Wikipedia:Village pump (miscellaneous)

 Policy Technical Proposals Idea lab WMF Miscellaneous 

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

The miscellaneous section of the village pump is used to post messages that do not fit into any other category. Please post on the policy, technical, or proposals sections when appropriate, or at the help desk for assistance. For general knowledge questions, please use the reference desk.

Discussions are automatically archived after remaining inactive for a week.

« Archives, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76

Obsolete policy proposals appearing in Google answers

If you do a Google search for "Do Wikipedia edits have to be approved?" the first search result is a link to Wikipedia:Edit Approval. Google displays a passage from it, highlighting "If an administrator or a person who is considered trustful by any two administrators believes the edit is a valuable contribution to Wikipedia, they will approve the edit." Does anyone know how to ask Google to stop coughing up historical policy proposals as answers to questions? Clayoquot (talk | contribs) 21:32, 9 October 2023 (UTC)

I see a "Feedback" link at the bottom right of the snippet, I was able to click that and select that it was "Inaccurate content" and then explain why. How Google actually take that feedback into account is another question, but maybe if enough VPM watchers do the same thing it could help. the wub "?!" 22:05, 9 October 2023 (UTC)

I'd be very surprised if Google took the slightest interest in a few complaints about their search engine. They've been misrepresenting AI-generated nonsense of their own as Wikipedia content for years, and clearly don't care about inaccuracies as long as they can generate search results that look superficially convincing - what AI does best. AndyTheGrump (talk) 22:20, 9 October 2023 (UTC)

@Clayoquot, I wonder if we could WP:NOINDEX that page. WhatamIdoing (talk) 00:06, 10 October 2023 (UTC)

Boldly  Done Edward-Woodrow • talk 00:22, 10 October 2023 (UTC)

Thanks everyone! Clayoquot (talk | contribs) 00:32, 10 October 2023 (UTC)

Should we NOINDEX all {{historical}} or {{failed proposal}}s? are there any disadvantages to doing this? Alpha3031 (t • c) 15:06, 14 October 2023 (UTC)

Can't think of any, tbh. Edward-Woodrow • talk 16:36, 14 October 2023 (UTC)

The disadvantage would be making it difficult for those who are specifically searching for particular pages to find them. CMD (talk) 03:07, 15 October 2023 (UTC)

They could just go on our site and search there. QuicoleJR (talk) 14:09, 25 October 2023 (UTC)

That's a difficult question. I would like to say that people who believe AI-generated nonsense only have themselves to blame, and that we should leave people who want to find out whether a proposal has been rejected before in peace, but lots of people seem to prefer to believe the AI-generated nonsense. Is it our job to protect people, and Google, from themselves? Phil Bridger (talk) 17:50, 14 October 2023 (UTC)

If you have a deep hole on your property, you put up a fence. It'll stop most people from falling in, and if someone really wants to climb under the fence--whether to study the hole or out of stupidity--you did your part and you're innocent. But if you don't put up a fence and someone falls in, that's completely your fault, because how were they supposed to know the hole was there?

Case in point: It'll stop most misinformation, and if someone really wants to find out about a rejected Wikipedia proposal, they could still look it up in, well, Wikipedia. 2603:8001:4542:28FB:C4A:7FE3:361D:7B96 (talk) 02:51, 15 October 2023 (UTC) (Please send talk page messages here instead)

I've reverted the NOINDEX. We shouldn't be in the SEO game. Pages that are failed proposals are just as interesting as other pages from a search perspective. Let's not second-guess what pages people want to find when they do a search, and let's not second-guess what google shows to their users. RoySmith (talk) 03:15, 15 October 2023 (UTC)

If we want pages to appear in normal search results, we can still exclude them from Google snippets using a "data-nosnippet" attribute. [1] Do we (or can we make) a template that would put this attribute on a page? Clayoquot (talk | contribs) 04:40, 15 October 2023 (UTC)

In theory, we should be neutral between search engines and can't hope to pander to the idiosyncrasies of all of them. In practice, Google has such a monopoly that we should probably do this. Certes (talk) 13:53, 15 October 2023 (UTC)

There's two distinct issues here. One is whether we should be hiding specific pages from search engines based on our evaluation of whether they're obsolete or not. The other is the mechanism used to do this. I'll stick to the first question, and my opinion is emphatically that we should not.

