space in particular, if deprecation goes ahead fully. Here I'll use |authorlink= as a proxy for all six, because I personally type it more often. There are three areas that concern me: templates that indirectly invoke CS1, those that transclude CS1-invoking templates, and clones. I'm concerned about where and when red error messages appear, and documentation, in which I include both /doc and TemplateData.

1. There are many CS1-friendly templates that invoke one of the canonical templates; {{Cite encyclopedia}} is often used for example. Some use parameter rewriting, converting |authorlink= to |author-link= before passing it on.
   1. Does their documentation always give precedence to |author-link=? I think we caught them all during the earlier discussion, but I can't guarantee the quality of the search.
   2. Because the CS1 code never sees the deprecated parameter, should this template emit a red error itself during the substitution? Or maybe just forget about the mapping and have CS1 do it? And how hard is it to find templates with the parameter substitution code?
   3. If |authorlink= ever moves from deprecated to invalid, then all those mappings and documentation need to be trimmed in one fell swoop. Well, I guess the mappings can stay in place although they could trip up inattentive longtime users.
2. There are templates that simply transclude a pre-filled CS1-style template: for example embedding {{Cite book}} as a citation to a specific volume that is relevant to any use of this template. Have they all been fixed? If not, their user will see a red error message (correct?) that has nothing to do with their own usage.
3. There may be templates that are modeled on CS1 but roll their own expansion, although I don't know of any such. If they happen to use |authorlink=, should they also be brought in line?

Two final comments: there are more than six, if you consider variants like |author1link= and |authorlink1=. And in principle the arguments above apply to any page outside Template space that gets transcluded by someone, although I know that feature is rarely used. David Brooks (talk) 16:51, 12 February 2021 (UTC)

Shoulda said: the above applies to A and to some extent B but you can probably figure that out. David Brooks (talk) 16:57, 12 February 2021 (UTC)

DavidBrooks: There is a tiny army of editors at the ready to fix these templates. As far as I know, there are no transcludable templates that pass |authorlink=, your example, to any CS1 templates, so using those templates would not generate an error message. |accessdate= and the two or three remaining unhyphenated parameters being passed from template transclusions to CS1 templates were being processed by the bot before the bot was paused to hold this discussion. Once this discussion is closed, the bot will be able to resume that work, with human assistance, if the closure is a reasonable one. – Jonesey95 (talk) 17:13, 21 February 2021 (UTC)

@Jonesey95: I just addressed a related issue in Help talk:Citation Style 1, where by modifying the example given I found 3 templates that use |authorlink= in a CS1-style template that's used to reference a specific source: {{Barmakids family tree}}, {{Citeer web}}, {{Alox2}}. But this less restrictive query shows a few more (ignore the "DYK" archive, I think). Not sure if you meant this type of usage in your comment; I may be missing the context. David Brooks (talk) 05:05, 22 February 2021 (UTC)

@Jonesey95: OK, looks like you fixed {{Alox2}}. {{Barmakids family tree}} and {{Citeer web}} still need to be fixed (I'll do them later today). The documentation of "Cite book (short)" in {{Quicktemplates}} lists the not-incorrect |authorlink=, so that comes under the "prefer the hyphen forms in documentation" rule. I didn't look for |authorlink2= or |author2link= because they are unlikely to appear alone. David Brooks (talk) 19:17, 22 February 2021 (UTC) ... Y, also {{Bach's compositions (sources)}}, which indeed had |author2link =. David Brooks (talk) 21:42, 22 February 2021 (UTC)

• I would like to clarify what I meant when I originally wrote the options, since there is some confusion. My intention with "removal" was to refer only to removal of usage of nonhyphenated parameters by transclusions of the template, not removal from the implementation of the template itself (I will call this "disabling the parameter"). This is also distinct from deprecation, which is designating something in documentation and warning messages as discouraged. Remember that the original reason for this RfC is to clarify whether we should have a bot going through the article space removing the parameter usage as its sole action. A, B, and C are respectively continue removing now, remove only as part of other non-cosmetic edits or in a situation where cosmetic-only edits are allowed, or do not remove. In case of A or B, I did not intend for them to make a statement about whether we should disable the parameter when this is done. This is an important question, but orthogonal to whether the parameter usage is removed now or later. — The Earwig _alt ⟨talk⟩ 22:11, 13 February 2021 (UTC)

  Thanks for clarifying (except the part where you formulated the options). Unfortunately, it now means I'm missing the option: Non-hyphenated parameters should be deprecated now and removed later. The actual status quo seems to be: Non-hyphenated parameters should be removed now and deprecated later, and the removal can occur by bot which does nothing else on the page (cosmetic only), but will revisit the page as often as necessary to re-remove the non-deprecated params that keep getting added. — JohnFromPinckney (talk) 23:01, 13 February 2021 (UTC)

  For context, this is how I originally phrased it; note B and C are swapped. Isn’t what you’re desiring exactly option B (in the current formulation)? It seems describing B as the "status quo" is confusing. Instead, view A as what the bot was doing before it was stopped, and B as what is currently happening with the bot stopped, but with clear, formal deprecation. — The Earwig _alt ⟨talk⟩ 00:19, 14 February 2021 (UTC)

  Mmm, maybe, and in any case I'm very grateful for your link to the pre-RfC coordination at Primefac's Talk. And I can say that your original formulation was clearer than the formulation we're now using for the RfC proper. However:

  1. There appears to be no consensus for the removal of non-hyphenated multiword parameters. The 2014 RfC determined only that the hyphenated version must exist, and the close explicitly allowed for non-hyphenates.
  2. If this RfC is intended to determine whether non-hyphenates should be deprecated, it's poorly formulated to the extent that it mentions the current bot activities.
  3. I can't view Option A as what the bot was doing before it was stopped, because the parameter has not been deprecated. (The statement by Nikkimaria that deprecation "in the context of CS1/CS2" means a red maintenance message will be shown is not something I can accept, as any reasonable software provider knows that advance notice is the first part of deprecation. Unfortunately, I don't usually work "in the context of CS1/CS2", so haven't argued before against this unhappy definition.)
  4. Option B, as written on this page, is a garbled mess, not only because of the "status quo" mention, but also because of the "Deprecation can be bundled into genfixes..." bit. Deprecation, as above, is (first) a documentation task, followed by a (presumably small) step to cause red messages to be generated (I'm not sure what's involved here). I don't see what would need to be "bundled into genfixes". I had the feeling that many !voting here saw "removal" as meaning elimination of usages, but maybe that's due to my own confusion. Your Option C (and your original formulations in general) were much clearer.

  There should never have been any bot activities, because (a) there's no consensus, yet, to deprecate non-hyphenates, (b) the params are not yet deprecated, and (c) the bots' edits have been largely accessdate => access-date only, so violate WP:COSMETICBOT. So I guess I'll be !voting for B, although I'd much rather choose your original Option C. This RfC as written has been too unclear (at least for me). — JohnFromPinckney (talk) 11:20, 14 February 2021 (UTC)

  Re The 2014 RfC determined only that the hyphenated version must exist: As called out in the proposal, it also said The documentation is to show this lowercase, hyphenated version as the one for "normal use". Hence my comment above: some of us recently tweaked the documentation of a few high-use templates to make the hyphenated version the privileged one, but I can't guarantee we got both /doc and TemplateData for every template that indirectly uses CS1/2. If we end up with both hyphen and no-hyphen equally valid (is that C?), then tweaking the documentation will be the only change visible to current editors. SHould there be a more valiant effort to track them all down? David Brooks (talk) 19:45, 14 February 2021 (UTC)

  If the consensus is for Option C, then nobody has to track down anything; we can leave the documentation as it is. BTW, the claim written there, This will also mean that the deprecated parameter list will need to be updated to remove the non-hyphenated parameters is another red herring and would, it seems, not apply at all (as accessdate isn't even listed there yet).

  If we chose Options A or B, then yes, the documentation should be immediately changed. I can't believe it will take too much valor to find the Template:Cite web documentation, as it doesn't seem terrible exotic, but it should be included in the tweaking in those cases. — JohnFromPinckney (talk) 20:07, 14 February 2021 (UTC)

  • The 2014 RFC had a participation of only seven people and was six or seven years ago. It is patiently absurd to update documentation with such a sweeping change based on that alone, and any such changes ought to be reverted pending the outcome of a more clear RFC. Furthermore, the 2014 RFC specifically indicated that nothing would be depreciated, so if you are relying on it as a justification for any changes then obviously no non-hyphenated versions can be depreciated until / unless a new RFC overturns it. --Aquillion (talk) 22:19, 15 February 2021 (UTC)

  Please read the summary at the top of this Discussion section. Here's the summary of that summary: Dozens of unhyphenated multi-word parameters have been deprecated, removed from pages, and then removed from the Citation Style 1 templates over the last seven years. There are only six left. During those seven years, many new, hyphenated multi-word parameters have been introduced, without unhyphenated aliases. At present, the situation is that unhyphenated multi-word parameters are the standard, and there is just a bit more work to do to remove the final six outliers. Unfortunately, it appears that many editors have not enabled the useful settings "Expand watchlist to show all changes, not just the most recent" and "Group changes by page in recent changes and watchlist" so that bot edits can be hidden without losing visibility of bad human edits, so people are complaining about a bot "clogging" their watchlists. – Jonesey95 (talk) 23:59, 15 February 2021 (UTC)

  None of that is however relevant to Aquillion's comment in the slightest. That a flimsy consensus (at best) has been used to justify removing functionality in the past does not mean that that removal was a good thing or that the bot edits (which clog both watchlists and page histories with unnecessary edits, whether they hide other edits or not) should continue. Indeed, I strongly suspect there would be support for re-enabling the parameters already removed. Thryduulf (talk) 00:27, 16 February 2021 (UTC)

Please note, folks, that this bot is a one-off; it may be processing some 2 or 3 million pages, but it's still a one-off, that is (unless reverted) it will only ever appear ONCE in any article history. So the idea that it's "clogging up" page histories is a big red herring. If you're bothered about it "clogging up" your watchlist, then please set the options recommended by Jonesey95 and others. So that objection falls by the wayside too. Please also note that this bot finishes the job. Once it's done, that's it – no more bots making this sort of change. To the charge that this bot is "disruptive" or that it's somehow "removing functionality", I can't do better than copy Rexx's observation above: "Once we have standardised on hyphenated parameters, future editors will look back and think how lame and time-wasting this sort of debate is." --NSH001 (talk) 19:30, 21 February 2021 (UTC)

Except that, no, that's completely inaccurate. Consider List of University of Pennsylvania people. When I look at the last 500 edits just now, I see that Monkbot has been there SEVEN times: here (21:07, 19 October 2020) and here (01:02, 28 December 2020) and here (17:21, 14 January 2021) and here (21:21, 16 January 2021) and here (01:53, 18 January 2021) and here (15:30, 18 January 2021) and here (20:54, 30 January 2021). While that first edit (from October) did nothing with |accessdate=, the others all did, as often as it found new additions to the article. Note that the fifth and sixth edits both occurred on the same day.

I don't ordinarily block bot edits from my watchlist because I want to know what's going on with "my" articles, and the bots generally do useful and interesting work. It's just that the (to me) sudden and unforeseen flurry of multitudinous edits (doing, it appears, nothing which really interests me) was quite irritating. The repetition on List of University of Pennsylvania people really got my blood pressure up, and that's not far off from "disruptive" to me. — JohnFromPinckney (talk) 21:15, 21 February 2021 (UTC)

Hmm... The October edit is Monkbot 17 (not 18), so doesn't fall into the scope of this RfC. The edit on 28 December is the main application of this bot. That leaves 5 (mostly quite small) unexpected additional edits by the bot. They are all caused by editors adding parms that don't conform to the canonical standard. So yes, I should have qualified my statement by allowing for that possibility, sorry for that. Understandable that this should happen in the absence of a clear deprecation (so far) of the unhyphenated form. Perhaps the editors concerned are using some cite-generating tool that doesn't generate the canonical form, in which case the tool should be updated. Whatever, that strengthens the case for deprecating the non-canonical forms ASAP, and for moving forward as fast as possible with the main task. It's late now, and I will be going to bed shortly. Will comment on Phil Bidger's intemperate remark below in the morning. --NSH001 (talk) 00:06, 22 February 2021 (UTC)

Naw, the editor concerned is/was just copying/pasting whatever they found in whatever articles they were looking at. (The user also hasn't learned that refs go after punctuation, or that a named ref copied from another article might not be declared on the target page, or that Wikipedia has a Preview function to allow checking for errors. Not that I'm bitter.) Agreed, without actual deprecation, nobody knows what they're supposed to use or not use, so watchlists get plagued with unnecessary repetition. And I can't point such a user to the deprecation or consensus not to use the non-hyphenated forms, because there aren't any yet. — JohnFromPinckney (talk) 00:21, 22 February 2021 (UTC)

Thanks for that. If it is the case that editors on that page are simply copy-pasting cites from the original wiki bios, then that's good news – the problem will go away if the bot is simply allowed to continue and fix the problem in the source articles. I don't think it's true that there is no consensus for this change: the desired style was settled in an RfC several years ago, and has been 90% implemented in the years since, with no substantial objection. So there is an effective consensus, and it makes no sense not to carry the task through to its logical conclusion. The objections here amount to a dislike of large numbers of bot edits appearing on watchlists, not on the actual merits of the case. --NSH001 (talk) 07:34, 22 February 2021 (UTC)

The RfC in question only had seven supports, and only concerned making sure a hyphenated version existed and was presented first in the documentation. I don't know how that could be read as effective consensus for deprecating a parameter that, prior to the bot run, seems to have been more commonly used than the hyphenated variant - and even if it could, it would be a limited one. The objection to the bot is not just that it edits a lot, but that it "fixes" something that isn't a problem. Nikkimaria (talk) 14:01, 22 February 2021 (UTC)

Exactly this. — JohnFromPinckney (talk) 02:05, 23 February 2021 (UTC)

Nikkimaria, the first fallacy in your argument is that you are assuming the wider community, if it participated in the original RfC, would disagree with its conclusion. That isn't (quite) the case – I'm pretty sure that, out of the various choices available for multi-word parameters (runon, camelCase, under_score, hyphenation, etc), hyphenation would still be chosen, simply because it is the easiest to read in wikitext. Possibly underscore might be better (it won't line-wrap) but most people, apart perhaps from those with a programming background, will be much more comfortable with the hyphen, so that RfC came to a very sensible conclusion. The detail of its implementation is another matter, though. The second fallacy in your argument concerns the reason for the objection to the bot. Firstly, the observable fact is that some editors are now kicking up a huge time-wasting fuss about this bot, but said nothing about the many other bot runs for all the other CS1/2 parameters doing exactly the same thing. Secondly, and I see this repeatedly in other RfCs as well, that editors tend to focus solely on a perceived short-term problem, without taking account of the bigger picture and the wider context. That context has already been well described in the introduction to this RfC, and in the links given there. It's astonishing to me that people don't see that that the whole point of this work is to make the citation templates simpler and easier to use. Part of that process is to make the parameter names more meaningful and consistent. It's ridiculous to leave the mess inherited from the early days of citation templates, with these few remaining parameters sticking out like sore thumbs. --NSH001 (talk) 08:25, 25 February 2021 (UTC)

Please explain how allowing e.g. "accessdate" is less meaningful, is less simple, is harder to use for regular editors. That is the bigger picture, the wider coontext: the benefits are actually only for a small group of people, who have every right to present their case and ask if their life can be made easier, but should not be astonished when it turns out that in some cases, their preferred solution is not supported by the larger group of people who don't do (or not as regularly) the template and bot stuff. Putting error messages on thousands of pages for things which are not an error but something which a few people decided is no longer allowed at all is losing sight of the bigger picture as well. Fram (talk) 09:24, 25 February 2021 (UTC)

you are assuming the wider community, if it participated in the original RfC, would disagree with its conclusion. I don't know that, and neither do you, because the wider community was never consulted. We can be reasonably sure that the discussion would have been far less one-sided, based on subsequent commentary. As to the second half of your point, as Fram notes, it's not at all clear why deprecating widely-used aliases would make the templates easier to use for the average editor. Nikkimaria (talk) 13:33, 25 February 2021 (UTC)

(replying to both). Yes, that's easy. There's a very simple rule for all the CS1/2 citation templates: all multi-word parameters use a hyphen to separate the words. That's it. Dead easy to remember. Moreover, it's already implemented for the vast majority of the parameters (only 5 are left to do). I don't buy the argument that having to type ONE extra character is an insufferable burden. The only valid objection I can see is that the bot may flood watchlists, but that is temporary until the bot run is finished. So my sympathies to those who feel irritated (I don't – I have over 6,000 pages on my watchlist, and bot spam doesn't bother me), but the irritation will be temporary. If it really does bother you, others have mentioned a way to configure your watchlist to avoid the problem. --NSH001 (talk) 07:35, 8 March 2021 (UTC)

I'm glad you personally are not irritated by this change - but others are, and others have explained why hiding the problem is not solving it. No one is proposing removing the hyphenated variation, simply supporting the (often more widely used) aliases. It's not appropriate to justify a change on the basis of it being mostly carried out. Nikkimaria (talk) 02:49, 12 March 2021 (UTC)

On watchlists, see #Worth noting below for a possible way of reducing the "disruption". I forgot to reply to Fram's point about the wider context: I was thinking about the overall naming convention, and why the bot makes it easier, but Trappist's latest contribution to the survey section explains also the wider consideration that we need to consider all the non-English Wikipedias that have borrowed our CS1 templates/modules. --NSH001 (talk) 10:40, 8 March 2021 (UTC)

So the solution to a bot editing against consensus is for those who object to stop watching it? You couldn't make this stuff up. Once again, this is a fucking encyclopedia, not a place for techies to dictate to editors. Find a playground elsewhere, or get a real job and you will find out that people can only get paid to write programs if thay do what their users want them to. Phil Bridger (talk) 21:27, 21 February 2021 (UTC)

Phil, your first sentence is false. This bot was approved by consensus, following standard procedures. Indeed its operator went to extraordinary lengths to shout from the rooftops that it is a "cosmetic" bot. I'll ignore your intemperate and baseless personal attack. Finally, on the question of bot edits and watchlists, I refer you to Bilorv's conribution at 11:34, 13 February above, which looks like a good solution (I'll just add the caveat that I haven't tested it yet). --NSH001 (talk) 08:25, 25 February 2021 (UTC)

Reading some of the above comments, I'm sorely tempted to add a new proposal here: every editor who deliberately makes a change which adds an error message to at least 10,000 pages is stripped from their template editor right. Excluding hidden cats of course, these aren't a problem; but no depreciation of any parameter justifies such mainspace disruption for readers. Fram (talk) 08:30, 24 February 2021 (UTC)

Three responses: First of all, the error messages displayed by CS1 templates are displayed by consensus, not by a single editor. Second, the error messages, such as those displayed in articles within Category:CS1 errors: unsupported parameter, are shown not because a single editor changed a template, but because individual editors made errors when they used CS1 templates. Third, the objection to error messages being splashed across article space is why the bot was operating before the deprecation error messages were displayed. People hate the display of hundreds of thousands of minor error messages, so the bot was fixing the articles before the CS1 modules were changed to display deprecated-parameter errors.

