The policy section of the village pump is used to discuss already-proposed policies and guidelines and to discuss changes to existing policies and guidelines. Change discussions often start on other pages and then move or get mentioned here for more visibility and broader participation.
If you want to propose something new that is not a policy or guideline, use Village pump (proposals). For drafting with a more focused group, you can also start on the talk page for a WikiProject, Manual of Style, or other relevant project page.
If you have a question about how to apply an existing policy or guideline, try one of the many Wikipedia:Noticeboards.
If you want to ask what the policy is on something, try the Help desk or the Teahouse.
This is not the place to resolve disputes over how a policy should be implemented. Please see Wikipedia:Dispute resolution for how to proceed in such cases.
Please see this FAQ page for a list of frequently rejected or ignored proposals. Discussions are automatically archived after remaining inactive for two weeks.
I took long hours to condense core content policies into one document. By my calculations, I can retain at least 95% of the meaning of the policies, including most of their intricacies and details, by using only about a third of space that the policies and guidelines, scattered among multiple pages, now take (I estimated from raw text that the policies and guidelines I summarised have about 530 KB of raw text but under the compilation have just 173 KB - still a lot but much better).
It was a hard task and appears about as hard as working in one person to recompile the Constitution of Alabama - a ridiculously long document running at 373K words - 420K words before 2022 (for comparison, War and Peace is 587K words, and there's a good reason they publish it in volumes), but I think it is more than worth it, as people will have a unified set of policies that will be easier to read for people because there's gonna be much less of that but reflecting the same meaning. The overabundance of policies is one reason we have few new editors - there are too many rules, and then folks just randomly throw WP or MOS shortcuts not immediately obvious to the bystander, and suddenly nobody wants to join a project with United States Code-long rules and obscure jargon.
I will appreciate all feedback from you - positive or negative - and preferably some help into condensing further policies and guidelines, such as those about conduct, legal, editing etc. into one page where everything belongs.
I wish you well in your endeavour, but am worried that it will fail in the end because everyone thinks that "their" sub-sub-clause is vitally important. I admit that I rarely look at policies or guidelines now, but find a few basic ideas, along with common sense, to work. Phil Bridger (talk) 12:14, 22 August 2024 (UTC)[reply]
That works for me as well (I normally look into the rules on an ad hoc basis), but then you wouldn't need all those volumes of tiny rules covering, like, 99% of cases, and yet here we are. Also, admins themselves need a clear set of rules for proper enforcement (even if you catch the gist that the persob is just NOTHERE - an essay btw - you still kinda need a more concrete reason that just "that's my hunch")
It's like with RL: pretty common-sense that you shouldn't kill or rob anyone, or what appears common-sense like not using the army or the government to finance/securre your own reelection campaign, and yet these are codified lest anyone have an idea to bend the rules. Szmenderowiecki (talk) 17:29, 22 August 2024 (UTC)[reply]
This exists WP:Nutshell and I think your effort is noble but better focused on improving accessible language and navigation of existing guidelines for newcomers. Twinkle has feature to welcome new users for example. WP:Mentor finds ways to automate assisting newbies. ~ 🦝 Shushugah (he/him • talk) 12:47, 22 August 2024 (UTC)[reply]
Just to pick one example. There's no policy that an article must have (any) sources, let alone one. Yet we consistently advise new users to create articles with multiple sources. Save the edge cases and careful readings of guidelines/policies for advanced users who want to push the margins or change consensus. ~ 🦝 Shushugah (he/him • talk) 12:49, 22 August 2024 (UTC)[reply]
Actually there very much are two policies that prohibit articles without sources:
Verifiability says that you should only add content that you can check against a reliable source, and that you can remove any unsourced content
No original research says you just can't make stuff up. The only way you can have some sort of content is if it can be supported by a reliable source. Technically just have to demonstrate that the source for the passage is somewhere but if you don't provide it in the article, you can totally expect it to be removed and it's gonna be your problem.
No, policy prohibits unverifiable articles, not unsourced articles. Sources are required for BLPs and anything that is likely to be challenged should include a source (but this doesn't have to be inline). An article List of uncontroversial statements of fact consisting of things like "The sky is blue", "Many people are Catholics", "The 1970s happened before the 1980s", etc could be completely fine (it would be deleted, but for reasons completely unrelated to not having sources). Thryduulf (talk) 17:28, 22 August 2024 (UTC)[reply]
Look, I'm not making this up:
Any material lacking an inline citation to a reliable source that directly supports[b] the material may be removed and should not be restored without an inline citation to a reliable source
Which de facto means that if you are adding unsourced content, you are wasting your time as any editor may come by and just remove what you wrote, tell you "lacks an inline citation, I don't care why, pics or it didn't happen" and you are gonna still default to having to add a source (and then again giving time to fix it is a courtesy you needn't, though probably should, extend; though if you have the means to fix it yourself, you should do it)
So there's no obligation to source an article only in the most literal reading of policies. Anyone can enforce this policy provision. WP:SKYISBLUE is just an essay, although one with a pretty large following (and which totally makes sense for me, which is why there is a footnote to that effect) Szmenderowiecki (talk) 17:47, 22 August 2024 (UTC)[reply]
... any editor may come by and just remove what you wrote ...: the operative word is may. The reality is there's loads of unsourced text on WP, much of which will take years for it to be challenged, if ever. —Bagumba (talk) 18:29, 22 August 2024 (UTC)[reply]
It has been tried before. Take a look at Wikipedia:Attribution, an attempt 18 years ago to consolidate some policies. Some very active and well regarded Wikipedians put a lot of time and effort into that proposal, but it was rejected by the community. Consensus can change, but I suspect the community remains just as resistant to change as it was then. Donald Albury20:07, 22 August 2024 (UTC)[reply]
So I did read the poll, and a lot of stuff that was probably relevant here doesn't apply:
Users were complaining about lack of participation/that proposal being forced down their throats as policy - not an issue here (yet)
Merger of NOR, V, RS didn't appeal to people - not abolishing them, just giving sections to these concepts, not a problem.
Users complained about one massive page, or that they preferred separate policies rather than a massive policy code - well that is an issue to discuss but again it's not something that should extinguish all debate before it even starts.
Disgusted that truth is deprioritised - kinda not applicable here, because I'm not changing the framing of policy, just condensing it.
Change is unnecessary - again, debatable but let's have that debate in the first place
WP-links - well, you will have them all you like. Again, something to be discussed.
Assessment of any changes and their impact on disputes - to be discussed, again. This is how rulemaking process should work.
"If it ain't broke, don't fix it" - my whole point is that it is, in fact, broke, so needs fixing.
And again, you can say "meh, we tried eons ago and it didn't work, why bother anymore" but that's gonna be a catch-22, because nothing will change without discussion, which you don't want to hold anyway.
I believe the attitude should be "OK, let's see what you did there and if it makes any sense". It would be another thing if you told me why what I did was bullshit, which is fine. Szmenderowiecki (talk) 18:45, 23 August 2024 (UTC)[reply]
I know, but that's not the point of the compilation. What you are pointing to has a different purpose.
Nutshell, WP:5P etc. is a post hoc summary of policies and guidelines that summarise the main goals in slogans. Just like a company saying "we want to increase the market share; we want good treatment of workers" but not saying how.
WP:SIMPLE is a very high-level summary of policies and guidelines. It's the company analogue of saying: Good treatment of workers means paying more than the minimum wage, giving them extra breaks, paid leave and some other perks, without telling much specifics.
The body of the policies and guidelines is like all internal company directives about pay grades, conditions of getting worker benefits, levels of compensation, powers of HR/executives etc. This page intends to clean up all this body of policies and codify them in a couple of places, grouped by category, so that we remove unnecessary bloat, as in too many redundancies and passages repeated across different policy pages, extraneous comments etc.
@Szmenderowiecki, I don't feel like you're hearing what people are telling you, so l'm going to try a completely different, un-Wikipedian way of explaining this, because the previous efforts haven't worked, and maybe this will get your attention. Here's my new way:
Hi, Szmenderowiecki, and welcome to Wikipedia! I see you've been editing for just four and a half years, and that you've made a few hundred edits at Wikipedia:Reliable sources/Noticeboard, which I really appreciate. I don't know if you knew this, but you're in the top half a percent of contributors for all time. I also notice that you've never edited a single policy, and you have only made one small, uncontroversial edit to a guideline.
Just so you know, most of the people who have responded to you in this discussion have been editing for 15 to 20 years, and have made between 50,000 and 170,000 edits. Also, relevantly, we've been much more active in developing Wikipedia's policy and guideline ecosystem. If you'd like to see an incomplete overview of my own policy-related work, then you can start at User:WhatamIdoing#Policies and guidelines you can ask me about.
Now that you understand who's at the table for this discussion, I want to point out that there is an English saying that fools rush in where angels fear to tread. Every person in this discussion has more experience than you, and every single one of them thinks that, even though your goal is laudable and praiseworthy and at least partially shared by everyone here, your approach is not likely to be successful. It is, of course, possible that you know better than any of us and that rushing ahead is a great idea, but I suggest to you that it is unlikely that all of us are wrong in urging caution and small steps.
If you think you could slow down and get some more experience, and if you're willing to consider doing this over the space of years, then I think we could help set you up for success. For example, if we implemented this idea, that would get about 300 words out of a policy. The next step is to write a good RFC question. If you're interested in this, you could get some practical experience by helping out. WhatamIdoing (talk) 22:26, 23 August 2024 (UTC)[reply]
I don't know if you realise it, and I guess you didn't mean it, but the comment has a very strong patronising "you are too young to understand" vibe which I have a hard time shaking off rn.
I asked you for specific input in opinion and help, and I just think the folks who suggested Nutshell etc. misunderstood my intentions. I apologise if I wasn't clear. My intent is to retain the same scope and level of detail but in fewer words.
If what you meant is to do it in increments, fine, that's an option, still I'd love some feedback if I fucked up with the text in the first place. That is valuable. I'm open to discuss it one-by-one. I will hear input from people who actually drafted policies. That was what I intended to do anyway. Szmenderowiecki (talk) 23:30, 24 August 2024 (UTC)[reply]
You've dropped things, rearranged things, and changed things, and I suspect you have done this without knowing what effects any of that will have.
For example, you've added the word secondary to the WP:GNG, and swapped in a description for the WP:SIGCOV language:
Original: "A topic is presumed to be suitable for a stand-alone article or list when it has received significant coverage in reliable sources that are independent of the subject."
Yours: A topic generally may have a stand-alone article or list when several reliablesecondary sources that are independent of the subject address the topic directly and in detail, so that editors do not have to resort to original research.
I've been trying to get a definition of SIGCOV for years (and years), and failing repeatedly because nobody wants to admit how long (or, perhaps more precisely, how low) "in detail" actually is, for fear that some "unworthy" subject might deliberately seek a qualifying level of independent media coverage. The NOR line in the GNG is basically worthless, and AFAICT removing it would have no effect whatsoever on AFD outcomes, but the fear of making changes to such a high-visibility sentence will likely prevent us from fixing that. Your [e] footnote requires a huge amount of work (e.g., primary sources aren't always about events, sources don't have to adhere to the neutral point of view, a smaller number of high-quality sources is not automatically less indicative of notability than a large number of worse sources).
At the time we started leaning on secondary sources (about 15 years ago), we had a lot of editors who thought that secondary was a fancy way to spell independent. You have added a requirement for secondary sources that does not actually appear in the GNG statement (though it is in the explanations). The GNG offers a conditional rebuttable presumption, which you have turned into a statement of permission (may/are allowed to have...). The GNG says that multiple sources are only "generally expected", rather than required, and you have changed that. Oh, and "several" is often interpreted, at least in American English, as meaning "four" (a=one, a couple=two, a few=three, several=four), whereas the GNG is usually looking for "two".
Among the things you haven't resolved is whether the sources for an article must be considered in isolation. For example:
If I have ten brief independent secondary sources, is that multiple+independent+secondary+SIGCOV, or just multiple+independent+secondary and no SIGCOV?
If I have SIGCOV in a very lengthy, extremely detailed independent primary source, and I have a non-SIGCOV secondary source, does that add up to multiple+independent+SIGCOV+secondary and therefore notability overall, or do I have to get all three key qualities (independent+secondary+SIGCOV) in each source separately?
@Szmenderowiecki, just so you know, I'm always going to be interested in this. If you want to talk about how to improve our written policies and guidelines, then feel free to drop by User talk:WhatamIdoing and tell me about your ideas. It doesn't matter to me if that's that's next month, or next year, or next decade – I'd be happy to hear your ideas whenever you want to share them. WhatamIdoing (talk) 23:19, 25 August 2024 (UTC)[reply]
OK, I'm finally back after fixing my phone with 2FA, I will respond to your suggestions above on the user talk page. It will be there in an hour or two. Szmenderowiecki (talk) 19:31, 7 September 2024 (UTC)[reply]
If anything, the project would benefit from more "fresh eyes"—editors with enough experience to speak somewhat intelligently about these issues, but without so much experience that they are heavily invested in the status quo. There is a strong, almost indisputable case that current PAGs are far too convoluted and complex for the project's good. As a practical matter, the core problem is that the self-selected self-governance model, which created the problem, is incapable of addressing it. Resistance is futile; hence my semi-retirement after about ten years of futile resistance. ―Mandruss☎02:54, 25 August 2024 (UTC)[reply]
I would add that it's useful to hear from editors who still remember their first few edits, because a sentence that makes sense to the "old hands" isn't necessarily any good for the majority of editors. An actual majority of editors has made five or fewer edits, total, ever. WhatamIdoing (talk) 03:02, 25 August 2024 (UTC)[reply]
It might be "useful" in supporting the position that current PAGs are far too convoluted and complex for the project's good (but I think that's self-evident). As for fixing the problem, not so much. Incremental change is never going to be enough; what's needed is massive overhaul, and that's just not going to happen under the current model. Meanwhile, the current model is sacrosanct. ―Mandruss☎04:45, 25 August 2024 (UTC)[reply]
I'll confess that I rarely look at the written policies & guidelines, & mostly use them as a citation when I need to emphasize a point to another editor. I consider what they say is basic common sense, but I've been around so long that I've probably internalized all of the important points. (This is not to say that the original poster is wasting their time. The written policies & guidelines have been considered a mess for countless years.) -- llywrch (talk) 01:49, 26 August 2024 (UTC)[reply]
This is part of the reason why English Wikipedia's guidance is sprawling: a lot of additions get made because a situation arises, some people say "we should have a rule for that", it gets added with a shortcut used for jargon, and editors brandish it in future discussions. As I've written before, it would be better to address problems without creating new specialized rules. But in English Wikipedia's current decision-making environment, there is little appetite to delegate to a working group to more effectively rewrite any pages. Amongst those who like to discuss these matters, there are enough editors who want to be able to weigh in on each sentence that it's hard to modify existing guidance pages, and instead we accrete more. Writing well is hard; writing well in a group is even more so. The irony of a crowd-sourced web site is that crowd-sourcing works best for making incremental changes, but consumes a lot of editor time in discussion for larger-scale changes. Which is why the path of least resistance for modifying guidance pages (and articles too) is to add a few sentences, rather than rework the pages. isaacl (talk) 23:22, 26 August 2024 (UTC)[reply]
As a new editor, adopting this would help me contribute more effectively. I would cautiously suggest that it seems like people know too many abbreviations. Support ForksForks (talk) 22:08, 26 August 2024 (UTC)[reply]
And we use different WP:UPPERCASE for the same page/section, sometimes resulting in one editor claiming that "WP:PAGE" supports his view and the next saying that "WP:SAMEPAGE" requires the opposite, and neither of them realize that they're talking about the same page. WhatamIdoing (talk) 00:35, 31 August 2024 (UTC)[reply]
I don't see any substantive argument that our current P&G structure / the ability for users to peruse it is actually problematic. It should be telling that those editors generally seeing potential in this are relatively new, and those who don't are relatively experienced. That's not (just) survivorship bias, that's experience indicating that what is perceived by some newer editors as a pedagogical issue is (as stated above) actually just the inherent difficulties in learning how to do something conceptually multifaceted and difficult.
