Wikipedia:Village pump (proposals)

There's been some discussion at WP:ANI#Return to potentially POINTy editing regarding an editor who's been mass-nominating "Comparison of X" articles like Comparison of 3D printers for deletion. The general standpoint seems to be that indeed most of these 'comparison' articles are dubiously notable, and even the arguments for notability seem to mostly point to things like tech blogs hosting such comparison articles.

The more I think about it, the more it seems to me that the concept of an article comparing products seems to be inherently unencyclopedic. We are not a product catalog, after all. Why is it that we say Listings to be avoided include, but are not limited to... products and services... Wikipedia is not a price comparison service to compare prices and availability of competing products or a single product from different vendors. But we are, apparently, a quasi-tech blog where one can see comparisons of the technical specs of phones or 3D printers or cars or what have you? The technical aspects of these things can always be discussed in the main article dedicated to them, perhaps even inclusing some limited comparative discussion; are standalone articles for this purpose really within scope?

Should we have a blanket ban on these kinds of articles? Athanelar (talk) 20:39, 14 April 2026 (UTC)

If the comparison of products is commonly done in reliable sources, including a discussion on those differences, such as the case for home video game consoles, that makes it fair for us to include. But if that comparison is just because we have specs for each and just putting them into a table, and RSes dont discuss these differences, that's likely beyond our scope. Masem (t) 21:08, 14 April 2026 (UTC)

My suggestion is that such comparison is out of scope even if it's being done in RSes. We don't necessarily need to include articles about everything that any RS writes about; we have carveouts everywhere for things like WP:CORPTRIV for example. I think product comparison articles should be out of scope per se regardless of whether someone else is already doing it. Athanelar (talk) 21:24, 14 April 2026 (UTC)

I'd argue that even if a comparison is covered in reliable sources, an article starting with "Comparison of" would be about the comparison and how they've historically been compared by others. Not comparing them ourselves. Thebiguglyalien (talk) 21:47, 14 April 2026 (UTC)

I agree. Actually listing the properties of products and performing the comparison comes under the "not a product catalog" rule, I'd think. But if there are sufficient sources for an article about the comparisons, not about the products, then that would be OK. I doubt many (or even any) would qualify, though. Mike Christie (talk - contribs - library) 23:29, 14 April 2026 (UTC)

I'm not a great fan of blanket bans (or, indeed, blanket anything) but I agree that most such articles are inherently unencyclopedic. Phil Bridger (talk) 21:11, 14 April 2026 (UTC)

I'd go a bit further than that; which such articles are encyclopedic, and which would meet the required SIGCOV element of discussing the topic itself? Ravenswing 21:59, 14 April 2026 (UTC)

Comparison of lions and tigers isn't... A sad day for anyone looking to invest in that market. Chaotic Enby (talk · contribs) 11:45, 15 April 2026 (UTC)

I strongly oppose any sort of blanket ban. Very often these comparison articles are the best way for readers to understand how two similar encyclopaedic subjects differ - a necessary part of providing complete encyclopaedic coverage of those topics. Thryduulf (talk) 22:06, 14 April 2026 (UTC)

As long as that comparison is done by RSes, or at least routinely done in RSes. If not RSes makes these comparisons, then that's OR for us to make them. Masem (t) 22:22, 14 April 2026 (UTC)

How is it be OR to list characteristics of products if characteristics are reliably sourced, as long as we don't judge one "better"? Hyperbolick (talk) 00:30, 15 April 2026 (UTC)

Which features are we picking to make a comparison? Just that choice itself can implicitly bias one of the products over another if we include or exclude certain features. Letting RSes show how they make comparisons removes that possible OR bias. Masem (t) 11:29, 15 April 2026 (UTC)

Comparisons are secondary source material, so that's partway to the WP:GNG right there. WhatamIdoing (talk) 00:45, 15 April 2026 (UTC)

Can you give some examples of good comparison articles? Orange sticker (talk) 07:14, 15 April 2026 (UTC)

Definitely not encyclopedic. Among other things, such articles would be ephemeral. Who is going to care a year later which printer some magazine or other rating service recommended? Donald Albury 23:19, 14 April 2026 (UTC)

I'm not sure that's such comparisons are ephemeral. Sure, "ExpertsRUs recommended X from 1994 to 1997" is unimportant, but I thought the point of the comparisons was less ephemeral.

For example, the Comparison of 3D printers compares these points:

• Year released
• Print technology
• Max build volume (mm)
• Max build volume (in)
• Min layer resolution (μm)
• Print speed (mm/s)
• Kit or assembled

We might agree or disagree on what's important, but none of that feels ephemeral to me. It sounds like stable, objective facts. WhatamIdoing (talk) 00:51, 15 April 2026 (UTC)

This can lead to bias by omission though, if criteria are left out it can be impossible for a reader to judge, say, which printer is the most efficient or high spec. But really, although "provider of retail advice editorial" isn't under WP:NOT I don't think it's our role to provide comparisons of consumer products. I'm less bothered about the articles that compare languages, road signs or alphabetic country code, but nearly every other "Comparison of" article is about a tech product. We're not CNET. Orange sticker (talk) 07:24, 15 April 2026 (UTC)

Which criteria are used in a comparison is not relevant to this discussion, the question here is only whether to have these sorts of comparisons at all. Detail about what is compared is something that can only be determined at the individual article level. Thryduulf (talk) 10:18, 15 April 2026 (UTC)

But the fact is there's nothing stopping editors cherry picking criteria and, by the nature of the article, using them as comparison indicators, if we allow these types of article. This is what's happened and why it's WP:OR specifically WP:SYNTH - though that's not to imply bad faith. It's just inevitable. Orange sticker (talk) 10:22, 15 April 2026 (UTC)

We do not and should not ban whole categories of articles because they might contain OR or SYNTH, we fix and/or delete specific articles that do contain OR or SYNTH. It is not at all inevitable. Thryduulf (talk) 10:46, 15 April 2026 (UTC)

