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

Intro and !votes

Should there be Foo (disambiguation) redirects that target non-disambiguation pages that contain a list of articles, for instance where Foo is a ((Surname)) page, or ((Set index article)), or a list article? Shhhnotsoloud (talk) 07:01, 26 June 2018 (UTC)

• No. Such pages should not be created, and if they exist then they should be deleted unless there is a page history to preserve. Such redirects are technically incorrect and infer that the target is a disambiguation page when it is not (and therefore is not subject to dab page restrictions such as references and redlinks unsupported by bluelinks). Even if the base name page has a similar function to a disambiguation page, a (disambiguation) redirect would be an orphan, and a user is only ever going to arrive at it via the search box (in which case if it wasn't there the user would navigate to the base name), or by some sort of direct navigation, for some reason deliberately trying to avoid the base name, which would result in "Wikipedia does not have an article with this exact name." and advice to use Search. There is a balance between avoiding those (I would say unlikely) occurences versus creating/maintaining thousands of INTDAB links to non-dab pages. Shhhnotsoloud (talk) 07:31, 26 June 2018 (UTC)

@Shhhnotsoloud: I realise that you're playing devil's advocate here, but are you "proposing" just that we tolerate existing redirects, that we create a full set with no further changes, or that we divert all legitimate links via them, as we do with dab pages, to aid detection of poor wikilinks such as She beat Murphy in the final or He served on HMS Abdiel? Certes (talk) 10:46, 26 June 2018 (UTC)

Hello @Certes: don't read too much between the lines. I'm just trying to see if there can be a consensus that leads to a change in the guidance, to stop repeated discussions at RfD. If the answer is Yes then new redirects could/should be created; if the answer is No then existing redirects could be deleted via RfD unless there is a reason for an exception to guidance; neither answer infers mandatory routing. Shhhnotsoloud (talk) 18:44, 26 June 2018 (UTC)

• No. Such pages should not be created and should be deleted where they exist. If they have article history to preserve, that history should be preserved under an appropriate title (i.e., not "XXX (disambiguation)"). SIAs (including anthroponymy lists) are valid link targets. Bad links to them should be detected as with bad links to any other article. -- JHunterJ (talk) 12:13, 26 June 2018 (UTC)
• No. Per the 2017 comments of Widefox at a relevant RFD, such redirects violate the principle of least astonishment in user interface design. Users will be surprised at being redirected to a non-dab page when the redirect is has "(disambiguation)" in the title. —hike395 (talk) 13:49, 26 June 2018 (UTC)
  • Do you have any evidence that anyone looking for a page that disambiguates X has ever been surprised at landing at a set index of X? I certainly never have! Indeed every single time I've found exactly what I was looking for. Thryduulf (talk) 23:34, 26 June 2018 (UTC)

That would require a user study, which has never been the standard of proof necessary for user interface decisions at Wikipedia. —hike395 (talk) 06:42, 27 June 2018 (UTC)

Is a user study really beyong our reach? Let's ask a knowledgable editor about what steps need to be taken so that for a limited period of time all (or some) wikipedia users who follow a certain kind of redirect will be asked, when arriving at the target, to answer a short questionnaire. – Uanfala (talk) 19:12, 28 June 2018 (UTC)

@Uanfala: this isn't an "easy" type of thing to do, and even more so if you want it to apply to "readers" as well (what about the ever popular mobile-only readers including mobileapp-users?). — xaosflux ^Talk 21:51, 28 June 2018 (UTC)

Only on an opt-in basis, kind of like WP:FRS. — SMcCandlish ☏ ¢ 😼 20:07, 28 June 2018 (UTC)

So why are you asserting without evidence that such redirects violate the principle of surprise, particularly when there is much anecdotal evidence to the contrary? Thryduulf (talk) 10:11, 27 June 2018 (UTC)

Thryduulf, I feel like you are bludgeoning the discussion. From reading the comments, I think that JHunterJ, Widefox, Tavix and I all believe that the difference between a dab page and a set index article is evident to readers as well as editors, because of the difference in formatting and content. You (and perhaps SMcCandlish)) believe otherwise. This is an inference about readers' mind state, and hence will be difficult to prove or disprove without user studies. We cannot satisfy your demands for evidence. Is there a compromise that can be worked out? —hike395 (talk) 02:58, 28 June 2018 (UTC)

That's tautological and meaningless. Of course if the format is different, readers can see a difference. They didn't have their eyes evaporate! But it has nothing to do with the substance of the matter: If you put in "Foo (disambiguation)" looking for people surnamed "Foo", and we have a name list article of notable people by the name, that's functionally identical at least in that aspect to a DAB page and it is the page you want. It's "user-hateful" for WP to spit "Wikipedia does not have an article with this exact name. Please search for ..." at you – to effectively lie to you and tell you that WP doesn't have what you're looking for. Thryduulf is asking the same questions of multiple parties because they're all making the same weak "only thinking about the editor side and not about readers" points, and all refusing to address the substantive one. — SMcCandlish ☏ ¢ 😼 03:17, 28 June 2018 (UTC)

User:SMcCandlish, when we have both a dab and an SIA, would you propose redirects with "..(disambiguation)" targeting both, or equally valid to target either? (depending on the primary topic) Does that common case, for example with surnames, illustrate just one unintended consequence of extending to SIAs? Widefox; talk 08:46, 28 June 2018 (UTC)

One redir can only point to one article (unless you make some kind of ugly, manual soft-redir page), so it would naturally go to the real DAB page, which would also list the SIA page in it. We're only talking about "(disambiguation)" pages with no actual DAB to point to, but which could usefully go to an SIA or 'nymy page, at least temporarily. — SMcCandlish ☏ ¢ 😼 10:52, 28 June 2018 (UTC)

When you say "naturally", no, it's not clear what your proposal involves apart from a general desire to extend redirects to pages that don't have a need for such a redirect. I'd use the term unnatural (or WP:ASTONISH) for a "xxx (disambiguation)" redirect to target anything other than a disambiguation page, and the term cargo cult for the concept. To help tease out, I've broken down the common scenarios of SIAs and dabs below, so we know more of what you're proposing. Widefox; talk 15:06, 28 June 2018 (UTC)

Continual denialism of the obvious, on which you've been called out repeatedly, is not going to make anything less obvious or change anyone's mind. Please stop. — SMcCandlish ☏ ¢ 😼 15:46, 29 June 2018 (UTC)

I'm not sure what you consider obvious (or have any idea what I've been "called out repeatedly" on, but if you explain I will reply), ...but belief isn't as convincing as reasoning. Let me be clear about the status quo - the burden is on those who wish to convince others of a change to the consensus - SIAs are not dabs. How you treat others just because they disagree with you is your choice. Graham's hierarchy useful for a more convincing argument? Widefox; talk 21:11, 29 June 2018 (UTC)

I'm not getting frustrated with you because your opinion differs. It's because you're playing "I can't hear you" games and turning the discussion endlessly circular. It's a waste of time. Both Thryduulf and I have been over every detail of this in multiple forms, and will not be goaded further into recycling the same material (well, I guess I can only speak for myself, but am firmly predicting). Graham's doesn't apply here; there's no convincing to do, because your mind's already made up and you're transparently filibustering with argumentum ad nauseam in the face of every argument presented to you which never address and just pretend you didn't see it. — SMcCandlish ☏ ¢ 😼 07:16, 1 July 2018 (UTC)

• No. Per reasoning above. Has the question been asked "why should they be created?" Yes, there's a reason to create them targeting dabs - to eliminate intentional incoming (say INTDAB) links from being flagged up as needing fixing. Any extension of that to other types of pages needs justification. A counter argument (ie keep old ones, irrespective of target or history) has come up at CSD (both for G6, and G8) based on one editor's desire to use the vestigial redirects to avoid primary topics or some such argument (which I confess I don't understand further than conservatism. That counter argument has made deletion controversial hence non eligible for CSD, although I believe the advocacy of a navigation hack shouldn't prevent previously uncontroversial maintenance after broader discussion/consensus.) Widefox; talk 15:15, 26 June 2018 (UTC)

  I note that Patar knight has expressed a view supporting the existence of these redirects, and want to be sure that they are aware of this RFC in case they wish to explain their position. --R'n'B (call me Russ) 17:04, 26 June 2018 (UTC)

• Yes, always for set index pages, allow but do not require for surname pages. People extensively use these links to find articles when they know or suspect that the article they are searching for is not the primary topic but do not know how it has been disambiguated. These links enable them to quickly find what they are looking for without being required to know whether we have chosen to call our list of articles a disambiguation page or a set index (the distinction is lost on almost everybody who is not a member of the disambiguation project) and without first going via a page they know they do not want (for people on slow and/or expensive connections this is a serious consideration). I'm a Wikimedian of nearly 15 years standing and I cannot predict which a given title will have, so we cannot assume readers will know or understand. I agree this was not the original purpose of (disambiguation) redirects, but it is an emergent feature of them and removing it will make things harder for the reader and thus harm the encyclopaedia. Thryduulf (talk) 21:05, 26 June 2018 (UTC)

  If the title of a set index needs disambiguation as well as set indexing, both pages should exist, not be redirected to the other. Set index articles don't follow disambiguation style for a reason: disambiguation pages are navigational pages, not articles. Set index are list articles. If the set index is in the way of navigation, use navigational tools to address the problem, rather than hindering the readers needed navigation by making them use a list article for it. -- JHunterJ (talk) 12:47, 27 June 2018 (UTC)

  Another straw man - if a dab page exists at Foo (disambiguation) nobody is arguing that it should be replaced with a redirect. If the base title is a dab page and there is also a set index for that title then they should both link to each other and the (disambiguation) redirect should point to whichever is the more appropriate (most likely the disambiguation page, but there might be exceptions). The style and formatting of dab pages vs set indexes is, I repeat, completely irrelevant to this discussion - they are both pages from which people navigate to other articles with similar titles, which is exactly what someone searching for a (disambiguation) title is looking for. Thryduulf (talk) 22:46, 27 June 2018 (UTC)

  Yep. All this "they're formatted differently" and "they're classified differently" and "we wrote them for different reasons" argument is the tail wagging the dog. Those are editor-facing arguments that don't have anything to do with the reader experience. — SMcCandlish ☏ ¢ 😼 23:25, 27 June 2018 (UTC)

  Yes SMcCandlish, the issue is caused by the seemingly inconsistent duality in the current definition of SIAs - should they be linked to or not? Are they for navigation or not? Are they for readers or projects? References or not? They exist in this murky world between lists and dabs. Treating them as dabs with regard to extending dab redirects logically follows the presumption that they shouldn't be linked to. (I presume we all agree that linking to surname articles is OK, as the counter reality.) It would be better to fix the definition than all of us guess and extend presuming things which are incompatible with one of the dualities. Of course, we can ignore that and continue extending SIAs increasing the mess! Widefox; talk 09:00, 28 June 2018 (UTC)

  Whether a SIA is and/or should be linked to from articles is completely independent of whether (disambiguation) links should point to them - the redirects exist for the benefit of readers, allowing them to find the page they are looking for, whether or not they are linked to anywhere (which is normal for redirects, there is nothing special about these particular ones). Thryduulf (talk) 09:30, 28 June 2018 (UTC)

  My understanding is that they came into existence to allow indirect linking to dabs. Please correct me if I'm wrong, as that fact should be easily verifiable. That doesn't denigrate any unintended usage, but let's be clear about the facts. Widefox; talk 09:45, 28 June 2018 (UTC)

  Yes, as I've noted elsewhere in this discussion, that was the original purpose of the (disambiguation) redirects. However they have since gained an additional purpose that is at least as valuable (for readers, arguably more so). Those who wish to restrict these redirects to only pages that meet the disambiguation page formatting rules are ignoring (or dismissing) this emergent use. Thryduulf (talk) 11:12, 28 June 2018 (UTC)

  Thryduulf, when you say 1. "completely independent", but agree 2. that's the reason they exist for dabs. How can the two logically be true? What is the evidence of demand for this use case? (as I've pointed out, we're missing dab redirects that nobody seems to miss, indicating this isn't a big thing, and of course it's a hack, so hacks are great while they work but it's a big change to have hacks as supported features). Widefox; talk 13:34, 28 June 2018 (UTC)

  (edit conflict) You mean evidence other than the fact that I use them in this manner, that others commenting when they've come to RfD say they use and page view statistics showing that they get used? Not every single one is necessarily going to be used, but why does that matter? If a redirect exists when somebody tries to use it then the person using it wins and the encyclopaedia wins because someone has found what they are looking for and nobody loses. If a redirect exists and nobody uses it nobody loses. Your description of this as a "hack" turning into a "supported feature" over-eggs the pudding about the effort required - supporting them (letting them exist) actually requires less effort than not (discussing and deleting them). Thryduulf (talk) 13:52, 28 June 2018 (UTC)

  re 1 and 2 both being true: It's very simple - just because (disambiguation) redirects to disambiguation pages were created for one reason doesn't mean that (disambiguation) redirects to set indexes need to be used for that same reason. It's possible to use these redirects as search targets completely independently of whether they are linked. Thryduulf (talk) 13:57, 28 June 2018 (UTC)

  I'm aware from previous discussions that you strongly believe in this, being something you personally use. The above is, however, anecdotal evidence of demand. Anything stronger? Forgive me if you read "hack" to be pejorative, it's not meant so, it's worse than that - it's the cold/hard correct term in the industry, and officially supporting hacks can be a costly mistake. Widefox; talk 21:31, 29 June 2018 (UTC)

  Look at pretty much any time these redirects have been nominated at RfD and you'll find people saying they find them useful (which is an explicit reason not to delete a redirect) and there is a ton of hard data that shows only a fraction of readers edit the encyclopaedia at all (for all its downsides the article feedback tool generated many comments from people who new how to navigate Wikipedia but had never used a talk page, let alone contributed to an XfD), if you look at the page view statistics for many, maybe even most, of these redirects you will find clear evidence of human use. Whether this is anecdotal or not is debatable, but it contrasts with the complete lack of any evidence (anecdotal or otherwise) that they are in any way shape or form harmful. Thryduulf (talk) 14:40, 30 June 2018 (UTC)

  A) Isn't that's an WP:OTHERSTUFF argument? B) editors at RfD need to take this broader RfC into consideration, clearly C) My definition of hard data for this proposal is, for example, %'s, click through rates, reader numbers, user surveys, use cases, usability studies, A/B testing -> none of this type of hard data has been presented on this proposal here or the handful of RfDs I've seen. There's not one of those, is there? D) In fairness, absence of evidence isn't evidence of absence, but so far as presented this is thoroughly unconvincing. Widefox; talk 21:09, 30 June 2018 (UTC)

  A) No it is not an OTHERSTUFF argument because it is using the information presented in multiple discussions about multiple individual redirects of this type to provide evidence in a discussion about the class of redirects of this type. B) that would be impossible as those RfD discussions happened before this RfC - and this RfC is not (at least yet) reflective of a wide consensus as it the only participants have been RfD regulars and those deeply involved the existing guideline. C and D) Firstly this is massively over the top, as has already been explained, use cases have been repeatedly explained, click-through rates are not relevant as that would be a measure of the utility of the target not of the redirect, I can't conceive of how A/B testing would even be possible (a redirect either exists for everybody or for nobody), and finally you are the one who is asserting these redirects are harmful and you need to provide evidence of that (which you and others have repeatedly failed to even attempt). Thryduulf (talk) 10:17, 6 July 2018 (UTC)

  Thryduulf Consensus is that SIAs are not dabs per Wikipedia:SIANOTDAB. The burden is actually on anyone that wants to overturn that consensus. A redirect that has the parenthesis to something that it is explicitly not is WP:ASTONISH. I don't have to provide evidence of that, as I'm not attempting to change consensus. I have, however, been clear about that harm, so let me repeat: UI consistency harm, so harm for readers. Any proponent of using them to target SIAs needs to at least consider there's a cost, and there's no consensus on the benefit. Most/all these are redirects left after conversion of a dab to an SIA, so they have ceased having functional purpose per guideline. In that context, more pertinent than "harm" is "vestigial" (aside: I once had the pleasure to be at a lecture of Stephen Jay Gould, where alongside vestigiality, he showed many photos of elements of European cathedrals (spandrels), so that would make dabs spandrels in this analogy, the lecture comes to mind each time we discuss this). Widefox; talk 00:08, 30 August 2018 (UTC)

