Template talk:P

Should we rethink changing this to redirect to Template:Wikidata property link?

Back in 2014, @Bluerasberry and Pigsonthewing both proposed changing this redirect from {{Smiley}} to {{Wikidata property link}} ([1] and [2] respectively). It was closed with consensus to keep the current redirect. I was just about to propose changing it again when I saw those discussions (as I keep trying to use this redirect to link to Wikidata and getting unexpected smileys - maybe not a bad thing! - and this is a frequently-used shortcut on other wikis). Given that it's been 4 years, and Wikidata is used a lot more now than it was then, I'd like to do a quick straw-poll to see if it's worth re-opening the discussion. So I'm pinging @Red Slash, Frietjes, Technical 13, and Thryduulf as those that commented on the past discussions: have your opinions changed here? Thanks. Mike Peel (talk) 22:38, 3 September 2018 (UTC)

• Given that the template has almost 5000 transclusions, any retargetting would need to be preceded by a bot run to subst or otherwise alter those. The template also appears to be in common use for its current purpose, e.g. it was used on 7 August at user talk:Double sharp#9 years of editing by user:Kpgjhpjm, so changing the purpose would be very disruptive to those users so I'm very strongly opposed to the suggested change. I do understand that it's annoying when templates on different projects do different things (e.g. for many years the "tl" template on Wiktionary related to the tagalog language) but sometimes you just have to live with that - {{prop}} is probably the best you'll get here unfortunately. Thryduulf (talk) 23:00, 3 September 2018 (UTC)

  Reading the old RfD discussion, I found mention of what is now T6547, that would resolve this issue, and note that it remains open and high priority but has had no activity since September 2015. You may wish to try getting some attention on that. I also note that the template is apparently meant to be substituted, so it is going to be even more widely used than the transclusion count alone suggests. Thryduulf (talk) 23:05, 3 September 2018 (UTC)

  @Thryduulf: I would love to see that phabricator ticket solved, but that's not going to happen any time soon regardless of how much attention we draw to it (maybe put it forward in the next community wishlist round?). The transclusion count doesn't provide any sort of useful indicator for how often this is being used unless you can say when the last substitution of all of them took place. Thanks. Mike Peel (talk) 23:39, 3 September 2018 (UTC)

  I don't know when the last substitution took place (is there any way to know that?), but when this was discussed 4 years ago there were circa 1500 transclusions, there are now circa 5000. That means there has been an average of around 875 new uses per year since 2014, assuming that there have been zero new substituted uses (which seems rather unlikely). That's a massive amount of use for a template redirect. Thryduulf (talk) 00:01, 4 September 2018 (UTC)

• Support use of P for Wikidata The current use of Wikidata surpasses any available evidence of the use of this shortcut for the smiley (Q11241). Wikidata is a priority project which already has great fundamental influence on the development of all Wikimedia projects. The need of the Wikidata community outweighs the need for smiley users to have a shortcut. I would like to make an eminent domain (Q166332) claim on this template. While I respect that there is a community base using it, Wikipedia seeks to use resources for the best purpose and not to dictate all future practice based on precedents set. Already the {{Q}} template is in place. Wikidata's basic functions are Q as a code for "item" and P as a code for property, so once those two are covered, there would be no need for more shortcodes. Blue Rasberry (talk) 13:43, 4 September 2018 (UTC)
  • Wikipedia is not Wikidata. What is the evidence that the templates on en.wp need to match the templates on Wikidata? Where is your evidence that this "need" (if it exists) to use this as a shortcut to the Wikidata property template on the English Wikipedia exceeds the disruption that will come from changing a very highly used template? What has the Wikidata item for a smiley got to do with anything? Why is the convenience of editors using smiley templates in the same way they've been used for 12 years less important the the convenience of editors of a different project? Thryduulf (talk) 17:39, 4 September 2018 (UTC)

  @Thryduulf: I will acknowledge that what you are saying is the conventional practice.

  Some wiki insiders emphasize the distinctness of Wikipedia versus Wikidata, but I think that the trend since the beginning has been for the projects to be more interconnected over time. The Wikimedia community collectively is investing millions of dollars to integrate Wikidata and Wikipedia. Lots of English Wikipedia articles, such as Telescope Array Project, use Wikidata infoboxes, so I think saying Wikipedia is not Wikidata is incorrect for these experimental cases and more inaccurate as more of these uses roll out. The need to match Wikidata templates is the need to provide the most good to the most users in the most instances of need. The utility of this template for smileys is relatively less than its utility as the most obvious way to bring Wikidata properties into various languages of Wikipedias.

