Wikidata

Wikidata:Project chat/Archive/2023/01

< Wikidata:Project chat‎ | Archive‎ | 2023


Influenced or influences - Q65933422

Hello. I am having troubles with Q9235, i would like to described the influence relations that Hegel has with posterior philosophers and thinkers, but the only property available is influenced by, the inverse relation, Q65933422, isnt a property. Does this mean I cant use it in statements? Shouldnt it be available as a property? JoaquimCebuano (talk) 04:37, 2 January 2023 (UTC)

I think a property for Influences could border on breaking Wikidata with our current technical limitations. A single human may have been notably influenced by a small number of people in their life. Such small numbers of notable statements are simple to map. Conversely, if we were to record all the notable people that Plato (Q859), Jesus (Q302) or Sun Tzu (Q37151) influenced/influences we would be adding millions of statements to each item. Perhaps this is something to consider in the distant future but I don't think we are ready yet. From Hill To Shore (talk) 08:59, 2 January 2023 (UTC)
We dont have millions of people in wikidata. The relation between Plato and Socrates is as important as the relation between Plato and Aristotle, I cant see why it would break Wikidata, its value is based on the sophistication that it can achieve. JoaquimCebuano (talk) 10:10, 2 January 2023 (UTC)
We do have millions of items for people in Wikidata. To be exact, as I'm writing it we have 10,404,058 of them. General ideas about databases suggest that it's better to store data like this at one place instead of storing it at two places and both items.
Data user can query for the inverse. ChristianKl ❪✉❫ 10:37, 2 January 2023 (UTC)
To express that Hegel influenced Karl Marx you just add the statement Karl Marx (Q9061)influenced by (P737)Georg Wilhelm Friedrich Hegel (Q9235) to Karl Marx (Q9061). The inverse statement Georg Wilhelm Friedrich Hegel (Q9235)influencedKarl Marx (Q9061) is not needed, as it is already implied by Karl Marx (Q9061)influenced by (P737)Georg Wilhelm Friedrich Hegel (Q9235). The proposition is in Wikidata, even though it is not directly shown as a statement on Georg Wilhelm Friedrich Hegel (Q9235). To see all statements linking to Georg Wilhelm Friedrich Hegel (Q9235) via influenced by (P737)Georg Wilhelm Friedrich Hegel (Q9235) there are several possibilities, just to name three:
  • scroll to the bottom of the item and open "show derived statements"
  • look at the item using an interface like Reasonator: Hegel in Reasonator ("from related items" section; the page takes a while to load)
  • query for all items influenced by Hegel according to Wikidata: pictures of people influenced by Hegel
BTW: Thanks for working on this relation!. - Valentina.Anitnelav (talk) 12:54, 2 January 2023 (UTC)
Since I don't see it mentioned here yet, there was a property for the inverse relationship (P738), deleted after this discussion. Jamie7687 (talk) 18:43, 2 January 2023 (UTC)

Proposal: Related property sets

While creating new properties, I noticed that there are many groups of properties that are similar, such as "Hotel IDs" or "Manga Author IDs". Instead of having to manually make all the links, I was thinking of making related property sets. These would be pages like Wikidata:Related property sets/Hotel IDs and the page would just be a list of properties to link up. Every day, a bot could go through and make the new links. There could also be functionality for removing an item and then a bot would go through and remove those links. The power comes in that a property can be part of multiple sets, so when a property is removed from one set it'll delete those links only and nothing else.

Additionally, the bot might add references to what set the item was imported at and from which time it was imported from. What are your thoughts on this? RPI2026F1 (talk) 21:15, 29 December 2022 (UTC)