I'm not arguing in favour of a carpet ban, but certainly a policy or guideline that discourages their creation and ensures no OR. Preventative measures are better than having to fix large volumes of problematic content which is the position that's been created now. Orange sticker (talk) 10:52, 15 April 2026 (UTC)

We don't need or want a policy or guideline that discourages the creation or adding of encyclopaedic articles or content, we already have policies and guidelines against creating or adding non-encyclopaedic articles/content, including original research. You haven't yet identified any content that is actually problematic, nor any evidence that there is a large volume of such that can't be handled by normal editing. Thryduulf (talk) 11:15, 15 April 2026 (UTC)

There's currently an ANI complaining about that an editor is nominating these at too large a volume, and those AfDs discuss the matter of OR and unencyclopedic content in them. This is a case where policy could be more explicit to save editors' time and improve the project. Orange sticker (talk) 12:37, 15 April 2026 (UTC)

Doing anything in “large volume”… be it creating articles, or nominating them for deletion… is always a negative behavioral issue. Our policies, guidelines and procedures don’t handle “large volume” well. Blueboar (talk) 21:06, 15 April 2026 (UTC)

there's nothing stopping editors cherry picking criteria ...except for our anti-cherry-picking policy, Wikipedia:Neutral point of view. WhatamIdoing (talk) 02:12, 16 April 2026 (UTC)

Exactly, by their very nature this type of article is not in keeping with WP:NPOV. They could be if they only relied on comparisons done by third parties, but there's very few examples of that. Orange sticker (talk) 06:58, 16 April 2026 (UTC)

I think there are a limited set of product comparison articles that are worthwhile: those that compare generations of a single product. Often these end up embedded in the products main page, but if there is enough text, it makes sense to break it out. Examples: List of iPhone models, List of Intel processors. Note they're labelled "lists", but the body is primarily comparison information. Sometimes the same product is produced by multiple companies: the comparison is still useful: Comparison of ARM processors. Note that I'd differentiate this from comparison of unrelated products with the same functions (e.g., Comparison of server-side web frameworks), which I wouldn't see as appropriate for Wikipedia. That said, there are some heterogenous categories where the comparison page appears useful, e.g., Comparison of programming languages. Note that the programming languages comparison is going over the various attributes that we'd normally use when describing a programming language, so it's not a marketing comparison, rather a look at the inherent characteristics of programming languages that differentiate one from another. — rsjaffe 🗣️ 02:15, 15 April 2026 (UTC)

Agree, I don't have a big problem with Comparison of Google Pixel smartphones in terms of WP:OR because there's been no curation by editors, it's an exhaustive list and the specs also seem comprehensive (much of the sourcing isn't independent though). However I don't see why "Comparison" is in the title. "List of Google Pixel smartphones" would be fine or "Technical specifications of Google Pixels" would be more accurate. Comparison implies we are telling readers which is better/worse, faster/slower, bigger/smaller etc. These articles just present the information for readers to use how they see fit. Orange sticker (talk) 07:28, 15 April 2026 (UTC)

Comparison implies we are telling readers which is better/worse, faster/slower, bigger/smaller etc. Why do you think that? With the exception of better/worse, why is it a problem if we tell readers objective facts about them? Thryduulf (talk) 09:49, 15 April 2026 (UTC)

It's just that laying out objective facts isn't making a comparison, so those articles should just be called lists. It stops being objective when the (rankable) categories of facts are curated and someone selects which criteria to use to describe a product. It's like me creating a "Comparison of UK Cities" article and using "Number of Beatles born there" as a ranked category :) Orange sticker (talk) 10:04, 15 April 2026 (UTC)

But we've got to make some "selection" of criteria, because there is eventually a hard limit of how much content can be put in a single page. WhatamIdoing (talk) 02:14, 16 April 2026 (UTC)

If product comparisons are to be banned (which imo they should not be, per my above comments) it will be necessary to define what a product is in this context. Obviously Comparison of word processor programs is (although given that it compares software currently underground active development with programs with a most recent release date in 1986 how much it is a "marketing" comparison is debatable) and Comparison of graphics file formats is not, but what about Comparison of text editors, Comparison of operating system kernels, Comparison of open source wireless drivers, Comparison of North American ski resorts or Comparison of orbital launch systems? Thryduulf (talk) 02:52, 15 April 2026 (UTC)

I'm less bothered about the articles that compare languages, road signs or alphabetic country code, but nearly every other "Comparison of" article is about a tech product. I categorised every article starting with a title starting "Comparison of"

More information Category, Count ...

Category	Count
Software (inc. operating systems, libraries)	185
Computer, smartphone, digital camera, etc hardware	37
Protocols, standards, encodings, formats, etc	33
Websites and online services	23
Languages, writing systems, accents, etc	33
Sports	15
Programming, scripting and markup languages	13
Medicine and biology topics	10
Vehicles	11
Other technology (e.g. battery types)	9
Religion topics	5
Other	34

Close

Even if everything in the first top two categories is a product (which they very much aren't) then they only account for 57% of comparison articles. Thryduulf (talk) 11:04, 15 April 2026 (UTC)

wikt:every other is usually taken to mean "every second" or "half of", so that number does check out. Some software might not count as products, but that might also be offset by some online services and other technologies, although we'll need more details to be sure. Chaotic Enby (talk · contribs) 11:47, 15 April 2026 (UTC)

It all goes back to my earlier question of what is a product? And even within things that are clearly products there is a big difference between e.g. Comparison of iOS devices, Comparison of DVD ripper software, Comparison of HTML parsers and Comparison of early word processors. I suppose it depends what the exact problem you have with these articles is, as all the reasons given above are vague gestures towards non-specifics and complaints that some articles might contain things that are already not allowed. Thryduulf (talk) 12:57, 15 April 2026 (UTC)

I don't personally have any problem with these articles, I was just hoping to clarify the terminology matter! Chaotic Enby (talk · contribs) 17:07, 15 April 2026 (UTC)

Adding to the point about defining products, when is a list a comparison? List of battery electric vehicles is not a "comparison of" but includes lengthy tables through which readers can ... compare various aspects of each. Is any table of [products] with more than their name a comparison? If it makes sense to include more than the name of something, is the key just not putting the information in a table? — Rhododendrites ^talk \\ 13:00, 15 April 2026 (UTC)