Editing is just hard intellectually and socially, and there's no shortcut to becoming familiar with it to be found in compiling one huge document people will not be able to digest immediately versus having several documents. Remsense ‥ 论22:18, 26 August 2024 (UTC)[reply]
I agree with the inherent difficulties in learning how to do something conceptually multifaceted and difficult, but I believe there are more effective and efficient ways to communicate the PAGs than our current structure. I also believe it is pretty much impossible to replace the current structure with a more effective one, having seen what it takes to add a sentence to a policy. Schazjmd(talk)23:32, 26 August 2024 (UTC)[reply]
I think there are substantial improvements that can be made throughout each policy, or even refactoring involving multiple policies, but I think on the broadest level the modular P&G structure has no actual downsides—this isn't Justinian's code. Remsense ‥ 论23:36, 26 August 2024 (UTC)[reply]
Remsense No, and neither was that my point. The whole structure remains modular, just the modules are larger. As in, when you open a typical English textbook for foreigners, you have six books (A1-C2), each having 8-12 units.
Right now you have a separate page for every unit. I propose a separate page for a whole book, and all units get codified in one book.
So no, it's not gonna be one mammoth document, but maybe 5-8 big ones. Again, not even touching the Manual of Style.
That is more or less the reason why civil law countries have law codes. Instead of the law being scattered around 20-30 acts, you have a big one, but it resolves like 90-95% of cases. Any additional laws just build up on the codes. And even for American folks, the United States Code is sorted into 53 titles, and that organisation is in some sense not unlike that of civil law codes.
The downside is that yes, you get a tl;dr document (if you need the short story, WP:SIMPLE is indeed what you are looking for). The upside though is that you actually don't have to repeat parts of other policies in other places. Note how WP:V repeats a lot of RS because without it, V would kinda fall apart. Or how WP:OR has to remind of "related policies". Look at any subject-specific reliability guideline - depending on the size, up to half of it is copypasta from WP:N, and the only reason, apart from making the PAGs internally structurally sound, is to try to show that the policies are interconnected. In that large document, the one thing that connects all these policies is at the very top: why we need them. Because right now it's scattered all over the place. Other than that, the idea is to avoid redundant repetition. Szmenderowiecki (talk) 19:49, 7 September 2024 (UTC)[reply]
No, because a navigation template, either the one at the top or at the bottom - not sure which you refer to - is not intended to answer what the policy says, only shows you the way to actual policies. It's just a signpost, and it should stay that way. Very useful, but are not the thing I propose.
Why do we need to start a complimentary, high-complexity codification of documents which are already all in the exact same place? Again, this isn't Justinian's code, it's simply not the case that policies are "scattered all over the place".Remsense ‥ 论20:45, 7 September 2024 (UTC)[reply]
It is not "complimentary", it is intended as a replacement. And I'd say the current system has way more complexity because you need a hundred pages or so, a couple of FAQs, explanatory essays that sometimes define scopes of guidelines (see WP:BIOMED) and a couple of essays that de facto have the force of a guideline.
And what's the "exact same place" anyway? I can think of no one such place other than a navigation template, and it's simply not good enough, because duplicating the same thing has numerous downsides. a) slightly different text will be interpreted slightly differently, and it's not like wikilawyering will disappear anytime soon, b) maintaining two instances of same text is harder than maintaining one, c) after a change on page A there may appear a contradiction either with same copy of policy on page B or with a different rule on page C, which the community may simply not notice at first.
And if you think there are substantial improvements that can be made throughout each policy, or even refactoring involving multiple policies, why haven't you proposed that yet? Like a rewrite of V or N or whatever? Szmenderowiecki (talk) 21:39, 7 September 2024 (UTC)[reply]
About WP:OR has to remind of "related policies": I wonder if we should pull that. It makes the pages longer and introduces some confusion (e.g., if someone cites WP:THISPAGE for a rule that actually belongs to WP:THATPAGE, but happens to get mentioned in WP:THISPAGE). WhatamIdoing (talk) 23:29, 8 September 2024 (UTC)[reply]
I'm all for condensing our guidance down where possible, but I have to agree with Remsense – I'm not sure the goal should be putting absolutely everything on a single page.
There are quite a few policies and guidelines which the average editor doesn't really need to understand in detail unless they're editing in the particular area where it applies. For example, you've included our guideline on reliable sources for medical topics in your compilation, but I'd say that's not a guideline every editor needs to know in detail. For the average editor, it's probably enough to know "medical topics have stricter standards on sourcing, and if medical topics ever come up in my editing there's a page out there I can consult for more information". I don't see the benefit in trying to squeeze the details of that guidance onto a single page together with every other policy.
Sometimes there's a good reason for guidance to be extremely detailed or unusually attentive to particular wording. Our notability guideline for companies comes to mind – it goes over what sourcing does and doesn't demonstrate notability in pretty exhaustive detail. But the details are there for good reason – there's a lot of bad actors out there who manipulate the system for commercial interests, so we need to be unusually strict and explicit about exactly how to determine notability.
Of course, there's still a lot of cases where these caveats don't apply and we really could condense things down without losing much. My advice would be a more targeted approach: rather than condensing everything at once, work out where the low-hanging fruit is and push for changes there. – Teratix₵13:11, 28 August 2024 (UTC)[reply]
Or rather, do you mean, the response to the proposal has reminded you of "On Exactitude in Science"?
The notion that one condenses a vast network policies to a simplified reference document, a map if you will, is well-founded (and as others have pointed out, a similar concept exists already in explanatory essays WP:Nutshell and WP:SIMPLE). Many objections posted here seem to be that currently the simplified essay does not accurately represent every single paragraph of every single policy accurately.
Fwiw, I think a complete from-scratch attempt at simplified reference document every few years, to be measured against whatever already exists, is a good thing. As people choose to direct new editors to one version or another, or as new editors opine on one or another, maybe something appears completely superior. Or not -- we have a bunch of independent ways of presenting the basic P&G. Currently you can choose to link from between 5 or more independent general editing tutorial portals for newbies (WP:NEWBIE, Help:Introduction, Wikipedia:GLAM/Beginner's_guide_to_Wikipedia, WP:MAN, etc.).
An explanatory essay is not new policy, and as long as it is in userspace it does not require consensus. I frankly think some of the behavior so far has been disappointing to the spirit of VP: when someone asks for feedback on a work in progress here, then we should at minimum be constructive. (And to be sure: any feedback with a tldr of "don't do this" or "nobody wants it" or "put your effort somewhere else" is the opposite of constructive.) SamuelRiv (talk) 20:59, 4 September 2024 (UTC)[reply]
Just for clarity, what I would mean by "simplified" in this case is not "this is the TLDR version" - you have pointed to enough of that but "this is the new comprehensive version, essentially same content, but we completely reviewed the text and made it clearer, shorter, neater, better organised etc.", with a dollop of advice from WP:POSA and a sprinkle of other advice about concise writing. Other than that, I agree. Szmenderowiecki (talk) 20:13, 7 September 2024 (UTC)[reply]
I guess I'd suggest to just keep your presentation delicate for a while. Presenting as an alternative/better tldr version might be received more openly, since you've seen the reaction here when talking about a total conceptual rules overhaul.
In evaluating the quality, since many P&G are separated into pages with separate Talk sections, you might take one section of your draft you think is pretty refined and present it to one of those pages to get feedback on whether it makes sense as a concise version, then make adjustments from there. Everyone here is aware of WP:Bloat, so after doing the refinement, go back to the section Talk page and say at some point, "what if this entire page were replaced with just this, would it break anything?", you might get a better feel for what's possible. SamuelRiv (talk) 21:58, 7 September 2024 (UTC)[reply]
The problem is, tl;drs are already there, and xkcd covers creating "this one tl;dr that solves everything" well. So I don't want to go that way.
As for your suggestions, I may be subjective in what I believe is refined, but let's try that. Just choose what you believe is the best summary out there. Just not the one that has cross-references to other policies in the text, because folks will cry wolf and claim that it's absolutely essential to understand the policy. Szmenderowiecki (talk) 22:25, 7 September 2024 (UTC)[reply]
I enjoy reading the news in the "in the news" section of the home page of Wikipedia. It has the kind of hard news I like: tornados, earthquakes, fires, hurricanes, etc. from around the world.
But new stories don't come out very often. The current "cycling" item is now 10 days old.
Can't we get a top story each day? Ifyoucrydon'tlose (talk) 20:06, 28 August 2024 (UTC)[reply]
There are some nominations in the queue (see WP:ITNC) but these haven't been posted as they either haven't found to be appropriate topics for ITN to feature, or the article lacks the quality needed to be featured on the main page. As WP is not a newspaper, ITN should not be considered as a news ticket and we don't have an obligation to keep fresh stories there. Of course, more nominations for stories that meet the criteria can always help, but it should be forewarned that there is usually a high bar related to significance of the stories to be posted. — Masem (t) 20:17, 28 August 2024 (UTC)[reply]
Discussions at In the news are largely composed of a group of regulars who have their own ideas about what is "significant" enough to be posted to the main page. I largely avoid ITN because my one experience attempting to nominate something there was extremely frustrating and I don't think it's worth wasting my time updating articles with breaking news stories that editors will randomly find "insignificant" for idiosyncratic reasons. voorts (talk/contributions) 21:42, 28 August 2024 (UTC)[reply]
Its primary purpose is to point readers to Wikipedia articles that discuss the people, places and events that are in the news. These articles will (ideally) give readers the background information to better contextualize and understand what they read/see in news media. Blueboar (talk) 12:30, 29 August 2024 (UTC)[reply]
That's a good description and I thank you for that.
One small note: If you hover over and click on "Ongoing", you can get a list of previous items.
It is highly subjective whether an event is considered significant enough, and ultimately each event should be discussed on its own merits. The consensus among those discussing the event is all that is necessary to decide if an event is significant enough for posting.
Anything else is subject to who shows up. If non-regulars chose to change the outcomes, there is no current guideline-based reason why present trends couldn't change. Until then, what you are seeing is the will of the "regulars". —Bagumba (talk) 10:29, 30 August 2024 (UTC)[reply]
Tags for this (fairly new) WikiProject are being mass-added to practically every article imaginable that is tangentially related to the 2010s — every sporting event in the 2010s; every film, TV show, and video game released in the 2010s; every Super Bowl, solar eclipse, election, notable death, and session of Congress during the 2010s, and so forth. Here's a sampling of the flood of edits that are flooding my watchlist (and likely others). Is this really necessary? To begin with, it's unclear whether a WikiProject should have a scope this large, or if this is attainable (if so, a bot should be assigned to handle mass-tagging). The WikiProject page explicitly states, The WikiProject is and will be involved in thousands of articles. There's no WikiProject equivalent for other decades or centuries, so we're in uncharted waters. InfiniteNexus (talk) 01:01, 1 September 2024 (UTC)[reply]
Apologies if my edits have been overbearing or troublesome, I'll make sure to stop with any further edits. To explain my reasoning, I found the concept of a WikiProject with goals like these to be interesting, so I thought it would help to participate by expanding its scope. I understand that there are countless articles that could qualify as relating to the 2010s, but it honestly made the challenge more enticing. Again I apologize for the flood of edits and clogging up your watchlists, I will make sure to end the expansion and mass edits and I wouldn't be opposed if the WikiProject ends up being removed. DukeOfDelTaco (talk) 01:31, 1 September 2024 (UTC)[reply]
To be clear, I am not targeting you specifically; you aren't the first and only person to mass-add these tags for WikiProject 2010s, but you are the most recent person to do so. InfiniteNexus (talk) 01:55, 1 September 2024 (UTC)[reply]
No worries. I think it's laudable to try to resurrect a WikiProject, but tagging random articles probably isn't the best way to do it. The best way to get the project up and running again would be to find some other editors (maybe @TrademarkedTWOrantula, who revamped the 2010s page a few months back) who are willing to help narrow the scope of the project and focus on crafting some initial tasks. voorts (talk/contributions) 02:03, 1 September 2024 (UTC)[reply]
Well, then ... if the project is inactive, then the tags should definitely not be mass-added to thousands of articles in a bot-like fashion. Even if the project weren't inactive, I still don't think a WikiProject should have a scope this broad. InfiniteNexus (talk) 19:08, 1 September 2024 (UTC)[reply]
I think given that it's inactive, and there's no prospect that participants in the project will come to a consensus regarding narrowing its scope, there would need to be some sort of higher level of consensus to deactivate it and avoid having people tag random articles with this WikiProject going forward. On the other hand, as long as mass tagging isn't occurring, I don't really see the harm with tagging this WikiProject inactive and moving on. voorts (talk/contributions) 19:11, 1 September 2024 (UTC)[reply]
Works, from films to books, usually have a "plot" section, detailing its plot. However, I believe that they should be used only if there are reliable, secondary sources that summarize it. Without them, such sections become original research. So, we should make a policy where plot summary sections are only allowed in an article in the presence of reliable sources. 2804:14D:72B3:98F5:0:0:0:1F51 (talk) 19:55, 2 September 2024 (UTC)[reply]
I do not support this. Almost everything we do is summarizing or selecting from other sources. The work we have an article on is a reliable source for its own content. -- Nat Gertler (talk) 19:58, 2 September 2024 (UTC)[reply]
I'm not particularly fond of the practice, but it was already embedded in WP when I started editing 19 years ago. Not worth tilting at that windmill. Donald Albury21:31, 2 September 2024 (UTC)[reply]
Formatting of election pages - external links and pseudoheads
After noticing US election pages having potential issues with WP:LINKFARM/WP:NOTDIR and MOS:PSEUDOHEAD, I crafted some personal WP:AutoWikiBrowser find/replace rules to replace the semicolon pseudoheads with apostrophe-bolded versions and to switch the polling external links to wikilinks with references. I thought my edits were perhaps a bit WP:BOLD, but consistent with policies, provided readers with internal links to learn about the pollsters, and preempted expansion of columns when {{webarchive}} was placed directly into the tables to address dead links. A small subset of recent edits were reverted (e.g., [1]) and my attempt to discuss the issue with the reverting editor was met with my message being blanked from their talk page. I have found no guideline or policy to permit the types of pseuodoheads being used nor to include extensive tables of external links of polls, and this formatting may be more pervasive than I first realized. I had seen some of the election pages tagged with {{External links}} and I have since found that another editor brought up similar concerns about external links on WT:WikiProject Elections and Referendums. The RfC appears to have some support that extensive external linking is contrary to policy, but the discussion had few participants and several suggested to discuss at Village Pump, hence why I'm first posting here. I agree with User:Mikeblas's assessment in the RfC that "No explicit consensus has been discovered here, so the status quo is apparently just replicated behaviour counter to the site's policies." Is there consensus or guidance to allow such formatting of election (or other) pages? Is this the appropriate place to discuss this? Thanks, Ost (talk) 16:40, 4 September 2024 (UTC)[reply]
I think there are some issues with that user regarding Ownership of content, WP:COMMUNICATE, and civility. It seems like you're experiencing some of the same issues. There's no telling why your edits were reverted. I think they enhanced the article by replacing raw links to references (particularly to PDF files) and corrected formatting issues. Given the policy-breaking behavior issues, maybe WP:ANI is the next step.
