Shortcuts: WD:PC, WD:CHAT, WD:?

Wikidata:Project chat

From Wikidata
(Redirected from Wikidata:Project Chat)
Jump to navigation Jump to search

Wikidata project chat
A place to discuss any and all aspects of Wikidata: the project itself, policy and proposals, individual data items, technical issues, etc.

Please use {{Q}} or {{P}} the first time you mention an item or property, respectively.
Other places to find help

On this page, old discussions are archived after 7 days. An overview of all archives can be found at this page's archive index. The current archive is located at 2021/06.

Check if redirect[edit]

Is there a template to check if an object is a redirect? Queryzo (talk) 08:36, 7 June 2021 (UTC)

You could add this tool to your own common.js https://www.wikidata.org/wiki/User:You/common.js via this code importScript('User:Matěj Suchánek/checkSitelinks.js'); : it will add a button and tell you whether the wikiarticle contains a redirect.--Bouzinac💬✒️💛 11:16, 7 June 2021 (UTC)
Nice tool too! But I am searching for a template, which gives true for {{Check if redirect|Q105966850}}. Queryzo (talk) 07:52, 11 June 2021 (UTC)
Some page information can only be accessed from Lua, which a template can use as a backend. For querying pages on the same project, there is the Lua Title library (mw.title.new('Q105966850').isRedirect). But this isn't possible for pages on another wiki. Then, there is a hack like mw.wikibase.getEntity(item).id ~= item, but it needs to load all item data. --Matěj Suchánek (talk) 09:46, 11 June 2021 (UTC)

Great, Matěj Suchánek can you insert the new code into Module:Wikidata, so that I can use in Wikitext?

function p.isRedirect(id)
	return mw.title.new(id).isRedirect
end

Queryzo (talk) 18:53, 11 June 2021 (UTC)

✓ Done --Matěj Suchánek (talk) 11:55, 16 June 2021 (UTC)
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. Matěj Suchánek (talk) 11:55, 16 June 2021 (UTC)

Provocation[edit]

In the case: provocation defence (Q3566104), I have merged all of provocation (Q257931) except "All entered languages" of the Q257931 Because i can't add and edit at Q3566104 more languages, how can i add, Thanks --Bmt3s (talk) 07:31, 8 June 2021 (UTC)

Are you trying to merge them? If so, see Help:Merge. We normally merge to the older item not the newer item. — Martin (MSGJ · talk) 08:42, 8 June 2021 (UTC)
It seems that provocation defence (Q3566104) was originally a legal term (with English label "provocation defence") and distinct from the general term. Perhaps a merge is incorrect here. — Martin (MSGJ · talk) 08:44, 8 June 2021 (UTC)
So the merge is done and most statements on the item are about provocation. The subclass one is about provocation defense, as is the enwiki link. It seems the original poster did not understand what was said above. What's more, there is also provocation (Q6132400) so we now have a duplicate and a missing item. --SCIdude (talk) 06:56, 9 June 2021 (UTC)
I have undone the merge. There were a couple of languages which were incorrectly identified, which I have fixed to the best of the ability. Apart from that I think we are done here. The three items are all for different purposes and should not be merged. — Martin (MSGJ · talk) 21:04, 13 June 2021 (UTC)

References for values openly supplied by the individual themselves[edit]

I am involved in a consultative process to learn how Indigenous artists feel about having information about them and their artistic practice stated in Wikidata. As part of this process we are guiding participants as they add statements on their own person item and we collect feedback from them over the course of the process. The response is so far positive and the process is very instructive, but we face certain challenges with references. Sometimes, we can find external references in support for their statements. Other times, these external references don't exist (either because Indigenous artists don't benefit from the same media coverage as mainstream artists, because the information is part of the oral knowledge within an Indigenous community, or because the information isn't documented anywhere, but the artist wishes to make this information available in Wikidata). Among other things, we lack often external references for gender, year of birth and ethnic group, but the artist are stating these values themselves. Is there a property that could be used to source a statement to the individual themselves? Fjjulien (talk) 20:29, 8 June 2021 (UTC)