  To consider numbers, the Smiley has about 5000 transclusions now after 12 years compared to Wikidata entity link having 25000 after about 4 years. The clunky name "Wikidata entity link" is not sustainable but has been the choice in the absence of access to Q and P templates. I take this extensive use as evidence of present need and a likelihood of increased use in the foreseeable future. Thoughts? Blue Rasberry (talk) 18:43, 4 September 2018 (UTC)

  You have explained why Wikidata is valuable, which was never in doubt. You have not explained why the templates need to be named the same on both projects - why will using a clunky name or a redirect like {{Prop}}, {{WDP}}, {{property}} or {{wpl}} (among others) prevent any of the benefits you say will come from integration? Why is the transclusion of information from Wikidata into infoboxes (which doesn't use this template) matter? If there wasn't a pre-existing use then I wouldn't have an objection, but changing an existing use is very disruptive - not just from changing 12 years worth of uses, but also caused when the template no longer does what everyone who uses it expects it to do - even more so if they substitute the template as intended. The presence or absence of a smiley can significantly alter the meaning of a message, so you have to include in the disruption all the unnecessary confusion and misunderstanding that will result. For the sake of two or three more characters the benefits just don't come close to exceeding the problems. Thryduulf (talk) 21:39, 4 September 2018 (UTC)

  @Thryduulf: "P" is the name which does not need to be taught. Anyone who uses Wikidata knows that properties are called "P" and that to call the template we say P.

  The point is not only about making the templates the same on Wikidata and English Wikipedia - this is about making the template the same on all Wikimedia projects in all languages. I admit this is premature. It would be an overstatement to say that the Wikidata community is organized to propagate out P and Q templates. Wikidata's Template:Property and Template:Wikidata entity link at the bottom show which Wikimedia projects have a community of Wikidata integrators, so the count is about 40 projects of about 800 projects at this time. Still, I foresee that more projects will adopt Wikidata integration, and that "P" and "Q" should be the universal way to signal this across languages.

  The disruption from changing the smiley on English Wikipedia is much less than the disruption of rerouting a fundamental function of Wikimedia projects to accommodate a such a small demographic in using a cosmetic function. I cannot demonstrate this with data but I believe that the number of users who have attempted to use "P" in English Wikipedia to link to Wikidata properties is already greater than the number of users who use the smiley emoji. Blue Rasberry (talk) 14:15, 5 September 2018 (UTC)

  That still does not explain why there is a need for the template to be the same on all projects. If you are discussing Wikidata properties on the English Wikipedia it takes seconds to learn that it is done with {{prop}} not {{p}} - I know this as it took me that long to learn that on Wiktionary templates are linked using {{temp}} not {{tl}} (which at the time was a template relevant only to Tagalog). Thryduulf (talk) 14:56, 5 September 2018 (UTC)

  @Thryduulf: I advocate for reforming the "tl" designation also to some universal norm. There might be exceptions, but in general any function which is cross-language and cross-project should be as simple as practical and uniform.

  The majority of stakeholders in English Wikipedia content do not speak English. The majority of current English Wikipedia users do not use English as a first language. You are imagining an English userbase on English Wikipedia when this is not the case. You are timing how long it takes you as an English speaker to navigate to an abbreviated customized redirect. A better test would be for you to go to an unfamiliar script and then find the template. Try to find the way to post Wikidata templates to the property for these -

  • Hebrew
  • Korean
  • Vietnamese

  If you tried, I expect you would succeed with Hebrew and Korean in seconds, despite not knowing the language. You just type "template:p" in the search box. With Vietnamese this fails. A person might be able to learn a different way for every Wikimedia project in only seconds but multiplying a few seconds times the entire userbase of Wikidata stakeholders is a huge amount of time and a barrier to participation. The smiley userbase is not anywhere near as large as the number of users who are seeing Wikidata templates and who are trying to use them. This larger community of use is the justification for the change here. Blue Rasberry (talk) 16:16, 5 September 2018 (UTC)