Issues with link farms in election articles remain. As the article age, the links are rotting often becoming usurped to nefarious purposes. It's not hard to find articles where half of the campaign sites are dead links. I don't think the RfC was ever closed; it was just moved to an archive in the talk page of the project. (It is my first and only RfC, so maybe I was meant to do something differently.) -- mikeblas (talk) 18:47, 4 September 2024 (UTC)[reply]
WP:ELCITE says not to add extraneous information (e.g., archive dates).
As for how I know, I helped write WP:ELLIST. We talked specifically about the benefits of replacing campaign websites with archived copies. Mass replacement of US election websites, in articles of the "2024 [office] election in [place]" type, on the day after the election, was something we thought would be great. It shouldn't be anything fancy, just a simple link showing the campaign website as it was at the end of the campaign election. (@GreenC, perhaps you'd be interested?) WhatamIdoing (talk) 23:15, 4 September 2024 (UTC)[reply]
@WhatamIdoing: Bots like WP:IABOT are typically not updating links to the simple archive link; they generally use {{webarchive}} or similar formatting like your second link. I think it would require a change in bot behavior or manual cleanup to change them to the formatting that you indicate is preferred. It would probably also be good to list your examples in a policy or guideline. —Ost (talk) 22:11, 4 September 2024 (UTC)[reply]
@WhatamIdoing: Thank you for the reminder of that Noticeboard; I was following the suggestions at the RfC and did not consider that location to start the discussion, especially as there was also MOS concerns. However, this discussion can be moved there or split if it is more appropriate. —Ost (talk) 22:03, 4 September 2024 (UTC)[reply]
Thank you for also pointing out WP:ELLIST, though I don't believe that it answers the question because it states that "This section does not apply if the external link is serving as a citation to a reliable source for a stand-alone list entry that otherwise meets that list's inclusion criteria." That guidance may be valid for a minority of my edits that included links to debates (and I was open to further discussing those links and changing my edits), but the vast majority of the links are to polls which appear to be verifying the information in the table (i.e., serving as a citation). To me, the guidance also isn't clear on when it is permissible to include a list or table external links; it explains how they can be formatted and gives examples of political candidates, software, and websites, but it does not explain what makes a list acceptable to contain a links for entries (although the websites one may be self-evident). I don't think that most lists of movies, for instance, contain lists to the official websites for the movies. —Ost (talk) 22:03, 4 September 2024 (UTC)[reply]
@Ost316, keep reading that section until you get to the last paragraph:
In some cases, these links may serve as both official links and as inline citations to primary sources. In the case of elections or other one-time events, it may be desirable to replace the original links with links to archived copies when the event is over.
Polls should be formatted with <ref>...</ref> tags (or whatever system is used in that article, e.g., {{sfn}}).
What we recommend with ELLIST is that these be combined, e.g., through the addition of a column in the middle of the table adds something like "Official website" or "Campaign website" or even just [2]. Then most of the ==External links== section can be removed. After the election, a link to "ElectAlice2024.com" can be replaced with a simple archive link.
I add that we assumed that, for most articles, only the official websites for the main candidates would be appropriate, but there is no actual rule prohibiting editors from using their judgement to include more, if they really believe that would constitute an improvement for the article. I mention this because every country is a bit different, and we did not want to make a one-size-fits-all rule for such a diverse system. WhatamIdoing (talk) 23:28, 4 September 2024 (UTC)[reply]
Linking to archived copies of campaign websites for long-past elections (as opposed to live campaign websites in current elections) makes them essentially a historical primary document, appropriate for a bibliographic-style entry in a Further Reading section.
If it is the case, as others are implying, that every modern election page should have the archived campaign websites of all candidates linked, then perhaps a specific repeatable template should be used for that, such as within an infobox. But that should have some sense to it, and not just be fit into columns of whatever table is available. SamuelRiv (talk) 07:21, 5 September 2024 (UTC)[reply]
When the page has many WP:ELOFFICIAL links (e.g., defunct campaign websites), then it's more sensible to put them in a list with the candidates' names, instead of duplicating the list. The options are fundamentally like this:
Thank you for all your insight and explanations. I especially appreciate that polls should use the pages' referencing format, but the clarity over other ELLIST recommendations are also very good to know and understand. —Ost (talk) 18:21, 5 September 2024 (UTC)[reply]
A question on changing policies would be more suitable for policy over at the Village Pump, but I see no reason for this to have any effect on us here whatsoever. The case here was very specifically about controlled digital lending. CoffeeCrumbs (talk) 22:23, 5 September 2024 (UTC)[reply]
No. This judgement is completely irrelevant to the Wayback Machine, and not significant for any of the other ways we use the Internet Archive. In practice all that it will mean is that a small number of links in references/external links to books in their library might be broken. In the case of external links we'll just remove them if there is no other copy available. In the case of references then it doesn't affect the content at all, it just becomes slightly harder (but not impossible) to verify - we cite offline references all the time, and they are even explicitly allowed by policy. Thryduulf (talk) 15:14, 6 September 2024 (UTC)[reply]
+1 This is a legal determination for the WMF Office to make. This is not legal advice, but I agree with Thryduulf's reading of the case. As an aside, I'm not hopeful SCOTUS will take this case up. voorts (talk/contributions) 15:43, 6 September 2024 (UTC)[reply]
This has some slight relevence, at least to the degree that we cannot trust IA's own evaluation of the legality of their efforts (although that has been true for a while.)
One thing we should consider doing is altering the reference templates so that the Archived link only appears if the url-status is set to "dead". It seems to my not-a-lawyerly eyes to strain fair use claims when we are giving people that link to serve as a way around a paywall for commercially available copyrighted material.
Wikipedia for too long has treated IA as a sole good, without concern to copyright statuses, nor to the fact that they have been paying editors to skew Wikipedia toward them, in part to encourage used book sales. -- Nat Gertler (talk) 15:57, 6 September 2024 (UTC)[reply]
This has some slight relevence, at least to the degree that we cannot trust IA's own evaluation of the legality of their efforts ... The issue in this case was one of first impression and IA had a lot of amici briefs filed in support of their legal position, including the Wikimedia Foundation.
In any event, this case is solely related to IA's digital book lending program, and the court will likely order IA to remove the offending books. Wayback has been around for a long time and as far as I know none of the big media orgs have ever sued over it or indicated that they will sue over it. It's premature to start deprecating Wayback links or limiting their use. voorts (talk/contributions) 16:14, 6 September 2024 (UTC)[reply]
So should we rewrite WP:COPYVIOEL to say it's a-okay to be linking to sites that violate the copyrights of others, so long as they have not yet been specifically sued for the copyright violation we are linking to? IA is a confirmed piracy site, it just seems to be the one we pretend is okay because they put "Archive" in their name. -- Nat Gertler (talk) 17:44, 6 September 2024 (UTC)[reply]
Confirmed piracy according to whom? The US copyvio suit was resolved with a negotiated judgement, and there's an injunction on them lending these classes of books. Violating the injunction results in strictly defined penalties. If believe they are violating the injunction, you may wish to contact the parties' counsel to confirm -- I'd guess that'd be a faster and more reliable confirmation of copyvio than WMF counsel, since the original parties are motivated financially to enforce each potential violation.
With the injunction, we can be reasonably sure that book-lending that was considered to violate copyright in this judgement has all been removed. Furthermore, linking to the book does not violate copyright nor can it reasonably be considered to facilitate violating copyright (certainly now). The IA's action that was violating copyright in the NEL was controlled digital lending, which usually requires some kind of login to the site, and regardless can be (and is) deactivated site-wide. The NEL is gone, and our existing NEL book links should be fine as long as they continue go to some bibliographic info and are not otherwise broken.
Furthermore, even if IA admitted that its NEL was a "confirmed piracy site" or piracy operation (which they didn't, and keep in mind IA has not been charged criminally unlike something like megaupload), it does not make sources outside the NEL part of that operation (the scope of the lawsuit is only NEL book lending). SamuelRiv (talk) 18:06, 6 September 2024 (UTC)[reply]
According to both the court decision that their use was not fair use, and the appeals court confirmation.
With the injunction, I can confirm for you that they are still offering unauthorized electronic editions to which I own the copyright, so no, we cannot reasonably assume that the problem has been addressed.
The fact that IA hasn't admitted themselves to what they did does not obviate the rulings, nor should our standard be "avoid linking only to sites that have announced themselves to be pirates, yar har har." The idea that we should assume that even on a site that has already had identified piracy, we should assume that all their other files are okay is ludicrous. -- Nat Gertler (talk) 18:53, 6 September 2024 (UTC)[reply]
If your copyrighted materials are not now or were being offered in context of the NEL or controlled digital lending, then the case this past month has no relevance to you. If you have a direct copyright violation complaint, US law allows you to request removal from host websites, and you can see IA's procedure for it. Every major content-hosting site has DMCA takedown procedures, and every such site has had copyright-violating material be struck, legitimately, after such procedures were followed per applicable law. Such procedures, followed with such violating content, do not make such sites (like Youtube, Twitter, etc.) a "confirmed piracy site" in any meaningful manner. SamuelRiv (talk) 19:27, 6 September 2024 (UTC)[reply]
Yes, I know that the pirate site has a set of hoops they've put up in order to request that they stop one little piece of their piracy, and I've also seen them ignore that before. -- Nat Gertler (talk) 20:56, 6 September 2024 (UTC)[reply]
If they ignored your request to remove your copyrighted content, then speak to a lawyer, because you can cash in. But if this thread is no longer about NEL content then I don't think there's anything to discuss. SamuelRiv (talk) 23:45, 6 September 2024 (UTC)[reply]
But the copyright laws do themselves cover my works. If we want to remove all restriction from directing people to pirate sites, we should say so up front. -- Nat Gertler (talk) 20:53, 6 September 2024 (UTC)[reply]
Nobody is proposing removing all restrictions regarding pirate sites, but nobody, including you, has provided any evidence that IA is a pirate site. Thryduulf (talk) 21:00, 6 September 2024 (UTC)[reply]
If large groups of rightsholders suing them, and multiple court rulings against IA is not sufficient for you, here is them offering an electronic edition of a book that I hold the compilation copyright as well as publishing rights on, my name is right there on the cover, and I am telling you that I have not licensed them nor anyone else to create an electronic edition of that book. I've seen a lot of links deleted for WP:COPYVIOEL before, and I have trouble thinking of any for which the evidence was greater than court rulings and copyright holders objections. -- Nat Gertler (talk) 21:11, 6 September 2024 (UTC)[reply]
If you're a publisher you should be aware of, or have counsel to advise you on, the relevant laws in your jurisdiction on libraries, inter-library loans of ebooks, and something if you are a member of a class settlement (or else whether you are covered by the AAP). Wikipedia talk pages are not the place to seek legal relief. (Comic relief, on the other hand, ...) SamuelRiv (talk) 00:01, 7 September 2024 (UTC)[reply]
Wikipedia talk pages are, however, the place to discuss Wikipedia applying our policies and guidelines, which do include WP:COPYVIOEL, much as everyone seems to want to whistle and look the other way. Attempting to silence me seems rather inappropriate. Folks are erecting exceptions for IA that I do not recall them every erecting in other copyvioel concerns. But thank you for explaining to me how I should run the business I've been running for a quarter century, or how easy it is to take on legally a business that's hundreds of times one's size. -- Nat Gertler (talk) 18:19, 7 September 2024 (UTC)[reply]
I have not licensed them nor anyone else to create an electronic edition of that book. Institutions like Google and Internet Archive don't need a license to digitize the book so long as they are only displaying short passages and search results, it is considered Fair Use. This was established and reestablished in the Google and Publishers cases. -- GreenC21:41, 7 September 2024 (UTC)[reply]
According to them, they are not just displaying short passages and search results. They are telling people that they can log in and borrow the book for an hour, which would be their display of their complete electronic edition through a web browser. And I cannot point to any other WP:COPYVIOEL concern where we required a completed lawsuit to notice that copyrighted material is being offered without license, can you? -- Nat Gertler (talk) 06:35, 8 September 2024 (UTC)[reply]
If IA is making your book available contrary to the settlement then you contact their office as I linked above. If you are concerned that WP is facilitating copyright infringement on your material, then you contact admins per WP DMCA policy. The lawsuit on CDL is complete -- DMCA requests (or whatever your lawyer advises) are how you get it enforced. (And I imagine if you were to file a lawsuit, the very first question that will be asked is, "Did you send them a DMCA request?".) But please ask your lawyer first. (And when on the internet someone says "If you have personal legal concerns you should speak to your lawyer and our legal department", that's not them silencing you, that's just common sense.) SamuelRiv (talk) 16:34, 8 September 2024 (UTC)[reply]
You are right, it appears to have CDL. Then we need to look at the Publisher's settlement. According to the official statement:
The lawsuit only concerns our book lending program ie. CDL. The injunction clarifies that the Publisher Plaintiffs will notify us of their commercially available books, and the Internet Archive will expeditiously remove them from lending. Additionally, Judge Koeltl also signed an order in favor of the Internet Archive, agreeing with our request that the injunction should only cover books available in electronic format, and not the publishers’ full catalog of books in print. Separately, we have come to agreement with the Association of American Publishers (AAP) .. if we follow the same takedown procedures for any AAP-member publisher.