Making the the categories rankable makes the list a comparison because it implies some factors are > than others and means entries can be easily grouped into the haves and have nots. But it would be petty not to allow users to rank tables. My solution would be to ensure the title says "list", otherwise we risk putting into Wikivoice "product A ranks higher than product B according to Wikipedia's comparison". Orange sticker (talk) 13:09, 15 April 2026 (UTC)

Eh? You want to ban whole categories of articles because some people might choose to compare encyclopaedic information about encyclopaedic topics in way that might suggest something was better or worse than something else? This is screaming WP:IDONTLIKEIT and trying to police the way readers use our content. Thryduulf (talk) 13:21, 15 April 2026 (UTC)

Eh indeed. I've already replied to you directly saying I'm not arguing in favour of a carpet ban. I'm saying we shouldn't be making unsourced comparisons in Wikivoice, and I'm explicitly arguing that readers be allowed to read and manipulate the tables how they see fit. However, it's up to us to ensure WP:NPOV and no WP:OR and we're currently failing to do that. Orange sticker (talk) 13:26, 15 April 2026 (UTC)

What unsourced comparisons are being made? If product X has a sourced list of features, and product Y has a sourced list of features, neither become magically unsourced by placing them in a table. This proposal completely fails to demonstrate that product comparisons are a failure of NPOV or OR in and of themselves and completely fails to demonstrate that any such problems that do exist would be solved by adopting it. Thryduulf (talk) 13:42, 15 April 2026 (UTC)

If no independent third party expert has ever published a work that compares product X to product Y, we are introducing WP:OR. If an editor decides to create a table and chooses which criteria should be included for comparison and what is left out, that is WP:SYNTH. And beyond hypotheticals, a lot of the existing articles have vast amounts of unsourced info, e.g.: Comparison of genealogy software, Comparison of CRM systems, Comparison of DVD ripper software. Orange sticker (talk) 13:53, 15 April 2026 (UTC)

Absolutely none of this is a problem with the type of article itself, its a problem (and I'm not convinced on the OR or SYNTH claims) with the individual articles. Thryduulf (talk) 13:57, 15 April 2026 (UTC)

And yet its evident on so, so many of of them. I chose those three at randon from Category:Software comparisons and they all had glaring problems. It's clear allowing this type of article without any clear guidelines has lead to vast amounts of unsourced, original research being added to the project over the years. We have the opportunity here to improve Wikipedia. Orange sticker (talk) 14:01, 15 April 2026 (UTC)

Even if that is true, you still need to show that this will actually improve Wikipedia, and removing or discouraging whole swathes of encyclopaedic content is not an improvement. Thryduulf (talk) 14:02, 15 April 2026 (UTC)

In re If no independent third party expert has ever published a work that compares product X to product Y: And yet the complaints focus on tech subjects, for which comparing product X to product Y is the everyday normal form of the reliable sources. WhatamIdoing (talk) 02:17, 16 April 2026 (UTC)

In the three examples I give above, I can only see once source that comes close to being a third party comparison, and oddly it does not explicily support the information it is cited for (that Hubspot CRM Free is a SaaS). Orange sticker (talk) 07:03, 16 April 2026 (UTC)

Are you looking only at the sources that are already cited in the articles? WhatamIdoing (talk) 02:43, 17 April 2026 (UTC)

Yes, the point I'm making is that for several years these articles have been created without proper sourcing and its likely that will continue to happen without better guidelines. Orange sticker (talk) 07:16, 17 April 2026 (UTC)

Are you familiar with the WP:NEXIST guideline? WhatamIdoing (talk) 17:51, 17 April 2026 (UTC)

@Orange sticker, Does "Making the the categories rankable" mean "making it possible to sort the table by columns"? WhatamIdoing (talk) 02:16, 16 April 2026 (UTC)

Yes. I think on its own it's nothing but a handy feature for readers, but when you combine it with an article called "Comparison of..." I think it introduces WP:NPOV. Orange sticker (talk) 06:55, 16 April 2026 (UTC)

Thanks for explaining. I disagree. WhatamIdoing (talk) 02:39, 17 April 2026 (UTC)

I'm rather surprised to see that count being used to say that not many such articles are about tech products. It shows that about 70% of them are directly related to computers. It may surprise some Wikipedia editors to know, but the world is not made up of 70% computers and 30% everything else. The table shows just how bad the systemic bias is that's caused by young first-world males being vastly over-represented among us. Phil Bridger (talk) 21:50, 15 April 2026 (UTC)

Not all of the first two categories are directly related to computers, unless you take a very broad definition of "computer". For example graphical calculators and digital cameras are included in the hardware category, and not everything related to computers is a product, even by a broad definition. Tech products are probably the largest single thing, but they don't account for 70%. It's also worth stressing that this only considered articles named "Comparison of", not "list of", "table of" or anything else, even if there was a redirect from a "Comparison of" title. Thryduulf (talk) 22:02, 15 April 2026 (UTC)

My take is that they are important but appear to be WP:SYNTHESIS. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 14:11, 15 April 2026 (UTC)

We can compare topics frequently compared in reliable sources but that doesn't mean that all comparison-cruft has a place in this encyclopedia. As you point out, the choice of comparison-criteria itself should be reliably sourced to avoid WP:OR. A ban seems over-the-top and like rule-creep. However, it may help to articulate your arguments against comparison articles in an essay and cite that in relevant discussions. Joe vom Titan (talk) 00:24, 16 April 2026 (UTC)

I understand the scepticism of these tabular comparison articles but I don't think that they are intrinsically illegitimate and would oppose a blanket ban. In many cases the tabular comparison articles contain content that nobody would be complaining about (apart from its size) if it was included within the main article about the topic. So, if no blanket ban, how do we sort the comparison of sheep from the comparison of goats and then put them through the sheep dip to get them clean?