  But why should one project bend to the conventions of another? In the exceedingly unlikely event that I needed to discuss Wikidata properties, using templates, on a wiki where I don't speak the language then I would go to wikidata:Template:P (and get redirected to Wikidata:Template:Property) then follow the interlanguage link in the sidebar rather than assuming that every wiki worked the same (because in my experience of being an admin at en.wp, en.wiktionary, and commons (lapsed) and an editor at Wikidata they differ in very significant ways, having their own culture, jargon, abbreviations, etc), and/or I would just look at what template people on that wiki were using. I would have strongly opposed making {{tl}} on Wiktionary the same as on en.wp because every other language used a template named for that language's ISO 639-1 code. Until it was superceded by a lua module, {{P}} on Wiktionary was used to indicate the plural of a word; {{Q}} there remains widely used as a way to consistently format quotations. Thryduulf (talk) 16:36, 5 September 2018 (UTC)

  @Thryduulf: The projects are a social construct which have the goal of supporting the community of readers and editors. The point of templates and other tools is to support actual use, and not to dictate rules to follow for the sake of precedent or anything other than stakeholder benefit. To answer your question, a project should bend to the conventions of another when it is in the benefit of both projects to do so.

  I am asserting that right now there is a large userbase interested in editing across Wikimedia projects using Wikidata content. There is not a strong line between projects; English Wikipedia like other Wikimedia projects integrates strongly with Wikidata and these ties are growing. Projects should be distinct in limited use cases to adapt to small userbases. The situation here is that there are a few dozen people who use shortcuts for an emoji and the competing demand is hundreds of users, likely to grow to thousands within a few years, who jump across Wikimedia projects to edit similar content without regard for language or project.

  If this template does not get modified now then in a few months someone else will come along and propose this again, with proposals likely to repeat until something changes. I expect the change to happen either after data demonstrates the need or this discourse reaches its end. Maybe the level is now or maybe it is in the future. To me, I see small and slowly growing use for the emoji, and big and growing use for cross-Wikimedia applications which English Wikipedia has adopted. Blue Rasberry (talk) 17:06, 5 September 2018 (UTC)

  I think we'd better leave this discussion here as it's clear we have fundamentally opposite views about the importance of pretty much every single relevant aspect here. I remain utterly unconvinced that the benefits you assert outweigh the disruption change will cause, that standardisation is necessary, that there is a need for users unfamiliar with en.wp to have any significant need to use templates to discuss Wikidata properties, and fundamentally disagree with you about the scale of differences between projects and the desirability of steamrollering them. Equally the arguments I'm presenting in favour of my position are clearly not convincing you to change your mind. Thryduulf (talk) 17:18, 5 September 2018 (UTC)

  @Thryduulf: Thanks. It is a easy for me to accept a difference of opinion because you asked me direct and thoughtful questions and you do a great job of repeating back what I say in your own words. I can accept the status quo for as long as the supporting evidence for a change is lacking. If I am mistaken then evidence will always be lacking, and if I am not mistaken then the evidence will appear when the time is ripe for change. I think you are right to want to see supporting evidence that the value of benefits presently outweighs the cost of change and I agree that most of what I have is claims without evidence. Thanks for pushing back to protect the wiki because that is what makes the system work. All of your examples and that smiley transclusion data were relevant and made a case for your position. Blue Rasberry (talk) 18:22, 5 September 2018 (UTC)

@Mike Peel: Why did you choose to start this discussion here and not at template talk:Smiley (as the banner at the top of the page instructs) or at WP:RFD where proposals to retarget redirects are normally discussed. Thryduulf (talk) 17:41, 4 September 2018 (UTC)

• @Thryduulf: This is meant as a prelude to see whether a discussion is worthwhile, not a discussion itself, which I was planning on posting to RfD if people's views have changed. I'm happy for you to move it to template talk:Smiley if you'd prefer. Thanks. Mike Peel (talk) 18:03, 4 September 2018 (UTC)
  • Moving would be disruptive, but I've invited people reading that page to comment. My views have not changed because the underlying facts have not changed - this is still highly used for its current purpose and the disruption change would cause would be significantly disproportionate to the benefits. Thryduulf (talk) 18:25, 4 September 2018 (UTC)

FYI

 – Pointer at Wikidata. Nikkimaria (talk) 00:11, 6 September 2018 (UTC)