Fram's feedback offer give something to think about, however; if the logical options A or B are chosen here so that the last 10% of the hyphenation of multi-word CS1 parameters can be completed, perhaps we should not display deprecated-parameter error messages (except maybe in preview mode) in articles until the vast majority of parameter name replacements are complete. – Jonesey95 (talk) 16:29, 24 February 2021 (UTC)

Thanks, but remember Wikipedia:Administrators' noticeboard/Archive313#Is there a semi-automated tool that could fix these annoying "Cite Web" errors? from 1 1/2 year ago? There also was "consensus" for that change, among a tiny group of editors, but disregarding the wider community. I hoped that that episode would have learned some of the people most active at these templates that, when they propose a change affecting many pages (and certainly when they propose a change adding error messages to many pages), they should get a much wider consensus first. Still, I see in the above discussion people arguing to activate the red error messages for this (most error messages we get now are either on very few pages or for actual errors, e.g. impossible dates), which isn't an error but something some bot operators and template builders don't like. Fram (talk) 17:06, 24 February 2021 (UTC)

Three responses: First of all, the error messages displayed by CS1 templates are displayed by consensus, not by a single editor. One thing that this discussion has made crystal-clear, I think, is that many of the "consensuses" used to make sweeping changes like this are not remotely sufficient in terms of scale per WP:LOCALCONSENSUS; again, if they were, we wouldn't be having this conversation. If you want to make a change that significantly affects hundreds of thousands of pages, you should need a consensus involving a very large number of users, and it should be properly broadcast on high-traffic boards - a seven-person "consensus" is patiently insufficient for a change at this scale, and turning around and using bots or template messages to then try to enforce it amounts to WP:FAITACCOMPLI. --Aquillion (talk) 10:03, 4 March 2021 (UTC)

I wish that people would stop repeating this "seven-person consensus" canard; it is a weak platform on which to base an argument. The consensus about hyphenated parameters has been in place for seven years, and has been reinforced by multiple discussions about deprecation of dozens of individual multi-word parameters during those seven years. The discussion page on which every single one of those discussions has occurred, Help talk:Citation Style 1, has 396 page watchers. – Jonesey95 (talk) 05:48, 11 March 2021 (UTC)

This page has ten times that, and I'm willing to bet there are more people opposing deprecation here than there are who supported it in the discussions you mention put together. Repeating a local consensus for less used parameters != an appropriate level of consensus to get rid of other parameters from literally millions of articles. Not to mention that the consensus that was supposedly established seven years ago wasn't even for deprecation, just prioritization. Nikkimaria (talk) 02:49, 12 March 2021 (UTC)

• I wish that people would stop repeating this "seven-person consensus" canard; it is a weak platform on which to base an argument. The consensus about hyphenated parameters has been in place for seven years To be clear, the consensus of that RFC was that non-hyphenated parameters are allowed, subject to the unambiguous restriction that no parameters can be depreciated. The RFC specifically did not favor non-hyphenated parameters over hyphenated ones, and the statement at its start specifically promised that no parameters would be depreciated as a result. There has never been any sort of consensus (not even a weak, highly-local one) to depreciate hyphenated parameters; nor, as a result, have any depreciations made on those grounds ever been valid. And it is clear from this discussion that such consensus would never have been reached if it had been sought (which it was not.) A discussion among a tiny number of people, without an RFC, which directly violated the very RFC they tried to use to justify their actions, with no further RFCs or any effort to get consensus from or even inform the wider community of what they were doing, is not a "consensus" in any way, shape, or form - longstanding consensuses get their weight from the large number of people who have seen and accepted them, and in this case the longstanding consensus is (and remains) to retain hyphenated parameters. --Aquillion (talk) 10:31, 16 March 2021 (UTC)

This is uh, not a good idea. In some cases many templates are used in error, though they previously did not detect such errors. Detecting and fixing such errors is a good thing. IAR is policy, and if someone is breaking tons of things, sure, remove their rights, but simply displaying error messages isn't a clear-cut issue.

As for non-silent parameter deprecation, yeah, replace first then deprecate. I don't think anyone disagrees with this. Elliot321 (talk | contribs) 20:30, 3 March 2021 (UTC)

Well, I certainly disagree (and have, several times, on this page already). If there's consensus to do away with certain parameters, then the very first thing to do is deprecate them, that is, tell everybody not to use them anymore, next, run the bots to replace the usage, then, eventually, show the red messages and, ultimately, completely remove support for the deprecated params. Any other path is crazy. — JohnFromPinckney (talk) 03:54, 4 March 2021 (UTC)

• One of the objections I raised with Monkbot18 which is not being addressed whatsoever in the "status quo" (option A) argument is that the bot also make other changes, including the removal of whitespace and line breaks from citation templates. This is unacceptable per MOS:STYLERET. In my mind, the time I have spent undoing these changes to the articles I focus on has been a far bigger burden than the lack or presence of a hyphen in a parameter. If this bot is tasked with swapping parameters, it should ONLY be changing "accessdate" to "access-date" (and vis a vis similar parameter hyphenations), and absolutely nothing outside of that. Option A should not be construed as approval of the bot code as currently written, but the task for which it was meant to be accomplishing. - Floydian ^τ _¢ 18:33, 7 March 2021 (UTC)

  Well, that's very odd, because Monkbot18 is very careful to preserve whitespace (with one very reasonable exception), so I don't see how it's going against STYLERET. It's true that it does remove empty/blank parameters, but that's a good thing, as it reduces clutter in the wikitext. It does a few other good things as well, see User:Monkbot/task_18: cosmetic cs1 template cleanup. Can you give specific examples of edits you think it's doing wrong, please? --NSH001 (talk) 19:13, 7 March 2021 (UTC)

  There's a nice list of reversions if you look at my edits for January 17, but here is an example. - Floydian ^τ _¢ 06:27, 8 March 2021 (UTC)

  Ah, that's the "one very reasonable exception" that I referred to. All the other changes that you felt you needed to make in your "cleanup and standardise first 150 refs. Revert stupid friggen MonkBot making stylistic changes to thousands of articles on a "consensus" of like 3 people, and block it from making further edits" edit are down to other editors/bots, not Monkbot18. I think a blank line within a cite template is always unnecessary, and uses up valuable space within the edit window, but if it bothers you that much, you can always ask Trappist nicely, and I'm sure he'll remove that tweak for you. --NSH001 (talk) 07:15, 8 March 2021 (UTC)

  In that case, Your Mileage May Vary, and STYLERET applies. The blank lines make it easier to pick out citations from text, and scroll bars overcome the "valuable space" argument. I tried to ask that behaviour to be removed. I was met with (*paraphrased) "bot code was approved, proceeding". So I just blocked the damn bot from the 300 articles I work on. Now if someone is that concerned about a hyphen in a parameter, they can go manually change it. Simple as that.

  Or, you know, the bot could stick to making the changes to parameters that it is supposed to. I shouldn't need to ask, it's not part of the bot's mandate, remove it. - Floydian ^τ _¢ 15:43, 8 March 2021 (UTC)

  First, you say I tried to ask that behaviour to be removed. I was met with (*paraphrased) "bot code was approved, proceeding". That is surprising, as I would expect Trappist to be amenable to such a request (it's important to get this job done, so a small change like this one to avoid pushback is worth making). Can you point me to the conversation where you made this request, and received that reply, please?

  (later) - ah, don't bother – found it myself User talk:Trappist the monk/Archive 17#Task 18 taking out reference spacing. Looks like you didn't explain yourself very well to Trappist. FWIW, I can see the rationale for the blank line for in-line cites within the article body, if it stops people from turning it into the (horrible) horizontal format. I touch on this again below, but I agree with you that Monkbot18 shouldn't be removing the blank line. (I hope Trappist is reading this). But the rest of the changes it's making are good, and should stay.(Actually, to be strictly correct, Monkbot should remove it within LDR or biblio listings, but not within the main body. For simplicity, I'd say simply don't bother trying to remove it.)

  Secondly, Or, you know, the bot could stick to making the changes to parameters that it is supposed to. That statement is false. The bot was approved to carry out the tasks listed at the link I've already given: User:Monkbot/task_18: cosmetic cs1 template cleanup, and that's exactly what the bot has been doing. Apart from the blank line issue, the other changes are good, and valuable, and will reduce the need for more bot runs in the future. One of the reasons I like this bot so much.

  The next bit is very interesting (to me) but is wandering mostly off-topic, so I'm putting it in small text. If you'd like to take it further, feel free to discuss it on my talk page. The problem of citation clutter in the main body of Wikipedia articles has been annoying me for years, and especially the huge problems caused by long, horizontally-formatted citation templates, which in my opinion make wikitext difficult or impossible to read and to edit. This is all set out on a very long "thread" (it isn't really a thread any longer): on my talk page. It is talking about a way of setting out citation templates that I call "ETVP" for "easy to visually parse", which is similar in many ways to the citations in your example, but also differs in some respects. The interesting point to mention here is that I had a difficult and bruising experience trying to introduce ETVP-formatted citations into article bodies. The excuse offered for reverting me was mostly "it takes up too many lines" in the edit box (hence "the one very reasoable exception" above), so in the end I gave up on that, and focused mainly on other solutions (ETVP within WP:LDR or ETVP within biblio listings using short-form referencing) which in most cases are actually a better solution anyway. It's a fascinating paradox that you seem to have gotten away with it by using more lines, not fewer, combined with a lavish use of white space – the exact opposite of what one would expect. So I am now thinking of adding a similar option to my ETVP script – and thank you for prompting that thought. Will need some thought, though.

  --NSH001 (talk) 10:02, 9 March 2021 (UTC)

  The blank lines is the only issue I'm having / pointing out here. I have no problem with updating deprecated parameters, I have no problem with my watchlist having a litany of bot edits. I do have a problem with going through 250 articles that have an average of 50 citations, to reinsert a blank line in each. This is not one of the 6 tasks listed at User:Monkbot/task_18: cosmetic cs1 template cleanup.
  There are also some automated tools that do similar nonsense that I revert on sight (e.g. Regex Citation Formatter). - Floydian ^τ _¢ 15:49, 10 March 2021 (UTC)

  We appear to be in agreement here, specifically that you support option A, but only if Monkbot 18 drops its removal of entirely blank lines within citation templates. If you could confirm that my understanding is correct, that would be very helpful. Thanks. --NSH001 (talk) 10:00, 11 March 2021 (UTC)

  You would be correct good sir! - Floydian ^τ _¢ 15:36, 11 March 2021 (UTC)

Worth noting

Trappist has kindly set out, very clearly, a suggestion on how to configure your watchlist to avoid the problems of large numbers of bot edits in watchlists. I copy it here, in case it is helpful to anybody:

• at Special:Preferences#mw-prefsection-watchlist:
  • check Advanced options → Expand watchlist to show all changes, not just the most recent
  • check Changes shown → Hide bot edits from the watchlist
• at Special:Preferences#mw-prefsection-rc:
  • check Advanced options → Group changes by page in recent changes and watchlist

Note that I haven't (yet) tried this myself, since I'm already satisfied with the way my watchlist is set up. --NSH001 (talk) 09:09, 8 March 2021 (UTC)

Note to RFC closer: the above instructions are a remedy for all of the people supporting Option C because of "watchlist clogging" and the alleged problems that the bot's edits cause for editors who watch for vandalism. As far as I can tell, no editor using the above settings is objecting to the bot's changes based on watchlist issues. Unless a valid objection is raised, I propose that all Option C support citing watchlist problems be discounted and guided to the above recommendation. – Jonesey95 (talk) 17:30, 8 March 2021 (UTC)

This is a workaround that (a) should not be necessary, (b) doesn't work for everybody, e.g. those who want to see bot edits (I do for example) and (c) doesn't fix the problem only a symptom. It is additionally highly inappropriate to suggest that large numbers of editor's valid and rationally expressed preferences are discounted. Thryduulf (talk) 17:41, 8 March 2021 (UTC)

I've been trialing this workaround for the last day or so. Back in December I cleared out my watchlist and took a two and half month wikibreak while the bot ran. I've just come back to Wiki and discovered that the bot has been stopped, but thought I'd try Trappist's suggestion. So far so good, it's a bit different but at least it makes the watchlist saner than during the bot attack! Martin of Sheffield (talk) 20:23, 8 March 2021 (UTC)

I've always been puzzled that (some) editors get so triggered by large numbers of bot edits. I have more than 6,000 pages on my watchlist (which I have set to show bot edits), and bot spam has never bothered me. I even welcome it, as they sometimes remind me of articles I did a lot of work on perhaps 7 or 10 years ago, and which I really ought to look at again. That said, I do understand the problems of editors who want to deal with vandalism, so this new setting looks to be very valuable, and should indeed remove many (but probably not all) of the "watchlist spam" objections. --NSH001 (talk) 10:40, 9 March 2021 (UTC)

And on the subject of editors like me who aren't bothered by bot spam, has anyone considered the silent majority who either aren't bothered by the "spam" or who, if they are, aren't concerned enough to come to this RfC to complain about it? Perhaps they ought to be weighed somehow in the balance when considering "consensus"? --NSH001 (talk) 10:51, 9 March 2021 (UTC)

Well given that most of them will not know that this discussion exists and will just be getting on with adding the parameters, with and without hyphens, as they always have done I don't think was can say one way or the other what their opinion is - especially given that the bot has not been spamming its unnecessary changes for quite a while now (the discussion has been open a month tomorrow, and I think it was stopped a day or so before then, and more than a few editors are of the opinion that arguing against bots/bot operators is pointless). Instead of grasping at straws to discredit or dismiss the opinions you disagree with, perhaps you could instead try listening to why they disagree with the changes, not just the manner of the changes. I also note that you have completely ignored my explanation of why this will not actually solve the problem for everybody and ignored that there is no reason why the problem should need to be solved in the first place. Thryduulf (talk) 11:49, 9 March 2021 (UTC)

None of this changes the fact that there is no consensus to depreciate non-hyphenated parameters, nor has any such consensus ever existed. Rather than trying to convince the RFC closer, you should be planning how you're going to get that consensus, if you intend to keep doubling down on that. --Aquillion (talk) 10:35, 16 March 2021 (UTC)

• Comment It's easier for authors if every possible format is accommodated, and recognized as a synonym. There's no reason to delete any unless they;re actually confusing. DGG ( talk ) 05:59, 21 March 2021 (UTC)

RfC: Disable minor edits on English Wikipedia

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

There is a strong consensus against limiting minor edits to anti-vandalism as suggested in the proposal. Alternate proposals of limiting minor edits to autoconfirmed or extended-confirmed editors have attracted considerable support in the discussion, but it's difficult to tell how extensive it is since many respondents did not mention the alternate proposals in their comments. Therefore, another RfC to limit the use of minor edits to autoconfirmed editors may be in order. (non-admin closure) (t · c) buidhe 10:52, 23 March 2021 (UTC)

---

Proposed: Effectively disable the "minor edits" functionality on the English Wikipedia. Its usage will be limited by policy for anti-vandalism reverts only. Technically, the permission to mark as minor will be limited to rollbackers, admins and bots. ProcrastinatingReader (talk) 19:56, 15 February 2021 (UTC)

Survey (minor edits)

• Support The point is apparently for some editors to ignore "m" edits on watchlists. No experienced editor would do so because "m" edits can be as wrong, disruptive, or destructive as a "major" edit. Vandals can use them for vandalism, newbie editors in good faith with problematic changes, and experienced editors may make something they believe is minor but other editors disagree. Really, all edits are subject to dispute. The point of a watchlist is to monitor articles. What's more, we apparently block editors for minor edits related issues. The functionality is not only useless, it's a net negative. Minor edits can be communicated via the edit summary and the diff count. ProcrastinatingReader (talk) 19:56, 15 February 2021 (UTC)
• Oppose. I mark edits as minor all the time when they are, in fact, minor. Fixing typos and the like need not bother editors who don't want to see those edits. If there is a problem with non-minor edits being marked as minor, figure out how to figure those out. We already flag, for example, possible changes in birth and death dates. BD2412 T 20:03, 15 February 2021 (UTC)
  • Note: I support the below proposals to limit minor edits to either autoconfirmed or extended confirmed editors (preferably the latter). BD2412 T 04:56, 3 March 2021 (UTC)
• Oppose a policy that minor should only be for vandalism reversion; I agree with BD2412 that actions such as small typo fixes, small whitespace adjustments, etc should still qualify as "minor" for patrolling and review purposes. Nuetral on an overlapping component of this, the removal of the minoredit permission from "All Users" - if it is being misused frequently by newer users might be helpful - but if so I think it should go to +extendedconfirmed and/or other groups (i.e. rollbacker/patroller) that otherwise help recognize that an editor has a modicum of experience here. — xaosflux ^Talk 20:23, 15 February 2021 (UTC)
  • A perfectly cromulent use case is by a bot that has to make a non-content clean up to user talk pages, by declaring the edit minor and leveraging their nominornewtalk permission - the operator can avoid bothering users with a new messages notification. While "bots" were in the original proposal, this use case fails the "anti-vandalism reverts only" criteria. — xaosflux ^Talk 20:41, 15 February 2021 (UTC)