As I see it, the tabulated comparison articles are pretty mechanical supplementary content that exist to aid the understanding of existing coverage in other articles. They are more like indexes than articles. They should only contain data drawn from the article covering each entry. That would automatically exclude any entries without articles or (at the very least) substantial sections in other articles. It would take most of the burden of sourcing off of the comparison articles (because we don't need to source the same information twice) and also eliminate most of the concerns about SYNTH and cruft. SYNTH or POV could still creep in when deciding what columns to put in the tables but if we stick to what is already covered in the entries' articles then it can't get too far out of whack.

This does not apply to other types of comparison articles, which are prose based. Those are not mechanical. They are normal articles and need to be referenced to sources making the comparisons in order to avoid SYNTH or POV in the choice of what aspects are compared and in what ways. --DanielRigal (talk) 16:34, 16 April 2026 (UTC)

For the AfDs I have open, some are deleted or keep based on just a lottery as they are very similar. It seems there is no policy. For an external observer this looks bad and Elon/Trump fans will see this as Wikipedia decadence and argue that Grokipedia is the future. Please, have some coherent policies! Don't let Wikipedia die! Dncmartins (talk) 07:47, 17 April 2026 (UTC)

In my view, the lottery comes on account of the varied people who turn up at each AFD. AFD is a greater problem than "comparisons". Thincat (talk) 09:43, 17 April 2026 (UTC)

One path for Wikipedia to die is for it to delete information that readers want to find here. WhatamIdoing (talk) 17:55, 17 April 2026 (UTC)

I agree with you. That is why Wikipedia needs consistent policies. Dncmartins (talk) 18:43, 17 April 2026 (UTC)

I don't think that's relevant. A consistent policy of "delete everything" would lower Wikipedia's value for readers. An inconsistent policy of "keep things when we feel like they're cool or interesting or helpful" would raise it. An individual's actions of "send things to AFD to protest inconsistent policies" is WP:POINTY and unhelpful to readers.

If you could explain your concern, we might be able to identify inconsistencies (the first step in fixing them), or explain why they're not actually inconsistent. WhatamIdoing (talk) 01:24, 18 April 2026 (UTC)

• Seems like a way too satisfy everyone here is to WP:TNT all comparison articles, start over, and find whether they are misnamed, actually OK, or whatever other outcome could be. ~2026-22071-94 (talk) 19:13, 17 April 2026 (UTC)

Given the excessive effort of redoing the potentially valid ones from scratch, I don't think that's the best idea but if they were all to be draftified instead then that might work. --DanielRigal (talk) 04:07, 18 April 2026 (UTC)

Maybe after a spot check to see if the article is just not named correctly and should be renamed? ~2026-22071-94 (talk) 04:57, 18 April 2026 (UTC)

I think that would be to extreme a reaction. As DanielRigal has pointed out, it would take an excessive amount of time to restore the legitimate article, which I believe are probably a majority anyway. CommonsKiwi (talk) 13:38, 18 April 2026 (UTC)

I don't think a full out ban is necessary, but I do agree that the state of these articles are varying. There are some useful comparison articles and some unnecessary ones. EvanTech10 (talk) 02:01, 26 April 2026 (UTC)

I am starting the deletion process for all those unnecessary ones. Dncmartins (talk) 17:58, 26 April 2026 (UTC)

These articles are very useful and within scope. The only issue is that some may grow outdated. Sources should exist that look into the criteria in the table columns for the product type, not the entire comparison. People could use Wikidata more often to keep such articles up-to-date. Prototyperspective (talk) 14:42, 29 April 2026 (UTC)

I'm uncomfortable with banning all such articles. After reading through this discussion, I'm convinced that they're generally not good article topics (Donald Albury on ephemeral) and should be strongly discouraged (Orange Sticker on "certainly a policy or guideline that discourages their creation and ensures no OR"), but an outright ban seems unwise to me, primarily because there might be some false positives. Imagine that there's a topic that we all agree to be article-worthy, but it necessarily includes "Comparison" in the title, and maybe it includes a few lists that are definitely in its scope. Confused people might see the blanket ban, the title, and the lists, and seek to have the article deleted as violating the ban. Even if we didn't end up losing this useful article by accident, we'd end up with a lot of controversy (and maybe some deletions and undeletions) by the time it finally got kept. Better to say "these articles are generally not good" and to encourage AFDs, and we can get rid of them more slowly; they're not harming anything, so they can wait for AFDs. Nyttend (talk) 02:41, 20 April 2026 (UTC)

Before deleting "Comparison of X" articles, should consider renaming to "List of Y", where "Y" may be same as X or slightly different. • Sbmeirow • Talk • 22:57, 25 April 2026 (UTC)

An article that is not meeting Wikipedia policies will still not meet them even if its name is changed. Dncmartins (talk) 17:56, 26 April 2026 (UTC)

But, it seems, an article that meets our policies and guidelines, is far more likely to be nominated and deleted if its title starts "Comparison". Thincat (talk) 19:11, 26 April 2026 (UTC)

• Can everyone look at Comparison of crewed space vehicles? It was also nominated for deletion, but speedy closed. Listing every crewed space vehicle, linking to their articles, and listing basic information about them is something the encyclopedia should have. You can discuss on the talk page which things should be listed. In lists of software of a certain type, common sense, list the company that made it, a date, and what type of software it is. The articles linked to, always have this information in them. In List of adaptations of works by Stephen King the columns of information listed are just what is common sense. List of national flags of sovereign states, List of sovereign states, etc. Thousands of list articles exist that have columns of information. To prevent pointless nominations and arguments, can we just add a rule somewhere saying that columns of information in a list article are fine? Whether an article is kept or deleted always involves a small number of people that notice and comment, and the most determined and active people seem to be those who wish to delete things.