This book is not available for sale in electronic form, it's not in the Plaintiffs catalog, and I guess(?) you are not a member of AAP. So the settlement and injunction is probably not applicable to this book. Where does that leave it? I have no idea so I reached out at IA to see if they have further information. -- GreenC03:48, 9 September 2024 (UTC)[reply]
@NatGertler:The book now says "Borrow Unavailable". It has been removed from CDL. They said it is unclear if there were any previous requests to remove CDL. Nevertheless, if you need further help with other books, ask them, and if you don't get a response, you can ask me. -- GreenC19:24, 9 September 2024 (UTC)[reply]
No. (Put aside that the proposition as worded fundamentally breaks WP itself.) The negotiated settlement has been in force for nearly a month now, with an injunction on the books at issue (which is a small subset of the total books that IA has available.) Furthermore, in my editing here, I have seen a full book-edition-in-copyright linked on IA library as a reference maybe... a couple times at most? The vast majority is non-lending books out of copyright and/or out of print. That is reinforced by how citations work in WP, and how our citation template works, passively encouraging more openly accessible means of verification (e.g. searchable passages of books as opposed to limited-time lending of entire books). SamuelRiv (talk) 16:33, 6 September 2024 (UTC)[reply]
GreenC is a paid agent of the Internet Archive, something which he should probably raise whenever he journeys into conversations like this. (It can be verified down at the bottom of his user page.) -- Nat Gertler (talk) 18:56, 6 September 2024 (UTC)[reply]
The factuality of his statement (to whatever degree it has; it's not something they're backing up with a reliable third-party source) does not obviate that they are a paid agent of the organization and should identify themself as such in such a discussion. As you have a conflict of interest, you must ensure everyone with whom you interact is aware of your paid status, in all discussions on Wikipedia pages within any namespace. (WP:UPE) -- Nat Gertler (talk) 21:01, 6 September 2024 (UTC)[reply]
I've thought about it but my relation to the org is so far removed it gives the wrong impression to say "I work for IA", I'm certainly not an Agent, or posting on their behalf, I'm not technically an employee, or even been to their offices. My primary role here is as a 21-year Wikipedian posting for the benefit of other Wikipedia editors. Maybe in these discussions I could say something like "disclosed COI but not posting on their behalf", and leave it there. Also, feel free to reach out if you want help navigating the organization, possibly I could help, but not sure your situation. -- GreenC17:49, 7 September 2024 (UTC)[reply]
I think that would be appropriate; you are certainly welcome in these discussions, but such a disclosure would serve both those who don't know you have a WP:COI (and editing in discussions related to a client does indeed fit in that, even if it is not specifically what they are a client for) as well as those who do know that you have that link and may assume you are making statements at their behest. (And I'm saying that someone who tries to remember to note my own COIs even when all I'm doing is a very simple MOS fix.) -- Nat Gertler (talk) 18:08, 7 September 2024 (UTC)[reply]
You seem to have a much greater COI with regards to the IA than GreenC does, yet you are giving the impression that your contributions to this discussion are more objective than GreenC's are. Thryduulf (talk) 18:22, 7 September 2024 (UTC)[reply]
I'm sorry, I must've missed where having ones material copied by a site gives one a WP:COI; if that's the case, then basically everyone who has ever put anything on the web has that same conflict with IA. And can you point to anywhere where I state that GreenC has said anything not objective? The brief question of "factuality" is questioning at most the accuracy of a statement. Had he said "the sun is three feet across", that would be an objective statement, but not factual. If you have some concerns about impressions, you may want to deal with the impression you are creating trying to shout me down rather than addressing actual concerns. -- Nat Gertler (talk) 06:30, 8 September 2024 (UTC)[reply]
I was in no active dispute before posting on this thread that I was someone whose books they had pirated. Indeed, it was pulling up the examples to answer those in the thread who wanted to deny what was going on that drove me to send a letter requesting that they stop hosting their unauthorized editions. -- Nat Gertler (talk) 16:22, 8 September 2024 (UTC)[reply]
If the office (see WP:OFFICE) decided to ban linking to a site (regardless of which site and why) they would communicate that to us officially by posting on Meta (as I can't imagine such a ban applying to only one project) and cross-post it to various places on large projects. Editors would then disseminate that message elsewhere - you can guarantee that someone (whether the WMF or an editor) would post it to WP:AN for example. If the site is notable then it wont take long before the article is updated to reflect the ban and you'll see the decision discussed in countless internet forums and probably news publications too. Thryduulf (talk) 00:14, 7 September 2024 (UTC)[reply]
Agree with Thrydulf's assessment here. Wayback (normal archiving of URLs) isn't touched by this, and this only makes book copies on Internet Archive in question, which I've always felt was going to be a problem. Masem (t) 00:10, 7 September 2024 (UTC)[reply]
Per WP:NSPORTS2022 Proposal 5 and the subsequent update to WP:SPORTSCRIT, there is consensus that sports biographies are not subject to WP:NEXIST and are automatically subject to deletion if they do not contain at least one source demonstrating significant coverage in the article. To this point, cleanup has been slow as a courtesy to AfD. I propose CSD X4 so the articles can be addressed without causing a significant backlog. X4 would only apply if there is unambiguously no source to demonstrate significant coverage in the article, meaning borderline sigcov would still need to go through AfD or PROD. Articles sourced entirely to databases or passing mentions would qualify for X4 deletion. X4 would be an alternative to mass AfD nominations, which will be the likely outcome if no action is taken. Thebiguglyalien (talk) 21:24, 6 September 2024 (UTC)[reply]
I don't read the close of NSPORTS2022 as saying that such articles must be deleted, but rather that for an article to be created it must have at least one source with SIGCOV. voorts (talk/contributions) 21:32, 6 September 2024 (UTC)[reply]
Thank you! I considered this the most "central" location and went to notify other places, but I got distracted after adding it to COIN. I have no objections if anyone thinks this should be moved to WT:CSD. Thebiguglyalien (talk) 21:51, 6 September 2024 (UTC)[reply]
(edit conflict) Firstly new CSD criteria should be proposed at Wikipedia talk:Criteria for speedy deletion (I'll notify them for you) and should demonstrate that they meet all four requirements listed at WP:NEWCSD. My first impression is that this would fail at least point 1 (objective) because what coverage counts as "significant" is subjective and whether coverage meets that standard is frequently a matter of disagreement at AfD. AIUI there is also frequent disagreement about what counts as a database and whether database entries are always non-trivial coverage. Thryduulf (talk) 21:52, 6 September 2024 (UTC)[reply]
Speedy deletions are for situations where the deletion criterion is clear and the deletion of articles meeting that criterion is uncontroversial. Past discussions of sports stubs have made it obvious that both of these conditions are not met here. —David Eppstein (talk) 21:57, 6 September 2024 (UTC)[reply]
I regularly see athlete articles prodded, in the list at WP:PRODSUM, often in large batches. I haven't been keeping track of how successful the prods are but I think it's a good thing to try first before resorting to an AfD. —David Eppstein (talk) 22:08, 6 September 2024 (UTC)[reply]
All right, thank you. I do feel a little silly now that the points above have been raised. I'm just working through the NPP backlog right now and trying to find some sort of way to make it more efficient when sorting the good from the bad. Thebiguglyalien (talk) 22:10, 6 September 2024 (UTC)[reply]
If your goal is efficiency or a minimum of drama, I think that you should not try to delete any "borderline" cases. Using WP:PROD on egregious examples, with a good explanation and a link to the requirements, isn't likely to earn you many wiki-enemies. However, trying to remove (by any means: AFD, PROD, Draft:, etc.) any subject for which SIGCOV is perhaps a matter of judgement is not likely to win friends and influence people. In particular, I suggest turning a blind eye towards any subject that someone could claim as SIGCOV under WP:100WORDS. WhatamIdoing (talk) 00:24, 7 September 2024 (UTC)[reply]
There are 3 things I want to see in any X series new criterion proposals. Firstly, I want to see a list of the pages subject to the new x criterion. This is typically a page compiled in user or project space somewhere. This both defines the scope (how many pages are we talking about), and lets people take random dip samples. The second thing I'm looking for is a consensus that many or most, but not all or nearly all, of the pages subject to the criterion should be deleted if the appropriate consensus based process was run in full. If all or almost all of the articles should be deleted, make a single mass nomination at the appropriate XFD, or else have a RFC linked from that XfD. The third is a limit date, or some other way to demonstrate that the flow of new articles into the X series eligible pool has stopped. This is to ensure that an x series criterion doesn't grow in scope. No opinion on the merits of these articles, but I do need this info before I can weigh if X4 is an appropriate response. Tazerdadog (talk) 01:42, 10 September 2024 (UTC)[reply]
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
WP:INVOLVED policy regulates permissible conduct of Admins and editors performing non-admin closures. This policy was created before the existence of WP:Contentious topics. Inside the first sentence In general, editors should not act as administrators in disputes in which they have been involved. (bold emphasis mine) the term dispute is not well defined, despite the fact that some WP:Contentious topics are exceptionally well defined, e.g WP:ARBPIA while others like WP:BLP are not narrow in scope. Inspired by the larger discussion at Wikipedia:Administrators'_noticeboard#Possible_involvement_of_Admin_in_ARBPIA_area
My proposed re-wording (added text in green) of above sentence for minimal change would be:
In general, editors should not act as administrators in disputes in which they have been involved. When the editing area is covered under existing WP:Contentious topics, the definition of dispute scope may be broadened to include a larger subset of said contentious topic area to be defined by the Arbitration Committee.
This would allow us to avoid changing wording of WP:INVOLVED to litigate each individual contentious topic, while retaining flexibility. For example ARBCOM could clarify that ARBPIA could be envisioned as 'one dispute area', while WP:ARBAP2 might involve several (Trump, US voting rights etc..) and something like WP:BLP might not define any for purpose of WP:INVOLVED. ~ 🦝 Shushugah (he/him • talk) 10:53, 7 September 2024 (UTC)[reply]
I believe it is already covered by the "construed broadly" and "may be seen to be involved" portions of INVOLVED:
Involvement is construed broadly by the community to include current or past conflicts with an editor (or editors), and disputes on topics, regardless of the nature, age, or outcome of the dispute ... it is still the best practice in cases where an administrator may be seen to be involved to pass the matter to another administrator via the relevant noticeboards.
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.
Can we discourage indiscriminate automatic addition of Wayback archive links on all external links?
There are a bunch of Wikipedian meat-bots running around with an automated tool which indiscriminately adds Wayback Machine links to every external link in every reference on one Wikipedia article after another, either as an archive-url parameter for CS1 or CS2 citation templates, or using the {{Webarchive}} template for ordinary wiki-markup links.
While the Wayback Machine is a wonderful and invaluable resource that I use multiple times a week, and I am all in favor of adding archives for dead links (such as User:InternetArchiveBot currently does), adding a wayback link for every external link on Wikipedia, including all of the still live and publicly accessible URLs, seems to me extremely spammy and moderately reader-hostile. It wastes a lot of space, is confusing to readers, wastes time/attention and causes confusion for editors trying to verify claims in articles, and seems mostly like a way of turning Wikipedia into a massive advertisement for the Internet Archive. I personally think the Internet Archive is wonderful and well worth supporting with money and use, but this method feels really icky and unjustified to me, not that different in spirit from Wikipedia adding banner ads for third-party charities or individual Wikipedians indiscriminately spamming links to their own published papers. It feels abusive of Wikipedia as a platform and of the Wikipedia community, and as far as I know there was never any consensus for it.
Can we formalize some kind of policy discouraging indiscriminate automatic addition of archive links? If human editors want to deliberately add individual specific wayback backups for still-living pages for some good reason (e.g. a website changed their content, substantially changed their formatting, or put up a paywall, in a way that makes the original content difficult to use for verifying the claims in the article) I have no problem with that. And again, as I said above I am all in favor of adding wayback backups (including automatically) to dead links.
One thing that I suggested above is that rather than discourage the creation of the archive link (which is quite understandable at least when the person originally constructing the reference wants to make it bulletproof), we have the cite templates only display the archive if the link status is "dead" or some similar status that indicates that the original link is no longer appropriate. (I also question whether we need the statement archived at Internet Archive displayed on references even when the original URL is dead; anyone following the archive will see the site it's at, it doesn't really inform the reader about the quality of the source material, and as you say, it does come across as advertising.) -- Nat Gertler (talk) 06:14, 8 September 2024 (UTC)[reply]
This doesn't seem like a good solution to me. There are perfectly legitimate reasons for editors to manually and intentionally add archive backups for specific still-living pages, e.g. if the site is flaky and routinely goes offline, is notorious for regularly changing URLs, or has substantially changed the content of a specific linked page. However, for other links, a "just in case" backup in the article source serves no purpose. The action that is useful, for still living pages, is to explicitly ask the IA to crawl and archive the content of the page (e.g. by requesting that it crawl all pages linked from a particular Wikipedia article), so that there will be a backup in case the link ever rots. But that can be done from the IA site or some external script and shouldn't involve changing the source markup of a Wikipedia article. –jacobolus(t)06:30, 8 September 2024 (UTC)[reply]
I am aware that this is done, but can you link an example bot or script that is doing this? (As you note, IAbot does not.) The code people are using may just be hastily written without checks. I am opposed to adding such archive links indiscriminately to every citation:
Archive links added after the fact, when not for rescue/CPR purposes, should be dated as close as possible to the |access-date= parameter of the citation, and not to the date of the crawl (which may or may not be the case for a given userscript.) If the new |archive-date= parameter must be different, the cited text must be checked manually.