  Such redirects aren't always vestigial: I've sometimes seen editors creating such redirects, at least to surname SIAs. But it's probably true that most of the redirects were originally targeted at dabs – but then these are often the situations where one editor has converted the dab to a SIA and another editor is likely to be found who believes that this SIA ought to be converted back into a dab. So I wouldn't want to caryy the DAB/SIA distinction so far up the hill here. – Uanfala (talk) 00:34, 30 August 2018 (UTC)

  @Widefox: As I'm sure has been pointed out many times elsewhere in this debate, that SIAs are not disambiguation pages is a technical distinction that very few readers (and not all editors) know about, even fewer care about and is not at all relevant to the question at hand. I'm also not attempting to change anything about the distinction (because it's not relevant here, as repeatedly explained). I know I have definitely explained several times why it is not astonishing in the slightest (and nobody has ever provided evidence to the contrary, despite repeated requests) to search for Foo (disambiguation) and arrive at an SIA because what someone is looking for is a list of articles related to the title "Foo", either for the list itself or because they know what they are looking for is (probably) not the primary topic but don't know what the article is titled. Thryduulf (talk) 07:26, 30 August 2018 (UTC)

  @Uanfala: agree, I said most. I've flipped some myself. I'm sure we all know the raison d'etre of the redirects is to enable the mandatory linking via "(disambiguation)" to dabs. As we allow direct linking to SIAs, it's vestigial by definition when targeting an SIA. Any secondary use doesn't have consensus, but has been thin end of the wedged by hanging on to these vestigial ones. The big picture is 1. many (tens?) thousands of dabs are missing a redirect, which would be a bigger gain for proponents of the secondary use, and I see no desire to fix that low hanging fruit despite this being pointed out, 2. many (majority?) SIAs cannot have such a redirect anyhow, as there's an existing dab. This is quite perverse considering. Widefox; talk 12:15, 30 August 2018 (UTC)

  @Thryduulf: as others have said, it's not that I'm not listening, I'm not convinced. If one rejects the consensus that SIAs are not to be treated as dabs (and downplays it as a "technical distinction", and in the absence of consensus to treat them equally with respect to creation of (disambiguation) redirects (and given that dabs need the redirects to enforce linking via them, SIAs can be directly linked to), then one is left preserving vestigial redirects, which isn't a useful state of play IMHO, it's somewhat POINTY considering the low hanging fruit of creating thousands of missing redirects to existing dabs that would serve your use case better. Initially you were under the false impression they are bot created, which I've attempted to inform you many are still missing. I will assist if you wish to take up extending/fixing the bot so these are automatically created on all the missing dabs. Widefox; talk 12:15, 30 August 2018 (UTC)

• @Widefox: Again with the straw men. I'm not rejecting the consensus that the SIAs and Dabs are not the same thing (it's completely irrelevant to this argument and I don't have a strong opinion about it). I'm not rejecting the consensus that dabs should be linked to only via (disambiguation) redirects (I agree with it). I'm not rejecting the consensus that SIAs can be linked to directly (I agree with it). I would like to see a bot create INTDABLINK redirects to disambiguation pages, but (a) I don't have the skills to do that and (b) it's not relevant to this discussion. This discussion is only about allowing (disambiguation) redirects to SIAs and nobody has provided any reasons why they should not be. Thryduulf (talk) 13:04, 30 August 2018 (UTC)

@Thryduulf: Rejecting or ignoring consensus WP:SIANOTDAB is a fair portrayal of your position, as underlined by you just saying "completely irrelevant to this argument". The topic is about treating them the same or similarly with respect to these redirects, a change to the consensus at WP:SIANOTDAB which you say is not relevant. I find it astonishing that much more missing redirects can be created to existing dabs than can be preserved by hanging on to vestigial redirects that now fail to target a dab. If this was such a big issue as claimed, then the low hanging fruit would be top priority. I highlight this, as it is direct evidence that there is no recognised issue cause by the missing redirects, which is direct evidence that there's no recognised problem that needs fixing from a readers perspective invalidating the secondary usage argument as proposed here. It's thoroughly unconvincing from a lack of evidence of demand. Widefox; talk 13:27, 30 August 2018 (UTC)

No matter how many times you try to claim otherwise, I'm still not rejecting or ignoring that consensus. Allowing (disambiguation) redirects to both of them does not mean involve treating them the same, it just means that readers can find them. Thryduulf (talk) 14:10, 30 August 2018 (UTC)

It treats them the same in this regard, redirects. There would be no difference, so that's the same. Disambiguation is clear to create them, they're functional and dependent on a dab page. There's nothing in SIA about creating them, and even if created they would not provide that essential raison detre function (linking). If it is a straw man then please correct me, it is factual and logically correct statement that you disregard that functional different and replace it with a search hack to justify proposing they be treated the same with regard to a redirect. Although not explicitly disallowed in WP:SIANOTDAB, it's implicit and bolded that they are not the same, not to be treated the same (formatting etc). You may consider them as you wish, but we go by consensus. At RfD, there is more a presumption of redirects are cheap and harmless which is why we have this lack of consistency of no consensus to create, but newly contested at deletion, when it was previously uncontroversial. It feels like routine maintenance is being held hostage to a minority idea that has no consensus. I've G6 and G8ed them in the past. Maintenance should be easy, and the vast majority were happy before this became controvercial and effort. Now, of course we can change it, but this is why this is going nowhere currently, as other editor's concerns and desires to understand creation are not being addressed so far. Anyone is allowed to create redirects. The problem is there's no convincing argument per cost/benefit. A rejection of the proposal even by no consensus is status quo ante SIANOTDAB. Widefox; talk 15:21, 30 August 2018 (UTC)

@Widefox: That's mostly just word salad - I can't understand almost all of what you are talking about, let alone provide a coherent response, so I'll just pick out the few comprehensible snippets. Yes the primary reason for "(disambiguation)" redirects existing to disambiguation pages is per WP:INTDABLINK, but that's not relevant at all (as repeatedly explained) - the secondary use for "(disambiguation)" redirects to disambiguation pages is to help readers find the content they are looking for without needing to navigate via content they know they are not looking for is equally valid to the primary use. That the same use for "(disambiguation)" redirects is the primary use when pointing at SIAs but the secondary use when pointing at dabs is true but utterly irrelevant to how useful they are. It's also not a "hack", it's an emergent use that wasn't anticipated when they were first created, but that's also totally irrelevant to everything we're discussing here. SIANOTDAB does not mention redirects at all, and even if "(disambiguation)" redirects do target both, that is not treating them the same in any way that has any impact on the differences between them. The status quo ante is that there is no consensus for or against these redirects, and no consensus that speedy deletion G6 applies to them (and by definition therefore it does not, as speedy deletion criteria only apply to pages that unambiguously meet the criteria). SIANOTDAB is not relevant to this. Thryduulf (talk) 10:28, 1 September 2018 (UTC)

@Thryduulf: So you'd be surprised if that comment received a thanks from another editor? I got one. Where's the consensus for "the secondary use" ? Have you ever considered that it's confirmation bias to assume your personal use case is in fact a widespread use case for general users? Where is the documentation or evidence about this navigation hack, either on or off wiki? Useful computing hacks are quite often listed on websites even, at least one of mine has been. I'm confident that "hack" is the correct term for an unsupported useful thing as proposed, yes. It can't work generally, so it's a hack, however useful you personally find it. Disambiguation is setup with the assumption of navigation flow via the primary topic, users are free to navigate how they like and nobody can take your personally useful hack away from you (or wants to), but it's not fair to misrepresent it as more than a hack when not all SIAs can have such a redirect, is it?! You also know very well that the old consensus was that they were deleted G6 or G8 in the past, as you've contested them recently to make them controversial and therefore no consensus for deletion. The answer is to get them officially classified as dependent pages on a dab, therefore eligible for G8. That is irrespective of whether this RfC says to extend them to SIAs, which, despite your claim has no consensus, as shown by edits and the majority of dab project editors comments here. It's pure misrepresentation to claim just because SIANOTDAB doesn't preclude them, that there's no consensus for or against, as shown logically that that argument could apply to lists, or any other page as they're not precluded anywhere else. Widefox; talk 13:53, 1 September 2018 (UTC)

Yes, I'm amazed somebody thanked you for something so incomprehensible. I may be the most vocal editor here (it seems almost everybody else has given up in the face of the IDHT and straw men), but it is not just me - look at every time that these get nominated at RfD. If it was just me challenging them (I can't remember the last time I challenged a speedy deletion) then they would still be deleted per G6. If the target page exists, even if it is irrelevant, then G8 does not apply and anyone deleting any such redirect under that criterion needs to stop it immediately. The rest of what you say has been answered multiple times already. Thryduulf (talk) 16:14, 1 September 2018 (UTC)

Clarity is good. How about this: After this RfC, but irrespective of the outcome, I can understand getting these redirects accepted as dependent pages. Then CSD would apply (whichever fits better G6 or G8). Many editors on this page agree they should be able to be deleted CSD if they fail to target a dab, as was the previous consensus. Widefox; talk 18:42, 1 September 2018 (UTC)

• No. Set indexes are not disambiguation pages, so labeling them as such using redirects is misleading. -- Tavix ^(talk) 21:12, 26 June 2018 (UTC)
  • How is someone being misled by arriving at a page that disambiguates articles that have or could have similar titles (which is exactly what both disambiguation pages and set indexes do) by the presence of a (disambiguation) redirect? In terms a reader will understand, what is the functional difference (i.e. ignoring what types of formatting our style guidelines permit) between a disambiguation page and a set index? Thryduulf (talk) 21:50, 26 June 2018 (UTC)
    • Because a set index is not a disambiguation. -- Tavix ^(talk) 22:01, 26 June 2018 (UTC)
      • That does not answer any of the questions I asked. What is the functional difference to a reader? How is arriving at a set index when looking for something that disambiguates misleading? Simply repeating that it is misleading does not actually make it so. Thryduulf (talk) 23:32, 26 June 2018 (UTC)
        • See WP:SIANOTDAB. -- Tavix ^(talk) 02:43, 27 June 2018 (UTC)
          • That still does not answer the questions. That page explains how the formatting differs (in some cases), but still does not explain how they are functionally different from a reader's perspective; nor does it even attempt to explain how arriving at either when looking for something that disambiguates is misleading. Thryduulf (talk) 10:11, 27 June 2018 (UTC)
            • Sorry, I'm not answering your questions to your satisfaction no matter how often you ask. Stop badgering me (and others for that matter). -- Tavix ^(talk) 12:38, 27 June 2018 (UTC)
              • You assert that (disambiguation) pages redirecting to set indexes is "misleading", asking you to explain how it is misleading could not be more relevant. Thryduulf (talk) 12:53, 27 June 2018 (UTC)

    Tavix is exactly right. Set index articles do not disambiguate titles. They list elements of a particular set and they follow list article style, not disambiguation style. -- JHunterJ (talk) 12:47, 27 June 2018 (UTC)

    What's the functional difference between listing articles that do or could share similar titles while being part of a particular set and listing articles that do or could share similar titles while not being part of a particular set? Style differences are irrelevant to a reader who wants to find a list of articles that do or could share similar titles to their search term. Thryduulf (talk) 12:53, 27 June 2018 (UTC)

    One functions as a list. The other functions as a navigational aid. -- JHunterJ (talk) 18:14, 27 June 2018 (UTC)

    One is a list of articles that people use to navigate to those articles. The other is a list of articles that people use to navigate to those articles. How are they not both navigational aids? Thryduulf (talk) 22:49, 27 June 2018 (UTC)

    I agree Thryduulf, to readers SIAs and dabs would be similar or indistinguishable. To writers, especially projects, they're crucially different. Wouldn't this writer bias be better fixed at the cause - the definition of SIAs to align with readers priorities not projects?... rather than treating SIAs as dabs for this redirect aspect? (Acting as a dab is only one of SIAs dualities, the other being acting as a list, which is incompatible with the proposed extension of this redirect to target them.) Widefox; talk 09:39, 28 June 2018 (UTC)

    There is no incompatibility. SIAs and disambiguation pages both work equally well for readers for navigation purposes, the (disambiguation) redirects therefore both work equally well whether they point to a disambiguation page or a set index article. That set index articles also function as articles while dab pages do not doesn't matter - people using these redirects find the page they are looking for whichever they end up at. Thryduulf (talk) 11:19, 28 June 2018 (UTC)

    What's the evidence that SIAs are dabs "both work equally well", because if they do, why do we bother putting so much effort into dabs?! The consensus surely is that dabs are the best way to navigate (and let's not forget the majority are on small mobile screens, making the design and strictness of dabs more crucial), and anything not a dab is, by definition, not as good. How much less good would need evidence, but that's logically the consensus, and logically needs a convincing argument to change. Widefox; talk 21:46, 29 June 2018 (UTC)

    This is a combination of "I don't like set index articles" and "set index articles should be converted into disambiguation pages". Even if either were relevant to this discussion (which they aren't) you do not back up any of your claims with evidence - e.g. where is this "consensus that dabs are the best way to navigate" you speak of? (and note to be valid this consensus needs to include those people who are not involved with the disambiguation project and/or disambiguation style guidelines). Disambiguation pages do one job: Navigation. Set index articles do two jobs: Navigation and information. There is no evidence presented that either functions better than the other for navigation. Thryduulf (talk) 09:00, 1 July 2018 (UTC)

    I see my question about evidence isn't answered in two months, but was just turned around into my motivations and evidence. The burden is on proponents to change consensus. Please refrain from inverting the burden. Consensus is WP:SIANOTDAB. It's clear to all the history is that SIAs are allowed to diverge from strict dab formatting as they are lists not dabs. As the dab format is solely and purely for navigation, logically anything else is less optimised for navigation. The assertion that they "both work equally well" hasn't been justified, and IMHO is illogical by design. Widefox; talk 17:19, 31 August 2018 (UTC)

• Yes, exactly as per Thryduulf, who hit every single point I would have made. Other than to respond to some of the above "labeling" and "misleading" assertions, which are fallacious. The name of Page A has no labeling effect on Page B. And there's nothing faintly confusing about looking for a disambiguation page and finding a set-index article, surname index, or other things that probably originated as a DAB page and for the average user isn't really distinguishable from one. — SMcCandlish ☏ ¢ 😼 22:55, 26 June 2018 (UTC)

  The labeling and misleading assertions are accurate. Calling articles not-articles or not-articles articles is fallacious. Disambiguation pages and SIAs serve different needs, and where both are needed, both need to exist. If the goal is really to make them indistinguishable, try getting consensus to apply disambiguation style to a storm or ship index. -- JHunterJ (talk) 12:47, 27 June 2018 (UTC)

  Please demonstrate (as repeatedly requested), with evidence, how it is misleading and how having a redirect from (disambiguation) to a set index page causes the latter to become something other than a set index page. The only reason set index articles exist is that the disambiguation style guidelines are too rigid for all purposes (which is entirely irrelevant to the reader, and indeed to most editors). If this wasn't true then there would not be arguments about whether a given page is actually a disambiguation page or a set index - it would be possible to tell unambiguously in all circumstances. Your average reader (and editor) really doesn't care (or in many cases even know) whether a given page they are reading is one or the other. Thryduulf (talk) 16:48, 27 June 2018 (UTC)

  Please demonstrate, with evidence, how an article is a non-article. Disambiguation pages are not articles. SIAs are articles. QED. You are right, though, the average reader doesn't care, and I didn't care before, back when we could format what would eventually become SIAs as if they were disambiguation pages. But the SIA editors wanted more and more information on those nav pages, and made them articles instead, and objected when they were edited back down with navigation in mind. They can't be both things. If they are functionally disambiguation pages, then let us format them as disambiguation pages. But they aren't, so we don't. -- JHunterJ (talk) 18:14, 27 June 2018 (UTC)

  I'm not claiming that an article is a non-article so asking me to prove it is a straw man argument. Lists of articles function as pages from which people navigate to other articles, regardless of whether they are a disambiguation page or a set index. If someone searches for Foo (disambiguation) (which they do) they are looking for a page which lists articles related to foo. If a page or redirect exists at that target they will find what they are looking for, regardless of whether the target is a disambiguation page or a set index - as I keep saying the formatting is completely irrelevant. If Foo (disambiguation) doensn't exist but a page listing Foos does then we, as editors, have failed as readers will not find the content they are searching for. Thryduulf (talk) 22:40, 27 June 2018 (UTC)