Also for many years now people have been nominating perfectly good list articles, then claiming they aren't notable, but they are usually kept anyway. Wikipedia:Notability#Stand-alone_lists says: There is no present consensus for how to assess the notability of more complex and cross-categorization lists (such as "Lists of X of Y") or what other criteria may justify the notability of stand-alone lists. Lists that fulfill recognized informational, navigation, or development purposes often are kept regardless of any demonstrated notability. We need to change that to say that list of things logically grouped together, that have their own Wikipedia articles, are fine. If its acceptable for a category, then its acceptable for a list, since lists show more information than a category does and are thus far more useful for finding what the viewer wishes to find. Dream Focus 15:55, 27 April 2026 (UTC)

• Phase out Wikipedia management, begin experiments with Wikidata management I support migrating our editorial process in this space to Wikidata. I want to keep these articles, but start pulling more of this information from Wikidata. Wikipedia quality control processes are designed for checking and verifying claims in prose. For articles like these
  • Comparison of web browsers
  • Comparison of Nikon 1 cameras
  • Comparison of commercial battery types
  • Comparison of Google Pixel smartphones

we do not now and never will have the human editorial capacity to keep this content fresh and fact-checked to the quality of our prose content. There are thousands of claims here and these are points of data for machines to manage, not for humans to review. I do not think we should have a quick migration to Wikidata but I do think that it is time for us to pick 1-3 of these articles and explore managing it as Wikidata content. Somehow, we need to have a way to easily fact-check tables with 100s of cells in Wikidata, then present the content in Wikipedia. Regarding the original proposal - I do not support a blanket ban as suggested because we have readers and editors who want this content, but also, our editorial community has quality expectations and our current editorial strategy in this space will not improve to meet community standards without something changing. Bluerasberry (talk) 16:35, 27 April 2026 (UTC)

• In general, I feel like these articles are inherently unencyclopedic. However, I have more concerns when the comparaison of products are across companies, rather than within a product line (so, List of iPhone and iPod Touch models is more acceptable than Comparison of ARM processors or Comparison of North American ski resorts). In any event, these articles should be primarily based on RS, not product or feature guides, and the coverage of the comparison should meet the expectation in WP:NCORP. --Enos733 (talk) 18:38, 27 April 2026 (UTC)
• Oppose relying on Wikidata, as verifiability there has often failed to be up to the same standards as on Wikipedia. We shouldn't risk sidestepping our verifiability requirements by porting data from Wikidata, and the need to have a way to easily fact-check tables with 100s of cells in Wikidata only pushes the same problem upstream. The data in these tables is already being presented in an organized format, and such a fact-checking tool could just as feasibly be developed on Wikipedia itself. Chaotic Enby (talk · contribs) 13:49, 28 April 2026 (UTC)

  The current problem with our articles is that they only have Wikidata-level verifiability. Though I also feel like making the goal Wikipedia presentation defeats the purpose and preserves the verifiability issue, I think this could be something to explore now that we have the Abstract Wikipedia... extremely buggy alpha. Aaron Liu (talk) 22:31, 28 April 2026 (UTC)

I think that many of these should be renamed as lists. All lists that include columned data inherently allow for comparison. The issue lies in the fact that so many of these are for software and products, and the individual items listed are not necessarily notable. For example, Comparison of web browsers could be merged with List of web browsers. Reywas92^Talk 19:02, 29 April 2026 (UTC)

That doesn't address the primary complaint that the selection of columns (of criteria to compare) is OriginalResearch/unDue. Aaron Liu (talk) 19:46, 29 April 2026 (UTC)

That's not what original research means. That's not even what undue means. This is a page-specific editorial decision for how much detail should be included on a particular topic and what sources should be used, not an overarching issue about the concept of listing things with information about them. I might agree that there's a lot of specific items in the web browers page, but that has little bearing on other lists. — Reywas92^Talk 20:09, 29 April 2026 (UTC)

It is. It might be verifiable/uncontroversial that a feature exists, but Due means that what is Verifiable should be covered in proportion to how much sources mention it. Without that, across like two-thirds of comparison articles, we are arbitrarily promoting specific features without sourcing to demonstrate their proportionate relevance, which is also the spirit behind Due.

How does this content policy extend to deletion? WikiNotability's purpose is to make sure it is possible for every page to conform to content policies, primarily Verifiability but also NPoV (which Due is a part of). Aaron Liu (talk) 23:18, 29 April 2026 (UTC)

I think my suggestion got lost in the noise. Part of what I suggested was that OR can be avoided by requiring that the tables only contain data drawn from, and referenced in, the articles about the items listed. --DanielRigal (talk) 22:50, 29 April 2026 (UTC)

I don't oppose that suggestion, although it sounds like a ton of work. It feels much easier to check just references (for the lists of columns) as opposed to watchlisting dozens of articles to see if a feature got removed for being unDue. Aaron Liu (talk) 23:07, 29 April 2026 (UTC)

I do not know if banning Comparison articles or not will be consistent with Wikipedia policies, but please, be consistent. If comparison articles like Wikipedia:Articles for deletion/Comparison of smartphone brands are removed while comparison articles like Wikipedia:Articles for deletion/Comparison of orbital launch systems are keep, there should be a very clear explanation for that incongruence. Please, do not archive this discussion without at least providing a very clear explanation. Editors are very confused right now, just take a look at the comments of editors in all the Wikipedia:Articles for deletion submissions I submitted. Dncmartins (talk) 09:08, 30 April 2026 (UTC)

There was no discussion at Wikipedia:Articles for deletion/Comparison of orbital launch systems, it was closed as a speedy keep by a non-admin and is now being reviewed at Wikipedia:Deletion_review/Log/2026_April_29#Comparison_of_orbital_launch_systems. It doesn't help us reach a conclusion if discussion is being shut down. Orange sticker (talk) 09:14, 30 April 2026 (UTC)

No participant other than you at that discussion disagreed that the vague rationale for a former featured list candidate met SK#1-absence of deletion rationale. Nominations should not be made without providing a very clear explanation. It was speedily kept because the nomination needed more explanation. Aaron Liu (talk) 14:23, 30 April 2026 (UTC)

Just because both articles start with 'Comparison of' does not make them equivalent. You keep claiming that the outcome of Wikipedia:Articles for deletion/Comparison of smartphone brands has created a precedent to delete all comparison articles, but this simply is not true. CommonsKiwi (talk) 09:32, 30 April 2026 (UTC)

It's honestly embarrassing that you can't tell the difference between your page, which was an undersourced table (mostly gsmarena.com) that just listed seven random features like "Install apps without account" and "Call recording" with a green yes or a red no, and a list that provided statistics, dates, and other data about notable rocket types with over 200 sources. Nothing is the same except the first two words of the title, and I'm concerned if that confuses you. — Reywas92^Talk 14:40, 30 April 2026 (UTC)

I think a reasonable compromise would be to require every column have at least two independent secondary RS-s supporting cells under it. The sources need not be comparison articles, so if two independent sources highlights a feature in their review, it could also mark the feature as important/Due enough.