If no |access-date= parameter is given, the cited text must be checked manually. (In practice it may not be justified to date it to the day the original edit was made, and I doubt it's computationally feasible to back-trace a section of wikitext to its last edit in a bot.)
Some very few citation links might be explicitly meant to reference the live version of a website, such as a webapp or "current status of..." footnote.
However, if the |access-date= parameter is given in the citation, I don't see an issue with adding an archive link with |url-status=live as applicable. (Although we might consider querying more archiving sites than just IA when doing this.) SamuelRiv (talk) 16:51, 8 September 2024 (UTC)[reply]
Here's an example edit, special:diff/1243928783, edit summary "Rescuing 13 sources and tagging 0 as dead.) #IABot (v2.0.9.5 Tag: IABotManagementConsole [1.3]". (This one is a particularly annoying common variant that adds wayback links for stuff like Google Books pages, bibcode links, etc.) I'm not trying to call out specific Wikipedians or publicly shame anyone. There are a bunch of people doing this type of edit using the same automated tool, often to many pages in quick succession. I think these kinds of edits are done in good faith, but I think we should have a policy page discouraging them. –jacobolus(t)17:46, 8 September 2024 (UTC)[reply]
@Jacobolus could you clarify if your objection is to the use per-se of IA, or to the fact that people are doing it indiscriminately in a bot-like manner? I routinely run IABot on my own articles and tell it to add links to anything it archives. I hope you're not objecting to that. RoySmith(talk)17:33, 8 September 2024 (UTC)[reply]
Yes, I am objecting to adding archive links to "anything it archives". This is, in my opinion, a type of spam. Please use archive links (whether from the internet archive or some other site) for links where the original is dead, changed, or unreliable for some reason, not just for fluffing out the citations section with distracting and unnecessary clutter. –jacobolus(t)17:37, 8 September 2024 (UTC)[reply]
This isn't purely rhetorical, but one could plainly search for URLs of IAing Google Books and quite quickly get a sense of how common this indiscriminate archiving is and what it looks like. Remsense ‥ 论20:18, 8 September 2024 (UTC)[reply]
OK, I guess I'll have to respectfully disagree. The problem is that by the time a link goes dead, it's too late to archive it, and then it's lost. I wouldn't mind if the citation templates treated this in a less "fluffy" way, but I'd hate to see the archive information omitted completely. RoySmith(talk)17:45, 8 September 2024 (UTC)[reply]
Here's an example from RoySmith: special:diff/1232611207 to Big Duck. Most (all?) of the archive links being added here go to still-living pages which are publicly accessible on the open internet, so the archive links are nothing more than clutter. "by the time a link goes dead, it's too late to archive it" – People should directly ask the IA to crawl and backup pages linked from Wikipedia. It is completely unnecessary to change the markup source of Wikipedia articles to accomplish this task. But to be honest, I don't mind that strongly if one of the primary authors of an article wants to go ham on the wayback links. Discussion of that choice could be contained to individual article talk pages and hashed out locally. But there are also folks doing large numbers of these edits to one page after another in a bot-like fashion, and at scale it's impossible for someone who disagrees to meaningfully revert or contest those changes without spending an even greater amount of time. I don't think there's any established consensus that such archive links should be added site-wide, so adding them automatically at scale is a kind of fait accompli. –jacobolus(t)17:52, 8 September 2024 (UTC)[reply]
Adding the IA link into the citation template makes it much easier to find later should it be necessary. If it bothers you to see that, it's easy enough to add some custom CSS to hide those sections of the citation. I'd be happy to have the templates add some classes to the markup to make that easier to implement. RoySmith(talk)18:32, 8 September 2024 (UTC)[reply]
It bothers me that we are spamming editors and readers in a way that seems abusive of Wikipedia's platform and community. Hiding what I personally see does not at all address my concerns. "much easier to find later should it be necessary" – this seems like nonsense to me. We are talking about the result of a fully automated tool without human inspection or intervention beyond clicking a submit button; the IABot is currently automatically "rescuing" dead links across the site without any human involvement at all, and will presumably continue to do so. Speaking as a human who constantly uses the Wayback Machine to find archived pages, it is trivial to go to web.archive.org, type a URL in the box, and then click through to see the various archived versions of that URL. –jacobolus(t)18:47, 8 September 2024 (UTC)[reply]
@Jacobolus: I just want to express the exact opposite position to your proposal.
Clicking on a link vs. going to the Wayback page, copy/pasting the url, possibly comparing alternative archive copies ... yet you say you're concerned for everybody else rather than just yourself? That makes no sense to me. Fabrickator (talk) 19:13, 8 September 2024 (UTC)[reply]
@Fabrickator You are not understanding my point. Let me try to be clearer. The claim is that we need to defensively add archive links to every external link because otherwise it is not "easy" enough to find the archive link in the event that the link rots and it becomes "necessary". My responses are: (1) it doesn't get any harder to manually trigger an archive bot/script years down line in the possible case that some links rot, (2) there's a continuously running automated process scanning the web for dead links and "rescuing" Wikipedia citations by adding wayback archive links, requiring no human intervention at all in the general case, and (3) even if this bot somehow doesn't catch a very occasional stray dead link and a human editor tries to verify a claim but ends up at a 404 page or whatever, it is quite trivial for a human editor to use the wayback machine from their browser (it takes no more than a few minutes to manually find the wayback link and add it to an article, and we are talking about a tiny number of links at that point). People running this script certainly aren't comparing various archived versions of every link. I have no problem at all if editors carefully manually inspect and think about how best to present every citation link to readers, sometimes making the specific individual decision to add an archive link. –jacobolus(t)20:34, 8 September 2024 (UTC)[reply]
@Jacobolus, if I understand you correctly, you're saying that the current system, which is:
The bot does everything automatically. Bonus: Since it's running soon after you added the source, it will (almost) always link to the correct archived page.
is inferior, because the alternative:
Wait until the link breaks
Trigger the bot manually.
Check the results.
Oops, it linked to an archived copy of the broken page.
Go to archive.org.
Paste the dead URL into the search box.
Manually check through various versions of the page to figure out which one ought to be given to readers.
Copy that URL.
Paste the archive URL into the Wikipedia article.
Figure out what the citation template requires for the date stamp (because it produces a red error message if you don't get it right).
is "very easy" and "quite trivial", right? Speaking just for myself, I don't feel like a 10-step process is "easy" or "trivial", especially compared to the option of "The bot does everything automatically". WhatamIdoing (talk) 05:32, 9 September 2024 (UTC)[reply]
No, this is a dramatic mischaracterization both of what I am saying and also of "the current system". I'm not sure what's going wrong with communication here, but I'll try briefly to clear things up before I have to go. The working parts of the "current system" is (1) the Internet Archive (IA) automatically checks nearly all external links added to Wikipedia and crawls them, adding snapshots of the pages they point at to the Wayback machine (henceforth WB), (2) anyone in the world can explicitly request that any arbitrary webpage and all of its outgoing links should be crawled and a new snapshot of each added to WB, (3) the IA runs some kind of automatic process checking the outbound links from Wikipedia and flagging them when they are no longer working, (4) User:InternetArchiveBot ("IABot") "rescues" dead links which appear on Wiki pages by adding WB snapshot links to Wikipedia article citations, without much need for human intervention (though sometimes the bot screws up somehow, or the archive contains the wrong content, or the website changed their URL scheme and there's still a living page at a new URL, etc., in which case a human can fix it). All of the above generally works great. It prevents a lot of link rot across Wikipedia, and I have no problem with it.
Additionally, there's a part of "the current system" that I find to be spammy, confusing, wasteful of space and attention in both output and source markup, unnecessary churn in page histories, without any obvious practical value, and not supported by current community consensus. Namely: (5) human "meatbots" run a semi-automated script, "IABotManagementConsole", which indiscriminately adds a WB snapshot link to every citation with an external link on it on a particular article, nearly all of which are working links to still-living pages accessible to anyone on the open internet. These editors are not checking anything manually or making careful choices, but just clicking one or two buttons, and their behavior is not fundamentally different from bots. Some of them are doing this repeatedly over and over again to dozens or hundreds of different pages which they have no other interaction with.
These edits are unnecessary, because in the event a link rots, the other parts of "the current system" already take care of it. Only a trivially tiny proportion of links rot and get clicked on by human editors trying to verify a claim before they have been "rescued", and in these cases, it takes about 1–2 minutes for the human editor to manually do the rescuing, or a few clicks to get a bot to do it. There's no "10-step process" involved. –jacobolus(t)05:57, 9 September 2024 (UTC)[reply]
Okay, the part I dispute is your claim that "in the event a link rots, the other parts of "the current system" already take care of it". I find that the post-rot "rescues" are worse quality (e.g., undetected 404 pages and redirects to the site's main page) than the archived links added pre-rot.
Why do you think that 1–2 minutes manual work per link – and realistically, we have to assume that they will all die some day, so for an article with ~15 sources, that's half an hour of manual time – better than clicking a button once and being done forever with all the (current) refs? WhatamIdoing (talk) 06:09, 9 September 2024 (UTC)[reply]
In my experience the "pre-rot archive" links also have a substantial proportion of mistakes, and would also routinely require manual attention. (Except we can skip it by just deleting those snapshot links as reader-hostile spam.) My own expectation is that the amount of manual work required for these is going to be nearly identical either way, except demanding backup links everywhere front-loads that manual effort and forces a human to double-check a wayback link for every link on the site instead of reserving manual checks for cases where it makes a difference. I imagine most of whatever difference you are seeing is selection bias: working links are relatively more likely to come from better managed websites which don't break their URLs as often and therefore tend to have working archived snapshot links as well. But if those pages do break at some point, they will still have working snapshot links. By comparison, broken links are relatively more likely to come from neglected or mismanaged websites. But if you find "post-rot rescue" links to often show the wrong thing, then that seems like a problem for bot authors to work on: for example, the bot could try to check for variants of the page and avoid 404 pages or versions where there is limited visible content, or could try to pick a snapshot date close to the access date described in a citation template, when presumably a human was able to view the content. –jacobolus(t)06:22, 9 September 2024 (UTC)[reply]
In the source of pages, this cavalier archiving behavior manifests as either apparent negligence or obsessive compulsion far too often in certain situations. "Indiscriminate" is the key word here. Remsense ‥ 论20:20, 8 September 2024 (UTC)[reply]
I very much agree with this. I used to be one of those editors that clicked the button and let IAbot run because I thought adding the archive links was an improvement to any article. I no longer think so, especially since I learned that Internet Archive archives all external links on Wikipedia anyway, so having the bot run doesn't actually archive the links, it just adds archive links to the citations, which is usually not helpful, unless the link is dead and IA has the full version (or to get around a paywall, but that's not really a kosher use). The archive links make citations harder to read and work with. It's even worse when the link being archived is useless, e.g. archiving Google Books pages because it's linked in a citation to the book (which doesn't archive the book!). Bottom line: I do not think the automatic addition of web archive links to citations is a good thing. I would support a change to IABot that it only adds archive links if |url-status=dead. I think I'm in the minority on this, though. Levivich (talk) 19:27, 8 September 2024 (UTC)[reply]
If what you say in your first bit of bolding is true (I have no reason to doubt it but don't have time to check now) then I am in the same minority, if it really is a minority. Phil Bridger (talk) 19:40, 8 September 2024 (UTC)[reply]
This being Wikipedia, I should cite my sources :-) IA in 2018: For more than 5 years, the Internet Archive has been archiving nearly every URL referenced in close to 300 wikipedia sites as soon as those links are added or changed at the rate of about 20 million URLs/week.WP:WAYBACK also says New URLs added to Wikipedia articles (but not other pages) are usually automatically archived by a bot. It also says Editors are encouraged to add an archive link as a part of each citation, which I think it should not say. BTW, another very easy solution to this would be to code the citation templates to not display archive links unless |url-status=dead. This could also probably be done as a CSS hack, so editors who didn't want to see those links for live URLs could turn them off as it were. Levivich (talk) 20:07, 8 September 2024 (UTC)[reply]
One of the reasons for linking to wayback is so that we can continue to cite webpages that are no longer available, or have been significantly changed. But this raises red flags for me… I want to know WHY the cited source is no longer available or has changed. This question should always be asked before we link to wayback. The answer may require us to reassess whether the information in our article is out of date, and thus no longer accurate and verifiable. Blueboar (talk) 21:19, 8 September 2024 (UTC)[reply]
In my experience, archive links for still-alive links could be useful, as the site may be temporarily down or region-blocked. I don't have a strong opinion on this, though, because for me personally it's not hard to manually open a link through the Web Archive. Levivich's solution seems good to me, but I also agree with jacobolus that there should be an option in IABot to archive all the links but not add them to the article. AstonishingTunesAdmirer連絡21:10, 8 September 2024 (UTC)[reply]
I just tested it, it's not that, it still adds archive links to the article. As far as I understand, the bot makes a list of all references in the selected article, then saves each link through the Web Archive, and then adds links to these archived pages to the article. What jacobolus was talking about is skipping that final part, so the bot would save all links through the Web Archive, but would not add them to the article. AstonishingTunesAdmirer連絡21:30, 8 September 2024 (UTC)[reply]
@Jacobolus: You wrote: "You are not understanding my point. Let me try to be clearer." I hear your point, but I think it's a bad one. I'm unclear whether you object to the archive link being visible, being present when you edit the article, or being stored in the wikitext. Certainly I don't object to it not showing up when you view an article, and the details of how this is stored is something I should like to see improved (e.g. along the lines of storing the data in Wikidata and accessing it using {{cite q}}).