• No. The sole purpose of X (disambiguation) redirects is to distinguish links which deliberately target a dab from those which should be fixed to point to one of the dab's entries. That system only works if all such links go via the redirect. I can see a case for extending that convention to name pages and set indices but, as clarified above, that is not being proposed here. Having just some of the links go via the redirect is of no benefit; they should link directly and the redirects should be deleted. Certes (talk) 23:46, 26 June 2018 (UTC)
  • Your first sentence notes the original purposes of (disambiguation) redirects, but completely ignores that they have gained other uses. Your second sentence is actually an argument in favour of more (disambiguation) redirects so that everything that intends to link to a list of articles can be distinguished from those that intend to link to one specific article. Extending the (disambiguation) convention to set indexes and name pages is exactly what is being proposed here. Your final sentence doesn't make any sense. Thryduulf (talk) 10:11, 27 June 2018 (UTC)
    • @Thryduulf: Can we agree on this summary?
      1. All deliberate links to base-name dabs should go via (disambiguation) redirects, and (thanks to some sterling efforts) nearly all actually do so
      2. Currently, a few links to name/list/SIA pages go via (disambiguation) redirects but most don't; this is not helpful and should be changed to all or none
      3. There is a reasonable case for all links to name/list/SIA pages to go via (disambiguation) redirects, but the established consensus is that none should
    • The last two points may also apply to a few other page types. — Certes (talk) 11:12, 27 June 2018 (UTC)
      • Your point 1 is completely correct, yes, and afaiaa is not disputed.
      • There is a reasonable case that all intentional links to base name SIA pages should go via (disambiguation) or List of redirects, but most currently do not.
      • There is a reasonable case that intentional links to base name pages about names (surname or other types of name) should go via disambiguated redirects (name), (surname), (first name) or (disambiguation).
      • Redirects ending in (disambiguation) should be allowed for all pages that disambiguate articles, whether those are disambiguation pages, set indexes or name pages, and whether or not those redirects have any incoming links. There is a reasonable case that these should be created where they do not currently exist. Thryduulf (talk) 11:33, 27 June 2018 (UTC)

        Set indexes articles and anthroponymy articles do not disambiguate articles. They are functionally lists, and styled for listing things, not for navigation. -- JHunterJ (talk) 18:14, 27 June 2018 (UTC)

        False dichotomy that ignores how readers actually use them and why. This is a confusion of intended purpose with the purpose to which something is commonly put. I.e., it's bureaucracy for its own sake. — SMcCandlish ☏ ¢ 😼 23:31, 27 June 2018 (UTC)

        • (the duality issues with the definition of SIAs are above, but ...) SMcCandlish - dabs are for navigation only, everything else in WP isn't, including SIAs. That's by intention and usage. There's many types of pages that may be used for navigation, including SIAs and lists but their raison d'etre isn't solely navigation. In assistance of that sole purpose, the raison d'etre for the redirects is to assist that one aim as a dependent page, which doesn't apply to all SIAs (if we hypothetically classify SIAs into those which are used essentially for navigation, and those which have other purposes). TLDR - fix SIA's definition rather than extend necessary redirects to unnecessary pages. That's bureaucracy. Widefox; talk 09:23, 28 June 2018 (UTC)
          • That's a mixture of irrelevant, incorrect and nonsense. It doesn't matter what a page's definition is, or whether it exists solely for navigation or for navigation and other purposes. The existence of a (disambiguation) redirect pointing to a page that is used for navigation does not indicate anything about it's purpose, definition or anything else. That some of these redirects are also necessary to avoid links to disambiguation pages is correct, but that is entirely unaffected by the existence of redirects that do not have that purpose. Thryduulf (talk) 11:26, 28 June 2018 (UTC)
            • You may characterise it as such, but I would characterise it as an attempt at detailing the consensus. The definition of what's a dab, and SIA are the consensus. The burden is on those who wish to change the consensus, with reasoned arguments, which I haven't seen yet here, so status quo ante. Widefox; talk 13:48, 28 June 2018 (UTC)
              • The definitions are irrelevant. A set index is a set index, however that is defined, whether or not a (disambiguation) redirect points to it. A disambiguation page doesn't stop becoming a disambiguation page, however that is defined, because some (disambiguation) redirects point to pages that are not disambiguation pages. I'm not seeking to change the definition of either. Thryduulf (talk) 14:01, 28 June 2018 (UTC)
                • Stating that you believe the definitions (the consensus) is irrelevant doesn't reason your argument for what change is desirable. It's in bold that SIAs are not dabs, and not to be treated as dabs, but this would conflate them, against the consensus. Go ahead. Widefox; talk 14:57, 28 June 2018 (UTC)
                  • But it will not conflate dabs and SIAs at all. Nothing about the pages themselves will change. A dab will still be a dab, an SIA will stil be an SIA, neither will be treated as the other in any way shape or form, neither definition will change in any way. The only difference is that people will be able to find what they want without being required to know in advance which has been used. Thryduulf (talk) 10:47, 29 June 2018 (UTC)
                    • Claiming it doesn't conflate is at odds with the fact that creating "(disambiguation)" redirects to SIAs is directly, irrefutably treating the SIA as a dab with respect to the redirect. That's conflation, and ASTONISHING users, and is a UI inconsistency. Users following the link will not know if the target is a dab or SIA due to the (incorrect) conflation, for instance. Writers will have to check if the target is a dab or SIA to know if it should be linked directly or not. There will still be many SIAs with no redirect, as the term already has a dab, resulting in an inconsistency for creating redirects to SIAs. Widefox; talk 20:49, 30 June 2018 (UTC)
                      • The only confusion exists in the minds of those who are opposed to this proposal as far as I can tell. Nobody has presented any evidence of anybody searching for Foo (disambiguation) and ending up at a set index of Foo being at all confused (as has been repeatedly shown elsewhere). Your argument that a redirect from a technically incorrect name somehow conflates disambiguation pages and set indexes is laughable - they will remain separarate pages with separate styles and separate guidelines, etc. nothing will change about the content of either type of page. As for your comments about linking are also irrelevant - editors currently have to know whether foo is a disambiguation page to or not to know if it should be directly linked, so absolutely nothing will change in that regard. In fact it will actually get slightly easier because while there is no requirement to link to an SIA via a redirect there is no harm in doing so, whereas linking directly to a dab page does cause (minor) harm, so if one gets changed to the other (as happens) then no links will need to be changed. Thryduulf (talk) 10:26, 6 July 2018 (UTC)
                        • If there's confusion, wouldn't that indicate that the argument of the proponents is too weak or confusing to gain consensus, rather than some ad hominem just because others disagree?! Absence of evidence isn't evidence of absence. I do understand the desire for SIAs to be more useful for navigation, the obvious answer is get consensus to change them! I am vocally against supporting hacks. Widefox; talk 17:40, 31 August 2018 (UTC)
• No. RfD should not be required either since G6 already covers them (routine cleanup). Note that 99.9% of these redirects were created by bots, have no useful page history, and have never been used. —Xezbeth (talk) 12:57, 27 June 2018 (UTC)
  • G6 does not very much not apply to redirects pointing to set indexes as this discussion demonstrates they have clear use - even that this discussion is happening demonstrates that deletion would not be uncontrversial. Thryduulf (talk) 14:02, 27 June 2018 (UTC)

    G6 covers this routine cleanup, as this discussion demonstrates. -- JHunterJ (talk) 18:14, 27 June 2018 (UTC)

    • Except it's not cleaning up anything - it's making navigation significantly harder for our readers while not benefiting anybody. Thryduulf (talk) 22:40, 27 June 2018 (UTC)
  • These sort of redirects come up at RFD from time to time and from what I've seen, they've tended to split opinions, with "no consensus" closes being the most common. Maybe G6 could have been applicable in the past, but with the current environment it's difficult to imagine how one can justify speedy deleting stuff that almost always fails to achieve consensus for deletion when discussed at the venues. – Uanfala (talk) 23:33, 27 June 2018 (UTC)
    • Well put. Sadly I've lost count of the number of times I've had to remind various administrators that pages may only be speedily deleted (under any criterion) when a deletion discussion of that page would stand no chance of having any outcome other than "delete". Thryduulf (talk) 09:30, 28 June 2018 (UTC)
  • Xezbeth, agree, but please see my comment above - it's been challenged at G6 and G8, making this RfC important, unfortunately. Widefox; talk 08:38, 28 June 2018 (UTC)
• Comment. There are apparently 17,712 redirects affected by this discussion (quarry:query/27886: this query, courtesy of Cryptic, lists all redirects with titles ending in (disambiguation) whose targets are not members of Category:All disambiguation pages). – Uanfala (talk) 23:43, 27 June 2018 (UTC)
  • Thank you, that is a very useful comment - it shows that there is likely a widespread acceptance of these redirects among the community. Thryduulf (talk) 09:30, 28 June 2018 (UTC)
    • It does nothing of the sort. Half of those are talk page redirects, which are even more worthless. The remainder are as useless as they have ever been. Bobko_(disambiguation) for example has been used a whopping 1 time in the last 90 days. If I didn't just link to it then I doubt anyone else would have ever visited it. —Xezbeth (talk) 10:11, 28 June 2018 (UTC)
      • • So? Just because one example has only been used once in 90 days (it was used 11 times last year, for comparison) doesn't mean that entire class is worthless. Should we delete all redirects from misspellings because Ode of Rememberance has only been viewed once since the start of May? Thryduulf (talk) 11:36, 28 June 2018 (UTC)
      • Thryduulf not at all - that's just confirmation bias, equivalent to WP:OTHERSTUFF. Widefox; talk 10:23, 28 June 2018 (UTC)
  • The basic top-level categories contain 75,015 set index articles, 69,014 surname articles, 13,981 given name articles and 90,399 list articles. Other pages will be affected too. If 17,712 8,030 have (disambiguation) redirects, that's 7% 3% of them, which I wouldn't describe as "widespread acceptance". Those counts also highlight that we'd need to create at least 250,000 new pages. To use the (disambiguation) suffix consistently, we'd also need to edit all incoming wikilinks to those 250,000+ pages. (Updated to limit (disambiguation) count to main namespace.) Certes (talk) 10:00, 28 June 2018 (UTC)
  • There's also many dabs missing "..(disambiguation)" redirects. How many missing dab redirects do we have? 20%? That would indicate this isn't some urgent, crucial, widely used and supported navigation path that needs extending to SIAs as it's not fixed for dabs even - I create missing ones all the time and I assume nobody noticed before or after! Widefox; talk 10:23, 28 June 2018 (UTC)
    • That logic doesn't track. The subway doesn't go to as many places as the streets, and the former is only very slowly being expanded despite a need for it to expand. That doesn't mean we should stop road maintenance (or eliminate the roads!) to quintuple our subway tunneling. — SMcCandlish ☏ ¢ 😼 10:56, 28 June 2018 (UTC)
      • To use that analogy, what's the point of putting road signs for subways?! Will that assist or confuse by conflation? My point being I see little evidence of demand, which that analogy doesn't address. Widefox; talk 13:30, 28 June 2018 (UTC)
    • A crucial difference is that we deprecate accidental links to disambiguation pages: over a million have been fixed, and the remaining 5,000 are actively being worked on. Dab pages with no (disambiguation) redirect are no problem unless the page has deliberate incoming links. Links to SIAs and name pages are accepted as perfectly proper, so no attempt is being made to divert them via redirects. There is no point in creating such redirects unless someone is happy to amend a lot of wikilinks (probably more than a million) to conform to a new standard. Certes (talk) 11:35, 28 June 2018 (UTC)
      • I can't tell whether that is a genuine misunderstanding or a deliberate straw man, but regardless it's irrelevant. This proposal is not to link to SIAs via (disambiguation) redirects, but to allow the redirects to exist for the benefit of those who use them to search. Thryduulf (talk) 11:39, 28 June 2018 (UTC)
        • It was a genuine misunderstanding. Thank you for clarifying. Now I understand that the proposal is to create 250,000 orphan pages, so that readers wanting to find a list of ships called HMS Abdiel can find HMS Abdiel (disambiguation). I don't think that's useful, so I still oppose. Certes (talk) 12:11, 28 June 2018 (UTC)
          • I genuinely don't understand why you think it is a good thing to require readers to know whether HMS Abdiel is an article about a specific ship, a disambiguation page or a set index before being able to find what they are looking for - i.e. a list of articles about ships with that name because they don't know the title article they want to read but suspect it is not the primary topic. Redirects being orphans is normal, not any sort of problem, and maintaining another 250,000 redirects creates significantly less work for editors than has already been expended in this one discussion. Thryduulf (talk) 14:07, 28 June 2018 (UTC)
• Comment. Disambiguation pages are styled for navigation, not providing information. List articles (including SIAs) are styled as encyclopedic lists, providing information. Articles are not non-articles. Non-articles are not articles. Without that tautological ground to start from, this discussion is inarguable for or against. -- JHunterJ (talk) 12:19, 28 June 2018 (UTC)
  • I think we can argue the question of whether certain types of page (which bear some similarity to disambiguation pages in that they also aid navigation) should have (disambiguation) redirects. It is understood that these pages differ from dabs in (a) being articles, (b) not being disambiguation pages and, following recent clarification, (c) that the proposed redirects would be orphans rather than marking deliberate links. Certes (talk) 13:53, 28 June 2018 (UTC)
    • Once again, the definition is irrelevant. Redirects being orphans is irrelevant. The difference between disambigation pages and set indexes is, for the purposes of navigation, irrelevant. Thryduulf (talk) 14:12, 28 June 2018 (UTC)
      • "The definition is irrelevant" is just wrong. The definitions and differences are extraordinarily relevant, and sufficient to disprove your position. -- JHunterJ (talk) 14:54, 28 June 2018 (UTC)
        • Why and how is the definition relevant then? This will not change any definitions, it will not conflate the two, combine the two, remove the distinction between the two, or anything of the sort. Thryduulf (talk) 10:47, 29 June 2018 (UTC)
          • The definitions are relevant because they define the things under discussion. Again, this is foundational. You cannot simply dismiss the parts of reality that do not suit you. You would like these definitions and differences to be changed, ignored, or overridden to a particular end, or you disagree with the consensus that what you would like does any of those things; nothing wrong with that. Consensus is against you (either your interpretation of whether this does anything of the sort, or whether any of those sorts of things should be done), though; it happens (and it's certainly happened to me). So we go with the consensus and continue the work of the encyclopedia. -- JHunterJ (talk) 12:44, 29 June 2018 (UTC)
            • No, I explicitly do not want the definitions to be changed - as I've repeatedly stated. The definitions of disambiguation pages and set index articles are not the things under discussion. The only thing that is under discussion is whether redirects ending in (disambiguation) should be allowed to point to set index articles when there is not a disambiguation page by that name. Also, there is no consensus in this discussion against the proposal - there is either no consensus or consensus in favour depending how the closing admin regards the strength of the arguments. Thryduulf (talk) 10:31, 6 July 2018 (UTC)
    • Certes, in that case, the SIA project should come up with a redirect that would be serve that purpose and not contradict reality. Like "(SIA)" or "(list article with information but also some wikilinks, like any other list article, so use this redirect if you're looking for those wikilinks only for their navigational use)". -- JHunterJ (talk) 14:54, 28 June 2018 (UTC)

This sounds like a seed for a possible compromise. If other editors found utility in redirects such as Africa (list of ships) to HMS Africa (or something similar), I would not oppose such redirects. Certainly, that is less surprising. —hike395 (talk) 15:19, 28 June 2018 (UTC)

I agree with JHunterJ. So is this a sort of "poor man's short description", along the same lines as Venus (planet) → Venus for the benefit of readers who might have thought that the article was about the goddess? Certes (talk) 19:04, 28 June 2018 (UTC)

Adding "(SIA)" redirects is fine, but it will not solve the problem of people being required to know in advance whether the page they are looking for is a disambiguation page or an SIA. The only way to do that is to have a single string that users can search on that leads to every page that serves as a list of articles with similar titles. Thryduulf (talk) 10:50, 29 June 2018 (UTC)

That string already exists, and is the blank string. The exception is when it leads to a primary topic such as HMS Victory, but that convention helps far more readers than it hinders. Certes (talk) 11:12, 29 June 2018 (UTC)

Once again you've missed the entire point. The point is that people are looking for an article that is not the primary topic. For example if I'm looking for the ship called HMS Victory that was built a Deptford but don't know the title of that article, I would search for HMS Victory (disambiguation) because I know that the ship preserved at Portsmouth is the primary topic and not the one I want to read. That title is a redirect to a set index article at List of ships named HMS Victory, this page is exactly the content I was looking for and so far from ASTONISHing me, the redirect was exactly what I wanted. There is no evidence that any of the 52 other people who have used the redirect this year found anything other than what they were looking for either. Thryduulf (talk) 14:31, 29 June 2018 (UTC)

If you're searching for info about the HMS Victory built at Deptford you can search for "Hms victory deptford". DexDor ^(talk) 16:40, 6 July 2018 (UTC)

• (ec) Oppose. The purported "usefulness" of these redirects is akin to developers who utilize unsupported features. Yes, they might find them useful and they might actually work sometimes, but there should be no expectation that they can be relied upon consistently. There is a difference between disambiguation pages and set index articles although some editors confuse them (and to be perfectly honest, many SIAs should be converted to disambiguation pages, but that's another discussion). The only supported use of such "(disambiguation)" redirects is to mark intentional links to disambiguation pages. There were long discussions to get the practice started and to get it accepted. And even so editors need some hand-holding to understand. Allowing occasional and inconsistent redirects to non-disambiguation pages only contributes further to the confusion. older ≠ wiser 12:33, 28 June 2018 (UTC)

  Yup, agree. Proposing to "support" a search hack, let alone extend it needs careful thought - what is the cost/benefit. Where's any of that here? Widefox; talk 13:50, 28 June 2018 (UTC)

  What is the cost? Essentially zero. What is the benefit? Readers find the articles they want to read without having to navigate via the search engine or articles they don't want to read. Thryduulf (talk) 14:11, 28 June 2018 (UTC)

The cost is not zero: the tens of thousands of potential redirects would need maintenance when their targets were moved, deleted or repurposed. The benefit is only to "disambiguation-aware" users who are looking for a list of articles and who navigate directly to a (disambiguation) page title without knowing if that page title exists or not. See my possible mitigation below. Shhhnotsoloud (talk) 09:12, 29 June 2018 (UTC)

If pages are moved then bots update redirect targets as now - Additional cost: Zero.