    For clarity, I meant anti-vandalism only for human editors (bots being able to use it as they do). ProcrastinatingReader (talk) 20:44, 15 February 2021 (UTC)

    @ProcrastinatingReader: OK thanks, any bot use should already be governed by a specific WP:BRFA for whatever it is they are doing (and this is almost always used in addition to the 'b'ot flag). — xaosflux ^Talk 20:52, 15 February 2021 (UTC)

    Re typos, a typo fix can still be disputed. Perhaps the editor got it wrong and the 'typo' was intentional. More importantly, Suffusion of Yellow's comment, especially about the evil bit, is a good way to phrase the issue imo. So long as some disputed edits can flow through it, the entire feature is worthless. The anti-vandalism allowance is because that's already governed by existing policy, WP:ROLLBACKUSE and WP:VAND etc, and allows pure 'junk' to be filtered out. Though I'm also okay with removing this exemption. ProcrastinatingReader (talk) 20:55, 15 February 2021 (UTC)

• Support. So long as this feature is available to spammers, vandals, and the clueless, it's as useless as the evil bit. It's also a source of needless drama, when users innocently overuse it. But if people think this proposal goes too far, limiting it to extendedconfirmed would be a good start. Suffusion of Yellow (talk) 20:33, 15 February 2021 (UTC)
• Oppose Why? Mike Peel (talk) 20:35, 15 February 2021 (UTC)
• Support, tentatively, disabling the option when making a manual edit. It's simply not that useful; even experienced users don't agree on what it means, and don't use it consistently. The byte count is a much more effective way to identify "minor" changes at a glance, since it's less prone to manipulation. Because the minor flag can be set inappropriately, it's almost never reasonable to completely filter out minor edits, so it simply acts as a weak signal of intent on page histories, redundant to edit summary and byte count. Not convinced about making it antivandalism-only, though. — The Earwig ⟨talk⟩ 20:37, 15 February 2021 (UTC)

  After reading the comments in opposition and consideration of alternate solutions for the problem, I don't think this proposal as written is the way to go. — The Earwig ⟨talk⟩ 19:13, 16 February 2021 (UTC)

  Support limiting to autoconfirmed as a first step. Oppose limiting to extended confirmed. I understand the arguments in favor, but I generally oppose expanding the role of EC by taking privileges away from autoconfirmed users (giving EC users privileges otherwise only available to more restricted users is another thing). Even with this we could not prevent every mistagging of edits as minor, nor should we strive to. The majority of minor edits tagged by someone with, say, 2 months of experience and 100 edits will be fine, and if they aren't, they're not likely to figure it out by the 500th edit. — The Earwig (talk) 06:46, 3 March 2021 (UTC)

• Oppose as written. It might make sense to further limit who can mark edits as minor, but eliminating them nearly wholesale as this proposal does goes too far. This is using a bunker buster to wipeout a few hornets. -- Calidum 21:11, 15 February 2021 (UTC)
• Support the idea of disabling the option to mark edits as minor when editing manually. I don’t really see the point of making the ‘minor’ flag CV only, or granting it to admins or rollbackers. If we think it’s useless, then let’s deprecate it fully, rather than trying to debate who is experienced enough to use it. The only truly clear use for minor edits is bots editing user talk pages as Xaosflux sagely said above, so bots should obviously keep the right for that reason. Bot edits can already be filtered from watchlists, so other than on user talk pages it doesn’t matter whether bot edits are minor or not. ƒirefly ( t · c ) 21:35, 15 February 2021 (UTC)

  Other uses of minor edits include (but are not limited to):

  • Spelling fixes
  • Capitalisation fixes
  • Typo fixes
  • Grammar fixes
  • bold/italic fixes
  • List order fixes
  • Markup fixes (templates, tables, etc)
  • Small corrections to comments (e.g. missing words)
  • Signing unsigned comments
  • Whitespace fixes
  • Addition of templates that have no or minimal impact on content (e.g. {{As of}}, {{Update after}}, {{convert}})
  • Adjusting links
  • Adding/adjusting hatnotes
  • Protecting/unprotecting pages (automatically marked minor). Thryduulf (talk) 21:53, 15 February 2021 (UTC)

• Very strong oppose. There is (probably) an issue that needs resolving behind this proposal, but there is no evidence presented for any of (a) the problem being widespread (i.e. that the solution lies with something other than educating and/or blocking individual users), (b) restricting minor edits will solve the underlying issue, (c) that restricting minor edits as much as is proposed is either necessary or proportionate, or (d) that the inevitable collateral damage will cause fewer and less significant problems than the original one. It is therefore way, way too premature to be bringing this to this board, let alone an RfC - it has very clearly not been thought through quickly, let alone repeatedly, fully and in-depth. Thryduulf (talk) 21:37, 15 February 2021 (UTC)

  it has very clearly not been thought through quickly, let alone repeatedly, fully and in-depth I've discussed this with other editors in multiple AN/ANIs last and this year, and this directly came from a VPT from this week. So that's not true. ProcrastinatingReader (talk) 22:26, 15 February 2021 (UTC)

  This is vague and lacking in so many ways I'm truly gobsmacked that multiple editors thought this was viable proposal. Thryduulf (talk) 23:02, 15 February 2021 (UTC)

• Oppose. Most experienced editors use minor a lot when performing wholly uncontroversial typo fixes, minor formatting changes, wikifying and the like. This feels like a baby–bathwater situation. Espresso Addict (talk) 00:36, 16 February 2021 (UTC)

  The thing is in order for some people to gain some benefit from ignoring edits flagged as minor, others must continue to monitor all edits. At present it's like some people just play with the baby without taking a turn dealing with the bathwater. Which is OK when there's enough people who don't filter out minor edits, but it means that minor edit filtering inherently fails to scale up, and that makes me uneasy about touting its benefits for all. isaacl (talk) 00:59, 16 February 2021 (UTC)

  I think we're talking about slightly different things here. I see it as a signal from experienced editor to experienced editor that says nothing to see here, which seems unambiguously useful. You, I think, are complaining that editors who filter out minor edits in watchlists (or, like me, don't bother with a watchlist at all) are placing a burden on others to check them. That seems more a problem in the way watchlists are configured, rather than with the principle of edits being marked as minor/not. Perhaps an edit filter could be devised to flag edits that are clearly not minor, and remove the designation? Espresso Addict (talk) 02:02, 16 February 2021 (UTC)

  The signal is unfortunately not a reliable one. It's an AI problem to automatically detect minor edits. If it works well enough, then there wouldn't be any need for manual flagging. That might be a good idea, but I'm not sure if it should be given priority over other developer tasks, particularly given the likely high effort-to-benefit ratio. (I appreciate the advantage of the signal, when accurate, and have written about it in the discussion section.) isaacl (talk) 03:13, 16 February 2021 (UTC)

  I wouldn't object strongly to restricting the ability to mark edits as minor to, say, extended confirmed editors; or more usefully, letting inexperienced editors continue to mark them (so that they learn to do it to community norms), but ignoring the mark (from inexperienced editors) in watchlist presentations. I don't think AI is going to differentiate them accurately enough any time soon. Espresso Addict (talk) 03:48, 16 February 2021 (UTC)

  It already has the ability to filter based on experience and minor edits (as well as a contribution quality prediction), but I don't think it can be set for all edits from inexperienced editors plus non-minor edits from experienced editors. The reliability issue remains, even with experienced editors, so someone ought to be checking all minor edits. isaacl (talk) 04:06, 16 February 2021 (UTC)

• Oppose, but open to reform. The minor edit designation is a piece of information, just like the edit summary, that has to be taken in context. When I see an "m" next to a reputable username, that saves me the effort of reviewing the edit. When it's next to a giant edit from an IP a redlinked user with a suspicious summary, that's different. For that reason, I don't filter them from my watchlist, but if someone else wants to do so, they can. That said, I agree misuse of the minor edit box is a problem. Here's a more modest proposal: limit it to autoconfirmed editors, and the first time someone checks it, have the software generate a popup that quickly explains what we mean by "minor". With that, we could take a harder stance on misuse of the box, since no one could claim they just didn't know. That might get us closer to a point where more editors feel comfortable filtering minor edits from their watchlist. {{u|Sdkb}} ^talk 07:19, 16 February 2021 (UTC)

  @Sdkb: unless I'm missing something, IP's can't tag edits as minor today (please point to any diff if you can see one). Unconfirmed users can - so the next "small step" up would be to remove that access from "All users" and give it to "autoconfirmed" users. — xaosflux ^Talk 14:30, 16 February 2021 (UTC)

  You're correct; I've fixed my hypothetical. {{u|Sdkb}} ^talk 18:34, 16 February 2021 (UTC)

  @Sdkb: those are both steps that I can support, along with perhaps a byte limit for the size of a change to still be marked minor. BD2412 T 17:57, 16 February 2021 (UTC)

  A byte limit might be tricky; reverting a vandal blanking a page is a very valid minor edit, as it's clearly uncontroversial. As an alternative, maybe disallow if ORES flags it as possibly unconstructive? {{u|Sdkb}} ^talk 18:38, 16 February 2021 (UTC)

  Something like that would definitely be useful, yes. BD2412 T 00:20, 17 February 2021 (UTC)

• Support The checkbox adds complexity to the process of making each edit, with very little benefit to other editors, and just a little anxiety at perhaps getting it wrong. On average, I guess I spend half a second to a second on this decision each time. Doing the math, that adds up to 2 to 4 working weeks of my life over the past decade. -- John of Reading (talk) 08:28, 16 February 2021 (UTC)

  @John of Reading: if it bothers you personally you can add this one line to your Special:MyPage/common.css and you won't have to see that box anymore:

  #mw-editpage-minoredit {display: none;}
  

  — xaosflux ^Talk 19:57, 16 February 2021 (UTC)

  @Xaosflux: The checkbox I use most is the one in AWB, not the one in the edit window. But if I hide the checkbox, or always leave it unchecked, and some other editors think the distinction is important, then I'll be failing to signal correctly to them that most of my edits are minor. -- John of Reading (talk) 20:09, 16 February 2021 (UTC)

• Strong oppose as unnecessary and counter-helpful. A large percentage of my own edits are minor, and I like the ability to keep the groups separated (if only for myself). No reason other people should have to pore through my edits when all I did was correct spelling or fix date formats or eliminate the spaces before <ref>, etc. — JohnFromPinckney (talk) 18:11, 16 February 2021 (UTC)
• Support limiting the "minor edit" checkmark to experienced users. It is very unreliable when used by newbies. - Vis M (talk) 21:04, 16 February 2021 (UTC)
• Oppose A supposed vandalism revert isn't necessarily minor. And the fact that the flag might not be accurate is just like everything else on Wikpiedia. Every single page, edit, summary or whatever might not be accurate. Per the disclaimer "WIKIPEDIA MAKES NO GUARANTEE OF VALIDITY". Andrew🐉(talk) 23:12, 16 February 2021 (UTC)
• Oppose per BD2412 and xaosflux. While I'm open to stopping non-autoconfirmed editors from marking edits as 'minor', I am in disagreement with the bulk of the proposal. Sdrqaz (talk) 20:42, 18 February 2021 (UTC)
• Limit to autoconfirmed Any potential damage from misusing the minor edit flag is minimal---on par with moving a page. It takes time to figure out what's minor and what's not, and we already restrict IPs. While we can certainly improve our guidance on using the minor edit flag, the threat level doesn't warrant restricting autoconfirmed editors who we already trust with much more powerful tools in their hands. — Wug·a·po·des​ 22:31, 18 February 2021 (UTC)
• Limit to autoconfirmed. We already prevent IPs from using the box, and new users can be assumed to be similarly inexperienced. However it is definitely useful for exconf+, and seeing as it's, well, a minor feature, autoconf should have access to it as well. Would be open to Sdkb's idea of having a popup the first time someone ticks the box, though I'm not sure as to the feasability of this on the software end. —pythoncoder (talk | contribs) 19:54, 19 February 2021 (UTC)
• Nein - minor edits are edits that well, are minor. If we mark them as major edits, then it will only waste more editors time in the RC patrol backlog. — Preceding unsigned comment added by Awesome Aasim (talk • contribs) 03:15, 21 February 2021 (UTC)
• Limit to autoconfirmed. For experienced users, especially when working with other editors, this is an important flag. Anyone who does not trust the minor edit flag can still look at all edits, so why could there possibly be a need to remove this functionality?ThoughtIdRetired (talk) 15:07, 21 February 2021 (UTC)
• I very strongly oppose the idea as proposed, but agree with Pythoncoder (this is the second RfC today I've agreed with him on) that limiting it to autoconfirmed is a good idea. JJP...MASTER!_{[talk to] JJP... master?} 01:55, 22 February 2021 (UTC)
• Support with comments: (1) Admins should be permitted to mark any of their edits, vandalism-related or not, as ‘minor’ if they they deem doing so beneficial to the community. For the rest of us (non-bot editors), rollback seems as good a threshold as any. (2) ‘Minor’ may cease to be the optimal name for the tag. ‘Procedural’ might be more apt? No idea if there’s anyway to change that for the sake of new editors. —jameslucas ^{▄▄▄ ▄ ▄▄▄ ▄▄▄ ▄} 03:14, 23 February 2021 (UTC)
• Oppose Minor edits are important - the little changes to markup, spellings, white space, capitalisation, changes/updates to numbers, correcting typos: they are the little things that make the bigger machine work. By removing "minor edits" as a tag, you run the risk of flooding timelines with clutter, and potentially dissuade editors from making necessary minor changes in the first place. doktorb _words^deeds 07:25, 23 February 2021 (UTC)
• Oppose per Mike Peel. --Jayron32 19:31, 24 February 2021 (UTC)
• Oppose. If someone is inappropriately flagging non-minor edits as minor, that's a user conduct issue, not a reason to deprecate minor edits. Speaking as someone who's apparently made 318,850 minor edits, it's of no benefit and causes obvious disruption if readers who don't need to be made aware every time I quietly search-and-replace the misspelling "targetted" are unable to filter it from their watchlist. It would also make reviewing editors' contribution histories (whether it be for processes like RFA, or just "I'm looking for the diff of that edit you made last week") orders of magnitude more complicated, for no obvious benefit. ‑ Iridescent 19:45, 24 February 2021 (UTC)
• Oppose - Personally I don't ever use it even if I'm indeed making minor edits... It really is too much effort clicking on one box everytime I want to save changes... that being said not using it isn't a reason to support and no reason has been given as to why this should be disabled in the first place. –Davey2010^Talk 23:12, 24 February 2021 (UTC)
• Oppose. I am somewhat sympathetic to the proposal but the problem and its scope need to be much better defined, and the proposed solution appears to be too radical. I mark copyedits as minor all the time; same for manual delsort edits. If the feature is used properly, I find it quite useful; if an edit is marked as minor in my watchlist or in a page history, I am much more likely to ignore that edit. I agree probably is some issue with the misuse of this feature, but at the moment I am unsure about the extent of the problem. Perhaps less radical solutions, such as limiting the class of users with the userrights to mark edits as minor, and trying to provide a more precise definition of what constitutes a minor edit would be useful and should be attempted first. Nsk92 (talk) 23:50, 24 February 2021 (UTC)
• Oppose minor edits are useful to distinguish, well, which edits are minor. If minor edits are disabled, we have no way to distinguish here, so we lose information. Potentially faulty information - as improperly marked minor edits are - is worse than no information. Maybe limit to autoconfirmed, and in the worst possible case limit to extended confirmed, but removing entirely is not a good idea. Elliot321 (talk | contribs) 01:53, 28 February 2021 (UTC)
• Oppose as proposed per Xaosflux. I am open to limiting minor edits to autoconfirmed or extendedconfirmed, but not as proposed. Twassman | Talk | Contribs 06:48, 28 February 2021 (UTC)
• Support this makes "request minor edits feature" a main value of the rollback permission. Once somebody knows enough to want it, we should give it to them. power~enwiki (π, ν) 03:03, 2 March 2021 (UTC)
• Oppose A solution which would be far worse than the problem it seeks to address. Minor edits are useful. Limiting it to autoconfirmed users might solve most abuse (since most vandals are non-AC); if there is such abuse in the first place. RandomCanadian (talk / contribs) 03:44, 2 March 2021 (UTC)
• Oppose. While marking edits as minor isn't super useful, it does have some benefits when going through the edit history of an article. Of course the "minor edit" flag is just as reliable as the edit summary there (i.e. not very) and you shouldn't do anything automated based on the presence or absence of the flag. —Kusma (t·c) 11:55, 2 March 2021 (UTC)
• Support. Should be a privilege. At present it's worse than useless for vandalism and just superfluous cognitive overload for good-faith editors. Rollo (talk) 22:06, 2 March 2021 (UTC)
• Limit to extended-confirmed (autoconfirmed is too easy to get around) because only experienced editors should be able to mark edits minor. Then if I ignore minor-flagged edits, I won't have to worry about missing any non-EC edits. Levivich ^harass/_hound 04:42, 3 March 2021 (UTC)
• Oppose I don't think there is an Issue that needs solving. I think minor edits are a useful tool for actual minor edits. jort⁹³|talk 19:56, 4 March 2021 (UTC)
• Oppose This is a solution in search of a problem. Many users, myself included, do not abuse minor edits. This is like not allowing anyone to eat food because some people eat too much. Thank you, 🐔 Chicdat  ^{Bawk to me!} 11:58, 6 March 2021 (UTC)
• Oppose Just sanction those who abuse it. Minor edits are beneficial to this project overall. ~ HAL333 22:16, 7 March 2021 (UTC)
• Oppose as written. Support limiting it to autoconfirmed or extended confirmed. --Ahecht (TALK
  PAGE) 20:23, 9 March 2021 (UTC)
• Oppose Minor edits are clearly defined in WP:MINOR, and they have a good purpose. Besides, many people who revert vandalism are not rollbackers (like me) — Preceding unsigned comment added by Bop34 (talk • contribs) 20:46, 9 March 2021 (UTC)
• Oppose Just becuase some people may abuse minor edits it does not mean we should essentially remove their use to editors almost entirely.  Spy-cicle💥  ^Talk? 18:49, 10 March 2021 (UTC)
• Oppose, but open to reform, if we just got rid of minor edits, then it would either discourage people from making edits that are simply just spelling or grammar fixes or it would make it harder for those who moderate editing to figure out what is and isn't spam. A way we could reform this would be my making it so all edits under a certain size (say 100 bytes, this is simply an example) will be marked as minor and cannot be unmarked as minor. When I edited on Fandom (similar to Wikipedia except more for games) I would mark edits that are just me making small edits to articles (like fixing spelling or grammar) as minor because, well, they are. If we just removed the option to mark edits as minor then it might make people feel like they have to radically change the article. I'm probably repeating myself so I'm going to give another way to reform it. Just have all edits marked as minor by default. Yes, there is an option for this but I think people don't know that. Another way to reform, is, like I said above, have all edits under a certain size marked as minor and if someone attempts to unmark it as minor and its still under that size, a warning message pops up, letting the user know what the point of it being marked as minor is. When it gets over a certain size, it cannot be unmarked as minor unless the user has special permissions to do so (I.e they're a trusted Wikipedia editor). This is just my opinion but after reading all the stuff above I felt like I should contribute. Blaze The Wolf | Proud Furry and Wikipedia Editor (talk) 01:08, 16 March 2021 (UTC)
  • Diff size is not a reliable way to determine what is and isn't a minor edit. A rewrite of an entire paragraph is not a minor edit but if it used the same number of characters it would have a diff size of 0 bytes. At the other end of the scale, adding an archive link to sources can add 50-100 bytes of text per URI but as it makes no change to the prose it would be valid to mark that as a minor edit. Thryduulf (talk) 01:23, 16 March 2021 (UTC)
• Oppose as proposed, Support for limiting minor edit to autoconfirmed/extendedconfirmed editors.--Vulphere 04:56, 16 March 2021 (UTC)
• Strongest Oppose There are other solutions to the dishonest use of the "minor" button. This, as proposed, should not be implemented. GenQuest ^"scribble" 16:38, 16 March 2021 (UTC)
• Support Bots, reverts have an obvious place for minor edits. While experienced users don't all agree on definition of a minor edit, they'll at least use it consistently, which can be useful for reviewing edits of a particular user. They are less useful however, when monitoring recent changes of a watchlist. In order to make them more useful, I'd recommend limiting the ability to enable minor edits to WP:ECP users who'd at least have a consistent rationale for what's minor/major. Shushugah (talk) 16:52, 16 March 2021 (UTC)
• Oppose Minor edits are useful. Not only for me to use when correcting typos, or to help me bypass minor edits by editors I recognize, but also because the "minor edit" checked plus a canned edit summary like "Fixed typo" is a clear red flag to check for vandalism or other disruptive editing. – Muboshgu (talk) 16:54, 16 March 2021 (UTC)
• Oppose - when looking at an article's history, it can be useful to see that an edit has been marked