How should be ensured that it's really the person making claims? If the person provides information in Wikidata, it goes public. So this info could be put on a web or Facebook page and that page referenced. Is that too much to ask for? --SCIdude (talk) 08:10, 9 June 2021 (UTC)
@Fjjulien: While Wikipedia looks down on sources that are controlled by the subject of an article, those sources are okay for Wikidata in circumstances like this. While an Indigenous artist might not have mainstream media coverage, in most cases they will have their own webpresence, whether that's Facebook/Twitter or Soundcloud. If the artist wansts gender/year of birth/ethnic group to be publically known it would make sense to have that information on their own webpresence.
If you discover during your consultative process that there are artist that currently don't have any kind of webpresence, can you tell us more about them? ChristianKl❫ 14:00, 9 June 2021 (UTC)
@SCIdude: "How should be ensured that it's really the person making claims?" – In this particular case, we are guiding the artists over videoconference, and they are indeed providing these values themselves. But I acknowledge there are many other circumstances where it would not be possible to assert if the person has indeed made the claim themselves. Fjjulien (talk) 15:26, 9 June 2021 (UTC)
@SCIdude: @ChristianKl: You are right. Most of the time, these artists have their own website from which certain statements can be referenced. This is entirely possible with properties such as "occupation", often possible for "ethnic group", but more difficult for "sex or gender", "date of birth" or "country of citizenship". Artists do not usually not share this kind of information on their website. However, when we asked in a safe context and explained why this information can be useful for disambiguating entities, we are finding that some are happy to disclose this information. I wish there was a way to source these statements as self-declared by the subject (even though such a reference would not be verifiable). I guess, I could create an item for this consultative process and then use stated in (P248). But that's a rather intricate solution that would only meet the need in these few particular items. If there's no right solution, then so be it. We'll leave the claims without references for now.
Please note – We have been reaching out to performing arts associations, inviting them to collect year of birth and self-identification from their members + seeking their consent for disclosing this information in Wikdata. This has worked really well with the catalog for the project of the Conseil québécois du théâtre (Q100443909). We are currently in discussion with the Indigenous Performing Arts Alliance. We hope that we can build enough trust with the association and their members to perform a similar batch upload. The consultative process described above is part of building trust (and gaining confidence that Wikidata is deemed by Indigenous performing arts community as an appropriate knowledge base for publishing this kind of information about them). Fjjulien (talk) 15:52, 9 June 2021 (UTC)
Can you tell us more about who suggested this project? While it may be acceptable in some languages of Wikipedia's that users write articles about themselves, it's not ok in most. Wikidata isn't designed either as place of primary publication. --- Jura 18:24, 9 June 2021 (UTC)
@Jura: This project was suggested by members of the advisory committee for the project grant Modelling and Populating Performing Arts Data in Wikidata. Members of this advisory committee felt it was not ok for Wikidata to feature mostly white men and they devised a project by which they could identify obstacles that led to the under-representation of Indigenous artists. They thought that, by involving Indigenous artists in the project, they might find a few champions who would then encourage other Indigenous artists to get involved with Wikidata. Wikidata is indeed not designed as a place for primary publication. But if that's an obstacle to greater participation by Indigenous Peoples, then perhaps it should be addressed. Fjjulien (talk) 01:55, 11 June 2021 (UTC)
@Fjjulien: Maybe. But this can only work if an account is linked to an identity. Just imagine I create such an artist account and make false claims. Would that be good for Wikidata? Note I'm not against such "identified" accounts: they would also allow data curation by specialists---their claims could have the same authority as referenced claims (the remaining problem being the inherent difficulty of extracting the author of a claim). --SCIdude (talk) 09:09, 11 June 2021 (UTC)
  • "advisory committee" is that Core Working Group Members? Many are from organizations with their own website. Why not publish the information there and then link that as reference? It also simplifies citations for researchers. --- Jura 09:32, 11 June 2021 (UTC)
@Jura: Indeed, the "advisory committee" is the same as the project's "core working group" (the terminology evolved over time – sorry about the confusion). Members of this group devised this project although not for themselves: for the hundreds of Indigenous artists who were not involved in the project. We put out a call for participation, and then we guided these artists as they created their own items. We used the "user acceptance test" methodology to collect their verbal and non-verbal feedback. We also took time to specifically inquire about their feelings vis-a-vis properties denoting aspects of identity/belonging to a group. So far, we created two items: Mike Alexander (Q107139441), Monday Blues (Q107155126). We were able to rely on these artists' websites to source the ethnic group (P172) statements, but not for citizenship, gender and year of birth: Indigenous artists are using their website to describe their practice, not their feeling about being a citizen of a nation state that has purposely and repeatedly attempted to assimilate and destroy their culture. Hence my question for finding other ways for referencing these statements as coming from the persons themselves. If no mechanism exists to meet this need, then so be it. This will be another finding from this project that we can share back with the Indigenous arts community (and possibly discuss at the WikidataCon). Thanks for your response. Fjjulien (talk) 21:31, 11 June 2021 (UTC)
  • @Fjjulien: There's an RfC about signed statements that's about creating functionality in Wikidata that proposes a way a person or organization can claim to create a statement. It might be worth to read it and comment how you think it interacts with your use-case. ChristianKl❫ 20:44, 12 June 2021 (UTC)
@ChristianKl: Thanks for pointing this RfC out. I will certainly read, and I will comment if I have anything relevant to contribute. Fjjulien (talk) 03:35, 15 June 2021 (UTC)

atomic mess[edit]

Looking at a few elements such as hydrogen (Q556) and lithium (Q568) I notice that the property mass (P2067) is being misused for atomic weight, a.k.a. relative atomic mass. What should be used is a property analog of the item relative atomic mass (Q41377). But I am unable to find an existing appropriate property. Jc3s5h (talk) 12:09, 9 June 2021 (UTC)