If pages are re-purposed from a dismabig to a set index currently then (disambiguation) redirects need to be created or deleted, in future they will not be - Additional cost: Non-trivial negative (i.e. it will require less work than now)

If pages converted from a set index to an article then it is likely that a set index or disambiguation will still exist at a different title and just need to be retargetted - additional cost: tiny

If pages are deleted without replacement then the redirect will just be deleted under G8 exactly as they are for disambiguation pages now - additional cost: trivial

If redirects are discovered pointing to SIA articles then currently some users feel the need to nominate them for deletion, in future they will have to do exactly nothing - additional cost: Significant negative.

So in summary, having these redirects will on balance have less cost than the current situation, especially as the changes that add small costs are the least frequent. Thryduulf (talk) 10:38, 29 June 2018 (UTC)

Thryduulf, a fact: the bots haven't created redirects for all dabs yet. Why? I don't know, but it would take effort to see why/fix that. I don't estimate your answer immediately above as being cost free. The details of the benefits aren't clear either (to me). Widefox; talk 19:43, 29 June 2018 (UTC)

• Question: Dab and an SIA proponents of having " (disambiguation)" target more than dabs, please detail your proposal for redirects for the common situation when there's both a dab and an SIA. There's only one title with "ambigterm (disambiguation)", so it can't target both. I believe the majority of SIAs are (surname/given) names, this is very common (anyone know the % ?). If there's no solution, then there's not even a complete proposal to cover common SIA. For example 1. WP:FURTHERDAB: Laing (surname) and Laing (disambiguation) (no primary topic), but also 2. when the primary topic is the SIA (which does also happen) and 3. when the primary topic not the SIA. Widefox; talk 14:42, 28 June 2018 (UTC)

  This was already addressed above. If there's a real DAB page, then of course "Foo (disambiguation)" would go there. And it's already standard practice for SIAs and anthroponomy lists to have entries in a DAB page. — SMcCandlish ☏ ¢ 😼 20:08, 28 June 2018 (UTC)

  It's acceptable to include a short list of disambiguation-styled entries for name-holding partial title matches on a disambiguation page, after all of the entries for the appropriate topics, until an anthroponymy list article is created. It is not acceptable to include such lists when the anthroponymy article exists. It's acceptable (and expected) that disambiguation pages list all topics (appropriately formatted for navigation) for the ambiguous title, whether those topics are also included on one or more SIAs, other lists, other articles, or anywhere else (where they'd be for encyclopedic information). If there's a dab page (the only dab pages are real dab pages), then "Foo (disambiguation)" would go there. If there's no dab page, the "Foo (disambiguation)" goes nowhere, since there's nowhere (or no real where) to go. But you're right, all this is covered above. -- JHunterJ (talk) 20:30, 28 June 2018 (UTC)

  I see there's no reply to the three common SIA/dab scenarios above, so Thryduulf & SMcCandlish (please point to where you detailed how 1,2,3 works?) if there's no solution for the stated goal of Yes, always for set index pages - it is completely unworkable! If there's no proposal that describes how this is possible from the proponents, then it appears impossible even for the proponents, and is dead in the water. Widefox; talk 13:09, 29 June 2018 (UTC)

  Except that has been addressed, multiple times. 1) If there's already a DAB page, then "Foo (disambiguation)" would obviously go there; "Foo (disambiguation)" would mention "Foo (surname)" if it existed. Your premise is that SIAs are articles (that part is true but meaningless; the off-kilter half of your premise is that the sky will fall if a "Foo (disambiguation)" page points to one because its list serves the same purpose – there's zero evidence of any problem, and obvious evidence that it's helpful for readers, since they are in fact using and creating such redirects).
  2) If an SIA is the primary topic, then it is not different from any other article that's a primary topic. If there are things with the same name that also have other articles (if there's not, then "primary topic" doesn't have any meaning in the context), then "Foo (disambiguation)" would not be a redirect at all but an actual disambiguation page, just like normal.
  3) Your last scenario isn't distinguishable; it again would result in an actual DAB page at "Foo (disambiguation)" because the "Foo (whatever)" SIA page isn't the only page to disambiguate. None of this is new or even a question, it's standard operating procedure.
  Your idea, just because no one pointed out the obvious to you, that an already-widespread actual practice that you don't like is "completely unworkable", "impossible", and "dead in the water" (with a bunch of excessive italics) is obstructionst, reality-denying, "I just don't get it", wikilawyering nonsense that is anti-collaborative and thwarts consensus formation. It's a debate tactic you need to stop whacking people in the face with. See also WP:SATISFY, WP:GASLIGHTING, and WP:RTFM. — SMcCandlish ☏ ¢ 😼 13:57, 29 June 2018 (UTC)

  SMCCandlish, you are engaging in the very obstructionst, reality-denying, "I just don't get it", wikilawyering nonsense that you're decrying. Along with an assumption of bad faith. Widefox appears to be trying to help you identify what you and Thryduulf are trying to get to, since it is currently contrary to consensus (and the definitions of the articles and nav pages in question). I'd just as soon end it here too, given the consensus for the current reality, but Widefox is putting in the effort to see what else might be done to address the hypothetical issue you and Thryduulf seem to see. -- JHunterJ (talk) 16:48, 29 June 2018 (UTC)

  I'm not assuming any bad faith, just being critical of unhelpful behavior, like asking the same thing over and over again after it's already been answered multiple times. That is absolutely not "putting in the effort to see what else might be done to address" the issue or "trying to help [someone] identify what [they] are trying to get to", it's annoying pretense that we don't know and can't explain what we've already very, very clearly explained. — SMcCandlish ☏ ¢ 😼 17:55, 29 June 2018 (UTC)

  SMcCandlish, if I understand Thryduulf's proposal (please correct me if wrong), it is to create redirects targeting all SIAs. Correct? If not, not all SIAs will have a redirect (as all dabs do, in theory), then this will not be a workable, reliable method for searching! I see no proposal above for the common cases 1.2.3. . If there is none, then this needs to be recognised as completely and utterly unworkable. The burden is on the proponent to convince others of changing consensus, so please go ahead (as the answers to this so far are not unconvincing (to be polite), more of a desire to extend without having thought through the consequences for common cases, I'm afraid to say). My tip: complaining about those that try to tease out the proposal doesn't help you convince them per WP:CLUE. Widefox; talk 19:37, 29 June 2018 (UTC)

  In the probably vain hope that different formatting will make you hear something you obviously don't want to hear:

  1. Laing (surname) and Laing (disambiguation) (no primary topic): The exisitng redirect would not be changed - i.e. it would go to the dismabiguation page at Laing. No existing pages with (disambiguation) in the title will change.
  2. Where the SIA is the primary topic but there is also a dismabiguation page: The (disambiguation) redirect would go to the disambiguation page. No existing pages with (disambiguation) in the title will change.
  3. Where there is a disambiguation page and an SIA but neither is the primary topic: The disambiguation page will remain at the (disambiguation) title. No existing pages with (disambiguation) in the title will change.

  The consequences for common cases have all been thought through, and have been answered many times previous to this, in the FAQ below. Thryduulf (talk) 19:57, 29 June 2018 (UTC)

  1. Thryduulf thank you for the answer (maybe I missed it above?) - so the SIA would not get a redirect. So not all SIAs would get redirects, and therefore (my point) that this isn't as you said Yes, always for set index pages. I see your !vote now has a provision ..., allow but do not require for surname pages so that just kicks the problem down the road (away from the most common SIAs!) - if not all SIAs are to have a redirect, then it's not even a simple rule like for dabs (where they're still missing). The maintenance with a rule that depends on several pages/aspects is more, especially when moving articles/dabs/SIAs move. I really don't like complex, and I vehemently disagree this is cost free to create, let alone maintain. Blurring dabs and SIAs is, IMHO, the cause of this, and this proposal exacerbates the problem, rather than being a solution to it. Widefox; talk 22:46, 29 June 2018 (UTC)

  Where both disambiguation and SIA pages exist the disambiguation page the (disambiguation) title is either the disambiguation page or a redirect to it, and the dab page links to the SIA. This means that anyone looking for an article listed on the disambiguation page has found what they want directly, and anyone who is looking for something only listed on the SIA is given a direct link there without having to know the title of the page in advance - i.e. everybody finds what they are looking for. The "allow but do not require" for surname pages is because not all surname pages are or have lists of articles about people with that name - e.g. Smith (surname), Jones (surname) (although these are not perfect examples as there are separate disambiguation pages that would remain the target of the (disambiguation) redirect - indeed in 2-3 minutes browsing I've yet to find an example of a surname page that doesn't contain a list of people where there isn't also a separate disambiguation page). You still have not provided any explanation (let alone evidence) for how these redirects would blur disambiguation pages and set index articles. Thryduulf (talk) 14:31, 30 June 2018 (UTC)

  So in terms of sets of users rather than empirical data, the bounds of user numbers of such a navigation path are (in the absence of any evidence of stats)... for A) a minority of SIAs and B) for a subset of a minority of users (per typical WP:PRIMARYTOPIC more "likely than all others combined", for instance). Agreed? You must concede that A * B is niche of a niche, so the stats must be vanishingly small by those bounds alone, and as presented lacking convincing data. Widefox; talk 21:47, 30 June 2018 (UTC)

  No, that doesn't make any sense. What you are trying to saying is that because only a small number of people who use this method explicitly comment in favour of this then the correct thing to do is to make things unnecessarily difficult for everybody who uses it just because you don't personally use it and don't like the fact that other people use it - an attitude that is directly harmful to Wikipedia. Thryduulf (talk) 08:39, 1 July 2018 (UTC)

• Possible mitigation. This RfC boils down to do we support this "undocumented feature" and "search hack", or not? and must be resolved to stop frequent re-runs of the argument at RfD. There are a number of arguments against, versus an argument for to support those users looking for a list of articles and who navigate directly to a (disambiguation) page title without knowing if that page title exists or not. If we can settle on No then one compromise could be for such "disambiguation-aware" users to be advised to stop navigating like that by guidance (somewhere) that says "Users wishing to find a list of articles on a particular topic are advised not to navigate directly to titles such as Foo (disambiguation): information about various uses of Foo might be found at a disambiguation page, or list article, or other index article located at the Foo page name, and if not then a hatnote is provided to the alternative location (which might not necessarily be a disambiguation page)." Shhhnotsoloud (talk) 09:12, 29 June 2018 (UTC)
  • How do you propose to advise people to change their behaviour when it is almost certainly something they have just discovered for themselves works, or have been told by others works? What benefit is there to changing the practice from: "You can find what you want at Foo, Foo (disambiguation)" to "What you want might be at Foo, List of Foo, Foo (disambiguation), Foo (SIA), Index of Foo or some other title"? How is that in any way shape of form better for anybody? Thryduulf (talk) 10:42, 29 June 2018 (UTC)
  • Yeah, Shhhnotsoloud's idea is utterly impractical. Guidelines a written for editors, not readers. Our readers usually never see any of them, nor even know where to look. Even most of our "Help:" namespace pages see very, very little use relative to the size of the readerbase. — SMcCandlish ☏ ¢ 😼 12:42, 29 June 2018 (UTC)
• No - mainly because this would further muddle the distinction between dabs and SIAs for little/no benefit. Most Wikipedia readers probably have little understanding of dab pages and SIAs, but they do understand the basics of searching for things on the internet (the more info you put into a search box the more likely it is that the thing you're looking for will be at the top of the results). So, for example, a reader interested in Portland (UK) who doesn't want to risk a slow download of the article about Portland (USA) would type "Portland uk", "Portland dorset" or similar into the searchbox - not "Portland (disambiguation)" which is more characters to type, unintuitive and guarantees that they'll have to go through another step before reaching the article they want. DexDor ^(talk) 12:52, 1 July 2018 (UTC)
  • @DexDor: The problem with that theory is that it doesn't match the evidence of what we see in practice. Sure some people, maybe most, go via the search engine (and these people will be entirely unaffected by this proposal). If people didn't use this method then nobody would have created these redirects and they wouldn't get any page views. The difference between going via search results and going via a disambiguation page or set index is not one of the number of pages that you need to view, but of what pages you need to view - either search results that may or may not contain your query on the first page or a list of articles that almost certainly will contain the article you want on the first page. Yes, only people who are familiar with Wikipedia's naming of disambiguation pages will use this method, but very many people are familiar with that - far more than are familiar with the difference between disambiguation pages and set indexes. The question being asked by this RfC boils down to: Some people use this method, do we want to go out of our way to hinder them for no benefit to anybody? Thryduulf (talk) 10:40, 6 July 2018 (UTC)
• Yes. A string like (disambiguation) might have been only an obscure element of out technical infrastructure 15 years ago, but now it's a generally known term: pretty much everyone who's ever used wikipedia will know of the existence of disambiguation pages, and it's likely that many people would have figured out that if they started typing some term in the search box and one of the drop-down suggestions ends in (disambiguation) then they could click on it and be taken straight to the disambiguation page bypassing any primary topics. For example, a user might be looking for the town named "London" which they know is somewhere in the Americas, but they don't know whether it's in the US or Canada or somewhere else, and they know they're on wikipedia so making google-type queries like "London America" won't return anything meaningful, so of course they would go straight for London (disambiguation). This is not some search "hack", but a reasonable and obvious way to navigate the encyclopedia, which becomes especially relevant for mobile users for whom the loading of a huge page they don't want (like London) is costly.
  The only question is whether this avenue should be restricted only to "actual" disambiguation pages, or whether it should extend to set index articles (SIA) or name pages. First off, the distinction between DAB and SIA articles is virtually unknown outside the dedicated core of members of the dab wikiproject, and it would be really surprising if our readers knew (or cared) about it. And really, SIAs (and to a large extent name articles) are lists of things sharing the same name: this is not encyclopedic by itself and its only function is to disambiguate between those things. These lists are functionally equivalent to dab pages. In the absence of a "proper" dab page, a user following a redirect like X (disambiguation) who lands on a page that lists things named X would have found exactly what they were looking for.
  A case for deprecating such redirects can be made only if it turns out that either they aren't used at all (the evidence presented so far suggests otherwise), or that they don't get readers where they want (so far, that's not a reasonable assumption). And really, if the issue is that the targets aren't technically speaking disambiguation pages, then the redirects can be tagged with ((R from incorrect disambiguation)); again, redirects are there to help readers and they don't need to be "correct", there are tens of thousands of perfectly good redirects in Category:Redirects from incorrect names. – Uanfala (talk) 11:50, 6 July 2018 (UTC)
• Meh Very weak yes. IMHO, both sides of this discussion have vastly overstated their positions. On one hand, it seems to me that the benefit of a "(disambiguation)" redirect to a set index article is fairly minuscule. Although some users might indeed search for the term "disambiguation" explicitly, those are likely to be a relatively small set of relatively sophisticated Wiki readers, nearly all active editors. Furthermore, the vast majority of the time, if the redirect "Random title (disambiguation)" turns out not to exist, then the article at "Random title" will provide this sophisticated reader with the guidance they were looking for. On the other hand, redirects are cheap, and no one is ever going to see these unless they have explicitly typed the word "disambiguation" into the search bar or some external search engine. So I don't think these sophisticated readers are going to be harmed in any material way if their search takes them to a SIA or list article instead of a disambiguation page, as long as it helps them find the topic they were trying to find. In fact, in writing out this comment, I've changed my mind from "couldn't care less" to a very weak support, since the proposal seems to do no harm to anyone even if it does have very little benefit. However, I only support the proposal insofar as it seeks to permit these redirects, not to require them. --R'n'B (call me Russ) 21:47, 11 July 2018 (UTC)