But even the possibility that the live url is temporarily not working (or permanently not working) is good enough reason to have the archive link handy. It's a bonus that there are other scenarios where it's useful ... e.g. some "multi-page" articles (requiring a different url to get to subsequent pages) may also have a link that displays all pages with a single url. You might suppose this is neither here nor there, but as an editor, I make such a decision as to which is preferred. I am unhappy about the nuisance of all this, but I am happy that we are not stuck with what unintelligent bots choose to do.
I would long for a bot that was smart enough to make good decisions, but I see no evidence that we have the appropriate technology to develop sufficiently intelligent bots, so we have a fallback of manually adding or editing archive links, yet the existing system is useful in that it makes a passable selection at least most of the time, yet somehow you are annoyed about the presence of that selection. I throw up my hands. Fabrickator (talk) 21:51, 8 September 2024 (UTC)[reply]
"I'm unclear whether you object to the archive link being visible, being present when you edit the article, or being stored in the wikitext." – I am strongly opposed to the first, which seems like spam and is confusing to readers, and moderately opposed to the other two, which still add a lot of clutter to the markup for no practical benefit. "may also have a link that displays all pages with a single url" if you want to link the article title to one URL and also add a second URL to a consolidated version of the content, there's nothing stopping you. It would be extremely confusing to use the "archive-url" parameter for this, so please don't do that. "as an editor, I make such a decision" – again, I am mostly unhappy with automated tools imposing poor choices not backed by site consensus, not by human editors making specific decisions about individual links on single articles. The latter is entirely possible to discuss and work out in local talk pages. –jacobolus(t)22:03, 8 September 2024 (UTC)[reply]
I guess my response to this would be "you never know when a publication/company could go down". Just as a recent example, Game Informer went down this year and there was a mad scramble to archive all of their website links and making sure all that was saved because GameStop couldn't be bothered to keep GI's history up. I'd rather be preventative and avoid stuff like that from occurring rather than being reactionary and having to scramble to archive a source if it goes down. ♠JCW555(talk)♠ 22:22, 8 September 2024 (UTC)[reply]
"You never know" simply isn't a compelling enough argument to applied literally 100% of the time in favor of obvious clutter that has clear concrete disadvantages in the meanwhile. What is the aversion editors have against literally any discretion being applied here? Remsense ‥ 论22:52, 8 September 2024 (UTC)[reply]
"obvious clutter that has clear concrete disadvantages in the meanwhile" this is your opinion, which is fine to hold of course, but I (and others) don't feel that way so don't try and speak for everyone. "Discretion" in this case would be "editors should have the same opinion as I do". I'd rather be proactive in archiving sources than be reactive. ♠JCW555(talk)♠ 23:05, 8 September 2024 (UTC)[reply]
My opinion is putting emphasis on what are demonstrable, objective issues caused by the practice. You can have a different emphasis, but let's not pretend whether the issues even exist (which I've just elaborated upon below) are a matter of opinion themselves. Remsense ‥ 论23:08, 8 September 2024 (UTC)[reply]
"what are demonstrable, objective issues caused by the practice" again, according to you. I, and others in this conversation don't agree with you. You can think it's a problem, others can think it's not a problem. That doesn't make either of us right of course, but I don't say my opinion is 100% true and everyone must agree with me. I only speak for me, not every editor on the platform. So again, don't speak for me or everyone on here. ♠JCW555(talk)♠ 23:19, 8 September 2024 (UTC)[reply]
My browser crashing when I try to turn on syntax highlighting isn't a figment of my imagination or a mere aesthetic preference I have. In the general case it is a problem for me and for you for all intents and purposes, so you're not going to force me to mince words as such. Remsense ‥ 论23:20, 8 September 2024 (UTC)[reply]
I've read this whole discussion a couple of times now and I'm still not understanding what is problematic about including the archive links for active links. They are useful for temporary outages, region locking, verification of pages that have changed content, usurped domains (which are hard to detect automatically, which is why the WP:JUDI domains have to be tracked manually for example) and for other uses too. I also don't understand why some people in the discussion find the presentation of both live and archived links confusing - is this an actual problem that readers have noted or is it just a theoretical somebody might potentially be confused? I'm not attached to changing the display, but I am opposed to removing it without evidence that the current presentation is actually problematic and that the proposed alternative is actually an improvement. If anything, I would prefer to encourage the inclusion of archive links with every citation, because they are significantly valuable to maintaining the integrity of the encyclopaedia and allowing continued verification of our articles. I completely disagree that they are "clutter" and nobody has presented any arguments here that convince me that inclusion has any disadvantages at all, indeed quite the opposite. Thryduulf (talk) 22:57, 8 September 2024 (UTC)[reply]
When they are clearly addressing one of your stated issues then of course they are useful. Otherwise, wikitext length itself plus its associated visual and technical complexity simply isn't free or essentially free: the additional bundle of parameters indiscriminately added to every citation adds visual clutter for editors to scroll through, which does make it harder to factor articles and understand their wikitext, and even the rendered article if the archive is not serving any clear function. It is not controversial that it can also lag the editor, especially when syntax highlighting is enabled. I am dead serious when I say there are articles I cannot comfortably edit in their present state, whereas I would be able to if every Google Books url didn't add nearly 1kB to the article size instead of 100B. If we're supposed to have archives on literally every external URL, then that should be added to the backend or something so we don't have to deal with it this way. Let's not treat it technically like a matter of discretion when it's actually not editorially. Remsense ‥ 论23:05, 8 September 2024 (UTC)[reply]
When they are clearly addressing one of your stated issues they always address those stated issues, because it is not possible to predict when a site is down (temporarily or otherwise), whether any editors will be geolocked (e.g. many US news sites are blocked to editors in Europe but it is completely invisible to US readers which they are and which they aren't), etc. Thryduulf (talk) 23:15, 8 September 2024 (UTC)[reply]
But it is actually possible to weigh some of these factors based on context. Plus, why are we a priori able to assume the IA is perfectly reliable in this way, but any other site isn't? Remsense ‥ 论23:19, 8 September 2024 (UTC)[reply]
It is possible for a human to weigh some of those factors in context, which means that as humans don't (and can't) evaluate every link in context in advance of every reader reading the article the links are always addressing the stated issues. We don't assume that IA is perfectly reliable, we know that the IA does not geoblock and is generally reliable, we also know that the combination of a link and an archive (IA isn't the only game in town, despite what this thread implies) is more reliable than a link alone. Thryduulf (talk) 00:03, 9 September 2024 (UTC)[reply]
Right—all I am arguing for is that discretion is possible, and that it is still expected in the use of automatic tools like with any other dimension of editing.
As a total tangent, I've always wondered why it isn't possible to collapse CS1 archiving into two parameters—date and liveness. Surely it's pretty feasible to derive |archive_url= from the |url= and |archive_date= (etc.) already given? Remsense ‥ 论00:11, 9 September 2024 (UTC)[reply]
Ah, that makes sense. Do you think it would be viable for the "default mode" to assume IA with no redirect, and other configurations could be further specified if needed? As strong as my opinion is here, reducing the wikitext footprint of archiving would make it essentially a nonissue! Remsense ‥ 论00:20, 9 September 2024 (UTC)[reply]
It's a good idea, in my opinion. I think combined with an option to disable rendering for archive links on live sites it should solve most of the complaints (wikicode would be less cluttered, there would be fewer archive links on screen). But I think this more of a question to people working on the citation templates, whether it's hard to implement, etc. Pinging @Trappist the monk and Izno: how hard would it be to generate an archive link for a URL, so instead of {{cite web|url=https://en.wikipedia.org|archive-url=https://web.archive.org/web/20240501005015/https://en.wikipedia.org|archive-date=May 1, 2024}} you could do just {{cite web|url=https://en.wikipedia.org|archive-date=May 1, 2024}}, but with an option to provide archive-url if it differs from url (or you need a specific snapshot, I guess)? AstonishingTunesAdmirer連絡00:49, 9 September 2024 (UTC)[reply]
You would need the full timestamp of the archive 20240501005015, so you need a separate parameter regardless |archive.org-id=20240501005015 or whatever. You would not be able to do it solely with a |archive-date=. Izno (talk) 01:02, 9 September 2024 (UTC)[reply]
Sure—I was envisioning a |archive_vendor= parameter (or something terser) in the de facto exceptional case when we're using something else, which would default to IA.Remsense ‥ 论 01:34, 9 September 2024 (UTC) Scratch that, it should've probably occurred to me sooner that this can remain minimally complicated by just assuming IA by default and specifying |archive-url= otherwise. Remsense ‥ 论05:01, 9 September 2024 (UTC)[reply]
I do not think that making changes to cs1|2 in an attempt to curb annoying or bad societal behavior is a good idea. For the narrow question of hiding the archive information in a rendered citation, I can imagine having Module:Citation/CS1 wrap the live form (Archived from the original on <archive date>) in a css class so that individual editors may hide that portion of the rendered citation with something like this:
Another possibility is to get someone at WP:US/R to write a script that hides live archive info but also provides a clickable something so that that individual editors can show a citation's archive info if the want to see it.
@Thryduulf, I talked to the IA folks about this a while ago. The problem we have is that what the bot does violates our guideline at WP:ELCITE. The problem they have is that the bot is too simple-minded to know how to stay out of the ==External links== section. It takes an all-or-nothing approach to adding archive links in an article. (It also has had problems with incorrectly claiming that sites are dead.)
For clarity, in the ==External links== section, either of these is Y good (depending whether the link is live):
WP:ELDEAD agrees with you. We tend to keep archived copies of official links, and occasionally something that an editor thought was particularly valuable, but otherwise, they should usually be removed/replaced with something else. WhatamIdoing (talk) 23:55, 8 September 2024 (UTC)[reply]
@WhatamIdoing that's an issue with the implementation of the specific bot, not with the addition of archive links.
@JacobolusSome links should remain archived (e.g. official websites), some can (and should) be fixed by replacing them with an updated link the bot is not aware of, others should be removed. However that determination needs to be made by a human, and marking a link as dead is a very good way of letting humans know that a determination needs to be made. Thryduulf (talk) 00:07, 9 September 2024 (UTC)[reply]
Adding {{dead link}} is not usually a good idea for the ==External links== section. However, in the specific case of WP:ELOFFICIAL links for which you believe a working website to be reasonable (e.g., a company that is still in existence), it may occasionally be appropriate as a temporary measure (e.g., until someone can find the new corporate website). WhatamIdoing (talk) 01:05, 9 September 2024 (UTC)[reply]
Why is it not a good idea? It signifies that a human needs to make a decision about whether to remove the link, replace it with a working alternative or replace it with an archived version. Thryduulf (talk) 01:14, 9 September 2024 (UTC)[reply]
I can understand the impulse to kick the can down the road to a hypothetical future editor. However, WP:ELDEAD says that instead of doing that, you should normally update or remove dead links right now. We don't need one human to ask another human to make a decision. Just make the decision yourself. WhatamIdoing (talk) 01:49, 9 September 2024 (UTC)[reply]
We aren't talking about humans adding the template (although it would be appropriate while a discussion is ongoing or while attempting to locate the new location), we're talking about it being added by a bot. Thryduulf (talk) 02:00, 9 September 2024 (UTC)[reply]
Are we? I thought this thread was primarily about edits like this one, which affects nothing outside of the ref tags, and that I was talking about my desire to prevent the bot from making edits like this one, which added:
{{Webarchive|url=https://web.archive.org/web/20180101214652/https://www.irishfa.com/ifa-international/squads/senior-men/chris-brunt |date=1 January 2018 }}
in the ==External links== section. It is the location of this edit, rather than the substance of it, that I believe is inappropriate. If this edit were made inside ref tags, I'd never complain.
I understand that edits like this one (adding the dead link template) can't be easily improved upon, because all the bot can do, when it can't figure out how to fix it, is either ignore it or tag it. I might prefer ignoring it in that case (the domain works, and the reader could search from there; also, this specific bot sometimes finds itself blocked when the URL is working correctly for humans), but bots have limited capabilities. WhatamIdoing (talk) 02:41, 9 September 2024 (UTC)[reply]
The first is policy compliant, but irritates some editors when they are working in a wikitext editor. That first edit is the problem described at the top of this thread. I don't know how to solve that problem, or even if we should solve that problem.
The second is a violation of the Wikipedia:External links guideline and should stop. However, that will reportedly require more than a trivial bit of coding effort.
The third is perhaps discouraged under WP:EL, but I don't think we can realistically hope for the bot to do any better than that.