We might need to start an actual proposal thread after this one—it's more of an idea lab thread right now. Aaron Liu (talk) 23:18, 29 April 2026 (UTC)

I agree that an RfC might be a useful next step; it seems relatively few people think all "Comparison of" articles ought to be treated identically, but I see a lack of consensus around how exactly our content policies should be applied to comparison articles. If nothing else, trying to make a concise list of the possible options might help the participants here clarify their viewpoints and arguments. ~ le 🌸 valyn (talk) 18:41, 30 April 2026 (UTC)

WP:XC says the following:

Membership in the extendedconfirmed group is typically revoked if a user is in another group with which it is redundant (for example, administrators)...

I think this is unhelpful, and I'd like to see this practice stop. I've been around for 20 years and have hundreds of thousands of edits, so I got the right automatically in 2023 (I assume this is when the right was created), some while after I lost admin rights for inactivity. I requested the restoration of admins rights in 2024, and in the process of restoring them, bureaucrat 28bytes complied with the policy and removed XC as redundant. This caused me a bit of confusion just now; see WP:VP/T#Template:Extended confirmed restriction wording. Now imagine that I lose admin uncontroversially in the future, either because of inactivity or resignation. Unless the bureaucrat remembers to give me this right, I'm suddenly unable to edit XC articles, so I have to go and request the right. Isn't this a bit absurd? Now, barring some egregious actions on my own part, obviously it would be granted, but it just seems inconvenient for everyone involved. Shouldn't we just ignore the redundancy, as we already do with rights such as rollback? Nyttend (talk) 02:50, 20 April 2026 (UTC)

Yes. Anything that stops editor time being wasted is a good thing. CheeseAndJamSamdwich (talk) 15:15, 21 April 2026 (UTC)

Of course that shouldn't happen. The current policy defies common sense. What's wrong with a bit of redundancy? Phil Bridger (talk) 16:58, 21 April 2026 (UTC)

• As someone who gets their sysop bit deactivated periodically (I think I hold the record for this, actually), it would be better to just leave the redundancy in so it doesn't have to be remembered each time. It isn't hurting anything by being redundant. Dennis Brown - 2¢ 04:35, 24 April 2026 (UTC)

  If you look at my rights log, you see that it was sometimes removed, sometimes not, when I regained the sysop bit. Barkeep49 kept the EC and IPBE active, which I would prefer both be kept, even though I have global IPBE, to make future changes simpler prevent any accidental leaving it off the next time I drop the sysop bit for a break. Dennis Brown - 2¢ 06:40, 25 April 2026 (UTC)

  I asked Barkeep about this case on his talk page. I don't know if we really need to do anything, but the current situation is that occasionally a 'crat will forget to add or remove the EC right. No big deal, but could theoretically be a minor inconvenience. Novem's bot idea would solve this, but whether it's worth it I don't know. Toadspike [Talk] 23:13, 25 April 2026 (UTC)

• I would just note that this is true of every permission that is included in the admin set (rollback, NPR, etc). As far as I am aware, though, ExCon is automatically restored when the admin bits are removed, and even then it is trivial for a bureaucrat to simply re-tick the ExCon box when removing the sysop bit. Not really "inconvenient". Primefac (talk) 08:47, 24 April 2026 (UTC)

  As far as I am aware, though, ExCon is automatically restored when the admin bits are removed. Can anyone confirm this? I wonder if the system only auto awards ac & xc once, before the person became an admin, to avoid edit warring with an intentional removal. –Novem Linguae (talk) 12:28, 24 April 2026 (UTC)

  Judging by the rights log for Firefangledfeathers, it doesn't appear to be automatic. However if the account became an admin before XC was created it does seem to grant automatically, at least per Smalljim's user rights log. 45dogs (they/them) (talk page) (contributions) 19:17, 24 April 2026 (UTC)

  The automatic granting is done only after they have made a subsequent edit; FFF has not edited since requesting a desysop. Primefac (talk) 19:29, 24 April 2026 (UTC)

  Automatic granting only happens if the group has never been removed from that user, specifically so as not to override an intentional removal as Novem Linguae describes. —Cryptic 09:37, 25 April 2026 (UTC)

  Good to know. I suppose the only time I've seen it is for that category of former admin. Primefac (talk) 10:15, 25 April 2026 (UTC)

• I’m not convinced by this proposal. One of the features of adminship is that it automatically bundles most of the other user rights. When you pass RFA you immediately become a page mover, a template editor, event coordinator and everything else, including extended confirmed. As such, I don’t think the rights you happened to hold beforehand are especially relevant, and we should avoid the notion that they should automatically define the set of rights you retain if you are later desysopped. By that point it may be years or decades later, and the circumstances may be very different. And if someone is desysopped for cause, it may or may not be appropriate for them to retain lesser rights, depending on the circumstances. IMHO it should be up to the crats who remove the bit, in discussion with the affected editor / Arbcom / the community where appropriate, to make that determination, rather than admins retaining a variable set of redundant rights underneath the sysop bit that silently reappear later. Now I realise XC is probably the least controversial of the additional rights, but the above logic still applies and I don't see a good reason to treat it separately. Cheers  — Amakuru (talk) 08:58, 24 April 2026 (UTC)

  I take the exact opposite view to Amakuru. If someone is desysopped for inactivity then all the other rights (that don't have their own activity requirements) should be automatically retained as they have done nothing wrong and it might give a slight psychological benefit to those who return (it certainly won't do the opposite).