Our job is to write an encyclopedia. Let the search engines worry about what they index and how they present it. Once we get into customizing things to optimize how the search engines present our stuff, it'll never end. If we hide obsolete policies, why not failed FA, GA, DYK, ITN, etc, nominations? Failed RFAs? How about AfDs which ended up getting overturned at DRV? Maybe search for every place somebody struck out text and figure out how to get the search engines to ignore that?

We already have ways to mark content as obsolete, for example the {{historical}} and {{Failed proposal}} templates used here. If the search engines want to use those to adjust their internal rankings, good on them. If not, then not our problem. RoySmith (talk) 14:54, 15 October 2023 (UTC)

The slippery slope argument? Really?

I'd support noindexing that page. Wikipedia should be giving accurate information. If we know a failed proposal is leading to inaccurate search results, and adding one line of text to the page will fix it, we should fix it. And if that happens again with another page in the future, we should fix that page, too.

It's better to fix it that to look down our noses at Google while not fixing it. Levivich (talk) 15:02, 15 October 2023 (UTC)

Fully agree with RoySmith. If Google produces inaccurate search results because it is misreading correctly tagged information on Wikipedia, that seems to be something Google could be interested in fixing. The search results for "questions" at Google often answer incorrectly or not the question I asked; it is not our business spending any effort on improving that. —Kusma (talk) 16:58, 15 October 2023 (UTC)

Should a failed proposal ever appear in a data snippet? If not then we might add "data-nosnippet" to the {{Failed proposal}} tag. Is that any worse than using the blunt instrument of __NOINDEX__, which we already use in appropriate templates? Certes (talk) 15:26, 15 October 2023 (UTC)

That would just prevent the banner from appearing in a snippet. We'd have to wrap the entire page in a div with data-nosnippet. Nardog (talk) 15:35, 15 October 2023 (UTC)

What about making {{nosnippet}} render <meta name="robots" content="max-snippet:0">? I'm not sure but I think that can be thrown in the body anywhere (so at the top of a page) without having to be part of a div/span/section. We could then add {{nosnippet}} to {{failed proposal}} and other similar templates. Levivich (talk) 16:10, 15 October 2023 (UTC)

<meta> can't be in wikitext. We have to request __NOSNIPPET__ on Phab. Nardog (talk) 16:47, 15 October 2023 (UTC)

So that means we'd need a closing template to close the div tag, right? Ugh. Idk, imo, all of WP: space should be no-snippeted. WP-space pages, including policy pages, are not good pages to snippet, as the snippet will undoubtedly be taken out of context (as demonstrated here). Maybe that should be the Phab request. Levivich (talk) 16:55, 15 October 2023 (UTC)

I'm opposed; perhaps a search engine comes along that does a good job at making snippets, and we shouldn't block that. —Kusma (talk) 16:59, 15 October 2023 (UTC)

That's highly speculative. If and when someone develops technology that will correctly snippet Wikipedia policies and failed proposals (mark my words: that will never happen), then we can remove the no snippet. Levivich (talk) 17:12, 15 October 2023 (UTC)

If people are running LLMs or similar on Wikipedia and ignore page headers that clearly state the page is obsolete, they clearly do not care about the quality of their results. Improving untrustworthy machines slightly can be bad if it increases trust. —Kusma (talk) 17:29, 15 October 2023 (UTC)

The Google snippet does not include the header templates. Edward-Woodrow • talk 17:48, 15 October 2023 (UTC)