"primarily about edits like this one (special:diff/1244763493)" – No, this is not the type of edits we are discussing here. I think this edit – which is adding archived snapshot backups for dead links – is just fine, and my impression is that almost everyone appreciates such edits. –jacobolus(t)06:07, 9 September 2024 (UTC)[reply]
This URL does not have a working wayback snapshot, so no the snapshot should not be added as a backup. (And if the wayback snapshot had been added by a human while the page was still alive, we'd still be seeing the same empty content, because this website served a broken page to the wayback crawler, and is still serving broken pages to me when I navigate there now.) If this link does not show the intended content it should just be deleted as a citation, and a human will have to find a working replacement source. Any human adding wayback snapshot links should be double-checking every single snapshot link and making sure to never add anything like this to Wikipedia. (For example, wayback backups of Google Books pages should never be added, because they are similarly broken.) However, that is not currently happening, because wayback links are being added indiscriminately – what I am complaining about here. –jacobolus(t)06:33, 9 September 2024 (UTC)[reply]
wayback backups of Google Books pages should never be added, because they are similarly broken – This is not always correct. Pre-2020 it archived just fine, but it's possible the new Google Books UI broke something. A magazine we extensively cited was removed from Google Books in 2022, leaving hundreds of 404 links. And I've personally fixed many of these using the Web Archive. See ref 5 in Respect (Shaquille O'Neal album) for an example. AstonishingTunesAdmirer連絡16:23, 9 September 2024 (UTC)[reply]
This is a different view of the same Google scan of the same magazine hosted on the same website (archive.org), just going through their front-end instead of a wayback snapshot of google books. If you think that either of the two links is an unacceptable copyright violation then the other link surely has the same status, and you should just remove the external link and leave readers to find a copy of the magazine on their own. (I don't have any idea why the magazine was removed from Google Books, or what the copyright status is of the IA-hosted scan.) –jacobolus(t)18:42, 9 September 2024 (UTC)[reply]
the other link surely has the same status – does it though? A snapshot of a web page is explicitly allowed per the policy. But I haven't seen any mentions of reuploading of the web page's content. I do agree with you, the full scan is infinitely more useful. But I want to link sources that: 1) another editor will not delete at some point in future (right now this area is much grayer than what is allowed explicitly); 2) Internet Archive itself will not delete at some point in future (they absolutely will the moment copyright holders notice it and/or decide that sending a DMCA request is worth the effort, but it's extremely unlikely they will go after the archived copy of Google Books). AstonishingTunesAdmirer連絡19:40, 9 September 2024 (UTC)[reply]
If you want to follow "However, if you know or reasonably suspect that an external Web site is carrying a work in violation of copyright, do not link to that copy of the work without the permission of the copyright holder." and don't think the current copyright owner of Vibe gave permission for a scan to be hosted by third parties, then you should probably delete the links. I have no idea whether this media conglomerate that bought and shut down the magazine 10 years ago and now calls it a "leading entertainment and lifestyle brand delivering content across multiple media platforms to a diverse audience around the world" (i.e. a fancy blog/email newsletter) is going to be trying to scrub 15-year-old magazine scans from the internet. Maybe? Perhaps the Internet Archive should be gating access to similar still-copyrighted magazines behind their "borrow" feature. Personally I would just link to the IA copy, assuming that the publisher will sort things out with the IA, and if the book later ends up unavailable or deleted from the IA then Wikipedia could just take out the external link at that point and leave readers to find the magazine on their own if they want to verify the quotation. YMMV. –jacobolus(t)19:54, 9 September 2024 (UTC)[reply]
Honestly, I was just as puzzled when they deleted it from Google Books. I tried emailing them to ask whether it's some mistake on their side and if they are planning on fixing it, but it looks like they only accept corporate email addresses. It appears that they also own Billboard magazine, which is still available at Google Books (it will be an even bigger problem if they delete that one), so who knows. But I think you are right, eventually all of these will be unavailable. A shame, really. AstonishingTunesAdmirer連絡20:51, 9 September 2024 (UTC)[reply]
I too agree that we should discourage the blanket archiving of live links. It feels we're adding extraneous text to the references (and extraneous wikitext. Like jacobolus', my browser crashes when I turn on syntax highlighting on large pages; I wish it didn't) that are unlikely to be intentionally read or clicked on. Internet Archive is already doing us and our readers the great service of archiving pages linked to from our articles. If those links go dead, a bot is already doing the great service of linking to an IA snapshot of the page. I don't think the other much less common (I dare to assume) use cases merit so much text clutter. I agree with jacobolus that the scale of the linking feels like advertising for the service IA provides. Ajpolino (talk) 00:10, 9 September 2024 (UTC)[reply]
From the description, it does not seem that the addition of such links is indiscriminate. As long as it is reasonable to believe that the AI links will persist, which I think we all hope they will, there are numerous good reasons to always have a backup for any external link that can be backed up. BD2412T03:43, 9 September 2024 (UTC)[reply]
But you're not actually arguing that "the situation as described isn't indiscriminate"—instead it's "we don't need to discriminate". I put such a fine point on it because that's rhetorically what we're all hinging around, degrees of totality versus discretion. Remsense ‥ 论03:48, 9 September 2024 (UTC)[reply]
That's not what "indiscriminate" means. Were someone to create an article titled List of interesting news articles archived in the Internet Archive, and to populate this list with a random selection of a few hundred internet archive links, that would be indiscriminate. If someone were to linkeverywordinanarticle and provide an online dictionary reference for each word with a consequent Internet Archive link, that would be indiscriminate. However, were someone to create an article on a discrete and notable subject, including references to reliable online sources relevant to that subject, and were to then provide Internet Archive links as backup links for that set of sources, that would be the opposite of indiscriminate. That would be providing backup links to exactly parallel the sources relevant and appropriate to the article. BD2412T12:38, 9 September 2024 (UTC)[reply]
Doing something without discriminating, such as adding archive links to every link without considering them individually, is precisely what the word "indiscriminate" means: it would make a good example for a dictionary. Doing something indiscriminately doesn't necessarily mean it is bad, inappropriate, irrelevant, or involves ill intent. For example, a parent might indiscriminately and unconditionally tell each of their children they love them. (Though I personally think indiscriminately adding archive links is bad and inappropriate, which is why I raised the issue here.) –jacobolus(t)14:36, 9 September 2024 (UTC)[reply]
I would say that fundamentally, if in a frequently-updating publication like a newspaper or magazine (or something else reasonably expected to possibly have updated content within a month or two), IABot uses an archive-date that does not exactly match the access-date, and the editor does not verify that the content sourced is still verified by that archived copy (or else the archived text still matches), then that is indiscriminate editing subject to concern.
Raising my hand as agreeing with Remsense in general here. My objection is that unnecessary archive links are a waste of reader time. If the IA was fast, it would be one thing, but it's not fast at all. --jpgordon𝄢𝄆𝄐𝄇04:39, 9 September 2024 (UTC)[reply]
That's a distinction we should be clear about.
Some editors object to displaying archive links when the original is working just fine.
Some editors object to huge long strings of wikitext, even if readers never see it.
The first group would be perfectly satisfied if this wikitext: {{cite book |url=https://example.com |title=Example Domain |website=example.com |access-date=2 September 2023 |archive-url=https://web.archive.org/web/20240908173056/http://example.com/ |archive-date=8 September 2024 |url-status=live}} was in the article, so long as all the readers saw was a simple "Example Domain". example.com. Retrieved 2 September 2023.
Thryduulf to their credit has also stressed above that "working for me right now" isn't always sufficient for "working just fine", as an ancillary point. Remsense ‥ 论05:25, 9 September 2024 (UTC)[reply]
I support, in general, the archiving of as many reference urls on Wikipedia as possible. I don't think the "clutter" affects readers very much, since the reference section is not meant to be read through as prose. I think readers are getting with ease all of the useful information out of the citations they check out, assuming the citations are properly formatted. The presence of text like "Archived from the original on December 26, 2021", usually presented at or near the end of the citation, is not unduly burdensome on readers who don't care about archive links, and is obviously useful to those who do. I do think local consensus should be able to determine that a given article's citation style is not to use archive links for live urls. It's currently the case at Donald Trump, mainly for size reasons. Article creators with a preference for less archive-related text should be able to establish that style and not have it overridden unless there's a discussion that leads to consensus for inclusion. Firefangledfeathers (talk / contribs) 12:23, 9 September 2024 (UTC)[reply]
How do we take them out once they're in? Manually? I'd like this idea except in practice it's fait accompli. Once someone pressed the button, the links are there to stay, unless someone else manually removes each one. Levivich (talk) 14:49, 9 September 2024 (UTC)[reply]
I've brought this up a few times over the years. Basically, IABot is already creating archive links for articles added to citations without the need to run IAbot manually. It stores them in an off-wiki database and adds them as needed (I don't know exactly what it's looking for to do this). My perspective is that while there's an obvious downside of increasing the size of an article, there are some good reasons to err on the side of including the archive links in the articles. Namely, because what happens off-wiki is opaque and wiki bots have a tendency to just stop without notice, and also because it provides a link for a savvy reader to circumvent a paywall (not that this reason should be documented anywhere, mind you -- IA is having enough troubles lately). Regardless, however, this system of users drive-by "rescuing" sources that don't need rescuing isn't great. We could use an RfC simply asking "should all links have an archive link by default". If so, a bot should be doing it rather than random users adding them randomly. If not, we'll have to decide if it should be prohibited or left at a case-by-case basis. If case-by-case, what are the considerations? I'll also just add that my pet peeve in all this is less the size of the addition but how that addition messes with the tools we have to quickly determine who the primary contributors of an article are. — Rhododendritestalk \\ 17:47, 9 September 2024 (UTC)[reply]
If there's real interest we could run an-house experiment on link clicks on a couple articles to see if there's any significant uptick in people viewing sources when the archive link is available versus not. That'd give the benefit for displaying the links to live urls, at the least.
If there's a client-side performance cost/benefit per page to consider, I don't know if that can be reasonably mitigated by something like passing an iarchive-id parameter instead of the full urls, or else only rendering the live parts of citations on mobile and limited connections, or keeping the citations section unexpanded except on clicks (as several academic publishers do). I don't think stopping a few characters here or there (that time will force upon a page eventually) addresses the legitimate issue of client-side performance that people are raising.
at Rhododendrites (above): since bots are flagged in edits, your tools should discount bot edits, no?
I'm more concerned with the examples given that it seems IABot is adding archive links that do not conform to |access-date=. It seems like this should be the default behavior (as opposed to adding the latest archive link), but I'm not sure whether it is according to IABot's documentation. Not sure I can tell from the API source main whether that's the default already, not using the tool myself. Can someone confirm? SamuelRiv (talk) 19:21, 9 September 2024 (UTC)[reply]
My understanding is that IABot uses the archive that is closest in time to the access date in the article, if one is present. The closest archive may be quite different to the access date in some cases. When an access date is not present (e.g. for bare URIs) then I think it uses the most recent archive available. Thryduulf (talk) 19:37, 9 September 2024 (UTC)[reply]
According to doi:10.1145/3366423.3380300, 99.7% of readers don't click on any links in the refs. To give a few examples from the traffic report in The Signpost:
Vinesh Phogat might have had ~5,500 people click on a ref link during the last year. There are 90 refs, of which 66 have archive links. If we assume conditions along the lines of the spherical frictionless chicken, that might be an average of 11 opportunities per day for choosing between an archived link and a current link.
Noah Lyles might have had ~8,000 people click on a ref link during the last year. There are 68 refs, of which 9 have archive links. That might be around 3 opportunities a day for people to choose between an archived link and a current link.
Armand Duplantis might have had ~7,000 people click on a ref link during the last year. There are 79 refs, of which 52 have archive links. That means someone had to choose between an archived link and a current link about 12 times a day.
These are super-high traffic pages (mostly during the last month), and recording just 1,000 events would take more than a month. While I think that it would ultimately be feasible to test whether archived links encourage people to open a source, I would not want anyone to think that this would be quick and easy. I think it is more likely to involve words like long-term, perseverance, and (unfortunately) indeterminateresults. WhatamIdoing (talk) 20:27, 9 September 2024 (UTC)[reply]
a reference popup screenshot I just took to illustrate this There's a problem I'd love someone to sink their UX teeth into. How can we make users read references period! @Valjean and some of the other editors on the Trump article were experimenting with lead ref targets. I wonder if Wikipedia should invest more in footnote overlay gadgets and the like. I use some kind of gadget or user script that makes a mouseover preview popup show up. It does work for footnote links, but you just get an overlay that shows you the footer. It could be a lot bigger, brighter, and with more information about the reference. Also, theoretically, it could show you other pages that cite that source or general information about its reliability, etc. There's also a great user script by @Novem Linguae that adds color highlighting to the references. Andre🚐20:37, 9 September 2024 (UTC)[reply]
I believe that one of the ideas considered in Vector 2022 was having the refs displayed on the right-hand side (in all that "unnecessary" white space, if you have a wider screen and don't expand to full width).
But: Why should "making users read references" be one of our values? Also, would pushing that upset the apple cart in terms of WP:NONENG and WP:PAYWALL?
I looked up a Wikipedia article earlier today. I needed to remind myself that Bratislava really was in Slovakia, and I hadn't misremembered it. Do you think it would have been better for me to read the references for that? I don't. WhatamIdoing (talk) 20:43, 9 September 2024 (UTC)[reply]
I'm not a fan of the extra white space, I like the popup approach better. The Vector whitespace was a disaster on my monitor until they fixed it and offered some adjustable option for a wider view. I was using Monobook for a while until they fixed it.
I think making users read references is pretty key to the mission of Wikipedia and verifiability. I think accessibility to open access journal articles and content literacy are key to the mission of education and inspiring critical thinking. As everyone knows, Wikipedia is NOT reliable. As someone seeking information that's a bit less common, anytime you look at an article you don't know whether everything on it is correct, especially controversial statements, unless you check the references. Hoaxes and vandalism exist due to the wiki model, and the chain of corroboration is key to making the whole thing work. Sure, looking up where Bratislava is, is not going to require looking at the references. I've come to see Wikipedia as basically a public source of information, while plenty of valuable stuff is locked behind paywalls or private, difficult to access archives. Wikipedia's goal is unlocking the world's information and making it available. Andre🚐20:51, 9 September 2024 (UTC)[reply]
This is a use case for having invisible template text or a special transclusion type. Basically the archive links should live somewhere else, not in the actual article. Andre🚐19:43, 9 September 2024 (UTC)[reply]
The concern brought up here, and brought up repeatedly at CS1, is site outage and region blocks. The solution, given time and time again, is to just use the archive link. (Adding explicit parameter options to notify bots is apparently unworkable, so undocumented html comments are recommended instead?) SamuelRiv (talk) 20:14, 9 September 2024 (UTC)[reply]
So yeah, the archive link is useful, as a permalink of the article at that time, and solves a legitimate problem as you say, don't get me wrong. But for large articles someone blindly rescuing all the links, even ones that will never probably go down bloats the article with a ton of extra size for no gain, and it also creates a usability issue because in many cases the archive link, while preferable to a 404, is less usable than the real website - either due to lacking some assets or images, special javascripts or what-have-you. We've all experienced a great deal of jank browsing archive links, right? And there's no real reason that the archive links themselves need to be in the article for highly-trafficked sites, where the output of feeding that URL into the archive API will deterministically return something usable. As Rhodo says there's actually a bot that will store a map of the links, which seems like a good direction, versus having humans doing it. Andre🚐20:19, 9 September 2024 (UTC)[reply]
Region locks and region throttles. That gets worse for big articles on regional subjects, especially those updated with local news (frequently region-throttled), Trump and other US pols being good examples. I also fail to see how a relatively small percentage increase in the tiny fraction of articles that are both utterly enormous and highly volatile should be what makes or breaks IABot (especially since it's those articles for which the first point of my paragraph here is most relevant). There are surely other ways to address performance issues on oversized over-newsed articles (including, but not limited to, those already enabled by default for mobile). SamuelRiv (talk) 20:54, 9 September 2024 (UTC)[reply]
If articles are getting so big that editing them is difficult then they should be split. Unfortunately this will lead to having to defend some spinout articles against deletionists at AfD but that's a separate issue. Thryduulf (talk) 11:14, 10 September 2024 (UTC)[reply]
Some articles are also more densely cited than others, which is a dimension that does not justify merely equating wikitext length with the abstract conceptual length withholding particular technical hiccups. Remsense ‥ 论11:23, 10 September 2024 (UTC)[reply]
I'm not sure I fully understand that comment, but splitting because it cannot be edited without problems is just as valid a reason for splitting as it being too long to read comfortably. Thryduulf (talk) 12:04, 10 September 2024 (UTC)[reply]
Sorry for the confusion—I suppose I would rephrase by saying I disagree with that. Articles should be split for the sake of readers; if there is a distinct temptation to split for the sake of editors, we should critically examine whether there are technical problems we can fix to avoid such a solution first. Remsense ‥ 论12:09, 10 September 2024 (UTC)[reply]
I don't think Levivich is in a minority, and I agree with Ajpolino. Here is a good-faith example by Susmuffin which has made the article miserable to work with. After the addition of archive links, 12% of the page content is still archive links; it is not necessary to add archive links on sources like The New York Times, the BBC, The Washington Post and multiple others whose links rarely go dead, and editing around the now excessively lengthy citation templates is made more complex. I don't feel I can revert a good-faith edit like that, but I wish I could, as having all of that template clutter isn't helpful. SandyGeorgia (Talk) 11:04, 10 September 2024 (UTC)[reply]
The Washington Post and The New York Times both limit access in various ways. Both at times also change the content on pages without changing the URL. While it might not be strictly necessary to add archives to such pages, they still add value to the page and I would revert removal of them for any reason other than them being broken (and even then it would be much better to replace them with ones that aren't broken). The extra wikitext is unfortunate, but nothing more than that, especially since tools like Visual Editor and syntax highlighting reduce the impact significantly. Thryduulf (talk) 11:12, 10 September 2024 (UTC)[reply]
I agree that using IA as a tool to bypass paywalls is unethical, but how exactly does that work? Does IA maintain subscriptions to paywalled sites and then expose the pages they retrieve? RoySmith(talk)15:04, 10 September 2024 (UTC)[reply]
Some paywalls only trigger after a certain number of reads. Typically news websites. By using IA you can access a website even when you have used up all your free reads. Jo-Jo Eumerus (talk) 15:53, 10 September 2024 (UTC)[reply]
Websites sometimes
offer an openly available article at one time and then (sometimes many years) later retroactively put it behind a paywall, or
offer an openly available article to crawlers such as the Internet Archive's or Google's, but present a paywall to human readers.