as minor, for example, changing a single letter to correct a typing error. This is useful because it helps to distinguish minor edits from more major changes in an article. Rollo August (talk) 20:15, 20 March 2021 (UTC)

• Oppose - may discourage future edits ParallaxVision222 (talk) 16:28, 22 March 2021 (UTC)

Discussion (minor edits)

Although the minor edits flag isn't a reliable one to use for filtering all edits, it can be used (visually) to filter edits from editors you know. This advantage is, of course, only available to you if some of your known editors actually use the flag. isaacl (talk) 21:13, 15 February 2021 (UTC)

@ProcrastinatingReader, would your actual problem be solved by changing the default prefs settings to show minor edits? Whatamidoing (WMF) (talk) 21:55, 15 February 2021 (UTC)

The issue I have is that the functionality doesn't make sense. I don't think it ever makes sense to actually hide minor edits in a watchlist. The indicator itself may be helpful nevertheless, but is still redundant to the summary + byte count. I think Suffusion of Yellow and The Earwig's comments put my concerns succinctly. My second issue is individual admins thinking blocking even a single editor for their use of the minor indicator (eg or blocking individual users) is appropriate. I think the issue is frequent enough, per it having a section in a policy page (Wikipedia:Vandalism#Gaming_the_system). ProcrastinatingReader (talk) 22:22, 15 February 2021 (UTC)

If you could configure specific editors for which minor edits would be filtered out, it might be more useful. But given the low amount of usage, and the amount of overhead processing required for that level of filtering, I don't think it's worth the development and ongoing sustaining cost. isaacl (talk) 22:40, 15 February 2021 (UTC)

How do you know which editors to filter out? I think it'd be difficult to create an exhaustive list of editors who are incapable of using the feature. If it's a local list, it'd be very difficult for editors to maintain it (besides, if you have minor edits hidden, how do you know which editors are misusing the minor feature to add to the list since, well, you won't see their edits?) If it's a global list, I sense more pointless ANI drama. Plus, it may just be one-off edits that require more scrutiny rather than continuous inability to determine what is minor or isn't. ProcrastinatingReader (talk) 22:44, 15 February 2021 (UTC)

It would be up to individual users to decide whom they trusted to properly flag minor edits, and then configure their watchlist accordingly. But like I said, I don't think the work to implement this warrants the benefit. I didn't realize the current filtering capability allowed for level of experience to be configured; I'm not sure if that is useful or not for one's watchlist. isaacl (talk) 00:03, 16 February 2021 (UTC)

@ProcrastinatingReader, if the function doesn't make sense to you, then why do you propose that certain editors should continue to use it?

I think that whether it makes sense depends upon which version of the watchlist you're looking at. If you have each edit displayed separately, then it could let you skip a lot of edits that you don't want to see (e.g., ClueBot reverting simple vandalism, which is marked 'minor' but not as a 'bot' edit).

The minor edit flag is displayed in Special:RecentChanges as well. This allows interested editors to filter (for example) for minor edits by less-experienced editors that are the most current version of the page. I imagine that this would be an interesting list for both anti-vandalism patrollers and people seeking out promising editors. Whatamidoing (WMF) (talk) 23:05, 15 February 2021 (UTC)

The answer to the first paragraph is your second paragraph: so that ClueBot, and Hugglers, reverting simple vandalism can continue to be filtered out of watchlists. It's less restricting it to certain editors, and rather restricting it for a certain purpose. Even rollbackers and admins wouldn't be allowed to mark as minor "grammar changes", for example, under this proposal. ProcrastinatingReader (talk) 23:21, 15 February 2021 (UTC)

That's the problem with this proposal - vandalism fixing is only one of many valid reasons to mark an editor minor, including grammar changes (a I wrote a non-exhaustive list above). You seem to be assuming that everybody uses Wikipedia in the exact same way you do, which is flat out wrong. Thryduulf (talk) 12:28, 16 February 2021 (UTC)

As an example, this edit is very clearly a minor edit yet it has nothing to do with vandalism. Thryduulf (talk) 12:39, 16 February 2021 (UTC)

• Since this proposal is unlikely to get consensus as written, here's a completely different suggestion, more narrowly targeted at what I see is the most egregious problem: the fact that we occasionally block otherwise productive editors for misuse of the minor edit feature. Allow admins to disable the minor edit flag for an editor as part of the blocking interface, in the same vein as partial blocks or disabling email, but without restricting the ability to edit. — The Earwig ⟨talk⟩ 15:21, 16 February 2021 (UTC)
  • Two thoughts. The first, is it actually technically possible? If not, I think it’s unlikely to get through phab and even if it did we run the issue of having people reporting each other even more than now to ANI for “minor edit marking violations” for a “minor edit block”. The second, the feature would still be useless imo, but that’s just my idealism talking I suppose; if editors aren’t being site blocked for it then I don’t care as much. ProcrastinatingReader (talk) 16:40, 16 February 2021 (UTC)

    Not currently possible, would require developer time. That alone makes it a bit hopeless, I suppose. But I'm still interested in how we can solve this problem in a more targeted way. — The Earwig ⟨talk⟩ 16:52, 16 February 2021 (UTC)

    Indeed it isn't currently technically possible, but a lot of work has recently been done on partial blocks and marking edits as minor is something that is given to only a subset of users so it strikes me that it is something that going to be possible with a reasonable amount of work (I'm not a developer though). Thinking of ways to solve the actual problem in a focused way that doesn't cause massive collateral damage is absolutely the right thing to be doing though - indeed this is the sort of thing that should have been done before coming here with a proposal, let alone an RFC. Thryduulf (talk) 17:25, 16 February 2021 (UTC)

    @The Earwig and ProcrastinatingReader: I've created a phab task for this, see phab:T274911. Thryduulf (talk) 17:29, 16 February 2021 (UTC)

    Since minoredit is a separate user right, if it were to be assigned by bot instead of being a permission set for all users, then it could be removed from anyone who used it inappropriately. isaacl (talk) 19:36, 16 February 2021 (UTC)

    @Isaacl: I think you might be mixing up some terms here? It could be possible to make a group that includes the minoredit permission; and it would be possible to make it auto-assign-once on a threshold (like extendedconfirmed) such that it could be revoked - but that is a lot of overhead for such a small permission (and I don't think we'd get "bots" involved with this process at all). — xaosflux ^Talk 19:50, 16 February 2021 (UTC)

    It's not something I know a lot about, so almost certainly. I elided talking about groups for simplicity, and I didn't know (or forgot) about auto-assign-once; my apologies. Yes, it would be overhead, but less than enhancing the partial block capability to support blocking a user from flagging minor edits. isaacl (talk) 19:55, 16 February 2021 (UTC)

    It would also be possible, without any overhead in the common case or coding effort, to create a group that removes the ability to make a minor edit using $wgRevokePermissions. This way of doing it also avoids the issue of blocks being seen as a stain on people's record that was pointed out by Suffusion of Yellow below. That said, I am far from convinced that any of this is actually a good idea. * Pppery * _{it has begun...} 23:38, 16 February 2021 (UTC)

    @Pppery: from a quick check we have exactly one project using that setting in all of WMF, and it looks like they haven't used it in over 10 years so I'm shocked it is still around (looks like may get merged in t pblocks - see phab:T227618). It does leave a "stain" on the user's rights log of course (e.g. w:ta:Special:Redirect/logid/58001. — xaosflux ^Talk 00:11, 17 February 2021 (UTC)

    As I recall when I was working with another Wiki implementation that had similar functionality, it can be useful in a corporate setting when defining user groups and managing their permissions. isaacl (talk) 00:20, 17 February 2021 (UTC)

  • This would be a fantastic source of drama. A block (even a partial block) isn't just the technical inability to do something, it's (for better or worse) going to be viewed by the user as a "stain" on their record. For something as trivial as this, that doesn't seem worth it. Suffusion of Yellow (talk) 20:00, 16 February 2021 (UTC)
• When you see a ±10k edit marked as minor, you know that the editor is almost certainly up to no good. Narky Blert (talk) 17:55, 16 February 2021 (UTC)
• Apparently more than seven thousand users have been warned for marking non-minor edits as minor. Has the opposite ever happened? I.E. "you're making typo fixes but not marking them as minor, please start using the flag or I'll drag you to AN/I". Suffusion of Yellow (talk) 19:52, 16 February 2021 (UTC)

  When training I always say that if you are in any doubt about whether and edit is minor assume it isn't as the worst that will happen is someone will give you friendly advice, but marking a major edit as minor could lead to people getting angry at you. Thryduulf (talk) 21:06, 16 February 2021 (UTC)

• I'm surprised that productive editors are being blocked for their use of minor edit marking. Are there more cases than the one mentioned by The Earwig above? That particular case seems highly unusual as the editor is apparently using an iOS app that does not receive notifications, and hence is not responding to feedback. Someone being blocked for vandalism marked as minor is clearly being blocked for vandalism. Espresso Addict (talk) 00:44, 17 February 2021 (UTC)

  Good question, Espresso Addict. I dug up some examples of editors being blocked for marking edits as minor. In most cases, there are other reasons for the blocks, of course: [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]. — The Earwig ⟨talk⟩ 07:37, 17 February 2021 (UTC)

• Comment. I voted oppose above and indicated I would be open to limiting who can mark edits as minor. I've noticed several people have since suggested only allowing autoconfirmed users to mark edits as minor, but I was under the impression it was already limited to autoconfirmed users. If not, I see no reason why that shouldn't be the case. -- Calidum 21:11, 24 February 2021 (UTC)

  @Calidum: The present situation is that you dn't need to be autoconforemed to mark edits as minor, just be logged-in. --Redrose64 🌹 (talk) 17:12, 25 February 2021 (UTC)

• Some of the topics I keep my eye on are prone to POV vandalism. Often the vandals attempt to hide their vandalism behind a “minor edit” tag. This is (counterintuitively) quite helpful when it comes to spotting and cleaning up such vandalism ... I have learned to pay extra close attention to edits that are marked as being “minor”. Thus, I would oppose doing away with the tag. Blueboar (talk) 23:03, 24 February 2021 (UTC)

  We should rename it to Honeytrap 😅 Shushugah (talk) 11:08, 22 March 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.

notifications of indef blocked users

This is something that just kind of bothers me and I'm not sure I have a good solution to it but thought a discussion might be productive. Sometimes, very long-term users end up indef blocked, and over time many things they created are nominated for deletion. Notifying someone who has been blocked for years seems unproductive and kind fo like kicking them when they are down. I'm not by any means suggesting these nominations or notifications are done in bad faith, in many cases the notifications are done by automated tools anyway. And that might be where a tweak could be made. I use a script that automatically puts a line at the top of any user page or talk page telling me how long a user has been registered, what user rights they have, and will also inform me if they are blocked. For example, right now on my own user page it displays as A checkuser, oversighter, and administrator, 13 years 7 months old, with 96,136 edits. Last edited 27 minutes ago. If that functionality is compatible with Twinkle or other such tools, they could pop up and ask if the user is sure they want to notify, something like <USER> is blocked and their last edit was 342 days ago, do you wish to proceed?. Does that make sense and/or seem reasonable to anyone else or is it just me? Beeblebrox (talk) 20:11, 9 March 2021 (UTC)

I've been bothered by the same thing, and I think it's a reasonable suggestion. There are also people who've formally left the project, so it should account for people who haven't been blocked but are just plain gone. Acroterion (talk) 20:16, 9 March 2021 (UTC)

(edit conflict) Seems reasonable to me. I prefer to nominate pages for deletion manually rather than using Twinkle, and skip notifying indefinitely blocked users, or users who have been inactive for a long time (I'm not consistent about how long). Although, I once got personally attacked on Meta for not notifying an indefinitely blocked user ... * Pppery * _{it has begun...} 20:19, 9 March 2021 (UTC)