You ignored the preceding talk. -DePiep (talk) 19:04, 9 June 2021 (UTC)
@Jc3s5h: These are merely missing the unit, which should be dalton (Q483261). See aluminium (Q663) for where this is done properly. They should also have determination method (P459) qualifiers to specify the conditions under which these (average) masses are determined. Presumably the issues are mainly because the chemical element properties were added very early in Wikidata's history and aren't up to the usual standards in some cases. ArthurPSmith (talk) 13:37, 9 June 2021 (UTC)
Why is "relative atomic mass" a 'misuse'? You mean to say that is not a mass? @ArthurPSmith:: no the unit is not "missing", the relative atomic mass is defined to be dimensionless (i.e., no unit). This is how the source defines it. (If old trem "atomic weight"is confusing, then forget that one. 'Relative atomic mass' is fine) -DePiep (talk) 19:00, 9 June 2021 (UTC)
Is this based on the question of whether a chemical element corresponds to an individual atom or some collection of them? I mean, the dalton mass unit is defined as "1/12 of the mass of an unbound neutral atom of carbon-12 ..." (see en:Dalton (unit)) so that seems the perfect unit for this purpose otherwise. ArthurPSmith (talk) 19:12, 9 June 2021 (UTC)
As indicated in the citation given the lithium item, https://www.degruyter.com/document/doi/10.1515/pac-2015-0305/html the atomic weight is based on the relative abundance of isotopes in natural terrestrial materials. The atomic weights are expressed in daltons. By stating the "mass" rather than the "relative atomic mass" or "atomic weight" we are being fuzzy about whether we are referring to the average atom on the surface of Earth, the most common isotope, or whatever. Jc3s5h (talk) 00:56, 10 June 2021 (UTC)
I don't think so. I think it is agreed that an element item represents the typical mixture of isotopes. For specific atoms create a separate item. --SCIdude (talk) 06:38, 10 June 2021 (UTC)
@Jc3s5h: determination method (P459) relative atomic mass (Q41377) might be suitable then as a qualifier. We generally don't add special properties for every version of how one might measure something. There is a Chemistry WikiProject that might be a better place to discuss this in more detail. ArthurPSmith (talk) 12:08, 10 June 2021 (UTC)

Abbreviation of a name - can it have its own item?[edit]

I created item Q106825653 for the name "Md.", which is an abbreviation used by many authors for the name Muhammad. Another editor merged this item into Muhammad (Q19693229). Is that correct? The author Md. Rizwanul Islam (Q106288113) does not use Muhammad in his works, he uses the abbreviated form Md. as his first given name. His item now states that his given name is Muhammad, but that seems incorrect to me. UWashPrincipalCataloger (talk) 16:32, 10 June 2021 (UTC)

We have items for nicknames like Bob (Q18105736) (which are often used directly as given names) so an abbreviated name that is used the same way seems like it should have an item also, yes. ArthurPSmith (talk) 17:28, 10 June 2021 (UTC)
@UWashPrincipalCataloger: There are even data objects for abbreviations of first names. Please refer Chr. (Q21146553) --Gymnicus (talk) 18:31, 10 June 2021 (UTC)
Is it possible then for an administrator to undo the deletion of the item I created, or do I have to create it again? UWashPrincipalCataloger (talk) 18:41, 10 June 2021 (UTC)
It doesn't need an administrator, but I have reverted @Quesotiotyo's change to Md. (Q106825653). Bovlb (talk) 18:56, 10 June 2021 (UTC)
"Md." belongs as a statement on Muhammad (Q19693229), and perhaps an alias as well. There is no need for it to have its own item. Chemical element symbols do not have their own items. Neither do U.S. state postal abbreviations. There is no separate item for "Jas." or "Tho.", even though those were common abbreviations for "James" and "Thomas" for many centuries, and undoubtedly there are individuals whose first name was only ever recorded that way. If Md. Rizwanul Islam's actual first name is not known, "unknown value" should be used.
--Quesotiotyo (talk) 19:12, 10 June 2021 (UTC)
@Quesotiotyo: I do not think so. Nowhere in the available data is it stated that his name is Muhammad (Q19693229). Instead, only the abbreviation Md. (Q106825653) is used. So we don't know what the abbreviation stands for. The abbreviation Chr. (Q21146553) for example does not necessarily stand for Christian (Q18001597), it can also stand for Christin (Q21132614), Christina (Q1083457) or Q17689481. So if you don't know exactly what the abbreviation stands for, it makes perfect sense to use the abbreviation data object. --Gymnicus (talk) 19:29, 10 June 2021 (UTC)
I agree with --Quesotiotyo. Others use Mhd. No need to make separate items for these. In Turkey we have something called "göbek adı" (belly name) which is a grandparent's name that they give to the newborn. As -generally- it is an old-fashioned name, the youngster does not use it, hiding it behind an initial like M. (It can be Mehmet, Mahmut, Mustafa, Mümin, etc etc etc). --E4024 (talk) 19:33, 10 June 2021 (UTC)
@Quesotiotyo: Chemical symbols and US state postal abbreviations are set by an authoritative oversight body; there is no such body for names generally nor for abbreviations of names (though some countries may have such things). Names are a very personal thing, and there is no reason why they need match anybody else's conception of length, pronunciation, character set or any other criterion you may choose. In the US we have Prince (Q7542) who at one point had a legal name that is not representable in Unicode, and X Æ A-XII Musk (Q93418989) for instance. ArthurPSmith (talk) 12:58, 11 June 2021 (UTC)
  • If his name is not "Md.", why given name (P735) is used at all? "Md." is not a name, so given name (P735) should not be used with a "Md." as a main value. Name is not known = somevalue with a string- or monolingual text-qualifier with a value "Md.". Wostr (talk) 20:19, 10 June 2021 (UTC)
  • If the form generally used is "Md.", I don't really see an issue of having an item for it. Let's not try to guess if it's Muhammad or Muhammed. --- Jura 09:46, 11 June 2021 (UTC)
  • when the given name (P735) is not known, only an abreviation, like here, the value of P735 should be set to "unknown", stated as (P1932) "Md.", as long as the real name is not found... using items like Md. (Q106825653) or A. (Q19803442) ; those are no names... just the way to write them ; linking to such an hypothetic item is wrong in my opinion --Hsarrazin (talk) 13:22, 11 June 2021 (UTC)

filtering the History of an element on IP contributors[edit]

Is there any way to easily filter the history of an element on IPs only ?