I don't personally have a problem with including archived snapshots of paywalled pages in Wikipedia articles: these provide an important service to editors trying to verify claims made in articles, but are unlikely to significantly affect the website's revenue. If the website owner has a problem they can ask the archive site to take down the snapshot. –jacobolus(t)15:53, 10 September 2024 (UTC)[reply]
@SandyGeorgia If you find the archive links on living pages there to not be helpful, you should absolutely feel free to revert the change or otherwise remove them. Edits being done in good faith does not mean that no one else gets to disagree. –jacobolus(t)14:50, 10 September 2024 (UTC)[reply]
Is it time to explore WikiCite? I know enwp doesn't like wikidata, but that project to take all the sources we cite and include them in wikidata has been going on for some time now. I imagine an archive link property is already in place. Time to have a bot replace citations with wikidata references and a bot to turn new citations into wikidata items? I don't expect so, but worth bringing up at least. — Rhododendritestalk \\ 12:34, 10 September 2024 (UTC)[reply]
Is it possible to have a smaller step, say one where archive links for live sites (ie. those created purely as a backup) can be stored at wikidata? That would keep the wikitext burden of the archive urls small (just a Qcode I suppose), while maintaining the backup functionality. CMD (talk) 12:43, 10 September 2024 (UTC)[reply]
I have read WP:!VOTE, and I still don't really understand what it means when a closer says I see around 22 !votes endorsing the closure and 15 saying to overturn. Can someone explain properly and maybe rewrite the wp:!vote section? Contrary to the section, I very often see people saying it as if there was actually a vote, as above. It seems logically inconsistent to say "this is not a vote, however 5 people voted (or !voted) for this and 10 people voted for this". No matter how you phrase it, if there's two numbers and one is higher than the other, you are heavily implying it is actually a vote. MarkiPoli (talk) 13:51, 9 September 2024 (UTC)[reply]
@MarkiPoli: It sounds like you have figured it out fairly accurately. The "!vote" indicates how people "would have" voted, if a given proposal were to be put up to a vote. But it's not actually being put up to a vote (i.e. the number of "!votes" does not determine the result), and hence they are called "!votes". Fabrickator (talk) 15:31, 9 September 2024 (UTC)[reply]
I've discussed previously how it's a bit of jargon that I personally find unnecessary. For programmers who understand that "!" means "not", it generally means "I know this isn't supposed to be a vote but I'm framing my comment as a support/oppose vote anyway", or sometimes "I know this isn't supposed to be a vote, but we all know that the closer is going to treat the supports/opposes as a straw poll". However in some instances it's just used as a synonym for participate/participation (for example, every single use at Wikipedia:Advice for RfA candidates) or for actual voting. For better or worse, jargon forms part of a group's culture, and a lot of people are accustomed to this bit of culture. isaacl (talk) 17:32, 9 September 2024 (UTC)[reply]
I share your frustration, which I think in part is the cognitive dissonance of always reading in my head "!vote" as "not-vote". Perhaps the better metaphor would be a dynamic object type for discussion participation, that can then be down-cast into a static 'vote' type? Then for example one could say, "I 'cast my vote' for option A." SamuelRiv (talk) 18:06, 9 September 2024 (UTC)[reply]
I see we're moving even further into the obscurity of the programming domain... The thing is, since the opinions aren't supposed to be interpreted in an individual binary manner, the object in question shouldn't support a conversion operator to support/oppose. isaacl (talk) 18:22, 9 September 2024 (UTC)[reply]
Throwing my 2¢ in... Essentially the closer should first weigh arguments based on their level of policy/guideline-reinforced reasoning. If the balance between opposing viewpoints is roughly equal at this stage (e.g., a trump card such as a BLP violation has not been played), then the closer essentially moves on to a head count and tallies the !votes. Strength of argument still matters, of course. Simple WP:JUSTDONTLIKEIT comments could be ignored. Although we don't start with the tally, we might end with one. In fact, that's just how it plays out a vast majority of the time. --GoneIn60 (talk) 21:13, 9 September 2024 (UTC)[reply]
Although this may be a difference without a distinction, it's not exactly the case that the evaluator of consensus might move onto a head count, but that in cases where there is no overriding policy, the numbers of supporters of a particular argument gets used as a proxy for evaluating its strength. And since the evaluator isn't supposed to evaluating the positions themselves, but evaluating the consensus of the participants regarding the positions, there aren't a lot of good options for weighing the arguments in other ways. Thus in practice, it can be difficult for an evaluator to go against the straw poll results, inevitably facing challenges for their reasoning, which exerts subtle pressure to follow the head count. isaacl (talk) 22:24, 9 September 2024 (UTC)[reply]
Or as WP:DISCARD frames it, judge "which view has the predominant number of responsible Wikipedians supporting it", which in the end can boil down to a head count depending on your perspective of things. The participant(s) may not realize the process involved – what factors were eliminated – before that point was reached. -- GoneIn60 (talk) 22:46, 9 September 2024 (UTC)[reply]
One of the main reasons that Wikipedia is not a democracy is that it's not a government or any kind of body with rights or responsibilities. It's an encyclopedia full of processes, it's more of a product than it is an organization, and organizationally, it's designed as a charitable nonprofit. Boards of directors of corporations have votes, with different weights, but nobody says the board is a democracy and that it's 1 person, 1 vote. Disenfranchisement, for example, is a crime in democracy, but is permissible in a board. Andre🚐23:21, 9 September 2024 (UTC)[reply]
Ah, here is one of those early Wikipedians now! Welcome back :-) Of course it's not a democracy, and I didn't say "a democracy," nor does WP:NOTDEMOCRACY say "a democracy". I said, and WP:NOTDEMOCRACY said, "not democracy," as in "not democratic." BTW, disenfranchisement is not a crime in all democracies. It was the law, part of the constitution, in the United States for a long time (and some say, in some parts, still is). Levivich (talk) 23:24, 9 September 2024 (UTC)[reply]
That's a good distinction to split, but still, I would argue, given the stakes are low, that Wikipedia is not democratic. Consensus is not a mechanism by which the rabble mob wins out by the numbers. The rabble mob could argumentum ad populum all day but if they are wrong according to the trusted closer/arbiter (used to be admins), well, they are wrong. Doesn't in practice happen often probably because the situations where a minority numerically stubbornly argued the correct interpretation and the closer found irregularities or non-policy abiding positions, do not occur in such significant number. Wikipedia is meritocratic at its heart, which has to do with its roots in the The Californian Ideology, namely techno-libertarians. Rule by reason, not rule by numbers. Andre🚐23:30, 9 September 2024 (UTC)[reply]
I respectfully submit that may be an artifact of the reference point and the unavoidable fact of most English Wikipedians inhabiting one of the major parliamentary or constitutional domains. Also, while there's an overlap with the Objectivist crew, the techno-libertarians of Silicon Valley borrow a lot from both the mainstream social liberalism as well as the anarchism and socialist movements of the punk and acid hippie world of California computer scientists. In 2004 at least, Wikipedia quite decidedly did not want to have politicking such as populist electoralism ("adminship is not a big deal"), legal systems ("wikilawyering"), or extensive ceremony ("IAR") of something like a legislature. "Rough" is the operative word in consensus. Andre🚐00:00, 10 September 2024 (UTC)[reply]
40%+ of the world's population live in federal states, more than half the world's population have democratic systems, we're just starting to look more like the rest of the world. Early Wikipedia from my experience was a weird amalgum of epistemological anarchism and tech bro libertarianism. Thankfully, the adults took over. :) Regards, Goldsztajn (talk) 00:35, 10 September 2024 (UTC)[reply]
I always thought we were an anarcho-syndicalist commune, taking it in turns to act as sort of executive officers for the week. (And we're an autonomous collective wrt the self-perpetuating autocracy that is the WMF.) (Jfc, this is way too close to the truth.) SamuelRiv (talk) 03:11, 10 September 2024 (UTC)[reply]
(...But all the decisions of that officer have to be ratified at a special biweekly meeting, by a simple majority in the case of purely internal affairs, but by a two-thirds majority in the case of more important matters, except in reality, everyone is just a peasant pushing mud around for no reason? Way too close to the truth.) Levivich (talk) 03:22, 10 September 2024 (UTC)[reply]
Ya know, in all seriousness, syndicalism by sortition, in which a limited number of admin positions (and/or other elevated rights) are distributed on a rotating basis among EC editors might actually be a really fantastic idea. (Other admins stay long-term as-is.) One way to boost engagement and responsible editing is to endow responsibility (and ofc take it away as soon as it's abused), and it's also a way to get motivated editors motivated even more, or to take on projects on the site they might otherwise not. (Admittedly, these are not the problems for which sortition has been studied to have an ameliorating effect -- see citations in article -- we don't have anything like a problem of a broken polarized electoral process that encourages extremist politics and anticompromise -- so the benefits I suggest seem largely unexplored in the literature.) SamuelRiv (talk) 04:46, 10 September 2024 (UTC)[reply]
nonsensusn.: One of the founding myths of Wikipedia, which claims that in between everybody writing their opinion, and the closer writing the closing statement, an invisible, undefinable, indescribable, but nevertheless real phenomenon called "consensus" arises. According to the Nonsensus Myth, when writing a closing statement, the closer is not counting votes, nor giving their own opinion, but rather observing and reporting on the "consensus" phenomenon. The Nonsensus Myth holds that in some cases, "consensus" is said to be clear enough that everybody can observe it; although this only happens in numerical supermajorities, the Nonsensus Myth insists that it is not the numerical supermajority that determines the clarity of the consensus. In cases in which it is said that the "consensus" is not clear, which are almost always cases where there is no numerical supermajority, the Nonsensus Myth teaches that only a person who never wrote an opinion about a matter under discussion, called an "uninvolved" person, can accurately observe the "consensus" phenomenon, while continuing to insist that the lack of numerical supermajority is not the reason that "consensus" cannot be clearly seen by everyone, but only be an "uninvolved" person.
In reality, what happens is that the votes are being weighted against certain objective criteria -- do they align with policies and guidelines, do they accurately reflect reliable sources, are they made by a blocked or banned user, etc. -- and then the outcome is based on which side is in the majority after the weighing is done. It's not really anything more complex than that. Levivich (talk) 23:56, 9 September 2024 (UTC)[reply]
For an invisible and indescribable phenomenon, quite a few people have described it adequately enough. I recognize you're being a bit good-naturedly snarky or tongue-in-cheek, but I sure hope that people closing discussions in 2024 at least have some understanding of the supermajority merely being an indicator or a rule of thumb that should not substitute for analysis. Andre🚐00:03, 10 September 2024 (UTC)[reply]
On occasion, you will see in closing statements something like, 'those voicing opposition agree with the supports that . . .' -- the general idea being, the process searches beyond the raw count. -- Alanscottwalker (talk) 23:57, 9 September 2024 (UTC)[reply]
In the early days of WP, RFCs were more focused on trying to cobble together compromises, rather than simply counting opinions on yes/no questions. Doing a quick (often informal) !vote tally helped editors see which direction people were leaning. Indeed you might see several !vote tallies in the same RFC (to see if we were inching towards compromise). They were part of the process, but not necessarily the end of the process. Blueboar (talk) 00:12, 10 September 2024 (UTC)[reply]
Must each entry of a disambiguation page link to an article?
Following the thread on Talk:Rawdog, I think it appropriate to change Rawdog from a redirect to a disambiguation page as the current viral phenomenon is different from the current redirected article, with a second entry describing the alternative meaning. Wikishovel reverted it.
As the phenomenon may be too transient to warrant a new article, it seems out of place in Air travel#Commercial_air_travel, and yet a reader looking up "rawdogging" might be confused by its earlier meaning, what's the best way to handle this?
First, not clear why this is at VPP. Second, the rawdogging on flights thing is a passing fad that will have no lasting significance or notability and we should not create an article on it. voorts (talk/contributions) 13:41, 10 September 2024 (UTC)[reply]