• Leaving the same courtesy notifications that you would on any unblocked user's talk page is not WP:GRAVEDANCING, not kicking them when they're down. And it's a mistake to assume that those notifications serve only that user. There have been numerous occasions when notices on an indef blocked user's talk page helped me find an old deletion discussion, or discern a pattern in their behavior. Such patterns have led to further cleanup and to the unmasking of sockpuppets. Perhaps such breadcrumbs aren't as important to admins, who I assume can see deleted pages and revisions. As a non-admin, I hope no one stops leaving notices on the talk pages of blocked users. --Worldbruce (talk) 23:13, 9 March 2021 (UTC)
• I agree with what Worldbruce said above. Also a surprising number of highly productive editors have been indef-blocked, and they have often created significant pages, templates, images, etc., where a wider notification pool may be good. Furthermore, an indef-block is not always forever, and they might return one day to address any issues, or catch up, even if they miss the relevant discussion at the time. And if a page is bothering you, you can always unwatch it. However I also agree with Beeblebrox that people should have the information in a convenient place to opt out of notifying users. Ideally people should be taking the extra few seconds to look at contribs anyway, before choosing the relevant tickbox in Twinkle, but it is what it is. I would guess that it would be relatively easy to implement in Twinkle if it isn't already - you would want to speak to the Twinkle developers. -- zzuuzz ^(talk) 23:56, 9 March 2021 (UTC)
• How many of these scripts respect {{nobots}} and might that be something indef'd editors can be made aware of? I get the reasons above about the benefits of allowing these notifications, but it strikes me as similar to don't template the regulars: it's not wrong or forbidden, but just kinda tone-deaf. We should be up-front about the option, and let current editors make their own decisions on whether to template indef'd editors. I'd be fine preventing it by default. — Wug·a·po·des​ 01:36, 10 March 2021 (UTC)
• The problem here is that productive users have been blocked. For example, I was just reviewing the history of a couple of articles and noticed that Hullaballoo Wolfowitz had made some useful contributions. But they are currently blocked for a period of six months – a remarkably draconian and unproductive punishment. In such cases of obvious injustice, I will typically put that user's talk page on my watchlist so that I may observe any such notifications and take appropropriate action. HW's talk page has over 300 watchers and the notifications are for them as well as the primary user. So far, in that case, we have had one of Gerda's Precious anniversaries which serves to remind us that this block is disruptive. So, the appropriate action in this case is to remove the block – the OP should please oblige. Andrew🐉(talk) 11:47, 10 March 2021 (UTC)
  • See the ANI discussion which led to the block and especially note the discussion following it where the length was explained and endorsed. Andrew you might want to read up on WP:ADMINSHOPping. — Wug·a·po·des​ 20:38, 10 March 2021 (UTC)
    • An overly harsh block for pretty insignificant abuse. The effect of the block is worse than what it was addressing. Anyway for a 6 month block, notifications are still worthwhile to get on the topic of this discussion. Graeme Bartlett (talk) 00:06, 11 March 2021 (UTC)
      • An ... block for ... abuse. fixed that for you Graeme. You might want to try reading the discussion I linked, especially the part where I said even people defending HW pointed out a pattern of incivility, they simply try to excuse it through various means. If you're interested in more than just taking cheap shots at me and want to challenge the close, you know where AN is. — Wug·a·po·des​ 00:53, 16 March 2021 (UTC)
• This is one I have struggled with, I don’t like the perception of grave dancing, but when nominating an article for discussion I have sometimes left a message for blocked users wondering if (hoping) one of their former collaborators might maintain a watch on their TP and have an interest in or knowledge of subject matter. Cavalryman (talk) 22:02, 10 March 2021 (UTC).
• This is an interesting discussion. I'm not so concerned about any perceived awkwardness or impoliteness in leaving such a notice on a blocked user's talkpage, as I am in the productiveness of it. As people are saying, it is better to leave such a message rather than none at all, because then the message is at least seen, but if the user has few or no talkpage watchers, or if the user is not blocked, but simply inactive, then notifying just the first editor may not be productive at all. And even if the first editor is still active, they may not be the person who is most interested in the article. Some first editors make just the initial edit or two, and then leave the article, and it may get built up by other editors. I think it may be useful to have an added functionality in the automated tool to notify the first editor AND the most productive editor (perhaps with some added clause such as that the most productive editor should be unblocked, and have made an edit on Wikipedia in the previous month). SilkTork (talk) 09:21, 12 March 2021 (UTC)
• What Silktork said. There are articles where one creates, but another fleshes it out and has more of a vested interest in the article, so expanding notifications to those people is worthwhile. As for notifying someone who is indef blocked, I certainly don't see it as gravedancing, it is simply treating them like anyone else here, which is technically what we are supposed to do unless they are actually banned. It may be less than ideal in some ways, but it is the lesser of the two available evils. Even if indef blocked, by being notified, they can at least know it up for deletion, and ask someone else to look into it (which isn't the same as proxy editing), or a talk page watcher can look into it, putting more eyes on it. Not every solution is perfect, but I think the current system is the least un-perfect. Dennis Brown - 2¢ 10:13, 12 March 2021 (UTC)

• Just to be clear, the only thing I was even thinking of proposing was to ask for Twinkle or other tools to just have the option to double check if you want to proceed if the user has been blocked and inactive for say, six months or longer. I don't think we need a rule or anything. An example is User talk:Richard Arthur Norton (1958- ). This user checks in once or twice year, clears like 20 deletion notices off their page, and goes quiet again. Again, I don't think anyone is doing this maliciously, it just seems rather pointless. Beeblebrox (talk) 06:17, 13 March 2021 (UTC)
  • Why not revisit a lot of those possibly unjust, indefinite blocks? These users might be thinking, "Now, I'll get over to Wikipedia now. Am I unblocked? Nope. Do I have the strength to request unblock? No, it'll just be declined. Is my talk page huge? Yes. Full of deletion notices, and I can't even give a rationale for keeping them. I have so many ideas for restarting, but the administrators act like I'm a fly they swatted. Makes me think I should get away from Wikipedia now. See you... never!" 🐔 Chicdat  ^{Bawk to me!} 11:22, 20 March 2021 (UTC)

Alfred Dreyfus's talk page, 1895

• This also bothers me when I see it. Especially since an indeffed user can't actually participate in the deletion process, it just seems unnecessarily cruel: "Hey asshole, that article you spent a bunch of time on is getting nuked, thought it might be funny for you to watch people debate about it without being able to respond in any way". Not nice! jp×g 18:43, 16 March 2021 (UTC)
• The premise here is well-meaning, which we could always use more of, but I feel like this has maybe been over-thought. First off, if you're indef-blocked, why are you still logging in here? Do people do that, just so they can keep trying to edit? That is, does the audience even see the messages you're worried they'll see? Second, we're talking about people who were such fine folks that they got themselves indef-blocked and such fine articles that they're up for deletion. I guess I'm just being crusty, but I'm having a hard time kindling a lot of sympathy. I'd much rather expend time/effort/resources on the people who aren't dicks. Finally, the deletion is gonna happen or not regardless of the notification; isn't it just as likely - and perhaps more so - that the "nice" thing to do is to let people know what's happening rather than leaving them in the dark? We see a lot of questions on the help desk from people wondering where the hell the such-and-such article went. Matt Deres (talk) 20:13, 26 March 2021 (UTC)

Deprecate linking to Wikipedia books in templates and articles

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

There's a strong consensus to deprecate these links on the grounds that they're not serving our readers well, as judged by the low number of pageviews. (non-admin closure) (t · c) buidhe 10:37, 23 March 2021 (UTC)

---

I'm proposing that we formally deprecate linking Wikipedia:Books in templates and articles. Wikipedia books have not worked for many years now but are still included in various templates (e.g. {{Star Wars universe}}, {{Mercury (planet)}}, {{Abraham Lincoln}}) and maybe some articles—I recall seeing some in various "see also" sections. It seems unhelpful and unproductive to lead readers to a page that says the service in question is unavailable and points the reader towards external (third-party) sources; I speculate that most readers will see this as a dead end and not pursue it further. Around two years ago {{Wikipedia books}} and the relevant sidebar were suppressed, so I don't see why we should contradict that practice by still providing useless links in templates (and supposably articles). Aza24 (talk) 06:48, 13 March 2021 (UTC)

More info available at Wikipedia:Wikipedia book creator status.

Past discussion on the template itself - Wikipedia:Templates_for_discussion/Log/2021_January_16#Template:Wikipedia_books

• Support. I'm not fully read up on the history of Wikipedia Books, but it seems like an experiment we tried and that failed. We should clean up after ourselves, which means not continuing to link to it after it's dead. {{u|Sdkb}} ^talk 05:04, 14 March 2021 (UTC)
• Support: as Sdkb says, let's clean-up after ourselves. GenQuest ^"scribble" 08:48, 14 March 2021 (UTC)
• Support As a loose end to failed project, as Aza points out, this is just a dead end that does no service to our readers. Beeblebrox (talk) 20:45, 14 March 2021 (UTC)
• Support Not a contentious suggestion and understandable. I like the principle of "cleaning up after ourselves." doktorb _words^deeds 21:37, 14 March 2021 (UTC)
• Support Wow, I had never even heard of these (which says a lot). ~ HAL333 00:18, 15 March 2021 (UTC)
• Support Failed experiment, should be deprecated.--Vulphere 04:52, 16 March 2021 (UTC)
• Support I have a ton of opinions about books and think it is a really difficult problem how to handle them. I think they have a non-negligible value if they ever get working properly again and even now there are a handful of people finding value in them. At the same time the namespace is filled with a lot of crap, the category system is in most cases a better tool for reading about a subject and linking to pages with large warning banners doesn't look good at all. I have previously proposed a technical way to hide these links without removing them so we could put them back if the issues are fixed at Wikipedia:Village pump (technical)/Archive 183#Mainspace book links but to be honest I am absolutley fine with just removing them as well. My solution, while workable, would definitely cause some confusion for non-technical editors and possibly weird formatting in places if someone touched whitespace around them. Straight up removal would also have the benefit of pruning the links if they ever need be re-added and should in theory only link to decent quality books. --Trialpears (talk) 16:19, 16 March 2021 (UTC)

  @Trialpears: The problem with the Category system is that its lists are essentially unstructured and can be very large; Books allow intelligent design. Also, book links are currently suppressed, much as you suggested. — Cheers, Steelpillow (Talk) 17:35, 16 March 2021 (UTC)

• Strong object. The OP's "service in question", as described at Wikipedia:Books, is the Collection extension (aka Book Creator). It is very much here and working. And w have a Books: namespace with lots of books in it. We deservedly ditched our useless in-house renderer, but the idea that it is nowadays "the service in question" is nonsense. As stated on Wikipedia:Books and elsewhere, "As a result of anticipated future solutions, template transclusions should not be removed from articles." We would first need to establish a consensus to abandon any thought of such future services, such as adopting the long-anticipated Collector, and also of supporting external services such as MediaWiki2LaTex at WMFlabs. Then we'd have to agree to abandoning the Book: namespace. Let's discuss those first and not go round breaking them. — Cheers, Steelpillow (Talk) 17:35, 16 March 2021 (UTC)
• Support per WP:NOTPAPER. If readers want to use an external service to print collections of articles, they may do so, but we have no need to maintain support indefinitely for services unrelated to our goal of building a wiki. As sdkb says, we tried books and it failed. Book:Star Wars and Book:Abraham Lincoln average 3 views a day each. Star Wars got more page views yesterday than the book gets in an entire year. Mark the namespace as historical, remove reader-facing links to it, and let it collect dust (I oppose deleting the namespace or pages per m:Keep history). We don't need to waste time maintaining support for a feature no one uses. — Wug·a·po·des​ 19:27, 16 March 2021 (UTC)
• Support I sometimes have seen the links at the bottom of articles and wondered what additional info they offer. Seems in some cases they are more of a fork to out of date content. Jtbobwaysf (talk) 05:59, 19 March 2021 (UTC)
• Support Essentially these are an external link. Reading the guidelines here Wikipedia:External links#What to link indicates that Wikibooks no longer meets most (if not all) of them. MarnetteD|Talk 06:46, 19 March 2021 (UTC)
• Support no more linking to dead namespaces. 🐔 Chicdat  ^{Bawk to me!} 11:10, 20 March 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.

RfC: make Template:Authority control more reader-friendly

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

There's a strong support for an overhaul of the authority control template that uses human-readable names of the resources, in the interest of being recognizable to more editors. There is general support that Fram's proposal is preferable to the current version, but not any consensus on the exact form that an improved version might take. An alternative proposal which attracted some support is to scrap the entire template or replace with a link to wikidata, which could be discussed at another RfC to gauge if that proposal has consensus. (non-admin closure) (t · c) buidhe 01:24, 1 April 2021 (UTC)

---

Should Template:Authority control be rewritten to make it more reader-friendly? Fram (talk) 10:12, 18 March 2021 (UTC)

Authority control background

Authority Control is a template that is included in 1.7 million enwiki articles by now: it shows as links a number of "unique identifiers" from organisations around the world. All the data is stored on Wikidata, nothing is stored here. The proposed RfCs won't change anything about how Wikidata deals with these: it is purely about how we want to show these at Enwiki.

With an example taken from Kenzaburō Ōe, the current result of the Authority Control template looks like this;

Authority control proposal

Change the links to authority IDs: currently they are an acronym plus a unique but often meaningless ID. In the proposal they would become a readable, textbased link. The links could also be organized to make their use clearer, and to avoid repetition.

The above is a very simple mockup of how it could look. This is an example, and not necessarily the end result. For other common groups of IDs (e.g. art-related websites), another line can be added. Fram (talk) 10:14, 18 March 2021 (UTC)

I would like to ask people not to focus too much on the look of the proposal: the RfC is about the principles, the actual new look can be decided by people with more knowledge of and talent for user experience design. Fram (talk) 10:16, 18 March 2021 (UTC)

• There's probably something I'm missing, but the lead reference "Integrated Authority File" leads to a German language website for the Deutsche Nationalbibliothek which I would have expected under "National Libraries"? Other than that issue it does seem a good and sensible proposal. Martin of Sheffield (talk) 10:25, 18 March 2021 (UTC)

  Yes, I wasn't certain where to put this one. It is, to my best knowledge, an international standard, but maintained by the German National Library, and the text on the page is in German. These details (well, it's an important one) will need to be decided upon if the general principle of the proposal gets accepted. Fram (talk) 10:43, 18 March 2021 (UTC)

• Comment: Is the term "Authority control" even reader-friendly? Few readers other than librarians will have any idea what it means, and it sounds ... authoritarian? Should we talk about "Unique identifiers" for clarity? PamD 11:29, 18 March 2021 (UTC)

  No not really. Its a data/library term not a reader term. Only in death does duty end (talk) 11:41, 18 March 2021 (UTC)

  That would be better but is a separate discussion, I think. Fram (talk) 11:54, 18 March 2021 (UTC)

  I don't fully mind the term, as it is wikilinked, but it's not a major deal either way. Most casual readers likely don't even make it that far on the page. ɱ (talk) 13:58, 20 March 2021 (UTC)

• Oppose removing WP links - I'm pretty sure this has been discussed before, but that hasn't stopped Fram before, either. The links to, for example, LCCN inform the user of the AC-granting institution, and the ID link(s) n81033861 provide the information. Both are needed.   ~ Tom.Reding (talk ⋅dgaf)  11:39, 18 March 2021 (UTC)
  • Why should you require a reader to click on obscure acronyms to get the information about who issued the ID string, if you can include it much more clearly in the template directly like above? Most readers of the Oë article will have no clue at all what "NLG: 71274" is. In the new template, they see directly that it is a link to the National Library of Greece. Do they need an actual link to National Library of Greece from the Oe article? I doubt it, it is not relevant information for that article. Do they still have the link to the NLG ID page? Yes, that is still included. But now at least they will have a better idea of what they are presented with, without needing to click either link. Your two reasons are "informing the reader of the granting institution", which is in the proposed version done directly instead of requiring a click: and the ID links to provide the information, which is in the proposed version also still there, as it still is a link which still leads to the same information. Oh, and feel free to link to previous discussions, perhaps without the personal comments though. Fram (talk) 11:53, 18 March 2021 (UTC)
    • You brought it up, here, at this Village pump (proposal) in June 2018. Search "Option 2Names", which spawned "Option 2Wikidata", "Option 2pencil", and "Option 2edit". I too, at first, and others, liked these options, before realizing my statement above. Nothing personal, just facts about your person. Unfortunately, they seem like WP:I didn't hear that and/or slow-motion WP:FORUMSHOPPING.   ~ Tom.Reding (talk ⋅dgaf)  15:43, 18 March 2021 (UTC)
      • Oh, that trainwreck discussion ending in no consensus, but with a lot of support for the type of changes I propose here? Yes, that kind of discussion doesn't stop me from bringing a new RfC two years later. I do like the "Nothing personal, just facts about your person." though. Fram (talk) 16:22, 18 March 2021 (UTC)
  • LCCN, etc., links serve the additional purpose of notifying the editor which parameter they may use (|LCCN=) to either override the WD ID, or to suppress the ID altogether. With the proposed scheme, this is much more difficult to determine. Such a lack of foresight could have been avoided if there had been any discussion on the talk page, instead resulting in this very poor RfC.   ~ Tom.Reding (talk ⋅dgaf)  00:42, 19 March 2021 (UTC)
    • Despit the template being used on 1.7 million articles, the possibility to suppress links was used on 43 pages before I created the ACArt template recently. Such a barely used possibility (which will remain available anyway, but may if the RfC is successful need better documentation) hardly justifies making the template much harder to understand (and thus less likely to be used) for our readers. This is not a "lack of foresight", this is getting your priorities right. Fram (talk) 08:23, 19 March 2021 (UTC)
      • You're ignoring the ~2400 pages using Category:Pages using authority control with parameters (1). Par for the course though.   ~ Tom.Reding (talk ⋅dgaf)  10:18, 19 March 2021 (UTC)
        • For a start, hundreds of those seem to be redirects, which shouldn't have an authority control templatein the first place. But let's look at some actual articles instead. First one I checked at random, Anthony St. John Baker, has the same ID in the parameter as in Wikidata. Siddharth Basrur, added by same editor on Wikidata and a minute later on enwiki, so not necessary either. And in any case, the possibility isn't removed, in the few cases where it is needed they will just have to check a bit harder to get it right. Fram (talk) 10:43, 19 March 2021 (UTC)
          • At least we agree that this proposal will hamper functionality.
            Also, 353 redirects, on which AC is encouraged per Wikipedia:Authority control "If necessary, create a new redirect to carry the template." For an RfC, you are stunningly misinformed.   ~ Tom.Reding (talk ⋅dgaf)  11:37, 19 March 2021 (UTC)

            Maybe it's time to stop attacking Fram ("IDHT", "FORUMSHOPPING", "stunningly misinformed", "just facts about your person", and others)? Read the room: consensus doesn't agree with you. ProcrastinatingReader (talk) 12:02, 19 March 2021 (UTC)

            All facts relevant to the discussion.
            "Read the room"? This isn't a vote; this is based on merit, of which there's none, other than "it looks pretty". Introduce a version where WP links are kept, then we can move on to discussing how much taller the template will get, but that's a better starting point than this.   ~ Tom.Reding (talk ⋅dgaf)  12:39, 19 March 2021 (UTC)

            If all you see in the proposal is "it looks pretty", then I think we are done discussing things. Uses of this template on redirects, which no one will ever see and which serve no practical purpose on enwiki, are hardly a reason to keep a template as uninformative and opaque as possible. Fram (talk) 13:13, 19 March 2021 (UTC)

• Support improvements - Fram is right. The weird acronyms that no reader can be reasonably expected to know are utterly useless. You can hover over them to see what they mean, since they link to the respective article, but readers shouldn't have to do that to make sense of it. The presentation of the template currently is useless, and Fram's mock-up above is a leap of an improvement. Still, I think more can be done with this template. The "National libraries" presentation is good, the rest is still iffy imo. ProcrastinatingReader (talk) 15:19, 18 March 2021 (UTC)
  • When I hover of them, I get "Bibsys (identifier)", "BNF (identifier)", "CiNii (identifier)" and so on. Not very helpful. Fram (talk) 15:10, 19 March 2021 (UTC)