  A couple of further points: (1) If this is accepted as a standard practice, it would be helpful to have a tracking category similar to Category:Redirects to disambiguation pages for them. Maybe Category:Redirects to set indices would work. (2) As far as I can tell from this discussion, there seems to be no disagreement that a "(disambiguation)" redirect to an article is unacceptable, and that deletion would be uncontroversial, if the target article is not a set index, a list, or something that similarly assists a reader in finding other, similarly-titled topics. --R'n'B (call me Russ) 21:49, 11 July 2018 (UTC)

I'm not sure if you're aware of how many pages could be included by that criteria. e.g. WP:TWODABS failures (e.g. no dab but two articles with hatnotes) or any page with a hatnote to another article - reductio ad absurdum a lot! I'd rather hope we'd go in the other direction - towards eliminating the dab redirects (which are a supported hack) in software / improving search. Widefox; talk 18:05, 31 August 2018 (UTC)

• Strong yes, per the thorough arguments of Thryduulf and others. We have all sorts of redirects from "wrong" titles (misspellings, former names, etc.). Plenty of readers are aware of pages ending in "(disambiguation)" and may type that into a search bar; almost no non-editors are aware of the DAB/SIA distinction, and it's not our job to teach them about it just so they can find the article they want. The biggest complaint casual readers have about Wikipedia is its idiosyncrasy – the "insider" stuff you're expected to know if you want to make full use of the encyclopedia. Here's an opportunity to smooth a small bit of that. If a redirect has plausible value, leave it! Nothing about these redirects threatens the maintenance of the DAB/SIA distinction. —swpb^{T go beyond} 15:49, 12 July 2018 (UTC)
• No. Eligible for WP:CSD#G8; these fit the definition of being dependent upon nonexistent pages. I do not see evidence that users use the search bar in this way, but if users are familiar enough with the interface to know that they might want to search for disambiguation pages, they should also be competent enough to know about the intitle search function. Dekimasuよ! 18:36, 15 July 2018 (UTC)
  • 1) G8 explicitly does not say that: it spells out exactly the types of redirects it applies to, and these aren't among them. 2) Knowing about disambiguation implies knowing about intitle searching? That's an unsupportable and IMO implausible leap. Casual readers see "disambiguation" constantly; they never see "intitle" unless they click on "All pages with a title containing" on a dab, and then pay attention to the filled-in search bar, or they go exploring Special pages. All the non-pedians in my life know about dab pages; I'd wager 50 times as many people know about dab pages as know about intitle searching. —swpb^{T go beyond} 12:53, 16 July 2018 (UTC)

• When a title with "(disambiguation)" redirects to something that's not a dab page, it is included in the listed examples ("such as non-existent targets... and bad titles"). Further, even were that not the case, G8 begins "Examples of this criterion include." It is not an exhaustive list of "pages dependent on a non-existent or deleted page." Dekimasuよ! 16:08, 16 July 2018 (UTC)
  • "Non existent targets" means red links and so is not relevant in the slightest. There is no consensus that these redirects are "Bad titles" so speedy deletion cannot apply as that is only permitted for pages with "no practical chance of surviving discussion.". WP:CSD also notes: "[The criteria] cover only the cases specified in the rules here." (emphasis mine) and "Administrators should take care not to speedy delete pages or media except in the most obvious cases.". Thryduulf (talk) 00:40, 17 July 2018 (UTC)

    Yeah, Dekimasu is just making stuff mean whatever they want it to mean. "redirecting to something that's not a dab page" is not redirecting to a non-existent target but to a existent target, by definition. It's also not a bad-title case; we get one of those when we try to title something and the system doesn't let us, with a big Bad title error (e.g. try creating an article at [[Foo}{bar]]). — SMcCandlish ☏ ¢ 😼 09:52, 17 July 2018 (UTC)

• • There's nothing made up about "redirecting to something that's not a dab page" when the guidelines for disambiguation and set index pages are clear that one is and one is not. That some editors want to propagate confusion between them it's another matter. older ≠ wiser 12:17, 17 July 2018 (UTC)

    • • Have you actually read any of this discussion? It's been explained, repeatedly and at length, why there is no actual confusion. The entire point of this RFC is that the guideline about (disambiguation) redirects does not mesh with current practice - and even if it did a guideline does not and cannot override a policy page like CSD. Thryduulf (talk) 13:27, 17 July 2018 (UTC)

        And "redirecting to something that's not a dab page" is nothing remotely like "redirecting to a non-existent target". There is no escaping that fact. "some editors want to propagate confusion between them" is just a distortion; what those "some editors" want to do is help readers find content. It doesn't confuse disambiguation pages with SIA or anthroponymy pages, it uses a redirect to get people to what they're looking for, which is why we have redirects. Bkonrad's argument isn't distinguishable from claiming that the Lord Mayor of Bristol redirect to List of mayors Bristol, a SAL page, must be deleted because it's "propagating confusion between" lists and non-lists. WP just doesn't work that way. The internal editorial classification and subclassification of page types does not wag the dog of getting readers to content. — SMcCandlish ☏ ¢ 😼 13:47, 17 July 2018 (UTC)

That's it in a nutshell: "Dekimasu is just making stuff mean whatever they want it to mean". There is zero wiggle room in the meanings of "non-existent target" and "bad title". To engage further on that point is to be pulled deeper into a rabbit hole of bad faith. —swpb^{T go beyond} 15:26, 17 July 2018 (UTC)

I wouldn't ascribe it to bad faith, more like single-mindedness or territorialism or WP:WINNING or forest-for-the-trees or one of our other habitual human foibles. I don't think anyone's trying to do wrong here, they just have radically different priorities (some of which should be reconsidered, given the reader-facing nature of this project). — SMcCandlish ☏ ¢ 😼 16:20, 17 July 2018 (UTC)

This is not just "different priorities". This is doubling down on an error, by an admin, in an attempt to save face. That's bad faith. Wouldn't the project be better if that kind of behavior had a meaningful cost? —swpb^{T go beyond} 19:34, 17 July 2018 (UTC)

I strongly agree it's doubling down on an error, but I think anyone from an anthropology background would agree with the following corollary to an old canard: Never attribute to malice or stupidity that which can be adequately explained by everyday hominid stubbornness and self-interest. >;-) No one likes to feel criticized or shown wrong, so they tend to rationalize and handwave. Admins aren't magically immune to it, though people in a position of responsibility and trust need to catch themselves and moderate this tendency, especially after they've been repeatedly called on it (including by another person of the same community-vetted trust level). — SMcCandlish ☏ ¢ 😼 00:48, 18 July 2018 (UTC)

This page is not on my watchlist, so I didn't realize my comment created heat. If it's helpful, I am happy to concede that this is not what is meant by "bad titles", and a look through my deletion log did not immediately reveal any instance in which I have ever applied G8 (or the old R1) to this sort of redirect. Perhaps I should have just referred to G6 in my original comment, as was my original inclination. However, I reasoned that if it were self-evident that G6 applied, there would be no reason for the RfC to be taking place. Thus I substituted a reading of G8, a criterion I rarely apply except to talk pages of deleted pages, for simply expressing an opinion that this might reasonably be considered "non-controversial cleanup". I doubt there is any use to my arguing further over whether these disambiguation pages are "nonexistent" as "specified in the rules". I am not much interested in saving face and I don't think I've had someone suggest I was acting in bad faith in my 12 years here, so it's bemusing to have that happen in a discussion I'm not much invested in. Dekimasuよ! 11:05, 21 July 2018 (UTC)

To editor Dekimasu: You say you are "happy to concede": please show your good faith by striking your errors above, so anyone reading this thread will immediately know not to rely on them. I would also think, having your arguments so thoroughly debunked, that you might want to consider re-evaluating your "no" vote (since of course you have no personal investment here), but that's up to you. —swpb^{T go beyond} 12:50, 27 July 2018 (UTC)

• Yes: Redirects like these are likely to help the reader to find a page which will help them get to their sought information sooner rather than later. They do no harm, and I suggest that readers will not be confused as they are unlikely to spot the difference between a dab page, a name page, and a SIA, as long as it offers them the information they need. The redirect will also alert an editor thinking of creating a new dab page, by showing them that there is already a non-dab non-article page which needs to be taken into consideration. PamD 10:05, 16 July 2018 (UTC)

  "non-dab non-article page"? That's not what these are. These are non-dab article set index articles. -- JHunterJ (talk) 12:16, 16 July 2018 (UTC)

  That's obviously a true correction of a some poor wording. Yet it strangely has no effect at all on the actual argument PamD presented. — SMcCandlish ☏ ¢ 😼 14:57, 16 July 2018 (UTC)

User:PamD There's a couple of gut assumptions there 1. likely help 2. no harm. Both assumptions have been explored now. Logically, if true, wouldn't all lists and many other types of pages benefit from such a free improvement? reductio ad absurdum all pages with hatnotes etc, essentially almost all pages! (I've put a question below about the number of SIAs this fails for, do you know if the majority of name SIAs have dabs?) Widefox; talk 09:19, 31 August 2018 (UTC)

• Slightly late to the discussion, but I was pinged earlier. In short, yes because among other reasons:
  1. Wikipedia's DAB pages have existed long enough that many of our casual readers are aware that they disambiguate between topics which can be referred to by the same name (e.g. [1], [2], [3], [4], [5]).
  2. SIAs also do this, but only include topics belonging to the same class. Almost no readers and only a small minority of Wikipedia users, usually those who work on DABs, are aware of the technical differences between DABs and SIAs.
  3. Users sometimes do want see only a list of non-primary topics for X. If they are aware of DAB pages, specifically searching for them is reasonable, especially when the method of disambiguation can be obscure (e.g. years of battles/treaties, ship's hull numbers, people's birth years and professions).
  4. An overwhelming supermajority of set indices (@ Category:Set indices) are functionally DABs, with minimal formatting or content differences. Some exceptions that are actually more developed as set index articles do exist (e.g. mountains, car models) but are the minority. So if readers are looking for a DAB about X, they will almost never be disappointed or surprised to find a set index on X instead.
  5. Wikipedia should aim for consistency, so there should be one universal way to navigate to pages that allow for quick differentiation between very similarly named topics.
  6. Strictly speaking, "X (disambiguation)" will get you to a page where topics named X are disambiguated if those topics belong to more than one class, no matter how few or how obscure those other topics are. However, "X (disambiguation)" will not get you to the same page, if all X topics belong to the same class, even if there is a very large number of X topics, many of which are quite prominent. This is inconsistent and arbitrary.
  7. Wikipedia is meant to serve readers who not be hindered merely to strictly adhere to guidelines.
  • So if we can agree that some readers will want to specifically search for DAB pages and there's no real, functional difference between most SIAs and DABs, then these redirects should be kept, and actively created (perhaps by a bot). This is without considering factors such as widespread behind the scenes categorization of SIAs as DABs and vice versa (e.g. click around in Category:Set indices on storms), not forcing people to waste data loading long primary topic articles they 100% do not want, identifying bad links to SIAs that are nearly identical to DABs (e.g. via Dab Solver), preventing people from accidentally creating duplicate DAB pages where a SIA exists (e.g. via a redlink from an edit preview), and avoiding busy work deleting "X (disambiguation)" redirects when they are technically no longer DABS. ---- Patar knight - ^chat/_{contributions} 08:12, 27 July 2018 (UTC)
• Yes, per persuasive recent comments of User:Uanfala and User:Swpb, and general arguments of Thyrrdulf. These can indeed be labelled ((R to set index article in absence of disambiguation page)) or whatever, and automatically tracked in a category, for those who want to keep analyzing these. Also Swpb's point is that we have zillions of redirects from misspellings and the like... these are functional (they help many readers) and not harmful in any way. --Doncram (talk) 00:24, 28 August 2018 (UTC)
• Question - how many SIAs does this fail for? What's not obvious from this discussion is that this proposal will never work for many (most?) SIAs as there is already a dab, so a redirect will be impossible (e.g. many name SIAs). Does anyone have the numbers of SIAs with matching dabs? I do sense a desire to have SIAs be treated in some sort of a consistent way with dabs for the benefit of readers. Would the simplest way to get consistency for readers be to fix tighten the definition of SIAs to be proper lists (inclusion criteria, references) and convert the rest to dabs? Can't get to the moon by climbing a tree. SIAs are the only pages I know of that evade WP:V. This proposal layers a hack to enable a search hack without addressing the underlying inconsistency. Fixing the root cause would be consistent. Widefox; talk 09:19, 31 August 2018 (UTC)
  • "How many SIAs does this fail for?" - Zero. Any SIA with a matching disambiguation page should be listed on the disambiguation page (entirely independently of this proposal), so searching for "(disambiguation)" will land on a page that either has the content the searcher is looking for or directly links to it. Whether SIAs should exist is spectacularly further outside the scope of this discussion than the usual FUD that is being thrown, well done. Thryduulf (talk) 10:17, 1 September 2018 (UTC)

Thryduulf, I'm actually asking if anyone knows the number of SIAs that cannot have such a redirect as they have a dab. If you don't know (I don't), then your comment isn't adding light as zero is the wrong number. We cannot bot generate them all either (unlike dabs), as the creation rule would be complicated with many SIAs e.g. List of people with surname Smith are titled so far from their dab Smith that it isn't obvious that the existing redirect Smith (disambiguation) excludes creation of a redirect Widefox; talk 14:32, 1 September 2018 (UTC)

Questions and answers

There is a lot of repetition and misunderstanding above, this is an attempt to clarify what is actually being proposed:

What will be created?

Redirects ending in (disambiguation) that point to set index pages, but only where no disambiguation page for this title already exists.

What will happen to existing pages that end in (disambiguation)?

Nothing.

What will happen if there is a disambiguation page and a set index page for the same title?

The title with (disambiguation) will either be the disambiguation page or a redirect to it, exactly as it currently is.

Will we have to change links to set index articles to go via these links?

No. No changes to any links are required.

What changes will be made to disambiguation pages?

None. There will be no changes to disambiguation pages as a result of this proposal.

What changes will be made to set index articles?

None. There will be no changes to disambiguation pages as a result of this proposal.

How will the definitions of disambiguation pages and/or set index articles change?

They won't. No changes to any definitions are being proposed because no changes are needed.

Will this conflate disambiguation pages and set index articles?

No. No changes will be made to either sort or page.

Will this mean more work for editors who maintain disambiguation pages and/or set index articles?

No - overall it will mean less work as when disambiguation pages become set index articles or vice versa then no changes will need to be made to redirects.

Will this mean more work when creating disambiguation pages?

No. There will be no changes to this at all.

Will this mean more work when creating set index articles?

Only if you want to - you will now be allowed to create a redirect to it ending in (disambiguation), but you will not be required to do so as it can be created by another editor and/or a bot.

Will this mean more work when deleting disambiguation pages?

No. If a disambiguation page is being deleted to make way for a set index article at the same title then there will be less work as the redirect can remain. If there is a set index article at a different title, then the redirect will need to be retargeted (about the same amount of work as deleting it). In all other circumstances there will be no change.