  For those who are desysopped voluntarily the standard is already to give them everything they had prior to being an admin unless they ask otherwise. This proposal would therefore have absolutely no impact on who holds what permission but means less work for the crats and fewer log entries.

  That leaves only those who are desysopped for cause. That is the best time to determine what other rights (if any) should also removed, as the context is known to everyone relevant. So I fully support the proposal. Thryduulf (talk) 09:23, 24 April 2026 (UTC)

  Well, I suppose you could add all the rights explicitly and as standard whenever anyone's made an admin, and the concrete effect would be similar. What I don't support is the idea that editors retain whatever motley collection of rights they previously held, and that is then taken as the default state on desysop. So if the approach were to make all bundled rights explicitly standard for all admins, rather than dependent on what they happened to hold beforehand, I’d be less concerned. I'd still lean to retaining the current system though. The key point is that adding these permissions is redundant, and we strive to avoid redundancy because it causes confusion. The proposal seems to be predicated on something that WP:AINTBROKE, particularly given Primefac’s point that, in practice, there appears to be little or no inconvenience.  — Amakuru (talk) 11:44, 24 April 2026 (UTC)

  Personally, I don't think it's a good idea to add administrators to every user group corresponding to the rights bundled into the sysop user group. It would result in a make-work project to assign all administrators to a new user group whenever a new right is bundled into the sysop user group. isaacl (talk) 17:37, 24 April 2026 (UTC)

  isaacl and Amakuru, you seem to misunderstand me. I'm not asking for admins to get additional user rights that would be redundant to our current rights; we don't need someone to go and give XC and rollback to all current admins. I'm asking that when an account is given admin rights, its pre-existing Extended Confirmed status (and, I suppose, other rights) be retained. This would take no additional work — probably a little less, since the bureaucrat wouldn't have to tick the XC box to remove that right — and would avoid the problem of a longstanding editor later being unable to perform certain actions, should the admin right be removed and the bureaucrat forget to restore all previous rights. Let's say I resign my admin rights tomorrow, and the bureaucrat forgets to give me extended confirmed. How is it a case of AINTBROKE that I can no longer edit pages that anyone with 500+ edits and 30 days' tenure can automatically edit, unless I go and request those rights? Nyttend (talk) 19:59, 24 April 2026 (UTC)

  Yes, I understood your proposal. I was responding to Amakuru's suggested alternative of assigning administrators to all the corresponding user groups. isaacl (talk) 03:02, 25 April 2026 (UTC)

  Ah okay, sorry I misunderstood you. Nyttend (talk) 04:57, 25 April 2026 (UTC)

If the goal here is to make sure that people who resign as administrators get their extended confirmed user group back, then maybe writing a bot to do this automatically would be the cleanest solution. Let me know if folks would be okay with this. Someone could use a positive consensus here to file a brfa for an admin bot. –Novem Linguae (talk) 12:54, 25 April 2026 (UTC)

The other option is just to remind 'crats that ExCon should be re-added when removing the sysop bit. Other than FFF (who as mentioned hasn't edited since requesting a desysop) I can't think of anyone who had to ask for the perm later on down the road. Per the NOTBROKE comment above, we seem to be dealing with this "issue" fairly well (but reminders never hurt). Primefac (talk) 12:59, 25 April 2026 (UTC)

I wonder if we ran a quarry for the number of former admins who are not extended confirmed and who are not blocked, if there'd be a couple more than just FFF. Will run this quarry when I'm back at a computer. –Novem Linguae (talk) 13:06, 25 April 2026 (UTC)

On it. There are 13, including two indefinitely blocked sitewide, three that have had RTV renames, at least three that are deceased, and one bot account. —Cryptic 13:25, 25 April 2026 (UTC)

There are four editors that do not fit into the above categories. Two (Jack and Jon) disappeared and were desysopped, one stopped editing four edits later, and one lost access to their account, had their mop transferred to another account, then lost it for inactivity, but has not hit the EC threshold on the newer account to get it automatically granted. Primefac (talk) 14:56, 25 April 2026 (UTC)

Novem Linguae, the goal here is to end the practice of removing this right. Your idea wouldn't hurt, but still it requires a bit of work, and maintenance too; if we stop saying that this should be removed upon granting admin, it won't have to be bothered with, at least with new admins. Nyttend (talk) 20:30, 25 April 2026 (UTC)

• I think the simplest option here is to simply add an instruction to WP:INACTIVITY that says crats will automatically grant EC to sysops desysopped via inactivity. For resignations, admins are already free to request rights (and typically are granted by default) so not much else should be done there I reckon. I'm still not sure where this lands us on WP:AINTBROKE but this seems to be... easy enough and not posing much danger to the wiki if implemented. --qedk (t 愛 c) 22:02, 25 April 2026 (UTC)

  I think that would be optimal as well, but would also add resignations to that. If an admin is desysopped solely due to inactivity or resignation, that's not because of any fault of their own (well, not any on-wiki fault, anyway). In that case, they should at the very least get XC unless they specifically asked not to have that. Otherwise, anyone who was made admin in the last 10 years will most likely have had XC beforehand, and would've had it revoked upon passing RfA, which means they won't automatically have XC granted to them when they're desysopped. – Epicgenius (talk) 23:20, 26 April 2026 (UTC)

• There's an additional issue here that I've been meaning to raise for ages, ever since I wrote Help:-show classes, which is that many checks for whether someone's extendedconfirmed look for the group, not for the permission. So for instance when you add <span class="extendedconfirmed-show">...</span> to something that you want to only be visible to people with the extendedconfirmed permission, what you actually do is make it visible only to people with the "extended confirmed" group, which excludes admins (and bots, and potentially other groups in the future). This can be fixed by adding sysop-show, but many people don't realize to do this, and for the inverse class, nonextendedconfirmed-show, there is no solution; messages intended for non-EC users like {{ECR}} just have to have a note telling admins to ignore them.
  These aren't massive downsides, but they add on to the downside of 'crats forgetting to re-EC ex-admins, while the only upside is things looking marginally tidier on the backend. The redundancy of 811 admins also being EC wouldn't hurt anyone, wouldn't add any maintenance burden, and would be consistent with the fact that admins remain autoconfirmed (although that's admittedly because the groups are implemented slightly differently). We should remove the quoted line from WP:XC and add the group to all current admins. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 00:52, 27 April 2026 (UTC)