On many elements, vandalisms are made by IPs, and on some IPs only vandalize... like Empire State Building (Q9188)

being able to quickly toggle between all edits, and IP edits in history would help analyse whether a specific item is a target for vandals... and thus, know which items need semiprotection...

is there a way or a tool to do so ? or should I propose it on Phabricator ? --Hsarrazin (talk) 13:16, 11 June 2021 (UTC)

I think this could be a good idea for a tool. Fralambert (talk) 15:49, 11 June 2021 (UTC)
I am not exactly sure what exactly you are looking for. If this is just about some IP-username highlighting on page histories in the web UI, you can get this done with some custom CSS in Special:MyPage/common.css. —MisterSynergy (talk) 16:13, 11 June 2021 (UTC)

Repeated informations[edit]

In Lina Medina (Q202307), number of children (P1971) have a problem: as you can see in the image, the number 2 is repeated. It should be "1 and 2", not "1, 2 and 2". Why there is this problem?

https://i.postimg.cc/MZPd8j68/Lina-Medina.jpg

--93.35.184.189 15:28, 11 June 2021 (UTC)

Entity untangling help[edit]

I just stumbled upon a bit of a mess with Q21517518. It seems as though an import of a VIAF ID (P214) went a bit wrong. James E. King (botanist born in 1940) had information about James A. King (sociologist? born in 1940) imported on top of him. After that, automated edits (and manual ones like mine) added ID's and statements like date of death (P570). I have seen some odd use of initials and interesting career moves before but I'm pretty sure that this isn't one of them and it just needs splitting in two. I thought I'd ask for another pair of eyes on it before that though. I also thought it would be useful to highlight an example what can go wrong when bots fight over each other! Aluxosm (talk) 06:56, 12 June 2021 (UTC)

@Jura1: Glad I asked; I would have just split one of them off and kept the "original". I did have a quick look but couldn't find those docs (think I searched for "split entity" or something). Thanks for the pointer! Aluxosm (talk) 12:40, 13 June 2021 (UTC)
Here are the two entities it was split into: James E. King (Q107212758) and James A. King (Q107212764). Aluxosm (talk) 12:46, 13 June 2021 (UTC)

Taxon merge?[edit]

Middle Spotted Woodpecker (Q275591) and Dendrocoptes medius (Q27074888) seem to be duplicates and should be merged, but since I am not a taxon expert and both have quite some history I somehow hesitate to merge them myself. Ahoerstemeier (talk) 09:28, 13 June 2021 (UTC)

I'm not an expert either. But one has parent taxon Dendrocoptes (Q24033503) and the other Dendrocopos (Q723059) which suggests they may not be the same. — Martin (MSGJ · talk) 12:04, 14 June 2021 (UTC)
Thanks, didn't spot that small spelling difference, so they are different taxons, but synonyms. Interestingly, while Dendrocoptes medius catches all the wikipedia links, some wikipedia articles have the wrong taxon name. Ahoerstemeier (talk) 18:53, 14 June 2021 (UTC)

Wikidata weekly summary #472[edit]

Wikidata Community/Diversity Survey 2021 has been published[edit]

Hello,

In March, many of you participated in our first community survey (see announcement). We analyzed and summarized the results -- you can have a look at the slides on Wikimedia Commons.

Your answers helped us to get a better understanding of where editors come from and which gender and age they have, as well as what activities you do on Wikidata.

The results are meant to serve as a baseline to see how the community might (not) change in the future. We also hope that the data might inspire community activities around increasing diversity and give some more insights in who your fellow editors are and what they do.

If you have any questions, please do not hesitate to ask. Cheers, -Mohammed Sadat (WMDE) (talk) 15:10, 14 June 2021 (UTC)

IP Masking Engagement – New Update[edit]

Hello Wikidata community,

The Anti-Harassment Team has updates on IP Masking here.

In this update, there are details on the proposal for sharing IP addresses with those who need access and there’s also an update on tool development. 

Please pass by and check the new details out and give your feedback.

STei (WMF) (talk) 15:19, 14 June 2021 (UTC)

Kazakh in Turkey (Q64427352)[edit]

Kazakh in Turkey (Q64427352). There are Kazakhs in Turkey who came quite recently from their home country. I wonder when they could develop a Turkish version of Kazakh language... Is there anybody here who can remove my ignorance on this issue? --E4024 (talk) 15:26, 14 June 2021 (UTC)

Notable Nigerian Sport Ambassadors[edit]

I have applied for a new project on the above subject matter. you suggestions will be appreciated. the link is https://meta.wikimedia.org/wiki/Grants:Project/Rapid/kid_keen_47/Notable_Nigerian_Sport_Ambassadors  – The preceding unsigned comment was added by Kid Keen 47 (talk • contribs) at 18:59, June 14, 2021‎ (UTC).