@Certes Should a failed proposal ever appear in a data snippet? Of course it should. If I was researching failed proposals on wikipedia, I might run this search. I would be pretty annoyed if I discovered that all the failed proposals were missing from the search results. RoySmith (talk) 17:08, 15 October 2023 (UTC)

No snippet does not remove them from search results. No snippet isn't the same as no index. Levivich (talk) 17:11, 15 October 2023 (UTC)

Yes, you would want it in your search results. But "data snippets" seems to be something other than search results: an AI-based or AI-like attempt to answer your question by writing an answer based on information from Wikipedia and elsewhere. Unless the question is "which failed Wikipedia proposal would have required two supporters for each edit", this isn't the right answer. Certes (talk) 17:14, 15 October 2023 (UTC)

+1 for RoySmith's reasoning. — Frostly (talk) 18:27, 15 October 2023 (UTC)

Just throwing out that there would still always be Category:Wikipedia failed proposals (for failed proposals), Category:Inactive project pages (for historical or superseded pages), and for full completeness, Category:Defunct WikiProjects for historical WikiProjects, Category:Deprecated templates/Category:Deprecated templates kept for historical reasons for historical templates, and Category:Obsolete images for historical images. Yes, it takes a small bit more searching than a simple Google search, but if someone really wants to find a historical page it's still more than doable (not to mention that Wikipedia will obviously have more complete coverage of those than Google ever will), and it wouldn't inconvenience those who are looking for current information and would get served the wrong info instead. 2603:8001:4542:28FB:8997:D75D:C1E9:927B (talk) 20:46, 15 October 2023 (UTC) (Please send talk messages here instead)

Some people prefer to use their favorite web search engine to find old conversations, etc., which can include "historical" pages. I have done it myself, though I'm generally happier with the internal search, as it correctly handles "shirt" -stripes searches, and others don't. (As a general rule, Duck Duck Go gives me pages that still include stripes, and Google gives me pages that don't include shirt.) I wouldn't want to break searching in general, but when we have a known problem, I don't mind doing what we can to fix it. It's more important to me that people understand how Wikipedia works than that we maintain a consistent system internally.

That said, I do sometimes wonder where all y'all are when the question is about "letting" Google index new articles. I see several people here saying it's not our job to protect Google from a known-bad page, but I feel pretty lonely in conversations about whether unknown-and-probably-okay articles should remain NOINDEXd to protect Google's search results from something that definitely doesn't qualify for speedy deletion but isn't in very good shape. WhatamIdoing (talk) 00:45, 16 October 2023 (UTC)

I've complained to Google as The wub suggested. Pinigng NPerry (WMF) in case he knows someone at Google to notify as well. In the meantime I'd support a NOINDEX of all failed proposal pages because misinformation is bad, and misinformation about Wikipedia is ultimately bad for Wikipedia, regardless of whose fault it is. Clayoquot (talk | contribs) 04:21, 17 October 2023 (UTC)

As an aside, I see google have disabled snippets related to the search query "who is caliph of Islam?" Folly Mox (talk) 15:50, 17 October 2023 (UTC)

(This box only shows for autoconfirmed users.) WTF? It's bad enough we're talking about manipulating how the search engines index particular pages, we've got warning boxes on our own site that we hide from unconfirmed readers? What's next, shadow banning? Enterprisey (my apologies for the tone) can you explain why you thought this was a good idea? RoySmith (talk) 16:07, 17 October 2023 (UTC)

Since the box says "Feel free to move to a better title", and only autoconfirmed editors can move pages, it's not unreasonable to restrict the box to people who can actually do the proposed task. WhatamIdoing (talk) 05:15, 23 October 2023 (UTC)

In fact, that page is extended confirm protected for both edits and moving for the next three years (log), so even most autoconfirmed editors can't move it! Caeciliusinhorto-public (talk) 08:46, 25 October 2023 (UTC)