Will this mean more work when deleting set index articles?

Yes, but only a small amount and not in all circumstances. If the set index article is being deleted to make way for a disambiguation page at the same title then there will be less work as the (disambiguation) redirect will not need to be created separately. In other cases there will be one additional redirect to speedy deletion under criterion WP:CSD#G8.

Will this mean more work when moving disambiguation pages?

No, nothing will change about this.

Will this mean more work when moving set index articles?

A small amount in some circumstances, none in other circumstances. If there is no disambiguation page for the title involved and the move is not to make way for a new disambiguation page at the old title then one redirect will need to be retargetted, this will be done by a bot if the old title will now redirect to the new one. In other circumstances then there will be no change.

Thryduulf (talk) 14:23, 29 June 2018 (UTC)

• Oppose. There is a lot of repetition and misunderstanding on your part above too. The clarify what is actually being counter-proposed: I suggest coming up with a (SIA) or similar qualifier for this purpose, if one is needed for the hypothetical problem perceived, rather than use one that indicates the target is something it isn't. -- JHunterJ (talk) 16:52, 29 June 2018 (UTC)

  Applicable reasons from the reasons for deletion on WP:RFD: #2: The redirect might cause confusion. #5 The redirect makes no sense, such as redirecting Apple to Orange. -- JHunterJ (talk) 16:54, 29 June 2018 (UTC)

  Note to closing admin: The bolded oppose immediately above duplicates JHunterJ's bolded "No" !vote in the main section above. Thryduulf (talk) 19:47, 29 June 2018 (UTC)

  All repetition on my part is because the same questions get asked over and over and over again and answer get ignored - as again here. (SIA) will not help at all because the entire point is that it would require people to know what sort of page it is before visiting it - this is not predictable and can change. There is no evidence presented anywhere by anybody (despite repeated requests) that these redirects actually mislead anyone or cause any confusion - rather there is evidence that they are useful: The existing redirects are frequently used (over 50 times a year in some cases), get created by many different people and every time they are nominated at RfD people state that they are useful to them. It has also been repeatedly explained above in multiple ways why this proposal does make sense - these redirects are used by people looking for a list of articles related to their search term from which they can navigate to the article they want to read, something that is the precise purpose of both disambiguation pages and set index articles. Also note I have struck your bolded oppose for two reasons: 1. this set of questions and answers isn't something you can support or oppose. 2: You have already expressed a bolded opinion above. Thryduulf (talk) 17:46, 29 June 2018 (UTC)

  This is obviously a waste of time in this venue. The issue should be reformulated (and keep that FAQ with it, perhaps in a collapse box), and posed to village pump where the audience is broader. — SMcCandlish ☏ ¢ 😼 17:58, 29 June 2018 (UTC)

  Your answers are unconvincing, that's not the same as being ignored. SIAs do not have the exact purpose of disambiguation pages. If they did, they would be disambiguation pages and formatted as such, and we wouldn't be having this discussion. There is no evidence that those inaccurate redirects are more useful than the accurate (SIA) ones (if they are useful to the readership at all). Kindly do not alter my talk page comments again; I do indeed oppose what is actually being proposed. If you would like, you are welcome to strike your own comments, since you have already expressed them above too. -- JHunterJ (talk) 17:59, 29 June 2018 (UTC)

  For at least the 10th time, because you are repeatedly refusing to listen to the answers, the difference between SIAs and dab pages is not relevant to the way people use these redirects. Any differences for other purposes is irrelevant - as has been explained in detail multiple times already. If you do not understand why (SIA) redirects will not solve the issue then you have completely failed to understand the entire issue and your opposition should be ignored. I'm inclined to agree with SMcCandlish that repeatedly talking to brick walls here is getting to be a waste of time, and moving to the village pump where people will actually take the time to understand what is actually being proposed (rather than throwing up straw men) would be a good move. Thryduulf (talk) 19:47, 29 June 2018 (UTC)

  Your failure to convince me doesn't mean I haven't understood it. You would like to declare that (SIA) will not solve the issue. That you repeat yourself so often can just as easily be used to claim that you have completely failed to understand the problem with your proposal, despite the repeated explanations of it that you refuse to hear. But that's not it. There's a proposal here, and it has no consensus, and rather than work with that, you and SMcCandlish would rather paint everyone else as bad faith using whatever alphabet soup of WP shortcuts are handy. We disagree. -- JHunterJ (talk) 19:53, 29 June 2018 (UTC)

  Quote either of us making an accusatoin of bad faith. And if you'd prefer that names of policies and guidelines be spelled out for you in long form, and their applicability spelled out to you as if you were a noob, that could be done, but I think everyone would hate it, including you. — SMcCandlish ☏ ¢ 😼 01:13, 3 July 2018 (UTC)

  Once again entire reason that (SIA) will not work because the entire point of the proposal is that there should be one string that enables users to find what they are looking for (a list of articles related to their search term) whether that list is a disambiguation page or a set index article without needing to know in advance what sort of page it is. If disambiguation pages are found via (disambiguation) and set index articles via (SIA) then one must know in advance whether the page you want is a disambiguation page or a set index article. This is extremely user unfriendly. Thryduulf (talk) 20:01, 29 June 2018 (UTC)

  If someone is looking for foo, then the string to use would be foo. That is about as user friendly as one can get. -- Tavix ^(talk) 21:35, 29 June 2018 (UTC)

  Yup Tavix, my understanding is that this is asking to support a navigation path for those that want to go against the flow of navigation per WP:PRIMARYTOPIC. Similar to adding signs for the wrong way in a one-way street, I, for one, do not think it sensible to put limited resources into discussing signage for the counter-flow. Widefox; talk 23:13, 29 June 2018 (UTC)

  You miss the (repeatedly explained) point again. These redirects are used by people who know (or suspect) the article they are looking for is not the primary topic and do not want to navigate via an article they know to be one they do not want to read. Especially, but not exclusively, for people who are using slow and/or expensive connections this is a serious consideration. Using the example above of someone wanting the article about the ship called HMS Victory that was launched in the early 17th century - I know that this is not the primary topic as that is a very famous mid 18th century ship, but I don't know the title of the article. Now I could load the 56k article I don't want to read, wait for it to load, follow the link at the top and get to a page that lists all the ships of that name. Or I could navigate direct to the page that lists all the ships directly - but is that a disambiguation page or is it a set index page? I don't know - but HMS Victory (disambiguation) will get me there regardless of what it is called. Or if that redirect is deleted I might get to where I want to go (if I guessed right that it was a disambiguation page), or (if I guessed wrong) then I might get a page telling me that Wikipedia doesn't have what I'm looking for (which is obviously incorrect) and that I might want to create an article, or I might get a page inviting me to search, or I might (if I'm lucky) get directly to search results, which might or might not actually help me find what I'm looking for. The one-way street analogy is completely incorrect because there is no counter flow involved and nobody is going the wrong way, the correct city street is this: A person who can only travel on streets they know the exact name of is standing at a road junction looking for a small side street, they know it has "HMS Victory" in the name but they don't know it's exact name, they can see there is a large arterial road ahead of them just called "HMS Victory", but there is no way to know if it connects to the side street they are looking for. They do know there is a handy square though that has connections to all streets with "Victory" in the name with handy signposts telling the names, but they don't know this square's name. Experience has told them though that it will either be called "HMS Victory (disambiguation)" or there will be a narrow alley with that name that has both a direct connection to the square and a signpost with its name on. So obviously you would go to that alley or direct to the square - especially as it costs both time and money to travel on streets and bigger streets cost more. However some people prefer to demolish the (disambiguation) alleys because they the alleys weren't originally intended to help people find things, and helping people is somehow bad - instead they insist that you have to pay for the alley, only to find that it's now a dead end, return to where you started from, pay more to travel on the big road unnecessarily before they are allowed to get to the square. Thryduulf (talk) 14:15, 30 June 2018 (UTC)

  Yes, that's exactly what I said already - a navigation path not per WP:PRIMARYTOPIC. ie not 1. majority: -> primary topic and not 2. minority: -> primary topic -> dab -> article), but 3. not documented (subset) of minority: -> dab -> article, so this is a subset of a minority of users. The assertion that this is a big issue is just that, by definition of primary topic this is a subset of a minority of users, with only anecdotal evidence of demand presented. WP:IDHT describes exactly this situation. Widefox; talk 20:32, 30 June 2018 (UTC)

  That is easily resolved by retargeting HMS Victory (disambiguation) to the relevant disambiguation page. I have done just that. -- Tavix ^(talk) 15:27, 30 June 2018 (UTC)

  Agree Tavix, HMS Victory (disambiguation) is now a routine WP:INCDAB per consensus - much better than the previous invalid target of the SIA. Great example of how the common case of an SIA and dab has no need at all for incorrectly targeted redirects. If I've understood correctly, we probably all agree on that case. Thryduulf, I think the assertion that editors who work on disambiguation are trying to make navigation harder is bad faith, I certainly see no evidence of that, and as someone who's created 100x more redirects than deleted, you should really strike that. Widefox; talk 10:33, 2 July 2018 (UTC)

  It's much more likely that a reader (who has no interest in dabs/SIAs) would put something like "Hms victory 17th century" in the search box and then click on the HMS Victory (1620) result. DexDor ^(talk) 08:40, 1 July 2018 (UTC)

• Comment I note that what was formerly routine CSD G6 (or G8) deletion of these redirects that fail to target dab pages, has more recently been made controversial, so therefore ineligible for CSD. I assume, logically, that the result of this RfC would be binding in deciding yes/no if they can return to uncontroversial CSD G6 (routine redirects with no substantial history). Widefox; talk 23:03, 29 June 2018 (UTC)
  • (ping Xezbeth re CSD comment) Widefox; talk 23:27, 29 June 2018 (UTC)
  • Spelled out consequences for the decision here: the recent against old consensus discussions at RfD (one is linked above) that have pushed this hack have been preventing previously uncontroversial routine redirect deletions. The old consensus is clear, and the result here will either confirm or change that consensus. The limited consensus at those RfCs should be overridden as WP:LOCALCONSENSUS with this. I repeat, those pushing a hack should not WP:IDHT (which is the tone this has descended to), and get in the way of the broader consensus. Widefox; talk 10:27, 30 June 2018 (UTC)
    • Yes and no. The discussions at the RfDs have shown that there is no current consensus for or against these redirects. Whether there ever was a consensus against them (outside the local consensus of the disambiguation project) is arguable, but the status quo ante of this discussion is no consensus and that will remain unless this RfC comes to a consensus (which is looking likely given the repeated IDHT from those opposed). Thryduulf (talk) 14:43, 30 June 2018 (UTC)
      • Thryduulf, there are other ways to reduce the page size/time to assist navigation generally that don't need a change to the database, and may have more general scope and use cases. For example:
      • 1. Mobile client already has the option to save pages for offline, so users need not download the primary topic a second time.
      • 1.1 Browsers can enable more aggressive cache revalidation generally to reduce/eliminate the need to download a primary topic that is a client cache hit (Chrome about:flags has options for that, which I use generally without issues, but haven't specifically tested this on WP server setup)
      • 2. Progressive load option (e.g. up to lead)
      • 2.1 Specific navigation version of that
      • 3. Search

Maybe some have been proposed before, I don't know. Widefox; talk 10:10, 2 July 2018 (UTC)

Extended content

Moved to User talk:Widefox#WP:AC/DS awareness notice. — SMcCandlish ☏ ¢ 😼 01:21, 5 July 2018 (UTC) Note to all participants here A WP:AC/DS awareness notice has been posted on my talk User talk:Widefox#WP:AC/DS awareness notice by User:SMcCandlish regarding this, so that applies to everyone here equally. Comments on the suitability and neutrality of that are welcome on my talk (or here). Widefox; talk 22:26, 2 July 2018 (UTC) Except it doesn't apply that way at all. See WP:AC/DS and the Template:Ds/alert documentation; that exact template must be delivered per-individual in user talk – despite the fact that pages like the one we're on now already have DS-related banners atop them. I've repeatedly challenged this at WT:ARBCOM and other places as ridiculous WP:BUREAUCRACY, to no avail. This is why I've opened up the "Wikipedia:Village pump (proposals)#Bot to deliver Template:Ds/alert" proposal today. Every time Ds/alert is used by anyone manually, it just leads to more drama. Consequently, hardly anyone will use the template, and DS is thus not working. — SMcCandlish ☏ ¢ 😼 00:37, 3 July 2018 (UTC) "Except it doesn't apply that way at all." as you are aware of the correct usage, why didn't you post it on everyone's page? What made you chose one editor you disagreed with? I have equally informed all above. Just posting to one editor's page in the knowledge it applies to all editors equally singles out one editor without treating all editors equally. Mind boggling that you do this while saying you don't agree with it. Widefox; talk 07:22, 3 July 2018 (UTC) This is off-topic here. Try User talk:Widefox#WP:AC/DS awareness notice. — SMcCandlish ☏ ¢ 😼 01:21, 5 July 2018 (UTC)

• I'll repeat what I said in mid-thread above, since it got rapidly buried there: The way to resolve this is with a more clearly-formulated RfC at WP:VPPOL, our broadest-input venue for this sort of thing. Some participants here calling a WP-wide process like WP:RFD a "local consensus" then denying that the handful of editors who spend any time at all at WT:DAB isn't a local consensus but overrides repeated consensus discussions at RfD, is just not tenable. The fact that RfD discussions a long time ago were in agreement with the WT:DAB regulars but now are not isn't an indication of anything other than that consensus can change. Whether it has is an open question, and it's not going to be settled by another week of people here pretending that their questions and concerns haven't been addressed, and turning the discussion cicular. Just have proper RfC about it, in a place that anyone actually pays attention to. Ping all previous respondents. Easy-peasy. If a broader RfC that's not an echo chamber agrees with the WT:DAB crowd and against all of Thryduulf's arguments and mine (which Thryduulf beat me to and articulated better), I'll happy go along with that. But not with a tiny local consensus telling one of our WP-wide XfD processes that it is the local consensus. That's raw nonsense. — SMcCandlish ☏ ¢ 😼 01:38, 5 July 2018 (UTC)

@SMcCandlish: No other editor has stated or implied that this RfC is poorly formulated or in the wrong venue. The RfC, placed by me, asks a straightforward Yes/No question and is on the Talk page of the guidance that will be modified as a result. The RfC has attracted substantive comment from a number of editors and, in my opinion, we need to wait a little longer to make sure all interested editors have had their say before asking for a close. There has not previously been definitive guidance on this matter: a short period of quiet and calm won't do any harm. Shhhnotsoloud (talk) 07:52, 5 July 2018 (UTC)

Last I looked, I'm perfectly capable of stating and implying things on my own without someone else doing it for me first. Thanks for your concern, though. — SMcCandlish ☏ ¢ 😼 08:00, 5 July 2018 (UTC)

@Shhhnotsoloud: As you seem to think it is needed, I endorse SMcCandlish's comments - RfD discussions are a wider consensus than WT:DAB, and this RfC is getting very circular due to a couple of WT dab regulars making it so. Thryduulf (talk) 10:45, 6 July 2018 (UTC)

Right. Your arguments were unpersuasive, and you kept repeating them when you were told they weren't convincing, and it's the audience that made you do it. OK. -- JHunterJ (talk) 12:45, 6 July 2018 (UTC)

Yes, because the "audience" as you put it, didn't say "I have heard your answers and disagree with them because XYZ" they said "but you haven't answered this question" every time it was answered and then for good measure asked it again with a few more straw men. It's absolutely classic refusal to hear what is being said. Thryduulf (talk) 13:08, 6 July 2018 (UTC)

We disagree.[6] Yes, yours is an absolute classic refusal to hear what is being said. -- JHunterJ (talk) 14:58, 6 July 2018 (UTC)

Except you aren't saying anything but repeats of arguments that have already been addressed, and asking the same questions over and over again after they've been answered. There's only one IDHT pattern happening here, and it's not coming from Thryduulf. — SMcCandlish ☏ ¢ 😼 07:52, 9 July 2018 (UTC)