Also at Wikidata:Project_chat/Archive/2021/05#Grant_request:_Nigerian_Sports_Ambassador_on_wikidata. --- Jura 08:16, 15 June 2021 (UTC)

Separating Hansard (Q524352) into separate concept items[edit]

Hansard is the name of a long running publication in the UK that provided transcripts of the proceedings of the Parliament of the United Kingdom (Q11010). The name was then used for similar publications in other territories of the British Empire, or the successor countries. Item Hansard (Q524352) represents the modern UK database of hansard and is tied to Hansard (2006–March 2016) ID (P2170). However, the site links and some of the statements link the item to the similarly named publications in other countries. We also have Hansard 1803–2005 (Q19204319) and Hansard (1803–2005) ID (P2015) for a database hosting historic copies of the UK Hansard. I think Hansard (Q524352) needs splitting into an item covering Hansard in general (with potential links to items for each country), Hansard the UK publication from 1812 and Hansard the modern UK database (to link to Hansard (2006–March 2016) ID (P2170)). Does this sound like a reasonable idea? Should I keep Q524352 as representing one of these concepts or move all statements to fresh items and delete the original? From Hill To Shore (talk) 00:42, 15 June 2021 (UTC)

It sounds as though splitting might not leave enough connection between the present-day British Hansard and the legacy version of the same thing; Hansard in general would not be enough to specify which country's parliament is referred to. Additionally, while there will be a great deal of cross-over in membership between the pre- and post-database aera at Westminster, that will not be the case between Canberra (say) and Westminster. I think there ought to be way of grouping the original Hansard with its legacy editions, data, people. while keeping separate the Australian, Canadian, etc. GPinkerton (talk) 07:52, 15 June 2021 (UTC)

Botched police raids[edit]

I am trying to aggregate "botched police raids" https://www.wikidata.org/wiki/Special:WhatLinksHere/Q107238003 I was searching for the phrase on Google and matching them to our entries, but it seems easier to just use my memory, and then find a find a reference using the wording "botched raid" in the news. Can anyone think of others to tag? --RAN (talk) 03:05, 15 June 2021 (UTC)