    When I hover over them, I get "BIBSYS is an administrative agency set up and organized by the Ministry of Education and Research in Norway. They provide the exchange, storage and retrieval of data pertaining to research, teaching and learning – historically metadata related to library resources." and other similar previews of the articles. I think it depends on whether you've got page previews (Preferences -> Appearance -> Enable page previews) or navigation popups (Preferences -> Gadgets -> Navigation) enabled. --rchard2scout (talk) 09:15, 22 March 2021 (UTC)

• As stated at the VPI discussion, I support the general change. I have some nits about the suggested appearance, but as requested, that's not the focus. --Izno (talk) 15:35, 18 March 2021 (UTC)
• Support, broadly speaking, making it more readable. It's currently just acronyms no one knows with numbers that mean nothing. I exaggerate, but if a general reader would want more encyclopedic "structured" information, these should be easy to identify for them. I get that the acronyms link to the articles about the authorities, but this seems a secondary goal to achieve. —  HELLKNOWZ   ▎TALK 17:00, 18 March 2021 (UTC)
• Support, to make the display more human-friendly. There is room for more improvement but the suggested display is vastly superior to the current example above. Schazjmd (talk) 17:02, 18 March 2021 (UTC)
• Support the general idea of making the template more accessible to readers. Breaking it into categories is a good step. I'm less sure about removing the identifiers and linked acronyms: the former is useful if a reader wants to copy and paste it, the latter for understanding what they actually are. But those are details that can be worked out incrementally. I also agree with the sentiment below that, now Wikidata is a mature project, this template (which is a few years older) is now largely redundant. It might be an idea to go further and suppress links to authorities that don't provide any useful extra information, essentially turning the template into an automated component of the external links section. – Joe (talk) 17:12, 18 March 2021 (UTC)
• Support of course, a no-brainer. Maybe add a TfD tag so more people see this discussion? Elli (talk | contribs) 17:15, 18 March 2021 (UTC)
  • TfD isn't a forum for RFCs. I'm listing it at WP:CENT. 力 (power~enwiki, π, ν) 17:18, 18 March 2021 (UTC)
    • I'm not suggesting to list it at TfD per se, but since the template is under discussion indicating that in the template would be helpful, and TfD has been used for this before (for example, for whether we should collapse {{WikiProject banner shell}}. Elli (talk | contribs) 18:06, 18 March 2021 (UTC)
      • I don't think advertising this RFC on 1 million articles is needed. I don't think the ‹See TfM› showing up with every {{tl}} is needed either ... 力 (power~enwiki, π, ν) 18:09, 18 March 2021 (UTC)
• Support in a very general sense, as I believe there's room for improvement of the template. I'm hesitant about some of the specific changes proposed, though. Converting to just external links without wikilinks or numbers looks great for an item like the example above with tons of identifiers, but might not for items with fewer identifiers. Adding sections makes the template longer and more likely to be confused with a navbox. {{u|Sdkb}} ^talk 17:37, 18 March 2021 (UTC)
• Oppose While the the template could definitely be made better looking, I don't trust Fram to do it. This RfC was started without any previous discussion that I've seen at Template talk:Authority control, which would have been the logical venue. Fram has nominated the template for deletion, campaigned for removal of properties from it (e.g., Musicbrainz), and most recently created {{ACArt}} to systematically remove content from it (it is currently up for deletion). They are generally hostile to Wikidata, and seek to reduce the use of Wikidata here. It feels like this RfC is a trojan horse to continue doing that. Thanks. Mike Peel (talk) 17:40, 18 March 2021 (UTC)

  That isn't a valid oppose reason. The proposal doesn't say who will implement it, and presumably the implementation itself will be subject to further discussions and/or scrutiny (+ AGF etc). The idea behind the proposal is sound (as you've said), which is the matter of discussion here. Fram left a link to this RfC at that template talk when the RfC was written (a local consensus cannot override community sentiment on this template, which is present on 1.75 million articles). ProcrastinatingReader (talk) 17:53, 18 March 2021 (UTC)

  If Fram's not going to implement it, then they should have found someone that would before starting the discussion here (like discussing it on the template talk page, not just posting a link to this discussion). I said that it can be made better looking, which is generally true of most templates, but not necessarily like this (some extra whitespace would make reading it easier). Thanks. Mike Peel (talk) 17:59, 18 March 2021 (UTC)

  It doesn't look like a particularly difficult LUA change. 力 (power~enwiki, π, ν) 18:06, 18 March 2021 (UTC)

  This opposition is unfounded. I too disagreed with Fram's decision to create and add AC art without seeking a community consensus (and as such, supported deletion at TFD) but this proposal a clear backtracking of Fram to go through the system properly; seeking community consensus for what is (demonstrably—by the amount of support) a reasonable suggestion. Its also worth noting reminding everyone that this proposal does not change the amount of links being shown, which was the cause for dispute at the AC art TFD in the first place. I would urge Mike to AGF. Aza24 (talk) 08:31, 19 March 2021 (UTC)

  @Aza24: I have interacted with Fram on numerous times about Wikidata-related issues. Assuming good faith with them has long been worn out, I stand by my opposition here. They are not the right person to do this. Thanks. Mike Peel (talk) 09:31, 19 March 2021 (UTC)

  Dude I hope you never propose an RFC. Levivich ^harass/_hound 17:22, 20 March 2021 (UTC)

  @Levivich: I propose them regularly, but I do the background work first (including discussion with at least some people) so that it can be implemented if passed. Thanks. Mike Peel (talk) 20:17, 20 March 2021 (UTC)

• Support, broadly, improving the template, though I'm not sure exactly what this would look like, I trust our technically minded editors. I somehow doubt that Fram is trying to secretly attack wikidata or something by proposing this at a highly visible page, certainly much more visible than the authority control talk itself. Eddie891 ^Talk _Work 17:58, 18 March 2021 (UTC)
• (edit conflict) Strong support These are inscrutable to readers -- you know, the people Wikipedia is supposed to serve -- in their current state. Honestly, I don't expect them to turn into the most earth-shatteringly popular part of the project after this either, but they'd be a lot better than "actively useless". Vaticidalprophet 18:00, 18 March 2021 (UTC)
• Support There may be users who need the ID numbers, but this is a definite improvement. SportingFlyer T·C 18:51, 18 March 2021 (UTC)
• Support Don't need wikilinks to the IDs for every instance of the template. That is what the general help link is for. Likewise the new version is more descriptive as to what the IDs are, and doesn't look like a binary dump of abstracted codes. -- GreenC 20:49, 18 March 2021 (UTC)
• Support anything that makes this business easier for readers to understand, or else why are we even doing it? Beeblebrox (talk) 23:54, 18 March 2021 (UTC)
• Support Common sense proposal, it's far easier to understand than the existing method. ThePlatypusofDoom (talk) 01:20, 19 March 2021 (UTC)
• Support Removing the link to the identifiers Wikipedia article is a tradeoff that I'm more than willing to make in order to get a more understandable template where you don't need to know obscure acronyms and look at meaningless IDs when using. Clear improvement for readers. --Trialpears (talk) 10:14, 19 March 2021 (UTC)

• Unsure - I think I support the basic question at the top -- that it could be more user friendly -- but regarding the actual example I'm not so sure... and that makes me unsure whether there is a straightforward way to make it more user friendly without significantly increasing the amount of space it takes up.
  For example, while the section for libraries seems helpful, why is a direct link to "SUDOC (France)" more helpful to a reader than wikilinking what that even is before linking to the identifier? I agree that displaying the identifiers themselves probably doesn't add all that much value, but it does seem valuable to have both a wikilink and an external link (in which case, why not link the identifier rather than something arbitrary like "external" or "link"). — Rhododendrites ^talk \\ 14:54, 19 March 2021 (UTC)
  • The SUDOC identifier is used on 235000 pages, but Sudoc only gets 12 pageviews a day (unclear how many of those through the current authority control template, and how many in another way). The proposed version (which can be improved upon, and the "other" section may be one opportunity for this) immediately indicates to our readers where the info comes from: perhaps it would be better to indicate in which language the linked page is instead, but the basics are the same: if you are interested in Norwegian authority control for Ôe, by all means click on Bibsys: but you don't need to visit our Bibsys page first to be aware of this. Fram (talk) 15:08, 19 March 2021 (UTC)
    • But this doesn't explain why a direct link which just adds "France" (or even "French") is more user friendly than wikilinking to what that resource is in addition to the direct link. It seems like removing the wikilinks is a major part of this proposal. What I'm worried about is that the wikilinks seem quite useful, and that retaining them and breaking the template into sections and/or spelling out what they are may increase the amount of space it takes up to a degree some find unacceptable. Here's another idea (one which I'm not entirely convinced of myself, but while we're spitballing a bit...): have a section for the non-English sources and collapse it by default? We obviously cover topics from outside the English-speaking world, so retaining them would be important, but it's true that the vast majority of users of the English Wikipedia would not be looking for non-English resources. — Rhododendrites ^talk \\ 15:22, 19 March 2021 (UTC)
• Support per proposal. ~ ToBeFree (talk) 13:53, 20 March 2021 (UTC)
• Support change so long as implementation is thoroughly discussed. ɱ (talk) 14:04, 20 March 2021 (UTC)
• Mixed - my first choice would be to shift the entire Authority Control process over to Wikidata (so all we would have here on WP would be a small unobtrusive box pointing editors to WD.)... however, if we are going to have Authority control links here on WP articles, I agree with Fram that we need to make them clearer for the average reader to understand. Unexplained abbreviations and ID strings are not helpful. So a more expansive template gets my hesitant support. Blueboar (talk) 16:34, 20 March 2021 (UTC)
• Alternative proposal, like User:Blueboar, I'd prefer if Wikidata handled this directly, with Authority Control being a default template, instead of something people manually add. That said, with the current template, I'd keep/use the wiki links, but improve the readability by adding sections (National libraries etc..) and more descriptive names. It would become a fatter template, and that's worth testing. We can measure its effectiveness, by tracking the view count of the Wikilinks before/after. It doesn't need to be useful to everyone, but for the few people it's useful for, experimenting is worth a try! Shushugah (talk) 19:26, 20 March 2021 (UTC)
• Support; removing numbers with no human-readable meaning improves the signal-to-noise ratio. I would also support going further and displaying the links only on Wikidata; pace librarians and archivists, these links are only of interest to such specialists. The links by and large do not help readers, do not meet the standards of usefulness, excessive length, etc. that we would require of external links in any other context, and do not merit blanket inclusion across over a million articles. Adumbrativus (talk) 22:29, 20 March 2021 (UTC)
• Alternative / support-ish: I agree with the general gist of this, including better layout and using more plain English. However, I agree with Blueboar that interfacing more directly with Wikidata would make sense, and with Tom.Reding that the wikilinks are important (they allow the reader to find out what these databases and other sources are and what RS say about them, i.e. why they should care about / trust what they're about to click an external link to.  — SMcCandlish ☏ ¢ 😼  21:11, 21 March 2021 (UTC)
• Support the proposal as written, but ... I am more supportive of scrapping the template altogether per several others below. I am always wary of anything that removes control of what appears in an enwiki article from enwiki editors (acknowledging the less preferred option to manually override the template’s parameters), a simple link to Wikidata would be more appropriate. Cavalryman (talk) 22:17, 21 March 2021 (UTC).
• Support the general principle. A bunch of random numbers and letters won't mean much to most readers, so I think adding context will help. I'm fine with using WikiData as well, as long as we still have the template for readers to easily click to. — Wug·a·po·des​ 22:20, 21 March 2021 (UTC)
• Support a change like this. I think that, from a reader's perspective, the current blob of data looks like something only relevant for computers, something Wikipedia-internal that's not relevant for me. With this change, I think readers would be more likely to click through to the libraries etc. for more information. For the links to our own articles on these systems (all the (identifier) links), I think they should be added to Help:Authority control. That'll only be one click more for interested readers, and they won't clutter up every single article. --rchard2scout (talk) 09:15, 22 March 2021 (UTC)
• Support I was today years old when I actually figured out... vaguely...what authority control is. If it took me years to figure it out, then it's practically meaningless to newbies or readers. Our pages shouldn't have cryptic gobbledegook, it should all be plainly useful to the reader. CaptainEek ^{Edits Ho Cap'n!}⚓ 22:01, 22 March 2021 (UTC)
• Support the concept or making this clearer. If it were collapsed by default, it would take less space and be more removed from the majority of readers who could care less while having the entire title bar to explain what it is to those that could make use of it: MB 03:45, 23 March 2021 (UTC)
• Strong support all proposed changes: the encyclopedia serves readers, not editors or bots. The template is very non-user friendly. If it makes the information twice as accessible then it's worth halving the amount of it. Someone looking for the identifiers can find it from hovering over the URL, visiting the URL or visiting Wikidata. Fram makes a good argument about low pageviews meaning no-one is clicking on these wikilinks for identifiers that are supposedly useful to the reader (from what I can tell this is true as a rule, not just the one example given), so it's not a loss to remove them. Even if all else fails, I hope reorganization into sections by provenance of identifier, rather than the current alphabet soup, will be implemented as I don't think the opposers are opposed to specifically that part of the proposal. — Bilorv (talk) 00:47, 24 March 2021 (UTC)
• Support I sympathize with those that want links the Wikipedia articles on the various databases, but that feature is not worth the inherit confusion that stems from giving readers a large set of unfamiliar acronyms, this is much clearer. Aza24 (talk) 02:06, 24 March 2021 (UTC)
• Support making the template more human-readable. The specifics can be worked out later, but even if the exact proposed template was adopted, I think that would be an improvement. Ajpolino (talk) 21:21, 25 March 2021 (UTC)
• Support anything to make the template more user-friendly, but prefer the suggestions of User:Blueboar and others of a more drastic overhaul and or replacement. TheEmeraldBeyond (talk) 23:04, 26 March 2021 (UTC)
• Support making the template readableto the average reader Asartea ^{Talk | Contribs} 09:01, 28 March 2021 (UTC)

Alternate formatting

Discussion re: Authority control itself

I am still confused by the purpose the Authority control template. I looked at the template page itself, and it explains WHAT it does, but not WHY it does it. Is there a page or discussion that explains the purpose of having such metadata in our articles? Blueboar (talk) 12:12, 18 March 2021 (UTC)

It's linked in the template... Help:Authority control. Izno (talk) 15:03, 18 March 2021 (UTC)

A lot of the information on that page is wrong though or severely outdated (compare the concise 5-link authority control template shown for Alexander Graham Bell with the current 27-link bloat). Basically, everything that is described there is done through Wikidata (and if there are outside applications that rely on the template on enwiki directly, they should be rewritten by someone more competent), not through Enwiki, except linking readers to some of the databases listed. Every researcher and application that needs unique IDs should do this through Wikidata, which is the source, and not through enwiki, which is a partial mirror for this information only. Bus this discussion, while necessary, is separate from the RfC. Fram (talk) 15:17, 18 March 2021 (UTC)

I feel like this template only exists so folks can mass-add it to articles. Pretty sure it offers 0 benefit to readers, at least in its current structure which is just an obscure bunch of numbers and abbreviations. ProcrastinatingReader (talk) 15:22, 18 March 2021 (UTC)

For what it's worth, I have found it useful in several cases, especially with regard to reconciling data about BLPs (not as a reliable source, of course). Occasionally I've had dates of births that didn't match across cross-wiki articles and I was able to use the various authority control to figure out what was most likely correct. I've basically treated it as an alternative to WikiData when needing structured information about somebody. The pitfall, of course, is that authority control isn't a reliable source in the same way that WikiData isn't. I would be interested to know what possible use cases it has for actual readers who aren't looking to edit the page it's on. Perryprog (talk) 13:05, 19 March 2021 (UTC)

Agree with those above saying that the template is pretty useless almost anywhere in Wikipedia. Wikidata should do the matching of unique identifiers. Mirroring that in Wikipedia is odd at best (that is, for those understanding what it means); and completely incomprehensible for the majority. I can't support Fram's OP proposal of this RfC (neatly organized redundancy... is still redundancy... it just occupies even more space while I would reduce that space to zero), but I'd gladly support any proposal that would discourage the widespread usage of the template in English Wikipedia. --Francis Schonken (talk) 16:00, 18 March 2021 (UTC)

It’s already been mass-added to every other article. With 1.75 million transclusions, I think the ship has sailed to discourage even more use. If we can’t get it blanked/removed (which I’d support), reforming it to not be useless is the next best thing imo. ProcrastinatingReader (talk) 16:09, 18 March 2021 (UTC)

This isn't really mirroring it, it's displaying it. You could make a similar argument for removal of infoboxes. Elli (talk | contribs) 16:47, 18 March 2021 (UTC)

Most infoboxes don't mirror Wikidata info though, but get their input locally: and more importantly, most infoboxes are self-explanatory, they add on their own understandable info with direct use for many readers (though there as well a significant group doesn't like them). Fram (talk) 17:01, 18 March 2021 (UTC)

The infoboxes getting their input locally is arguable more of a "mirroring" than authority control pulling from Wikidata is. I do generally like your idea for reforming the template. Elli (talk | contribs) 17:03, 18 March 2021 (UTC)

It's useful in the same way that references are. You can find more info in the links, or confirm content in the article. Thanks. Mike Peel (talk) 17:43, 18 March 2021 (UTC)

References verify information written in the article. External links, which do not, are bound by the WP:EL guideline, which states: it is not Wikipedia's purpose to include a lengthy or comprehensive list of external links related to each topic. No page should be linked from a Wikipedia article unless its inclusion is justifiable according to this guideline and common sense. The burden of providing this justification is on the person who wants to include an external link. Authority control is the exemption to the rule, not the rule. ProcrastinatingReader (talk) 18:24, 18 March 2021 (UTC)