• Opposed. We already have a template for property, and this one is under full page protection as a heavily used template. The fact that the template is often substituted means the usage is significantly higher than the already large and growing transclusion count.
  Bluerasberry: Suggesting that the little Wikidata project has "eminent domain" over other projects is rather offensive. There was initially a strong consensus to enable Wikidata on Wikipedia to try out the grand promises made about it. After years of experience with Wikidata, the community now has one RFC consensus against linking to Wikidata in the body of an article, and a second RFC where half were against Wikidata in Infoboxes. Arguing that Wikidata has some sort of "eminent domain" over Wikipedia is an excellent way to rally a consensus for a Phab task to completely rollback deployment of Wikidata calls from Wikitext. I understand that you and some others imagine a wonderful future where much of Wikipedia is converted to be based on Wikidata. However there are others who do not view that as wonderful or even acceptable, and who have no intention of heading in that direction. Alsee (talk) 07:00, 28 September 2018 (UTC)

  @Alsee: I apologize if I spoke in an offensive way. I did not intend for this to become an emotionally tense issue, and if you think of something I can do or some commitment I can make to make for a friendlier and more civil conversation then suggest it.

  I do claim that English Wikipedia has a higher duty to be accessible beyond only editing users for whom English Wikipedia is their home project. I feel that English Wikipedia should also to present itself for translation to other language Wikipedias and partner with Commons, Wikidata, and the rest. If a few people on English Wikipedia would be inconvenienced by a changing smiley icon versus the rest of the Wiki userbase finding the interface inaccessible, then I think an argument for greater benefit to the larger wiki community is fair. I understand that it is not fair to compare a collaboration with Wikidata to benefit to the entire off-ENWP world, but I do think that Wikidata greatly increases inter-wiki information exchange in a way that ENWP alone is not attempting to do.

  You state no opinion on the smiley. So far as I can and that your interest here is not about the smiley, but about maintaining any barriers which exist to make integration between Wikidata and English Wikipedia more difficult. If you wanted to start an RfC on rolling back Wikidata then I would be among the first to participate. I think any conversation that anyone starts, regardless of the outcome, will lead to the eventual compromise which will bring the most benefit to the most people. I do not expect you or any other individual to engage in Wikidata but I would like to maintain the mutual respect that exists among Wikimedia projects. Blue Rasberry (talk) 14:17, 1 October 2018 (UTC)

  Bluerasberry thanx, and I apologies if my comment sounded like we were going in an uncivil direction. I do understand why some people see increasing Wikidata integration as a beneficial future. I respectfully disagree, and I'm more than happy to work together in other areas.<br\>I understand why you think taking over this template for Wikidata would be a good thing. I understand how you could jump to the conclusion that I was only opposing to "maintain barriers" and make things "difficult". However if you can Assume(Better)Faith, I see an unclear but sizable percentage chance that Wikidata-on-Wikipedia rolls backwards rather than rolls forwards. In that case, hijacking a heavily used template for Wikidata would just be another case of pointless disruption caused by good-faith, persistent, but failed efforts to push Wikidata here. So how about this: Neither of us has a reliable Crystal ball. How about we set aside this template-initiative until the community can actually manage to reach a consensus on what we want to do about Wikidata-on-Wikipedia? If and when there is a consensus that Wikidata is the future of Infoboxes, *then* I support taking over this template to make that work easier. Alsee (talk) 15:21, 1 October 2018 (UTC)

• Support, per Bluerasberry. However, if there was some evidence of consistency, I'd oppose, but there is none. All singular-use alphabetical characters in either the emojis themselves, or the code, are "D", "O", "P", and "S", per Template:Smiley#Usage. {{D}} redirects to {{Db}}, {{O}} doesn't exist, {{P}} is the outlier we're discussing, and {{S}} redirects to {{Strikethrough}}. If "the letter 'P' is in one of the emoticons" is the only reason for its creation and existence (is there a better reason?), then creating and using {{O}} is an acceptable alternative. Also, a 5k botreq is small as far as bot jobs go. As a 'feeler', I think this proposal has merit and I would like to see the broader consensus.