That looks like a very subjective item. I can see a lot of future arguments about whether a raid that doesn't meet all of its objectives should be called "botched." There isn't any practical definition of when a raid goes from being not-botched to being botched, so views could differ significantly. We will likely end up with cases where one journalist says a raid was botched and another who says it wasn't botched. Is it possible to come up with a more precise concept that covers what you are wanting to map? From Hill To Shore (talk) 07:25, 15 June 2021 (UTC)
Hmm, botched police raid (Q107238003) = failed police raid, when a police raid (apart from the fact that this term is already imprecise, seems to be US-centric and may have little reflection in the techniques of police work in other countries) can be called "failed"? Wostr (talk) 11:51, 15 June 2021 (UTC)
  • When the phrase appears in a news article it is 100% objective. GHits are 230,000 for "botched raid" with quote marks. Sorting through these hits to see if they correspond to Wikidata entries is hard, it is easier to use your memory then search for the phrase "botched raid" in conjunction with that event. See for instance "botched raid"+"ruby ridge" to get Q1604442. The early ones I added have the phrase in the corresponding Wikipedia article. It is no different than the aggregation we do for "police brutality" which can be argued is subjective but when a reliable source uses the phrase it is objective. --RAN (talk) 12:53, 15 June 2021 (UTC)
  • That's a very poor understanding of news media you have there, RAN. As a simple for instance, where a news article reports source A saying "x was a botched police raid" and source B saying (in effect) "x was not a botched police raid", the news story is not 100% objective proof of a botched police raid. There's very clearly a wide and complex spectrum from successful/good police raid through to botched/unsuccessful/bad police raid. You've coined an item which has a ridiculous combination of label and description (does botched=failed? does failed=botched?). Your item has no references. Your "but it seems easier to just use my memory, and then find a find a reference using the wording 'botched raid' in the news" is nothing more than the Garbage In element of GIGO. There is no definition of scope, no criteria whatsoever for evaluating whether instances are in scope of out of scope. Really, the whole thing is risible. --Tagishsimon (talk) 13:18, 15 June 2021 (UTC)
And I see there's more misunderstanding on your part. You have coined raid (Q107238062) with has part (P527) values of "botched raid", "military raid" and "police raid". THIS IS NOT HOW WIKIDATA WORKS. "botched raid", "military raid" and "police raid" may indeed be subclasses of a raid, but WD does not list subclasses using has part (P527), but rather by the subclass having a P279 pointing to the class. I fear to explore how much further you've gone with this line of work; how much more damage you've inflicted on WD. I see worrying signs of complete nonsense on police raid (Q97368680), where that item is now both a subclass of military raid (Q476807) and different from military raid (Q476807). The impression left is that you are well meaning & enthusiastic, but lack basic competence, and so unwittingly leave a trail of mess in your wake. --Tagishsimon (talk) 13:50, 15 June 2021 (UTC)
  • STOP YELLING! And you can be less pompous too by saying I "lack basic competence", keep your personal animosity toward me to yourself. Give me an example of one I listed where a reliable source used "botched raid" where it is contradicted by another reliable source saying it "not a botched raid". We also already have Category:Police misconduct (Q8768020) and Category:Police brutality (Q8768009) where I am sure you can find someone arguing that there was no misconduct or no brutality. No one else demands that 100% of the sources have to agree to have an entry. If you hate the phrase "botched raid" remove it from each Wikipedia article, or better yet, make your argument in the Wikipedia article on Breonna Taylor on why it was not a botched raid, remember to add in your reliable sources. Data integrity is important, but just a reminder, we have over 100,000 data points from reliable sources that had to be deprecated and marked as "typographical error" or "implausible" or "contradiction". We also have over 100,000 data points from reliable source for date of birth or date of death where two different years or two different dates in the same year are given. We average five entries a week where someone died before they were born. We have a list of humans living over 120 years waiting to have corrections made, or marked somehow as having a fictional lifespan. Despite your concern, I don't see you participating in any of the projects that make corrections for these errors. --RAN (talk) 12:52, 16 June 2021 (UTC)
  • When the phrase appears in a news article it is 100% objective – no, it's just source X stated that incident Y was "botched", nothing more. News articles are far from being objective. Also I don't think that such imprecise and poorly-sourced data should be added to WD, it's far from the concept of structured data, it looks like a tag, not a proper class. Police actions or operations could be classified more precisely with classes like police operation ended with the death of a civilian(s) – such classes can be properly structured, do not carry subjective opinions with them. However, (1) not every police operation ended with the death of a civilian could be classified as a police misconduct or police brutality; (2) with botched police raids class, I really don't know what kind of information you want to add to WD? What happened during the operations classied as botched police raids that allows them to be called botched or failed? Without more precise definition, the only thing that botched police raid (Q107238003) means now is police operations which have been called "botched" in the media, i.e. it means really nothing. Wostr (talk) 14:18, 16 June 2021 (UTC)
I think you are trying to distinguish the difference between absolute truth and the information found in reliable sources. We can only record information found in reliable sources. Again, they are called "botched" because that word appears in the reliable source and the corresponding Wikipedia article. You are welcome to add in reliable synonyms for the concept if you do not like use of the word "botched". Some concept entries in Wikidata have a half-dozen synonyms. You are welcome to edit each Wikipedia article which contains the phrase and remove it, if you think the source is unreliable. It is the concept that is important, even if reliable sources use synonyms. You can write The New York Times and complain of their use in the following articles: https://www.nytimes.com/search?query=botched+raid and the Washington Post: https://www.washingtonpost.com/newssearch/?query=botched%20raid&btn-search=&sort=Relevance&datefilter=All%20Since%202005 which contains both military raids and police raids. If you do not want the concept in the WikiUniverse, you have to stop the use by reliable sources first. You wrote: "it's just source X stated that incident Y was …" but isn't that how every data point in Wikidata is added? That is why some entries have three contradicting years of birth. Again this is no different than how we add entries to Category:Police misconduct (Q8768020) and Category:Police brutality (Q8768009). -RAN (talk) 16:01, 16 June 2021 (UTC)
Yes, reliable sources! News articles are not such sources in most situations. Also, we are not writing an encyclopedia here; in Wikipedia citing a source is allowed and you can write that X, Y and Z called something "botched". In Wikidata we are trying to establish a structured database with data that is objective (neutral), accurate and unambiguous, some information may not be needed here or have to be written differently than in Wikipedia. So botched is not an adjective that has any meaning in this situation in relation to the situation. It can be "botched" in so many different ways, that – as I said – the only thing that this class means in Wikidata is that someone in source X said that this operation was botched, nothing more. I don't think that we need such information in Wikidata. What we need, in my opinion, is a set of objective, structurally linked information why someone might think an operation was botched. And I don't have a problem with botched police raid term at all if it appears in the media, even in Wikipedia in a proper context, however, I have pointed out to you what ambiguities I see with using this term in Wikidata, I tried to show you a better, in my opinion, approach to classification of police operations which did not go as planned. Sadly, you seem to have some kind of "mission" to add this information exactly like this, no matter what the arguments against. Also, you failed to answer my questions and I don't see any willingness to find a common ground here, so I don't think there's a room for a constructive discussion here. Wostr (talk) 16:43, 16 June 2021 (UTC) PS I don't know what do you mean by linking Category:Police misconduct (Q8768020) and Category:Police brutality (Q8768009) – these are just Wikipedia categories and I see no connection to this problem; every project has its own rules, guidelines and – the most importantly – way of presenting knowledge. Wostr (talk) 16:53, 16 June 2021 (UTC)
@Richard Arthur Norton (1958- ): The key difference with misconduct is that there is a factual standard to apply. There will be a standard of conduct that has either been breached or not breached. We can have accusations of misconduct, investigations of misconduct, charges of misconduct, trials of misconduct and verdicts about a misconduct trial. All are factual statements with objective criteria. Similarly, brutality is excessive use of force beyond what is normally allowed. We will normally have a set of rules on a level of force allowed in a society, which are either breached or not breached. There can be accusations of brutality, investigations of brutality, charges related to brutality, trials regarding brutality and verdicts about the trial. Again, these are factual statements linked to objective criteria.
For botched police raid, it is unclear what standard is being applied. There is no objective standard for judging success or failure of the raid. Each raid will have a unique set of goals to be achieved and witnesses will apply their own subjective assessment of the outcome.
Here is a hypothetical scenario. A police raid is conducted on a drugs cartel and $1m worth of drugs are seized, 10 gang members are arrested but 3 escape and another person related to a gang member (but not an active participant in the gang) is killed.
Source A says it was a botched raid as some of the gang members escaped and an innocent person was killed.
Source B says it was a successful raid as a large amount of drugs and many of the gang members were captured. The killed person was involved in the crime (even if it was to the extent of not revealing the activities to the police) so was not an innocent party.
Source C says the results were inconclusive as many but not all of the goals were met.
We create an item about the incident and mark it simultaneously as an instance of botched police raid and an instance of successful police raid. A later editor creates an item for inconclusive police raid and adds that to the item.
What useful data can we draw from that item? From Hill To Shore (talk) 18:18, 16 June 2021 (UTC)
  • Can you give an example of the raids listed here at Wikidata that are inherently ambiguous? It seems people do not like the term "botched" because it sounds colloquial, even though it is the exact word used to describe the events in reliable sources. --RAN (talk) 18:48, 16 June 2021 (UTC)
    • A rename alone won't fix the problem. "Failed police raid" is equally ambiguous and is a subjective judgement by the source. The best we can map from that is "this source thinks it failed." Your earlier comparison to age data doesn't really fit here. A birth date is a fixed value; while sources disagree, we can assess the quality of the sourcing and come to a preferred value (that preferred value may even be a date range covering the sourced dates). The same situation doesn't apply here where there is no preference; an item could be equally failed, succeeded and inconclusive all at the same time. I'll take a look later to see if any of the current raid examples are flawed but I'd like to declare now that it is a straw man argument. It is not the current minimal implementation that is my primary concern but the potential expansion of an item with unclear parameters. You will probably have applied some level of consistency to the current data as you have a clear picture of what you want to map. The problem comes when future editors try to interpret the item and apply their own subjective judgement. If you can rationalise the concept you want to map into objective statements then there won't be an issue. From Hill To Shore (talk) 20:03, 16 June 2021 (UTC)
      • Looking at Pecan Park raid (Q65083136) there are several objective statements that could be made. 1) Police raid involving an illegal warrant, 2) Police raid involving a fatality, 3) Police raid based on a false report of a crime, 4) Police raid involving injury of police personnel, 5) Police raid using "No knock" tactics, 6) Police raid involving falsified evidence. While quite a few sources do use the word "botched," that doesn't change the fact that a decision to classify items using this or a similar term relies to a high degree on the subjectivity of sources. Sooner or later there will be an item where the sources are divided on the success or failure of a raid and we will end up in a quagmire. It is better to change tack now while the concept is linked to a few items and agree a less subjective concept to apply. From Hill To Shore (talk) 21:29, 16 June 2021 (UTC)