"External links... are bound by the WP:EL guideline" Nope. It's a guideline. Nothing is "bound" by it. Besides, Wikipedia guidelines are supposed to describe, not proscribe, current practice. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:41, 19 March 2021 (UTC)

As usual, WP:JUSTAGUIDELINE is the weakest of all possible rationales. How about WP:NOT#LINK? ProcrastinatingReader (talk) 15:52, 19 March 2021 (UTC)

• Thank you for the pointer. Wouldn’t a simple (small) box with a link to Wikidata (perhaps saying: “Wikidata has a page on this subject: see here”) serve the same purpose? Blueboar (talk) 18:05, 18 March 2021 (UTC)

Much like the link to Commons or Wikiquote? Thats sounds like a good idea too. Lugnuts ^{Fire Walk with Me} 08:04, 19 March 2021 (UTC)

Wikidata is not exactly meant to be browsed, and it shows in its design, so directing readers that way is not ideal. It's meant for retrieval, which is what the AC template accomplishes. – Finnusertop (talk ⋅ contribs) 10:35, 19 March 2021 (UTC)

• Beyond doing what's in the name, authority control puts a Wikipedia article about a subject in connection with other resources about the same subject. It basically does what our external links section does. But whereas lots of links in an EL section quickly bloat and take up too much page space, this template packs a ton of resources into a [relatively] small box tucked away at the bottom of a page. While most users probably don't notice it (ditto categories and other elements on the fringes of an article), I've talked to an awful lot of librarians, researchers, archivists, etc. who really value it. (This is a response to this sub-section, not an opinion about the RfC itself). — Rhododendrites ^talk \\ 14:41, 19 March 2021 (UTC)
• Isn't itpart of the idea behind Wikidata that this sort of information woould be moved there? DGG ( talk ) 06:02, 21 March 2021 (UTC)

  @DGG: the data lives on Wikidata, this just makes it visible on Wikipedia. Elli (talk | contribs) 05:36, 22 March 2021 (UTC)

  • There is no question that the data is useful for librarians, etc. The question is: Do we actually need/want that data visible here on WP... or do we want the data to be visible at Wikidata. Blueboar (talk) 13:02, 22 March 2021 (UTC)

there are times when such identifiers are needed to make it clear what the article is actually about, and they can be the best entry point into further information. But they are sometimes inconsistent, sometime erroneous, notably OCLC, and often reflect mere format variations, notably ISBN,. The practice of libraries confronted with this chaos is to record all available identifiers without trying to harmonize them--that's more in the realm of bibliographers. (Though I have sometimes resorted to analyzing them on WP, to elucidate the publication history of a periodical, usually on a talk p, because it is in a sense OR,). I have never tried to work with this on wikidata, as in the past the quality has been so erratic as to be discouraging, but I know the editors there are trying to clean it up as they can). The question then , as was said just above, is what part of this belongs in WP. For journal articles, I think the doi does--it's quite reliable. For books, I usually enter one ISBN, if possible for the print edition which tends to be more stable (unless OCLC record only th electronic version of what is actually a print book also). For identifying individuals, nothing that I know of is reliable. The one source for authors I know and can prove to be totally unreliable is LC after around 1960-70, and OCLC records or international cataloging records derived from LC, for if the information about the identity of the author is not immediately obvious from the title page of the book, they simply copy it from Wikipedia--and they seem to be doing so increasingly. . DGG ( talk ) 20:45, 22 March 2021 (UTC)

Remember WP:SAYWHEREYOUREADIT If you've read it in a book, then use the ISBN printed in the book. Martin of Sheffield (talk) 22:23, 22 March 2021 (UTC)

Authority control in mobile view

This is another thing to consider about changing Authority control to be more reader friendly. Currently the template is not visible in mobile view at all. A template used in over 1.7 million articles must be visible to mobile device readers, who form a significant part of readership. current result of the Authority Control template looks like this; I am reading this from mobile and all I see is a blank line. AVSmalnad77 ^talk 17:47, 18 March 2021 (UTC)

This is under discussion at the authority control talk page. That said, this is true of all navboxes today, which is a hard problem to solve. Izno (talk) 18:01, 18 March 2021 (UTC)

{{Authority control}} is a navbox, and all navboxes are disabled in the mobile view. It's been an outstanding bug for several years; see T124168. – Joe (talk) 18:08, 18 March 2021 (UTC)

Sad that it has been tagged as low priority. That explains why it has been pending for years. I hope they will provide more support for mobile users. AVSmalnad77 ^talk 02:55, 19 March 2021 (UTC)

It is trivial to enable again, but if we did it would look like in this video https://www.youtube.com/watch?v=eaos1s3UfLs. I think that is worse than the status quo and the template would need a significant redesign to solve it. If we had an agreed direction to go in to improve it I guess it would be doable. --Trialpears (talk) 10:10, 19 March 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.

What should the thanks confirmation be like?

What should the message and interface that pop up when you click "thank" on a diff or history be like? On the talk page of Help:Notifications/Thanks, it was suggested that rewording the message and adding a link to the help page would make it less confusing. So we have:

• Status quo: Publicly send thanks?  Thank Cancel
• Option A: Send thanks?  Thank Cancel Help
• Option B: Send thanks (help)?  Thank Cancel

See here for the preliminary discussion (whose participants I'll notify). Note this has no binding force and there's no guarantee the result will be acted upon; it's just an informal poll to see if a change is due. Nardog (talk) 07:53, 21 March 2021 (UTC)

• A or B, preferably A. While it is the case that the thanks are logged, the word "publicly" is highly misleading since only who thanked whom when, not even for which edit or on which page, is public, and only if one goes to Special:Log, which no casual user would even know the existence of. The following from the discussion illustrate the demand for the change. Diegodlh said:

  It wasn't clear to me whether my thanks had been sent already and if it was asking me if I wanted to publicly thank the editor on top of that, or if publicly thanking was the only way to go.

  Pdebee said:

  A few years ago, a new joiner wanted to thank me for my initial help with her first steps as an editor, but changed her mind because she thought that “publicly” meant that tens of millions of other registered editors were going to see what she’d hoped would be a simple, private signal of personal appreciation.

  Removing "publicly" will eradicate such confusion, and adding the help link will help newcomers know what the function does and hesitate less to use it, which will lead to increased editor retention if these studies are to be believed. Nardog (talk) 07:53, 21 March 2021 (UTC)
• Support change, preferably A - I’ve always thought the ‘publicly’ wording doesn’t really match up with reality. ƒirefly ( t · c ) 08:20, 21 March 2021 (UTC)
• Support change, preferably A - Thank you for considering this effort; if implemented, your solution should prove less confusing. Patrick. ツ Pdebee.^{(talk)(become old-fashioned!)} 10:45, 21 March 2021 (UTC)
• Comment: First off, I would have preferred the three given options to be given similar labels, i.e. Options A, B, and C instead of the cumbersome "Status Quo", making A and B more attractive simply to type. Secondly, the help page is considerably improved since the original discussion back in 2019 (if I may say so myself). The original discussion hones in on the word "publicly" and what that means. Go read the help page, I hope you will find it much more instructive than back in the day. Finally, here are the relevant issues I raised in the original discussion that never got addressed really: a) "Thank" and Cancel" should be buttons not links - they execute commands, they don't take you to new pages. This would make it easier to distinguish a help link. b) the help page is for the entire Thanks fuction, not merely to answer questions on "publicly" which suggests Option B as the more logical placement. c) As I understand PrimeHunter's argument (in the earlier discussion) option A requires developer intervention while option B can be done by only tweaking system messages. d) the degree of "publicness" of your given thanks is debatable - see the earlier discussion. Just so we make a change for the right reasons. CapnZapp (talk) 10:56, 21 March 2021 (UTC)

  To that end, here's an alternate poll:

  • Option A: Publicly send thanks?  Thank Cancel
  • Option B: Send thanks?  Thank Cancel Help
  • Option C: Send thanks (help)?  Thank Cancel
  • Option D: Send thanks (help)?  Thank Cancel

  CapnZapp (talk) 11:01, 21 March 2021 (UTC)

  You're throwing a monkey wrench in the middle of this discussion. Overriding the options when others have already participated will all but confuse people, and Option D isn't even comparable to the others! We can call the status quo "Option 0" if we want, and I don't mind you or anyone adding to the options (as long as it's a change on the equal level). Nardog (talk) 11:19, 21 March 2021 (UTC)

  I brought up the button/link issue in the 2019 discussion. Don't dismiss my reminder as if I had a chance to give my input any sooner, but blew it. (If you feel there is a better place where I should have brought up this directly-relevant-to-the-topic issue, you will have to tell me where) CapnZapp (talk) 14:49, 21 March 2021 (UTC)

  while option B can be done by only tweaking system messages That turned out not to be the case. See Xaosflux's comments. Both A and B require "developer intervention" (though B is easier).

  Okay, thanks. CapnZapp (talk) 14:50, 21 March 2021 (UTC)

  the help page is for the entire Thanks fuction, not merely to answer questions on "publicly" which suggests Option B as the more logical placement I don't follow this argument. What part of Option A suggests it's "merely to answer questions on 'publicly'" (when the word does not even appear in it)?

  buttons not links That would require even more of a drastic change than Option A. For now let's focus on what is within our reach (which both A and B are). You can request that change separately.

  the degree of "publicness" of your given thanks is debatable Um... so? Are you disputing my characterization ("only who thanked whom when, not even for which edit or on which page, is public")? Nardog (talk) 11:08, 21 March 2021 (UTC)

• See technical notes at phab:T229168 for software changes that would be required upstream. — xaosflux ^Talk 12:25, 21 March 2021 (UTC)
  • @Xaosflux: isn't option B already possible? Can't we just create MediaWiki:thanks-confirmation2 with the content "Publicly {{GENDER:$1|send}} thanks? ([[Help:Notifications/Thanks|help]]) " ProcrastinatingReader (talk) 12:41, 21 March 2021 (UTC)
    • @ProcrastinatingReader: no, as the interface message does not support links (see [https://test.wikipedia.org/w/index.php?title=Main_Page&action=history the example here on testwiki). Enabling this would require a software change, I think this discussion is to help gather support to lobby for such change to be enabled. — xaosflux ^Talk 12:54, 21 March 2021 (UTC)
      • Oh I see. Indeed, that’s a problem. Shouldn’t the function that renders messages parse wikitext? For example in the FlaggedRevs page the standard msg functions are used and that template has wikitext working. Looking at the source for thanks, the message is rendered using JS (using mw.msg) ([17]). Presumably this function doesn’t support wikitext? Is there an alternate JS function that does? ProcrastinatingReader (talk) 13:16, 21 March 2021 (UTC)
      • Did some digging. See here. If mw.message().parse is used instead, it should parse the wikitext, right? So should be a quick change? ProcrastinatingReader (talk) 13:20, 21 March 2021 (UTC)
        • No, this also needs to change. I'm realizing Option A might actually be easier. Either way it's going to involve a change to jquery.confirmable.js. .parse() could raise XSS concerns, so it can be tricky. Nardog (talk) 13:23, 21 March 2021 (UTC)
          • Still, a minor change there. What XSS concerns? ProcrastinatingReader (talk) 14:02, 21 March 2021 (UTC)
• Prefer Option B. The interface usually has (help) in brackets after the text, before the action, I think. For example the editnotice here about pending changes. ProcrastinatingReader (talk) 12:33, 21 March 2021 (UTC)
• Status quo. It's important to note that thanks are public, not private (anyone can look at the thanks log), and it's good to keep consistency with other wikis (Wikidata, Commons, etc.). Having the help link always visible isn't particularly helpful, it's one more link to mis-click on, and if you don't know what 'thanks' is then you can easily find out (just google 'wiki thanks'). Thanks. Mike Peel (talk) 14:26, 21 March 2021 (UTC)

  @Mike Peel: The proposal will require a software change, so it will be "consistent with other wikis", with the link by default pointing to mw:Special:MyLanguage/Help:Notifications/Thanks (cf. the " Help" links seen in many parts of MW). Nardog (talk) 14:31, 21 March 2021 (UTC)

  @Nardog: So this should be a cross-wiki proposal, perhaps on meta, or the output is a suggestion to the developers? Enwp can't decide on its own to change MediaWiki's default config. Thanks. Mike Peel (talk) 14:45, 21 March 2021 (UTC)

  Yeah, the latter. As I said at the beginning, this is just to get a grasp of how many people would think it's an improvement. phab:T229168 has already been marked as "patch welcome", but I fear they wouldn't accept my patch right off if there wasn't "any research saying this will be a positive improvement", as they put it. Nardog (talk) 14:52, 21 March 2021 (UTC)

• comment I think requiring a confirmation at all is unnecessary. "Thank" should be a one-and-done click. Schazjmd (talk) 16:10, 21 March 2021 (UTC)
• As I recall, several people expressed concerns about warning users that thanks were publicly logged. Thus at present I support keeping the current message, and adding a help link to explain the thank you notification mechanism. isaacl (talk) 16:27, 21 March 2021 (UTC)
• Support change, preferably B—the information about it being public is unnecessary and confusing, especially considering that almost everything is public on Wikipedia. When I was just starting on WP, I didn't thank people for a while, because I assumed the weirdly phrased/ambiguous "Publicly send thanks" would be an automated talk page post or something. Aza24 (talk) 17:58, 21 March 2021 (UTC)
• Status quo regarding the word 'publicly', per Isaacl et al. No comment on the rest. --Izno (talk) 18:11, 21 March 2021 (UTC)
• Status quo Really? This is even worth discussing? GenQuest ^"scribble" 18:53, 21 March 2021 (UTC)

  Quite. The "thanks" feature is simply a frippery to make us look like a social media site, not anything that should exercise our brain cells, which should be used to improve the encyclopedia. Phil Bridger (talk) 19:15, 21 March 2021 (UTC)

• Status quo Total non-issue. ~ HAL333 02:30, 22 March 2021 (UTC)
• Status quo I'd support option A or B but the help link is unnecessary imo. Just "Send thanks? Thank Cancel" would be fine. Elli (talk | contribs) 05:35, 22 March 2021 (UTC)
• Support change, preferably B I have always thought this was confusing. You get there by clicking on "thank", and then you are asked if you want to "Publicly send thanks" sort of implying there is a private alternative. I'd even be OK with no confirmation on this one at all. MB 18:54, 22 March 2021 (UTC)
• Status quo. The wording was changed for a reason. I sent an entire years worth of thanks into the bit bucket because the old query (which is closer to A/B above) made it sound like "you're sending thanks, do you want it to be public or private?" I of course just wanted the person I was thanking to be thanked, so clicked on the private option, which was really the "discard sending thanks at all" option. It's important to make clear that thanks are public, and there's no other option, so the status quo wording is fine. SnowFire (talk) 21:11, 25 March 2021 (UTC)

Adding the MediaWiki Tabs Extension

Hi, I'd like to propose adding the Tabs extension to Wikipedia. I was editing this article which contains projected data for 2021, 2022, 2023 and so on. The article also has a nice map, which makes the data easier to visualize, but it can only show the data from one year at the time. I wanted to add a map for each year to show how countries have changed/will change throughout the years, but obviously adding that many maps to the page would just clutter it. A tabbed structure with years for tabs, each having a map would work well here, something like this. This way someone could easily switch between years and instantly see the evolution of various economies.

I looked for a way to create tabs but only found page tabs which require creating multiple subpages. Not exactly elegant. Instead of tabs, this could also be done using Template:Hidden, however it would be a pretty clunky way to do it. In-page tabs work best in this case. But currently there isn't any functionality that allows editors to create in-page tabs on the English Wikipedia. I found the MediaWiki Tabs Extension which seems to be well adapted for this purpose but when I checked Special:Version it didn't show up in the list of installed extensions.

I searched the Village pump for any previous proposals discussing this and found this proposal from last year which also suggested adding the Tabber extension but that proposal didn't reach consensus. I want submit a second proposal for adding this functionality to Wikipedia since tabs can be used effectively in all sorts of situations and are a great tool for displaying data. In-page tabs are particularly useful for articles that with deal large amounts of data and can help make that data easier to interpret and to navigate. The only options those pages have right now in terms of sorting are limited: they can either split the data with HTML anchors which require you to scroll back to the top every time you want to go to another section or divide the content into multiple subpages which isn't always optimal or possible. Or even have to use both in some cases.

Pages like Lists of films for example could be greatly simplified, and would remove the necessity of having to go from page to page through sublists (ex: Lists of films → Lists of Hong Kong films → List of Hong Kong films of the 1960s → List of Hong Kong films of 1963). All those sublists would become irrelevant, and the end pages could be accessed easier using nested tabs, like this.

I'm honestly a bit surprised that the English language Wikipedia lacks this functionality, since tabs are a basic and well established interface element that a majority of internet users are familiar with. Page tabs feel like a strange workaround which in some cases just complicates things both for editing the content and navigating it (a separate page needs to load every single time a user switches to another tab, wasting bandwidth needlessly - for example switching between 4-5 tabs wastes around 1MB just for reloading the WP interface over and over again).

Another reason and the most common one they've been requested before was because editors want to use them in meta or user pages, which I also think is a good idea. A few people had some valid concerns about the usage of tabs, but applying the same rules that apply to subpages to tabs solves most of those issues. This extension also doesn't have the same problem as collapsible content has regarding hidden content disappearing on page load for users using Google Web Lite - I tested a page using tabs and all the content was displayed, even with JS and CSS disabled.

^OUT 06:34, 22 March 2021 (UTC)

RfC: No Nazis

There's an ongoing RfC at WT:NAZI#Proposal. Feel free to participate. Firestar464 (talk) 01:54, 23 March 2021 (UTC)

That was closed with an oppose. Emir of Wikipedia (talk) 14:32, 28 March 2021 (UTC)

Discussion on COI article-space templates

 You are invited to join the discussion at Wikipedia:Templates for discussion/Log/2021 March 19 § COI article-space templates. {{u|Sdkb}} ^talk 01:55, 23 March 2021 (UTC)Template:Z48

Edit conflict mitigation: early-warning tool

I'm proposing an idea for a tool for reducing the number of Edit conflicts, and mitigating the annoyance and other ill effects of them when they are unavoidable, including the number of inadvertent errors resulting from them (or abandoned edits due to not being able to deal with them), by providing a tool that gives early-warning of an impending edit conflict, plus some hints or links about what to do to avoid or minimize the effects.