Is it surprising that there's strong opposition to treating SIAs the same as dabs (with this respect), when the consensus is in bold WP:SIANOTDAB. To me, hanging on to these vestigial redirects is just pointy. I don't buy the arguments for a minute, and they have no consensus. The fact that the redirects don't seem to be officially recognised as a dependent page is the issue. That should be fixed anyhow. It's actually quite a minor, obscure issue and as such redirects can never be created to all SIAs, this is really an example of WP:BURO or bike-shed effect considering the number of dabs missing redirects, which could be bot fixed today. Widefox; talk 13:01, 30 August 2018 (UTC)

"Is it surprising that there's strong opposition to treating SIAs the same as dabs" yes, because for reasons repeatedly explained, having a (disambiguation) redirect to a SIA is not treating SIAs and DABs the same in any way that has any impact what so ever on any of the differences between SIAs and dabs mentioned on any page about either of them. Your "redirects can never be created to all SIAs" argument is another example of a deliberately misleading statement - for the purposes of this argument there are two categories of SIAs - ones which share their name with a dab page and ones which don't. As repeatedly explained, where there is a disambiguation page and a non-disambiguation page with the same name the title with "(disambiguation)" should be either the location of the dab page or a redirect to it - the SIA will be linked from the disambiguation page so anyone searching using "(disambiguation)" will find what they are looking for whether it is the dab or the SIA. Such redirects can easily be created by a bot. Where there is not a dab page with the same name then the "(disambiguation)" title can be created as a redirect to the SIA allowing people to find what they are looking for using "(disambiguation)" - these can also be easily created by a bot - using trivial logic once the redirects to dab pages are created (if "Foo (disambiguation)" exists, do nothing) or more complicated logic before then. Not that this has ever been about requiring such redirects, merely allowing them. Thryduulf (talk) 10:13, 1 September 2018 (UTC)

Is you proposal to create all these redirects manually? please see "Question - how many SIAs does this fail for?" there are SIAs that have titles far from the ambiguous term e.g. Smith and the SIA listing the surnames List of people with surname Smith. I don't think it's a fatal complication, but it misleads others to incorrectly state that this is a simple/zero cost like adding new redirects with a bot. A bot would create List of people with surname Smith (disambiguation). e.g. 2 Rangers F.C. (disambiguation) and SIA List of association football clubs named Rangers - what do you propose?. All dabs can have redirects (or be at that title), SIAs being different can't or wouldn't be desirable. Widefox; talk 14:56, 1 September 2018 (UTC)

This proposal is not (and has never been) to create these redirects by bot, it is to allow them to exist however they are created. I agree that redirects like List of people with surname Smith (disambiguation) seem silly but how to determine which (if any) should be created by a bot is a question for a separate discussion after we've established that these redirects are useful in at least some cases. Thryduulf (talk) 15:49, 1 September 2018 (UTC)

My point being the concept of using redirects to discover all lists that may be used for navigation is generally unwieldy and unworkable. If it won't work generally, one can't rely on it, so it's a hack. (I'm willing to concede if SIAs were split into dab like, and list like, then the dab like ones may be useful to be considered as dabs, possibly converting them to dabs is easier. Maybe that could be achieved by tightening up on SIA - inclusion criteria / minimum length / arrangement with a matching dab - I don't know. Instead of proposing solutions, the problem needs defining IMHO.) Widefox; talk 13:33, 16 September 2018 (UTC)

Proposed mitigation 2

If the answer to this RfC is Yes then I propose the following guidance:

A redirect Foo (disambiguation) may target a page that is not a disambiguation page but an article that could perform a disambiguation function, such as a set index article or a list, but such a redirect is not always required. Where there is such a redirect, place an appropriate ((redirect-notdab)) hatnote at the target article:

((redirect-notdab|Foo|SIA)) "Foo (disambiguation)" redirects here, although this page is a set index article and not a disambiguation page.

((redirect-notdab|Foo|list)) "Foo (disambiguation)" redirects here, although this page is a list article and not a disambiguation page.

This mitigation is offered in the spirit of compromise, taking into account that some contributors to the discussion think users find such redirects useful, but others think they are not necessary and their existence creates confusion. Short support/oppose comments (perhaps with improvements) are welcome, without re-running the arguments above. Shhhnotsoloud (talk) 07:35, 9 July 2018 (UTC)

• Support. Seems fine to me. The goal is to get readers to the content they seek, not force them to learn nit-picks of our internal page classification system and dump them at a directly misleading "Wikipedia does not have an article ..." red-link notice. — SMcCandlish ☏ ¢ 😼 07:52, 9 July 2018 (UTC)
• Why would a redirect that will never be used require a hatnote? Hatnotes are ugly and deface articles, they should be used sparingly and not just for the sake of it. —Xezbeth (talk) 12:09, 9 July 2018 (UTC)

  Hatnotes serve multiple purposes. The fact that most of them are for disambiguatory navigation doesn't mean that's the only purpose to which they're put. See, e.g., Category:Hatnote templates for names; ((Correct title)), etc. And "are ugly and deface articles" is obviously an extremist view no shared by most of the community or we would not use hatnotes. — SMcCandlish ☏ ¢ 😼 21:21, 9 July 2018 (UTC)

• Strong oppose, that goes against the purpose of WP:HATNOTES since there is not another article we would be guiding them to. Besides, it just makes the problem worse because it suddenly becomes "reader-facing", when it wasn't before. -- Tavix ^(talk) 13:31, 9 July 2018 (UTC)
• Comment. The first sentence of the proposal sounds good, but the idea of adding hatnotes is problematic. To what Tavix had said I could add that such hatnotes would confuse the vast majority of readers: even if the text didn't involve obscure metawiki distinctions, a hatnote is a surprising place to be declaring the type of the article. But what problem is that proposal addressing? It's the potential for readers following the redirect to be confused by arriving at a page that is not technically a dab page, right? Again, I don't believe these readers know or care about the distinction, but if we do want to reach out to them without giving everybody else a head scratch, then the simplest thing would be to turn the relevant redirects into soft redirects and add the text there; this will, however, add an additional step in the navigation process, so I don't think it will result in a net benefit. – Uanfala (talk) 22:46, 9 July 2018 (UTC)

  Not dumping readers at a directly misleading "Wikipedia does not have an article ..." red-link notice would, actually, be a net benefit. — SMcCandlish ☏ ¢ 😼 23:45, 9 July 2018 (UTC)

• Oppose, meh, there's been many opportunities to convince of a problem, so in the absence of convincing reasoning there's no need for a solution. Consensus is clear WP:SIANOTDAB. My gut feeling is improved search is the answer, not ASTONISHING redirect hacks. Widefox; talk 13:48, 30 August 2018 (UTC)

Proposed mitigation 3

I propose the following guidance:

A redirect Foo (disambiguation) may only target a disambiguation page. To assist reader navigation for the issue where a qualified redirect is needed to indicate in the search box that a list of similarly-named topics exist, but that list is an article and not a disambiguation page, use Foo (set index). The savvy reader who picked up on the existence of (disambiguation) as a qualifier for disambiguation pages will also pick up on the existence of (set index) as a qualifier for set indexes. This has the benefit of maintaining the reasons for redirects: making sense and not causing confusion (see WP:RFD#DELETE) while still providing the mnemonic search box assistance provided by the current redirects. -- JHunterJ (talk) 14:24, 9 July 2018 (UTC)

• Oppose. Unlike Foo (disambiguation), Foo (set index) is unheard of. Set indexes are disambiguated by type when necessary, not due to the fact they are set indexes (eg: Foo (ship), Foo (surname)). Furthermore, unlike WP:INTDABLINK, there is no need for links to set indexes to be fixed, so there is no need for an "overarching" disambiguator for set indexes. This would result in the creation of thousands of unnecesssary redirects for very little to no benefit. -- Tavix ^(talk) 16:10, 9 July 2018 (UTC)

  Before (disambiguation), "Foo (disambiguation)" was unheard of too. True, set indexes are disambiguated by type or better by "List of Xs named Y"; this would not be the article title, but the title of a redirect used for the readers who supposedly rely on incorrect "(disambiguation)" redirects. This would at least be "less incorrect" than targeting "(disambiguation)" redirects at non-disambiguation page list articles. I agree with you that there is little to no benefit; this mitigation is a compromise with those who claim that there is a benefit. It would also only result in the replacement of (disambiguation) redirects to set index articles, not the creation of new ones. -- JHunterJ (talk) 12:17, 10 July 2018 (UTC)

  I'm confused why you feel a compromise is necessary? -- Tavix ^(talk) 13:18, 10 July 2018 (UTC)

  To collaborate on the encyclopedia. I can agree that there is little to no benefit and then also agree to a mitigation with editors who agree there is a benefit, if one can be found that doesn't lie to the reader. There's no benefit to putting falsehoods in the encyclopedia, even if falsehoods might benefit some readers. -- JHunterJ (talk) 15:20, 10 July 2018 (UTC)

  You're mistaking the argument, JHunterJ; it's not whether it's "unheard of" in titles, its that there are too few (and will remain too few) pages in the class for it to ever become "heard of". — SMcCandlish ☏ ¢ 😼 22:03, 11 July 2018 (UTC)

  As long as we're just basing the arguments on anecdotal assumptions, I've heard of it. -- JHunterJ (talk) 13:39, 12 July 2018 (UTC)

  Of course you have; you're a long-term editor deeply involved in DAB and other aspects of page titling and management, not a random non-editor reader. You're the guy who works at Hatcha's Cafe in Mora, New Mexico, who's convinced that Hatcha's is famous because he's heard of it. Heh. — SMcCandlish ☏ ¢ 😼 01:32, 13 July 2018 (UTC)

  You've hit my point without realizing it. This whole hubbub is based on the anecdotal reports of someone seeking out (disambiguation) redirects to reach non-disambiguation-page set index list articles. The proposal is thus from your Hatcha's employee. Heh heh. -- JHunterJ (talk) 12:22, 13 July 2018 (UTC)

  But no one has very proposed such a rationale. Users quickly learn "(disambiguation)", but they are not trying "to reach non-disambiguation-page set index list articles"; they probably don't even know those exist, because they're rare and obscure. They're trying to reach a disambiguation page. If we don't have one but do have an SIA that provides the actual topic they're looking for, the responsible and user-friendly thing is to give it to them, since SIA serve both a content and navigation purpose. It also has the side benefit of introducing them to this thing called a set-index article, probably for the first time. On the main point, consider this analogy: I need a Philips screwdriver. You have a screwdriver that has a reversible head with both Philips and flat-head. But you refuse to let me use it because I asked for a Philips one, not a reversible two-way one. See the problem? — SMcCandlish ☏ ¢ 😼 12:29, 13 July 2018 (UTC)

• Oppose per Tavix; there are too few set index articles for "(set index)" to ever enter the awareness and memory of the average WP reader so the entire premise is faulty. I've been here over 12 years and have only run into a few SIAs; basically, no one knows about them but the small clan who work on them. — SMcCandlish ☏ ¢ 😼 20:35, 9 July 2018 (UTC)

  Right, no one knows about them yet, just as no one knew about disambiguation until Wikipedia disambiguation pages proliferated. The argument for using the incorrect disambiguation redirects is that savvy Wikipedia readers learned about them. The argument here is in agreement, that those Wikipedia readers are still savvy and would also recognize and then use the (set index) when needed. -- JHunterJ (talk) 12:17, 10 July 2018 (UTC)

  Addressed this above already. [7] — SMcCandlish ☏ ¢ 😼 22:04, 11 July 2018 (UTC)

• I like the first sentence. As to the rest why can't a redirect point to the type of article as in the examples provided by Tavix eg: Foo (ships), Foo (surnames)) --Bejnar (talk) 23:40, 11 July 2018 (UTC)
• Oppose This is worse than pointless - the first sentence harm the encyclopaedia by unnecessarily requiring people to know what type of page they are looking for before they have read it, meaning it is harder for people to find the content they are looking for. The rest of it offers absolutely no benefits at all (although to be fair it wont harm anything either). Thryduulf (talk) 09:13, 15 July 2018 (UTC)
• Comment meh, yes it's at least a logical solution regarding the redirect name. It's also a compromise which fits with the compromise that is SIAs. It's maybe more consistent to make it mandatory to disambiguate by type per Tavis, but in the absence of a well defined problem (like dabs not allowing direct linking which doesn't apply to SIAs, it's just a solution looking for a problem. The logic of extending (disambiguation) redirects to lists as they may also be useful is flawed, that's all I know in the absence of a simple, convincing argument. I can't help that there's vocal proponents of an unsupported hack, but let's not enshrine hacks into supported features. Widefox; talk 13:42, 30 August 2018 (UTC)

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

Is a disambiguation page necessary if it consists of multiple entries without a Wikipedia article that is directly related to the disambiguated term? E.g. High on Life consists of multiple entries but none of which has a Wikipedia article. The editor whose username is Z0 06:51, 27 July 2018 (UTC)

Suppose you remember a song title called “High on Life”, but you don’t remember who made it or when. That page would present you with possibilities to explore, even if there is no stand-alone article on the song you’re thinking of. That’s a perfectly fine disambiguation page, as long as there’s a blue-linked article for each entry that at least mentions the disambiguated term. — Gorthian (talk) 07:33, 27 July 2018 (UTC)

@Gorthian: Sounds reasonable but that could lead to the addition of non-notable (or promotional) content. I mean if it's okay to add anything without a WP article, that's going to create a mess. In the RFD for High on Life, I brought it up but editors seem to agree with you that disambiguation pages could be created even when the entries don't consist of a correlated article. The editor whose username is Z0 09:49, 27 July 2018 (UTC)

It's not a disambiguation page, it's a redirect that has been hijacked. The redirect is under discussion at Wikipedia:Redirects for discussion/Log/2018 July 25#High on Life, please respect WP:MULTI. --Redrose64 🌹 (talk) 11:17, 27 July 2018 (UTC)

And please see MOS:DABMENTION, the part that covers those entries while prohibiting non-notable ones. -- JHunterJ (talk) 12:39, 27 July 2018 (UTC)

Is it possible for someone to clarify in the guideline page that creating a disambiguation page based on just 2-3 entries that do not have a WIkipedia article matching the term, is acceptable? Courtesy ping @Ss112: @Hayman30: @Station1: The editor whose username is Z0 08:14, 2 August 2018 (UTC)

WP:DABSTYLE: "Each bulleted entry should have a navigable (blue) link, normally as the entry itself (see the previous bullet), or in the description if the entry is red-linked or unlinked.