Usecases for qualifiers defining exceptions to constraints[edit]

Please help expand https://phabricator.wikimedia.org/T236295 --- Jura 06:08, 15 June 2021 (UTC)

Incorrect Merge[edit]

It looks like the item Linearity Q1753188 [1] was erroneously merged into laser beam welding (Q937468). Will somebody please help revert? — The Erinaceous One 🦔 06:47, 15 June 2021 (UTC)

@Burakzdd, The-erinaceous-one: I have reverted the merge but that has left tr:Doğrusallık without a Wikidata item. Is anyone able to read the language and able to confirm where it should be linked? Some of the content suggests linearity (Q1753188) may be a good match but the article also seems to cover other topics, so it may need to sit on a separate item. From Hill To Shore (talk) 07:17, 15 June 2021 (UTC)
There is a Google Translate plugin for Chrome that offers translation of any page you are on. This one translates to: "Linearity is the property of a mathematical relationship ( function ) that can be represented graphically as a straight line" and the other meanings are mostly just applications of the principle (that's why math is fundamental btw), so I added it to Q1753188. --SCIdude (talk) 16:12, 15 June 2021 (UTC)
So sorry. That's my fault. Linearity must be related to Doğrusallık. But I couldn't. --Burakzdd (talk)

Nature News: scholarly article (Q13442814) or article (Q191067)?[edit]

I've just disambiguated a writer for the News division of Nature (Q180445) by the name of Rex Dalton (Q107241085). As far as I can tell, all of his works (query) are news articles, which (if I'm understanding this correctly) are not peer reviewed. Apart from Q28302723 (which I've already changed) they're currently all instances of scholarly article (Q13442814). Would it make sense to do a batch edit and change them all to article (Q191067)? In this case that would be easy enough to do, but unfortunately only a few (like this one) have the string "NEWS" as part of the DOI so I don't see an easy way to fully automate this for Nature news articles from different authors. Aluxosm (talk) 09:35, 15 June 2021 (UTC)

I think it would be fine to add such a P31 statement. I also think "scholarly article" should be changed to "article in scholarly journal". --SCIdude (talk) 16:27, 15 June 2021 (UTC)
@SCIdude: Thanks for your input. To be clear though, I would be replacing the P31 statement, not adding to it. Also, I'm not sure about the renaming, a similar entity already exists — academic journal article (Q18918145). All very confusing! Aluxosm (talk) 11:50, 16 June 2021 (UTC)
Yes, article (Q191067)? is more appropriate for these and, by extension, many more. Looking at the results to the linked query, I also noticed that there seem to be quite a few duplicates? For the one example I checked out, California approves $1.5 billion campus (Q58926806) seems to be correct while California approves $1.5 billion campus. (Q46529753) contains a DOI that leads to an entirely different article? (Edit: it’s fewer than I first thought. They just happened to cluster at the top of the list)Karl Oblique (talk) 21:37, 15 June 2021 (UTC)
@Karl Oblique: Good spot! I'll have to add that to the to-do list. To further complicate matters, I've just come across news article (Q5707594) 😬. Need to have a think before I go much further. Aluxosm (talk) 11:50, 16 June 2021 (UTC)

Subway modelling on adjacent station (P197)[edit]