Let's create a colored, rectangular badge that appears in Preview mode, initially green background, which means, "nobody else has changed this section" (or whatever is inside the Preview window). As soon as its recognized that something has changed (I'm thinking of the Watchlist notice, "View new changes since Timestamp" message that appears at the top of my Watchlist), then the badge background goes amber, and the message inside it changes appropriately (t.b.d. later), along with maybe 2 or three bulleted links or radio buttons or something, for possible actions to be taken. If the section I'm editing disappears entirely (as just happened yesterday, when a bot archived a Talk discussion, while I was replying to it), then the badge goes red.

That would be for changes on disk; but we can go further. Let's have a checkbox in the badge, that says, "Let others know I'm working on this section (or page)" (Maybe also a second checkbox: "and let them know my identity".) Then, if they edit that section as well, their box will immediately go orange, letting them know somebody is working on the same section (and optionally, who it is, if I've allowed that). This is inspired by Pending changes review, which gives you the option to let others know that you are working on a particular change, presumably to avoid conflicts. If you tick the box, then if others also select that same change to work on, they get a little orange message that someone is working on it (maybe even their identity; I don't recall). Enhancement: a Preferences option, "Share my identity by default in the Preview badge when I'm editing a section."

I see all sorts of possible advantages to this, and I have more ideas about what text to place in the badge, and what options might be possible, but this is getting longish for a proposal so I'll stop here for now, and open it to the floor. Mathglot (talk) 21:55, 23 March 2021 (UTC)

• Support That would be useful. ~ HAL333 21:59, 24 March 2021 (UTC)

• Support It would be great to be warned of a conflict when I hit "preview" instead of when I hit "publish". Not sure about the second part of letting people know an edit is in progress, perhaps implement it in two steps with first being the preview warning and the next step being enhancements. RudolfRed (talk) 22:56, 24 March 2021 (UTC)
• Comment doesn't Paragraph-based edit conflict in beta features of preferences kind of address this? Aza24 (talk) 23:04, 24 March 2021 (UTC)
• @Mathglot: I literally starting working on a script for the first part of this a few weeks ago. My long-term plan is to somehow highlight the words in the edit window that cause the conflict; I haven't even started on that part, yet. If anyone wants I can publish my (horribly ugly and buggy) work-in-progress. Suffusion of Yellow (talk) 23:13, 24 March 2021 (UTC)

  But I'd like to clear up a misconception. The automatic merging when you don't get an edit conflict is based on lines, not sections. Last I checked, it's just passing the three revisions off to diff3 (which was meant for merging code, not English text). What I want to attempt is something based on words, so there should be fewer conflicts. Suffusion of Yellow (talk) 23:17, 24 March 2021 (UTC)

Mass-deletion of more than 5500 stubs

Main page: Wikipedia:Administrators' noticeboard § Large batch deletion probably needed

There has been a mass-creation of more than 5500 articles which contain nothing but a coördinate, a Farsi name, and a statement that something is a "village". Unfortunately, the source database that was used to mass-create these actually includes pumps (fa:تلمبه), wells (fa:چاه), farms, and so forth; and the one prose fact in the resultant article is actually false, making these bad stubs that do not even give correct context for expansion. We have a specific list of articles that are likely these, based upon the article then asserting that "no population is reported" for the database entry; and editors are seeking consensus for an administrator to mass-delete them. There is also a separate discussion of the article creator. Please see, and discuss at, the hyperlinked noticeboard. Uncle G (talk) 08:47, 28 March 2021 (UTC)

Shooters

I'd like to propose that Wikipedia stop publishing the names of mass shooters. They do it for notoriety. They want to be remembered as a martyr for violence. Don't give it to them. It also stigmatizes their family members that had nothing to do with their vile acts. Publish where they're from, age, gender, race, background, but not their name. Don't give them the satisfaction of continuing to torture people. — Preceding unsigned comment added by Turtle595 (talk • contribs) 21:00, 30 March 2021 (UTC)

• While I am sympathetic to your cause, Wikipedia is not censored. We do avoid using the names of minors and take other steps to protect the victims, it is our job as an encyclopedia to document notable events in a neutral manner. If we avoid using the names of guilty parties, it is unlikely it would make a difference in the amount of violence in the world. Dennis Brown - 2¢ 21:37, 30 March 2021 (UTC)
• I'd also like to add that we almost always follow the reliable sources (in this case the media) on naming the accused. In basically all shootings we've covered, the shooter's name is already public knowledge. I think Turtle595's desire to deny recognition to mass murderers is admirable, but this really isn't something we can do as an encyclopedia. Spirit of Eagle (talk) 23:56, 31 March 2021 (UTC)

Make access date automatically inserted for visual edit citation tool whenever we enter a new citation

As a VisualEditor user I found it quite enjoying to use their visual citation services, but quite annoying to fill out the access date every time I create a new citation. It would be better for it to be automatically inserted whenever we made a new citation with the citation. --Regards, Jeromi Mikhael 02:08, 31 March 2021 (UTC)

The accessdate refers to when you checked the page the URL points to, not to when you made the citation or edited the article. How would a tool know when you last checked the URL? Martin of Sheffield (talk) 06:04, 31 March 2021 (UTC)

@Martin of Sheffield: Defaults to current day, I guess? When we make a citation we always see the page on the same day. --Regards, Jeromi Mikhael 07:32, 31 March 2021 (UTC)

I don't think this will work. For instance, there are many instances where people just copy over a citation from one place to another, often without checking it anew. If it didn't have an access date in its original location, your mechanism would now give it the appearance of having been freshly checked, which in many cases would be false. Fut.Perf. ☼ 07:42, 31 March 2021 (UTC)

Exactly. Also if I have a citation coming from Zotero that needs to reflect the date that the copy of the PDF (or whatever) was saved. Martin of Sheffield (talk) 07:45, 31 March 2021 (UTC)

Alas, defaulting to current date is just what happens with WP:ReFill (sample edit). Just one more reason why that tool should be killed.

—Trappist the monk (talk) 10:35, 31 March 2021 (UTC)

Union between IP address and account

Before joining Wikipedia I changed with the IP address. I would like "merge" the contributions of the IP address to my account. Dr Salvus 13:04, 31 March 2021 (UTC)

• @Dr Salvus: There is no technical ability to do this. GMG^talk 13:06, 31 March 2021 (UTC)
  • @GreenMeansGo: My proposal is to make this possible Dr Salvus 13:11, 31 March 2021 (UTC)
    • Dr Salvus, this was proposed in the 2019 wishlist & got nowhere. Cabayi (talk) 13:17, 31 March 2021 (UTC)
      • It's not feasible anyway, because there's no way to determine that all of the edits from an IP definitely came from the same person. GMG^talk 13:43, 31 March 2021 (UTC)
        • @GreenMeansGo: Can't we use Check User? Dr Salvus 18:49, 31 March 2021 (UTC)
          • Not really. The m:CheckUser policy states that data is only held for 90 days. I would add that because IP addresses are almost always being re-assigned, and can be shared among several people, having that data still wouldn't be sufficient proof. -- zzuuzz ^(talk) 18:59, 31 March 2021 (UTC)
            • ^ GMG^talk 19:44, 31 March 2021 (UTC)
• If your intent is to claim responsibility for bad consequences of your previous edits then that can be done simply by linking IP addresses that you have used on your user page. If your intent is to use previous edits for the purposes of hat-collecting then that can only end in tears. Phil Bridger (talk) 19:57, 31 March 2021 (UTC)

Should we sell Wikipedia while the Wikimedia Foundation isn’t looking and pocket the proceeds?

So I was recently in contact with a friend and mentioned that I edit Wikipedia in my spare time. He mistakenly assumed that I owned Wikipedia and offered to purchase the project off of me. I’m positive I can get $5 billion dollars from all of this, and I’m happy to split this with our ~150,000 active editors (which comes in at around $3 millions dollars and change per person). Wikipedia is a community first and foremost, so it would be unethical to knowingly sell false title to the project without first getting consensus from my fellow editors. So what say you: do you want to become a millionaire? Answer quickly please: I’m pretty sure its illegal to sell things you don’t own, so we should probably close this deal within the next 24 hours. Spirit of Eagle (talk) 00:00, 1 April 2021 (UTC)

We definitely should. As well, we should probably round up all the $3 donations Wikipedia has received and spilt that among us, so we make the most money possible, because the more money we get, the better. All us editors bust our butts all day as unpaid volunteers making an encyclopedia better, it’s time we get some money from it. Now quick, before the WMF finds out. D🐶ggy54321 ^{(let's chat!)} 00:11, 1 April 2021 (UTC)

• This is an excellent idea. The WMF had their chance and failed. However, my friend William just made me a counter offer of $6 billion on the condition that we put the encyclopedia on wheels. Should we go for it? —pythoncoder (talk | contribs) 00:23, 1 April 2021 (UTC)

I did the math and that should just about double our take home pay. Do it! Spirit of Eagle (talk) 00:32, 1 April 2021 (UTC)

Uh oh https://en.wikipedia.org/wiki/Wikipedia:Sockpuppet_investigations/Pythoncoder Natureium (talk) 00:32, 1 April 2021 (UTC)

• Help:Buying Wikipedia may be of use. {{u|Sdkb}} ^talk 00:40, 1 April 2021 (UTC)
• Strong abstain per nom. JJP_Master 00:49, 1 April 2021 (UTC)
• Definitely not! We aren’t being offered enough. I struck through my previous comment after reading Help:Buying Wikipedia. I am offended that someone would only offer $6 billion when the price of Wikipedia is $63,469,965,175,772,420.69 and counting!!! We need actual wealthy investors (not like the cheapskates you provided) who are willing to buy Wikipedia at the price it is valued. D🐶ggy54321 ^{(let's chat!)} 00:52, 1 April 2021 (UTC)

RfC on limiting minor edits

Question: Should the minor edit functionality be limited to a group of users (such as autoconfirmed or extended-confirmed users)?

• Option 0 (status quo): Limit minor edits to registered users.
• Option A: Limit minor edits to autoconfirmed (or confirmed) users.
• Option B: Limit minor edits to extended-confirmed users.

This is a follow-up to this RfC on effectively disabling minor edits. As there was no consensus then, this is to establish clearer consensus regarding an alternative proposal. (This is my first time requesting comment, please let me know if I'm doing anything wrong.) Tol | Talk | Contribs (formerly Twassman) 00:02, 1 April 2021 (UTC)

• Question What about the Gordian-knot solution, just get rid of the concept of "minor edit" entirely? Was that considered? --Trovatore (talk) 00:23, 1 April 2021 (UTC)

  The aforementioned RfC demonstrates fairly clear opposition to removing minor edits entirely. — ⊥ɥǝ Ǝɐɹʍıƃ (ʇɐlʞ) 00:30, 1 April 2021 (UTC)

  Well, that one says something like "limit by policy to anti-vandalism reverts", which is not really a simplification and adds another layer of things people can and can't do. What I'm talking about is, the whole concept of "minor edit" goes away entirely; it's not that some people can do it and some people can't; it's not that there are rules about when you can do it; it's that it just doesn't exist, period. Even historical edits would no longer distinguish minor vs non-minor. Like it never existed at all. I would support that. --Trovatore (talk) 00:37, 1 April 2021 (UTC)

  You are right that the RfC was specifically about restricting minor edits rather than removing entirely, but what I meant is many of the arguments in opposition would apply just as well for a proposal to remove them entirely, so I don't see that reaching consensus. — ⊥ɥǝ Ǝɐɹʍıƃ (ʇɐlʞ) 00:45, 1 April 2021 (UTC)

  The !voters may not have taken into consideration the advantage of a genuine simplification, for once, given that none was offered. --Trovatore (talk) 00:47, 1 April 2021 (UTC)

  Given that the title was to "Disable minor edits" I doubt this. Tol | Talk | Contribs (formerly Twassman) 01:03, 1 April 2021 (UTC)

• Support B, then A. As I said in the other RFC, currently it's about as useful as the evil bit. The point of the minor edit checkbox is to say "you can safely ignore this edit"; so long as vandals, spammers, and "what does this button do?" types can use it, "minor" edits still need to be reviewed. For new users, this will be one less thing to worry about, and one less thing to get yelled at for misusing. Suffusion of Yellow (talk) 00:34, 1 April 2021 (UTC)
  • Which strikes me as a good reason to remove the concept entirely, rather than to restrict who can use it. --Trovatore (talk) 00:42, 1 April 2021 (UTC)
    • Indeed, I would support that too. But in the other RFC there was significant opposition to this idea, so let's at least remove it for the users most likely to be spammers, vandals, or clueless. Suffusion of Yellow (talk) 00:49, 1 April 2021 (UTC)
      • Meh, can't get excited about that. If the concept is useless, remove it. If we're not going to remove it, then since I'm pretty much going to ignore it anyway, I don't really care who can use it. --Trovatore (talk) 00:52, 1 April 2021 (UTC)
• Support B, then A Johnbod (talk) 00:39, 1 April 2021 (UTC)
• Support A, Oppose B per my arguments in the previous RfC. I am not in favor of expanding the scope of EC if it means taking rights away from autoconfirmed users. The threshold for EC is too high for a feature as minor (no pun intended) as this one. — ⊥ɥǝ Ǝɐɹʍıƃ (ʇɐlʞ) 00:53, 1 April 2021 (UTC)
• Option A/B (as requester): minor edits are frequently abused or incorrectly used by new users, and this could limit this feature to users who can be better trusted not to misuse it. Tol | Talk | Contribs (formerly Twassman) 01:09, 1 April 2021 (UTC)
• Support 0, strong oppose B: I'm shocked to find out the VPR discussion, which I'd filed away mentally as "some weird discussion between people with outlier opinions that's probably fizzling out", made its way here, and I at first thought this was an April Fool's RfC that hadn't been tagged properly. I can grudgingly recognize an argument for autoconfirmed (selects away some-but-not-all of the people with actively negative amounts of clue) but overall think that an example of over-bureaucracy at the terror of Spammers And Vandals taking away a minor, useful, and mostly noncontentious function. Limiting it to extended confirmed is beyond parody -- do you know what 500 edits sounds like to someone who isn't Into Wikipedia? We use extended confirmed protection as a last resort for the most massively controversial and abused articles to make sure they're only edited by people who are in too deep to continue it. I do not see the point of restricting something to vested contributors (B) or the patient (A) that can very well be better explained with replacing "This is a [[subtle link|minor edit]]" with "This is a minor edit [[less subtle link|(when to check this box)]]". Vaticidalprophet 01:12, 1 April 2021 (UTC)

  Also, should this have a watchlist notification? Vaticidalprophet 01:15, 1 April 2021 (UTC)

  I'd see your point if it was a feature that had any direct benefit to the user. But (like autopatrolled) it can only benefit other people. How is the non-extendedconfirmed user harmed by the removal? Suffusion of Yellow (talk) 01:21, 1 April 2021 (UTC)

RfC: Modification of Agatheira for syntactic compliance with linguistic precision guidelines

Following workshopping at the most recent Wikimania and extensive consultation with the WMF, Blippers, and an external archeological consultant I happen to know off-wiki, I am pleased to present this landmark RfC regarding a significant stylistic overhaul of Agatheira that could establish an enduring precedent for our ancient Turkish city articles and reaffirm the Manual of Style's authoritative status vis a vis matters of punctuation.

Background

First, some essential context. Agatheira has long been known as one of WikiProject Classical Greece and Rome's articles. Although it arguably suffered an indignity or two on its talk page early on, its five primary editors have worked to develop it to its current status as a standout example of our coverage of ancient Lydian towns located near Halitpaşa. It has averaged pageviews per day and experienced a particular surge of popularity last August for obvious reasons. Many of you who edit in the area may also recall the major overhaul it underwent last December.

Guiding principles

Wikipedia's Manual of Style (MOS, Mos, or MoS) has provided authoritative guidance on grammatical matters on Wikipedia since its creation in October 2001. During this time, it has served as a crucial force for standardization of capitalization, punctuation, and more. One of the most comprehensive sections of the Mos, the section on dashes, stretches to more than 14,000 characters and covers a variety of dash-related concerns. Of particular relevance to this RfC is the clause within the "In ranges that might otherwise be expressed with to or through" level-5 subsection, which states the following: For ranges between numbers, dates, or times, use an en dash. The precise date when this clause was added is not currently known, but it has been present at least since last April.

Proposal

In the current revision of Agatheira, the second sentence of the second paragraph includes the phrase during the reign of Eumenes II (188-158 BC). This RfC asks whether it should be changed to during the reign of Eumenes II (188–158 BC), replacing the current hyphen with an en-dash. Feel free to share your thoughts on that question below. Given the potentially charged and controversial nature of this area, please remember to keep personal attacks to a minimum and to ground your arguments in Wikipedia policy (including IAR as needed). Please also remember to keep your comments concise to respect the volunteer nature of the project. - {{u|Sdkb}} ^talk 01:00, 1 April 2021 (UTC)

Survey

• Weak support as nominator. I have serious reservations about this due to WP:BROKEN and WP:COSMETIC, but on balance I feel it would be an improvement. {{u|Sdkb}} ^talk 01:00, 1 April 2021 (UTC)
• Support: It should follow MOS; using en dashes for ranges is generally considered correct. Tol | Talk | Contribs (formerly Twassman) 01:06, 1 April 2021 (UTC)
• Question: En-dash? Is that an emoji? Suffusion of Yellow (talk) 01:07, 1 April 2021 (UTC)