The more I think about it, the more I think that Wikipedia should nosnippet the entire website. I don't trust snippet algorithms to accurately summarize Wikipedia articles (nevermind our policy pages), or to grab the best excerpt from articles--even when it's grabbing a portion of the lead, it might mislead readers by grabbing the wrong portion and presenting something out of context. Since nosnippet is not the same as noindex, it would still allow people to search for and find information from Wikipedia, it would just stop the unreliable snippets from being presented to readers who are searching for stuff. Let them click on the link and read the Wikipedia page in context. Levivich (talk) 16:10, 17 October 2023 (UTC)

If you did that, I'll bet you 100 quatloos that Google will just start ignoring the no-snippet hint. SEO is a losing game. RoySmith (talk) 16:17, 17 October 2023 (UTC)

Same. jp×g 08:46, 21 October 2023 (UTC)

• There's a surprising (to me) amount of controversy here. Given the broader policy implications, I strongly suggest that before anybody takes any action on this, start a formal RFC and advertise it in the usual places so we don't fall into the WP:LOCALCONSENSUS trap. RoySmith (talk) 14:51, 17 October 2023 (UTC)
  • I concur. — Frostly (talk) 21:56, 17 October 2023 (UTC)

About vermiculite

Where does this mineral come from? Geko72290 (talk) 22:15, 16 October 2023 (UTC)

@Geko72290, if Vermiculite or its sources doesn't have your answer, the next place you could ask is WP:RDS. The village pumps are for discussing Wikipedia itself and related matters, not for general knowledge questions. 57.140.16.12 (talk) 14:00, 17 October 2023 (UTC)

@Geko72290: from the lead of Vermiculite: Vermiculite forms by the weathering or hydrothermal alteration of biotite or phlogopite. Edward-Woodrow • talk 12:04, 25 October 2023 (UTC)

Review and comment on the 2024 Wikimedia Foundation Board of Trustees selection rules package

You can find this message translated into additional languages on Meta-wiki.

More languages • Please help translate to other languages.

Dear all,

Please review and comment on the Wikimedia Foundation Board of Trustees selection rules package from now until 29 October 2023. The selection rules package was based on older versions by the Elections Committee and will be used in the 2024 Board of Trustees selection. Providing your comments now will help them provide a smoother, better Board selection process. More on the Meta-wiki page.

Best,

Katie Chan
Chair of the Elections Committee

01:12, 17 October 2023 (UTC)

Wikimedia Foundation PageTriage project: Final update

This is a quick notice that I've just posted the final update for the Wikimedia Foundation Moderator Tools' team's work on PageTriage for New Pages Patrol. Samwalton9 (WMF) (talk) 11:00, 20 October 2023 (UTC)

Norwegian names

Not sure where I should bring this up. I guess this is the best place? This is about Norwegian geographical names, specifically Svarthuken in Svalbard. Once named Negro Point by English sailors, changed in two steps to the current Norwegian name... but does the English name for a location follow name changes in Norwegian automatically? The impact and interest for the place should be minimal for English speakers so maybe it does not matter. If you have an opinion, please comment in Talk:Svarthuken. BRG Hubba (talk) 13:15, 24 October 2023 (UTC)

The commonly used English name should be used, see WP:COMMONNAME. I'm sure there isn't going to be any regular usage of the old name, given it's so obscure. So changing the name would seem appropriate. -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 21:11, 24 October 2023 (UTC)

Too many edits allowed

I believe this site is losing its credibility. There are too many people making edits on subjects they know nothing about. One deceased musician's page has been edited countless times by know nothing's. They changed his entire history, removed notable achievements. The Wikipedia information doesn't match media articles. It makes for too many questions. I'm sure this man's page is not the only one altered by people who have no business editing finished the pages. 2600:8805:C03:800:9095:FA06:91E6:506 (talk) 02:22, 26 October 2023 (UTC)

Feel free to link to the page in question here, and experienced editors will likely take a look and improve it. –Novem Linguae (talk) 06:12, 26 October 2023 (UTC)

Talk page villages in the city of Wikipedia

I have notice a phenomenon. Talk pages are sort of organic. One conversation can spark the idea for a new conversation. Or a conversation started 20 years ago can still have occasional additions, with a gap of 10 years between posts. Even simply having a longish talk page can encourage more posts, the page has a party momentum.