Hello, I have been unfriendly reverted on this https://www.wikidata.org/w/index.php?title=Q1340440&oldid=1442360329 edit. Adding qualifyers to adjacent station (P197) is useful : the qualifyer might be redundant but very useful to distinguish whether there are multiples lines served on a station (eg on other examples : https://www.wikidata.org/wiki/Talk:Q462201). I have been reverted because I did on one item, not on all 200+ Moscow subways. I don't have time to edit all of them very quickly before being reverted again by an individual. Bouzinac💬✒️💛 20:48, 15 June 2021 (UTC)

@Michgrig: Per P197#P1855, there is consensus that connecting line (P81) is a valid and necessary qualifier for adjacent station (P197) and so your revert of Bouzinac was not helpful. --Tagishsimon (talk) 21:19, 15 June 2021 (UTC)
@Tagishsimon, these are just examples, not the clear instruction. As I see it, if a station belongs to one line, adding the line as a qualifier to adjacent station (P197) is simply redundant. Any automation must be flexible enough to get a single value from the connecting line (P81) property. If a station belongs to multiple lines, then qualifiers are needed. You can find examples of this in Moscow Metro (e.g., Park Pobedy (Q1810822)) and on railways in Moscow Oblast (e.g., Dmitrov (Q4162925)). Michgrig (talk) 21:29, 15 June 2021 (UTC)
@Tagishsimon P.S. Please note that in the examples that you gave the link to, all of the stations belong to more than one line. Therefore, using the line qualifier fits into my logic. Michgrig (talk) 21:33, 15 June 2021 (UTC)
The trouble with your logic, @Michgrig:, is that it takes no account whatsoever of expectation by others of the data model. The connecting line is unambiguously an attribute of any reasonable model for adjacent stations, and it is completely unreasonable to suggest that the adjacent station <--> connecting line model should differ according to whether there is a single connecting line or multiple connecting lines - that one model should hold the data as a qualifier of P197, the other as a main property of the station, and that those querying for station data, for instance to draw maps, must make already complex queries more complicated because of your prejudice against storing valid data. I can point you at evidence that the connecting line qualifier is an expectation of all of the examples of this art that I've seen & been involved in in the last few years. I can point you at the property definition which specifies that the connecting line qualifier is an expected or sanctioned qualifier for the property. What, your prejudice aside, can you point me at to support your view? --Tagishsimon (talk) 22:00, 15 June 2021 (UTC)
I do not have any prejudice. In my support, I can point you at the Occam's razor principle: entities should not be multiplied without necessity. Please give me links to the discussions that support your point of view and to examples of ready maps that use it. Michgrig (talk) 22:10, 15 June 2021 (UTC)
Does Occam have anything to say about why two distinct models are preferable to one? Because that - a single model - seems to be highly desirable. I'm not aware of any discussion for or against the idea that WD should have two distinct models for adjacent station; no-one before you has suggested that would be a sensible way to go. Ready maps that use it: [2], for instance. --Tagishsimon (talk) 22:43, 15 June 2021 (UTC)
@Tagishsimon OK, at least I've seen a live example of using this. I agree to follow this "single model" and will try to add qualifiers to all stations of Moscow Metro and railway.
However I still have one question to @Bouzinac: why in the middle of tons of edits in stations of China, you decided to edit one Moscow Metro station? Michgrig (talk) 21:07, 16 June 2021 (UTC)
Some rapid transit (Q5503) and tram system (Q15640053) eg : Strasbourg have been modeled this way and it's perfect. But it's true it is not yet the case for every subways/tramways systems. And modeling takes some time (it took me 3 days to build the Shanghai map, in finding, cleaning, modeling). Bouzinac💬✒️💛 05:28, 16 June 2021 (UTC)

Sixth chord[edit]

Could I please have some assistance with Sixth chord (Q28089342) and sixth chord (Q515989)? I'm thinking that we ought to split this into three items: the classical music sense; the modern popular music sense; the conflation of the two as a disambiguation page. I'm having a little trouble assigning the various sitelinks and redirects to these three buckets. Thanks, Bovlb (talk) 00:05, 16 June 2021 (UTC)

Are VIAF entries supposed to be unique?[edit]

We have had Edmund Glover (Q21455684) and Edmund Glover (Q5339502) with the same VIAF and WorldCat_Identities_ID. I think both belong exclusively to the actor, the DNB entry does not mention that the actor was also a painter, so they seem to be different people with coincidental birth and death dates. If we had a reminder that a VIAF is used in another entry, like we do for Findagrave entries, it would help find potential duplicate entries, and help identify inadvertent conflations. --RAN (talk) 13:32, 16 June 2021 (UTC)

I assigned them exclusively to the actor. --RAN (talk) 15:30, 16 June 2021 (UTC)

Q822835 Bernarda Morin[edit]

I have some questions about this person:

  • her birth name is Vénérance. She was born in Quebec, Canada. When I try to add her birth name to wikidata; "Canadian French" is not an option for (mandatory) language. How to proceed?
  • her english wikipedia page is titled "Mother Bernard Morin"; but "Bernard" is a masculine name; surely this should be "Bernarda"?

Thanks. Fbax (talk) 14:09, 16 June 2021 (UTC)

@Fbax: :
  • in the dropdown menu, you can select "fr-ca", I did a test here: Q4115189#P1477.
  • yes, to be checked with references but it's likely a spelling mistake
Cheers, VIGNERON (talk) 16:48, 16 June 2021 (UTC)