Rarely should a bulleted entry have more than one navigable link; including more than one link can confuse the reader." I believe that covers it. Nobody's saying or going to promote the idea that creating a disambiguation page where there is nothing navigable for any of the entries is acceptable. Ss112 08:20, 2 August 2018 (UTC)

Some editors think there should not be a dab page if there are zero articles on WP for a term, which indicates the guideline isn't clear enough. It should be amended to include a list of requirements for creating a disambiguation page just for clarity. The editor whose username is Z0 09:56, 2 August 2018 (UTC)

Yeah, that would probably be harmless and likely helpful. I would watch out that the summary does not subtly but substantively change the requirements, though (intentionally or otherwise). — SMcCandlish ☏ ¢ 😼 04:17, 3 August 2018 (UTC)

RFC

• I propose adding the following (in bold) to Wikipedia:Disambiguation#Disambiguation_pages: "...focused on getting the reader to their desired article. A disambiguation page is warranted when there are two or more blue-linked entries and there is no primary topic." The editor whose username is Z0 07:58, 3 August 2018 (UTC)

• Comment Should we amend the wording to state explicitly that the blue link may occur anywhere in the entry? That would clarify that items such as
  • High on Life, an album by QFX

on dab High on Life do count as blue-linked entries, even though the initial words aren't linked. — Certes (talk) 12:27, 3 August 2018 (UTC)

Sounds reasonable. The editor whose username is Z0 15:00, 8 August 2018 (UTC)

Yes; that definitely qualifies. — SMcCandlish ☏ ¢ 😼 15:51, 3 August 2018 (UTC)

• But this questions is already treated, with much more detail, in the section Wikipedia:Disambiguation#Disambiguation page or hatnotes? which is further up the page. – Uanfala (talk) 12:53, 3 August 2018 (UTC)

  That section is about primary topic and suggests a disambiguation page may be created when there is no primary topic, but it does not say exactly what is needed to warrant a disambiguation page. Hence, causing some editors (1) to think a dab shouldn't be created if two or more title-matching WP articles do not exist. In some way I agree with their sentiment that non-notable entries shouldn't be added but then it could be helpful, so why not? @Uanfala: The editor whose username is Z0 14:58, 8 August 2018 (UTC)

  May be better to cross-reference that section, though it doesn't actually hurt to reiterate the central matter. We have to remember that people enter these page sections via shortcut links and do not read them top-to-bottom. So, something like: "... is warranted when there are two or more entries with blue links and there is no primary topic (see § Disambiguation page or hatnotes?, above)." No need for boldfacing. — SMcCandlish ☏ ¢ 😼 15:51, 3 August 2018 (UTC)

  That could work. The editor whose username is Z0 15:01, 8 August 2018 (UTC)

• "Disambiguation in Wikipedia is the process of resolving conflicts that arise when a potential article title is ambiguous..." Non-notable topics are not potential article titles, and therefore should not appear on dab pages. Dab pages were never intended to be indexes or search engines. In fact, a bad dab page can get in the way of the search engine. DABMENTION always used to be understood to be for cases where a notable topic was covered in another article. It's only recently that a few editors have been adding non-notable songs that often are barely mentioned at all in the article, sometimes with no more info than what appears on the dab page itself. If we start adding non-notable entries, we would need to add hatnotes or dab pages for every spouse of a state legislator who happens to be mentioned in passing and shares a name with a notable person, not to mention every Main Street and every bit actor and fictional character who shows up in a cast list. I recommend we change "topic" to "notable topic" in DABMENTION. Station1 (talk) 19:27, 4 August 2018 (UTC)

  That's one interpretation. But the other (and mine) is that any non-notable information should be removed from the article that mentions it. Disambiguation pages are not the place to make the value judgment of when a mention in the encyclopedia is notable and when it isn't. If any given song title were unambiguous, would it be a possible ((R from song))? Yes. So if ambiguity exists instead, it should be listed on the disambiguation page. If it's truly non-notable, it shouldn't be mentioned (or "noted") in the encyclopedia. -- JHunterJ (talk) 12:16, 10 August 2018 (UTC)

  And there is a third view, the one that's practiced by the majority of wikipedians: whether to create a redirect or add an entry to a dab page is a decision whose basis lies somewhere between the two extremes of notability ( = deserving a separate article) and noteworthiness ( = deserving a mention within an article). If we have meaningful encyclopedic content on a topic, then it doesn't matter whether the topic is treated in an article of its own, or together with other topics as part of an artile – this warrants having an entry on a dab page. Conversely, every reasonably developed article is likely to make passing mentions of a large number of entities, about which we will never have any content apart from these passing mentions – almost no-one would dream of adding these to dab pages. – Uanfala (talk) 13:07, 10 August 2018 (UTC)

  I suppose that's effectively my view as well. If it could have been a redirect except for the ambiguity, then it can be on the disambiguation page with the the ambiguity, because that's the technical problem disambiguation pages exist for in the first place. The passing mentions that no one would think to add to dab pages (or think to put into the Search Wikipedia box) wouldn't be redirects or disambiguation page entries. But song on albums would be things that people with think to make redirects for (and think to put into the Search Wikipedia box), so would be things people would think to put on disambiguation pages. Certainly if no one thinks to put it on a disambiguation page, the question of whether to remove it from that disambiguation page won't come up. -- JHunterJ (talk) 15:49, 10 August 2018 (UTC)

  I agree that we should use the same criterion for creating redirects as for adding dab entries. The disagreement, I guess, comes from the application: I don't think that a song that is merely mentioned in an article about an album warrants a redirect. If we don't have any meaningful content about the song, and if we can't expect out article to ever say anything about it other than the fact that it was on a certain album, then this is not the kind of thing that readers will be putting in the search box and it is not the kind of thing we would want to integrate in the encyclopedia's navigational system. – Uanfala (talk) 20:34, 10 August 2018 (UTC)

  Do such redirects, when created, get deleted through RfD? (I don't know, I don't follow RfD closely, but the existence of ((R from song)) indicates to me that they're viable without expectation of going anywhere else.) -- JHunterJ (talk) 20:44, 10 August 2018 (UTC)

  WP:R#PLA seems to indicate that that there should generally be content at the target but unlike WP:DABMENTION doesn't appear to mandate it. I'd suggest treating the criteria for a redirect as the same as DABMENTION. Crouch, Swale (talk) 12:57, 29 August 2018 (UTC)

• Disambiguation pages when there are only two pages is bloat. Totally unnecessary, even if there is no clear primary topic. Choose the one that was created first if you have to, but put one at the base name, and a dab link from it to the other one at disambiguated title. A dab page with two entries serves no purpose. This is what WP:TWODABS used to say, and that guidance should be restored. --В²C ☎ 18:33, 10 August 2018 (UTC)

  A dab page with only two entries when neither of them is primary serves the purpose of disambiguating two topics when neither of them is primary. WP:TWODABS, when it said that, did not have consensus. -- JHunterJ (talk) 19:04, 10 August 2018 (UTC)

  But User:Born2cycle is right though. It's bloat because instead of OneTopicX and OtherTopixX (2 pages) we have TopicX, TopicX (disambiguation) (an INTDAB redirect), OneTopicX and OtherTopixX (4 pages). Not only are there now twice as many pages to maintain, but every searcher is a loser. If a user searches for "TopicX" when they wanted OneTopicX, they are diverted via the dab page. If a user searches for "TopicX" when they wanted OtherTopicX, they are diverted via the dab page. 100% of searchers have to click through a dab page. Any mechanism to determine a primary topic (page views, Google hits, page creation date, toss of a coin) and the consequent use of hatnotes is better than a 2 dab page with no primary topic. (That's clearly my opinion, not what the guidance says). Shhhnotsoloud (talk) 10:20, 11 August 2018 (UTC)

  But that's wrong, because it's another click but a needed click because neither topic is primary. There are twice as many pages to "maintain", but of course "maintaining" a two-entry dab and its redirect is pretty light work (until there are 3+, and then it's no longer the subject of this discussion). And every disambiguation page without a primary topic means that every searcher arriving is a "loser" in your terms; should we randomly or otherwise compel a primary topic for every title? Since we're not ready to abolish base-name disambiguation pages in the general case, there's no reason to abolish them in the two-entry case. If one of the two entries is primary, great; otherwise, there's no problem here. -- JHunterJ (talk) 12:19, 13 August 2018 (UTC)

  Borderline primary topics already inconvenience readers when wikilinks to the wrong targets creep in. (For example, I recently diverted several soldiers from a company to a company (military unit), and species from a family to a family (biology).) Such errors would be more widespread if we put ambiguous topics at the base name just because there is only one other meaning. Certes (talk) 14:43, 13 August 2018 (UTC)

  JHunterJ, when there are only two topics and neither is clearly primary than both are close to being primary. That is, about half the people searching with that name are seeking one or the other. An article with a hat link to the other article serves as a disambiguation function just as well as a dab page for disambiguating. The distinction between a primary topic and a nearly-primary topic is, by definition, minor. So treating either nearly-primary topic as primary should not be a big deal. But, in the general case, with more than two entries, the odds of landing someone on the sought page drops from about 50% rapidly to around 33%, 25% and lower, and now the "losers" are usually redirected to a dab page from which they have to click again to get to their desired page. That's why TWODABS was originally treated as a special case. I don't think most participants understood and appreciated this critical distinction when it was changed. Certes, yes, topics that are nearly-primary have the same problem as topics that are actually primary - the possibility of linking to the wrong topic exists when you have an article rather than a dab page at a base name. I've seen others use this as an argument to not have primary topics at all. I think the nearly-primary topic in a TWODABS case having this issue is just a minor side issue. If we tolerate it for all genuine primary topics, tolerating it for nearly primary topics as well is relatively minor. --В²C ☎ 18:16, 13 August 2018 (UTC)

  Born2cycle, when there are only two topics and neither is primary (clearly primary is just primary), then both are far from being primary. That is, neither is close to being much more likely than any other single topic (the other single topic), since we just stipulated that they are close to each other. The distinction between being a primary topic and a non-primary topic (there is no such thing as "nearly primary") is, by definition, major. -- JHunterJ (talk) 18:48, 13 August 2018 (UTC)

  JHunterJ, okay, point taken, but I think we need to reasonably interpret that wording in a 2-dab situation. In a 3-dab situation we could easily have a primary topic having 51% of the likelihood being hit (e.g. 51/25/24). But in a 2-dab 51/49 situation the same 51% likelihood is not good enough to be PT? I don't see how it makes sense that a topic with 51% likelihood can be primary topic if it's the only use, or one of three or more uses, but not if there are exactly two uses. Seems absurd to me.

  Okay, I know, in a 3-dab 51/48/1 situation the 51 also wouldn't technically meet the PT criteria, but I really don't get the point. I mean, the whole point of primary topics is to take people directly to the article they are likely to be seeking, and to minimize those who are not immediately taken to the article they are seeking. I guess it comes down to whether one thinks being sent to a different article (with a hatnote to the article being sought) is less desirable than being taken to a dab page. I see those two results as being functionally equivalent in terms of desirability. I personally don't mind being taken to the wrong article if that article has a hat link to the article I am seeking; I mean, that's no worse than being taken to a dab page. Or, if you prefer, they are equally annoying. So I don't get the motivation to prefer sending everyone to a dab page with links to their sought articles rather than taking about half to the article they seek and the other half to an article with links to their sought articles (hatnote macros support links to multiple alternatives and thus can serve as mini dab pages at the top of any article at a base name). --В²C ☎ 20:46, 13 August 2018 (UTC)

  "wouldn't technically meet the PT critera" = "wouldn't be a primary topic". "Technically" is unneeded there. See Certes' note of 14:43, 13 August 2018 (UTC) above for part of the point. -- JHunterJ (talk) 12:20, 14 August 2018 (UTC)

  Yes, but we can still treat a nearly-primary topic as if it was primary and put it at the base name. As to Certes' point, I addressed that already, just above, in the latter part of my 18:16, 13 August 2018 comment. I, for one, think serving readers better is more important than making it easier to identify some very small percentage of imperfect links. Taking about half of those searching with a given term directly to the article they seek is better than taking none to the article they seek, even if it also means taking the half not seeking that article there. That half is no worse off than they would be had they been taking to a dab page; either way they are just one easy click to the article they seek, while the first half is clearly worse off at the dab page than if taken directly to the article they sought. Similarly, for the imperfect links that take a user to the wrong one of two nearly-primary topics, that user is also only one hatnote click from the right article, which is why I refer to the original link as being merely "imperfect".

  More generally, why the affinity for dab pages? Any search that lands a user on a dab page is a failure in our system. Note that Google searches don't take users to dab pages because they know that's not what users are seeking. For example, enter "obama presidential campaign" at Google.com and click on feeling lucky (their equivalent of our "Go"). You don't get taken to our TWODABS dab page; you get taken to Barack Obama presidential campaign, 2012. But in WP search hitting Go on the same search string takes us to the TWODABS dab page at Barack Obama presidential campaign. Putting one of two nearly-primary topics at the ambiguous base name helps us avoid sending users unnecessarily to dab pages like this - that's a good thing. --В²C ☎ 17:29, 14 August 2018 (UTC)

  There are two things that are good to avoid:

  1. Unnecessarily sending users to disambiguation pages
  2. Unnecessarily sending users to articles about topics they weren't seeking

  The primary topic guidelines, including those for base-name two-element disambiguation pages, balance those two. Landing on a two-element disambiguation page when neither is primary is not unnecessarily landing there. -- JHunterJ (talk) 19:52, 14 August 2018 (UTC)

  Landing on a two-element dab page when neither is primary is unnecessarily landing there, for at least about half of those that land there. It's unnecessarily landing there for them because if we put the one of the two articles they are seeking at the basename, they will land on the article they are seeking. As to the second half, the only alternative is also "good to avoid", unnecessarily landing on an article they're not seeking, which is no more important to avoid than unnecessarily landing on a dab page, so it's a wash from their perspective. That is, the second half is unnecessarily landing on one or the other; it's unnecessarily either way because no matter which it is, it could be the other one. However, it is necessary to land the second half on the article they are not seeking in order to improve the experience of the users in the first half, while not compromising the experience of the second half. When we can reduce the number of users unnecessarily landing somewhere good to avoid by about half of those in question, that seems like an obvious improvement to the encyclopedia user experience. After all, the users in the second half for a given topic might very well be in the first half when searching with other terms with two uses. On these encyclopedia-improving grounds, I think a strong WP:IAR argument can be made for ignoring the current wording at WP:PRIMARYTOPIC and WP:TWODABS in these two-near-primary-topic situations. Of course, it would be better to make explicit changes in the relevant policy and guidelines accordingly. --В²C ☎ 21:13, 14 August 2018 (UTC)

  I suggest that sending a reader to the wrong page is worse than sending them to a dab. They may not realise that they're on the wrong page, even with a hatnote (which won't appear on all devices). Articles also tend to take more data and time to load than dabs. No one can quantify this difference and say that sending one reader to the wrong page is as bad as sending exactly 3.142 readers to a dab, but there will be some such ratio ${\displaystyle r}$ and I suggest that ${\displaystyle r>1}$. Let ${\displaystyle p}$ be the proportion of readers seeking the most popular topic (best candidate for PT). Then a primary topic page will cause no inconvenience to ${\displaystyle p}$ of the readers, and ${\displaystyle r}$ times more inconvenience than a dab for the other ${\displaystyle 1-p}$. So we should prefer a dab where ${\displaystyle r(1-p)>1}$, i.e. ${\displaystyle p<1-1/r}$. For example, if you think that ${\displaystyle r=3}$ (a number I plucked out of thin air) then we should have a dab page when we estimate that fewer than ${\displaystyle 2/3}$ of readers seek the primary topic. Of course, there may be other factors to consider when making this editorial judgement, but that logic may be useful for simple cases. Certes (talk) 00:07, 15 August 2018 (UTC)

  Great analysis. Thank you! We disagree about sending a user to the wrong page being significantly worse than sending them to a dab (assuming the "wrong page" has a hat note link to the "right" page). Or, if you prefer, I think ${\displaystyle r=1.1}$ or so. But as long as ${\displaystyle r<2}$ then placing the article for either of two topics at the base name is at worst a wash if both of their ${\displaystyle p}$ values are near ${\displaystyle 0.5}$, which is the situation we're addressing, and is progressively better than a wash the lower below ${\displaystyle 2}$ the value of ${\displaystyle r}$ is, and the higher above ${\displaystyle 0.5}$ the value of ${\displaystyle p}$ is for the article placed at the base name in a given case (which it will be in most cases - both topics being exactly ${\displaystyle p=0.5}$ is highly unlikely). --В²C ☎ 17:07, 15 August 2018 (UTC)

two dabs history

• FWIW, WP:TWODABS "used to say" exactly this, that when neither topic was considered primary, a disambiguation page is used.[8] If someone else wants to identify whatever other language was put in there after what it used to say, we can go look to see if there was any consensus for that change. I believe there was not. -- JHunterJ (talk) 18:54, 13 August 2018 (UTC)

  This change and this discussion seem relevant. Certes (talk) 19:17, 13 August 2018 (UTC)

  This is what I mean by what it used to say, and what I think should be restored: "If there are two topics for a term but neither is considered the primary topic, then a disambiguation page may be used; an alternative is to set up a redirect from the term to one of the topics, and use disambiguation links only." [9] --В²C ☎ 20:58, 13 August 2018 (UTC)

  Let's highlight that appropriately: "If there are two topics for a term but neither is considered the primary topic, then a disambiguation page may be used; an alternative is to set up a redirect from the term to one of the topics, and use disambiguation links only." -- JHunterJ (talk) 12:20, 14 August 2018 (UTC)

  • Yes, that's the situation we are talking about: two topics neither of which is primary. What TWODABS used to say was that a dab page may (not must) be used. --В²C ☎ 16:48, 14 August 2018 (UTC)
  • Hmm, the concept was apparently originally introduced at MOSDABS on July 21, 2007, and there it clearly applied only when one of the two was primary[10]. --В²C ☎ 17:00, 14 August 2018 (UTC)

• Personally per WP:2DABPRIMARY I would give a lower threshold to PT status since the DAB means readers seeking either have a click, rather than just those looking for the alternative topic. However I would note that DAB pages are easier to load and probably should be used in situations where its unclear which topic is primary. Crouch, Swale (talk) 07:30, 28 August 2018 (UTC)

1

Is there consensus to make the proposed change? The editor whose username is Z0 04:49, 26 August 2018 (UTC)