I have noticed cases, a hypothetical example, where the talk page had a lot of activity, say between 2003-2013. Then it was archived. From 2013-present, there is very little activity. It's like a village that grew organically, and is fragile, then "destroyed" in the name of progress/cleanup/clutter. I'm not making any suggestions. only observing this phenomenon. Check old archive pages and see how busy they used to be prior to archiving, and what happened after. Of course some pages continue at good pace, but some never recovered the same momentum.

This is cool sociology stuff that is easily studied because it can be quantified, and there is a long history, and large amount of data. It reminds me of The Death and Life of Great American Cities where urban planners destroyed organic communities in the name of progress and the city never recovered. Talk pages are villages in the city of Wikipedia.

(there are ways to reignite old conversations, etc.. but in the main, most users don't go into the archives to contribute because those conversations are locked and copying it back to the primary is not done very often). -- GreenC 15:43, 26 October 2023 (UTC)

Archiving can be setup to leave a certain amount of threads on the talk page. I've have my talk page to leave at least five threads unarchived for exactly the reasons you mention. I don't think it would be helpful to not archive anything on talk pages, as the page would become so long as to discourage new activity. -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 19:53, 26 October 2023 (UTC)

I have seen many cases where someone replies to years-old comments but I don't recall seeing anything constructive. Just NOTFORUM problems or complaints about living people and similar. Johnuniq (talk) 22:12, 26 October 2023 (UTC)

I wonder how much the general reduction in talk page activity is related to the fact that anonymous users on mobile devices have historically been unable to interact with talk pages (permalink) ... though that link may well be out-of-date (see this discussion and m:Talk pages project/Mobile. Graham87 (talk) 10:13, 27 October 2023 (UTC)

Everyone seems to have different archiving preferences. I agree that setting minthreadsleft:0 is bad, and that there should always be some recent conversations left on a talk page, so that visitors can see what has been discussed recently. –Novem Linguae (talk) 10:17, 27 October 2023 (UTC)

Or maybe it simply means that there was much to be said and done in 2013, but that the topic itself (not the Wikipedia article) has stagnated since.--User:Khajidha (talk) (contributions) 16:09, 27 October 2023 (UTC)

I've been off on several rants recently about how the fundamental structure of wikipedia is broken, and this is another example. One of the important problems (perhaps the most important one from a technology standpoint) with wikipedia is that the information is stored as plain text. Any attempt to layer structure on top of that is based on parsing the text, which is fraught with peril because most of it is generated by hand with essentially random markup. There is no real concept of a "conversation" in the sense of a tree (or even a forest) of posts with clear metadata about who wrote what, and which posts are in response to which other posts. RoySmith (talk) 16:51, 27 October 2023 (UTC)

As others have, I keep threads on the page, usually as many as makes the ToC equal the length of the archive box (~6). I have encountered, without reverting, instances of people archiving, either by hand or automatically, and leaving nothing on the page. If the threads are very old, I let them go. However, leaving even very old threads on the talk page does contribute to a feeling of liveliness as well as showing what issues have been addressed in the past, which in itself can be helpful. Dhtwiki (talk) 23:52, 27 October 2023 (UTC)

There is a current proposal at BOTREQ to archive all talk pages automatically, to keep them clear of "clutter" ie. talk pages are mostly composed of "clutter". In need of "cleaning". By a bot algorithm. It will never get approval, but I suspect a lot of people would vote for it. There's nothing in the WP docs that talks about the downsides of archiving, or upsides of keeping pages intact. -- GreenC 04:15, 28 October 2023 (UTC)