Since one of the concerns is how disruptive this would be, I think I'll run through transclusions and count the number of unique users.   ~ Tom.Reding (talk ⋅dgaf)  13:55, 28 September 2018 (UTC)

Some rough usage statistics on 90% of transcluded pages (~10% weren't counted due to PUA chars, protected, or couldn't find sig after use):

• # of unique users: ~539
• # of total uses: ~7083
• # of users accounting for the top ~25% of usage: ~    3
• # of users accounting for the top ~50% of usage: ~  10
• # of users accounting for the top ~75% of usage: ~  22
• # of users accounting for the top ~90% of usage: ~  66
• # of users accounting for the top ~95% of usage: ~188
• # of users accounting for the top ~99% of usage: ~467
• the top half of users (270) account for ~96% of usage

So the template has a steep 'usage curve', with a few 'heavy users' and many more intermittent/infrequent users. If/when this RfC/RfD/etc. is reformalized, at least the top 22 users (top 75%) should be notified. To me, it doesn't seem like this is as disruptive as others have claimed. If the usage distribution was quite a bit flatter, then it might be considered disruptive.   ~ Tom.Reding (talk ⋅dgaf)  16:27, 28 September 2018 (UTC)

@Tom.Reding: This data is great! Is there a brief explanation for how you generated this? Blue Rasberry (talk) 13:59, 1 October 2018 (UTC)

Bluerasberry, I used AWB to do a quick-n-dirty search for all instances of \{\{\s*[Pp]\s*[\|\}][\d\D]*?\[\[[Uu][Ss][Ee][Rr]:([^\|\]]+)[^\r\n]{2,255}\d\d:\d\d, \d+ [JFMAMJJASOND]\w+ \d\d\d\d \(UTC\) on all transcluded pages, where the capture group $1 is the user's name, saved the usernames to file, then tallied them up in Excel.   ~ Tom.Reding (talk ⋅dgaf)  16:23, 1 October 2018 (UTC)

Tom.Reding nice work. However your data appears to weigh heavily against the proposal here. I took a look at {{Wikidata_property_link}}, as well as every template that redirect to it. Assuming I didn't botch the job, all of them combined have a total transclusion count of 676. If you're willing to do an equivalent analysis of number-of-unique-users and users-accounting-for-top-x%, I'd be willing to wager the number of people served by converting {{P}} to Wikidata isn't that far off from zero. And that's without considering that your analysis above missed anyone currently using subst:P. So your stats were surely low. Alsee (talk) 15:41, 1 October 2018 (UTC)

Alsee, I don't think it would be a surprise to anyone that {{Wikidata property link}} has fewer users, unique or otherwise, since it's only ~1.5 years old, compared to {{P}}'s 12 year history. And that could very well be an argument in favor of keeping {{P}}. Or it might not; 'just because something is, doesn't mean it should continue to be'. The point is, I think consensus is worth gauging.   ~ Tom.Reding (talk ⋅dgaf)  16:30, 1 October 2018 (UTC)

Correct, I didn't consider subst'd templates, but I just ran a scan and found 0 pages with the case-insensitive \{\{\s*subst\s*:\s*P\s*[\|\}] (nevermind - I think it's impossible to search for subst'd redirects). A handful of pages were protected, so I couldn't check them automatically, but that isn't worth worrying about.   ~ Tom.Reding (talk ⋅dgaf)  21:16, 1 October 2018 (UTC)

• Note that I'm not planning on taking this further myself, so if anyone else wants to start a formal discussion process, please feel free to go ahead. Thanks. Mike Peel (talk) 21:32, 1 October 2018 (UTC)
• Suppot, per the previous proposal; noting that the need to be able to easily link to Wikidata property pages in discussions is increasing. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:14, 14 October 2018 (UTC)

I think it should go to Template:Ping E Super Maker (😲 shout) 01:09, 25 December 2019 (UTC)

What about {{param}}? JsfasdF252 (talk) 15:32, 1 March 2021 (UTC)

Template-protected edit request on 18 January 2021

Use module:RfD/sandbox instead of module:RfD per Template talk:Rfd#Template-protected edit request on 17 January 2021. –MJL ‐Talk‐^☖ 18:32, 18 January 2021 (UTC)

 Not done: We don't use sandboxes in production. File an edit request there when the sandbox actually has the updated content, and it will be handled at that module instead. Elliot321 (talk | contribs) 00:00, 20 January 2021 (UTC)