  This begs the question, why is EC even part of the sysop tool kit? Has there ever been a sysop elected without being EC? So someone thought "oh, we should make this part of the kits of tools for sysop, as an added bonus..."? The inclusion of EC to the sysop kit seems to be what is redundant here. Dennis Brown - 2¢ 02:29, 27 April 2026 (UTC)

  At this wiki, yes. WhatamIdoing (talk) 03:37, 27 April 2026 (UTC)

  It's tricky because the MediaWiki software sees extendedconfimred as both a user group and a user right. Special:UserGroupRights goes into a bit more detail -- the column on the left is user group, and the column on the right is user rights. The idea behind removing the extendedconfimred user group from admins is that they already have the user right, so what's the point of having them in the user group. But Tamzin is saying that there's some code that checks for the user group rather than the user right, so maybe removing the user group isn't such a good idea after all. –Novem Linguae (talk) 10:46, 27 April 2026 (UTC)

  Pedantically speaking, yes, everyone who went through RFA before extendedconfirmed existed. A little less so, a few of the earliest RFAs were for users with fewer than 500 edits at the time, and I'm sure - though I haven't checked them - some of the previous mailing list appointments were too. And there are currently four admin accounts with under 500: one reserved username, two bots, and one WMF account. —Cryptic 12:02, 27 April 2026 (UTC)

  Why can't this be fixed by copying the rules in Mediawiki:Group-extendedconfirmed.css into Mediawiki:Group-sysop.css? Are there any cases where we actually want to check for the group but not the permission? —Cryptic 11:31, 27 April 2026 (UTC)

  This of course is only a hypothetical, but it could open to some kind of malicious behaviour where someone could add for example defaming text in a long article with the nonextendedconfirmed-show class, that now could not be identified by pretty much anyone. And would only be seen by readers. Squawk7700 (talk) 12:17, 30 April 2026 (UTC)

  A stronger version of that exploit is already possible. Now is an opportunity for anyone with a temp account to look very smart by looking at the source code of the comment and seeing why you can see this sentence but registered users can't. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 12:56, 30 April 2026 (UTC)

  Oh well then I have nothing more to say Squawk7700 (talk) 13:03, 30 April 2026 (UTC)

  Ah yes, <span class=anon-show>~2026-21916-69 (talk) 14:52, 30 April 2026 (UTC)

  This wouldn't affect diffs, though, so you'd need a lot of excon users to not view the diff and then miss a) the span, and b) the vandalism. Primefac (talk) 21:12, 30 April 2026 (UTC)

  And the addition of that span would be trivial to track via an edit filter. Thryduulf (talk) 23:37, 30 April 2026 (UTC)

• Meh, as someone who actually deals with this - I can say that dealing with it is not onerous at all, it basically only affects crats when already updating the page where this is used. Also, I don't think we need to go manually GIVE it to all the admins who are exempt from autopromoteonce for the sake of some css hiding hacks. — xaosflux ^Talk 00:28, 1 May 2026 (UTC)

I was asking for help (Question from The goat with a horn on his head (17:27, 14 April 2026))on how to find sources on not a well known subject and I got sent to this article and it looked like an idea long lost. Can I turn this into a policy, however the system works? Thanks The goat with a horn on his head (talk) 00:02, 30 April 2026 (UTC)

The page you link is indeed a very helpful guide about how to deal with offline sources! I don't think the core idea (that offline sources are just as valuable) is long lost, as I regularly see them in good and featured articles! Our policy on verifiability has a passage that mentions this and explicitly links to this essay, so you're certainly on the right path. Not sure if it should be turned into a policy, as explanatory essays are just as valuable and can usually go into more detail, as they complement the policy "basics" to provide more concrete guidance.

We also have quite a few tools to help editors acquire offline sources, such as Wikipedia:WikiProject Resource Exchange/Resource Request and my favorite: the Wikipedia:Resource support pilot spearheaded by the amazing @RAdimer-WMF to buy books to editors in need! Chaotic Enby (talk · contribs) 00:57, 30 April 2026 (UTC)

OK, your point is good, but does the tools you mentioned are mentioner on either of the articles? The goat with a horn on his head (talk) 15:16, 30 April 2026 (UTC)

One was already linked at Wikipedia:Offline sources, I just added the other! Chaotic Enby (talk · contribs) 15:23, 30 April 2026 (UTC)

Thanks for the help. The goat with a horn on his head (talk) 15:55, 30 April 2026 (UTC)

Suicide prevention, which the current hatnote of the Suicide article, gets around 235 page views per day, Suicide methods on the other hand, gets around 9,269 page views per day.

So viewers are far more likely to be interested in methods than prevention.

That means from:

To:

TyphoonHurricaneCyclone 12:11, 1 May 2026 (UTC)

That hatnote is there after an RFC in 2019 settled on that as the best balance between WP:No disclaimers in articles and some people's desire to add an "if you're considering suicide, seek help" disclaimer to that article. You (or others reading this) might also find Wikipedia:Perennial proposals#Proposals related to suicide and self-harm useful reading before starting another big debate on the topic. Anomie⚔ 12:24, 1 May 2026 (UTC)

This is an exception, established by consensus, to the standard hatnote usage outlined in WP:HATNOTE. The underlying principle of hatnotes is to help readers with navigation. In nearly every situation this means basing hatnote content off what readers are likely to think they're getting to when they search for some term. This is a somewhat unique situation where that rule of thumb is not the best way to serve the underlying principle of helping readers navigate.

At least that's a formal policies-and-guidelines justification/explanation of the consensus. A more straightforward of putting it is that consensus favours having the hatnote potentially prevent suicide over having it potentially facilitate suicide. ⹃Maltazarian ^{ᚾparley${\displaystyle \lor }$investigateᛅ} 16:46, 1 May 2026 (UTC)