There is a type of Wikidata property (Q107649491) class tree which - to the extent it has been developed, caters for this need - e.g. Wikidata property to identify manga (Q101083593). Is that insufficient? --Tagishsimon (talk) 21:20, 29 December 2022 (UTC)
I'm talking about the related property (P1659) property. RPI2026F1 (talk) 21:30, 29 December 2022 (UTC)
This may also be redundant to property navboxes. GZWDer (talk) 22:01, 29 December 2022 (UTC)
The problem is that related property (P1659) also groups together properties of different subjects from the same website (such as an author ID and a publisher ID and a book ID from the same website), which I don't think the property navboxes do just yet. RPI2026F1 (talk) 22:46, 29 December 2022 (UTC)
This may be solved by a new property "properties from this website". GZWDer (talk) 22:57, 29 December 2022 (UTC)
RPI2026F1, if properties are described using P31 values in the type of Wikidata property (Q107649491) class tree, then the contents of related property (P1659) would be the set of properties that have matching P31 values. The EasyQuery gadget will provide this sort of list in one click. Still unclear to me what advantages your proposed bot & page infrastructure would add. --Tagishsimon (talk) 00:36, 30 December 2022 (UTC)
I'm curious as to why related property (P1659) exists if it would be fulfilled with instance of (P31) values in that case. Seems like a needless duplication of data. RPI2026F1 (talk) 01:33, 30 December 2022 (UTC)
Agree. --Tagishsimon (talk) 01:47, 30 December 2022 (UTC)
@RPI2026F1: If you are curious about why something on Wikidata exists, you should look at the thing. If you do that for related property (P1659), you find that one of the examples is father (P22). If you look further than you see that father (P22) is used there only for a subset of the properties that are instance of (P31) Wikidata property for human relationships (Q22964231). It's not used for all the properties.
related property (P1659) is not about listing all possible properties that have some relation but only the most important ones. A set of all "hotel IDs" would be very different than the current usage and purpose of related property (P1659). ChristianKl ❪✉❫ 14:28, 30 December 2022 (UTC)
I'm trying to start moving away from using the class tree to categorize similar properties and instead creating items for identifier properties and using class of non-item property value (P10726) and identifies (P10476) to relate properties that similar. For example if you want to relate all of the hotel external identifiers, you would create items for all their identifiers and then use identifies (P10476)hotel (Q27686) on all of them and then link them to their property with class of non-item property value (P10726). Then you can query for all the identifiers with identifies (P10476)hotel (Q27686) and get all the hotel identifiers and their properties. Lectrician1 (talk) 02:48, 30 December 2022 (UTC)
The problem is that there's multiple different ways to do the same thing. One thing I've noticed about Wikidata is that unless something is decided uniformly (like as a requirement due to a bot/external tool or due to an existing discussion), everybody does something their own ay. RPI2026F1 (talk) 03:12, 30 December 2022 (UTC)
Wikidata was from the start designed in a way to often allow different people to do things in their own way. This does produce sometimes problems because there's no uniform modeling. To get to more uniform modeling we need to have discussions at places where people can later find the conclusions of the discussions and that unfortunately isn't easy on Wikidata. ChristianKl ❪✉❫ 16:50, 30 December 2022 (UTC)
The problem with Wikidata is that it's trying to be a easily queryable and searchable dataset. With a good dataset you shouldn't need to "clean up" multiple representations into a single common representation, and so I think the door to those discussions should be opened. Wikiprojects are a good way of doing this sort of thing, so maybe the solution to this particular case is a Wikiproject Property Data. RPI2026F1 (talk) 20:55, 30 December 2022 (UTC)
@RPI2026F1 for Wikiprojects to work well and come to modeling that many people use, it's necessary for people to engage more with the existing structures instead of having a first impulse to solve problems by suggestions new structures.
In this case, we basically have four ways how related properties are grouped right now, infoboxes, shared P31, shared P10476 and P1659. Adding a fifth way would not be helpful for the goal of a single common representation. ChristianKl ❪✉❫ 10:27, 2 January 2023 (UTC)
I was thinking the Wikiproject could turn the 4 ways into a single way. My biggest problem with talk pages is that most people do not open the talk pages. Also, it's much harder to edit one of the navbox templates than it is to edit a statement. RPI2026F1 (talk) 13:37, 2 January 2023 (UTC)
Having discussions in Wikiprojects is useful but there's no good reason to create a new Wikiproject instead of using the existing one's for properties. Having more discussions in existing Wikiprojects is a way to get more people to look at them.
It's indeed currently a huge problem of Wikidata that most people do not open talk pages where policy discussions happen. For that it's helpful to reuse existing pages instead of creating new ones. Creative solutions of how to get more people to look at them might also be helpful. ChristianKl ❪✉❫ 10:39, 3 January 2023 (UTC)
I wish it were possible to make "header" and/or "footer" sections on items/properties/lexemes that could be arbitrary Wikitext. RPI2026F1 (talk) 11:52, 3 January 2023 (UTC)
Wikidata:List of properties (e.g., Wikidata:List of properties/transport) could be the thing. --Matěj Suchánek (talk) 09:50, 31 December 2022 (UTC)

P6 Data is not updated with latest value for Q1191

Hi. The new chief minister of Maharashtra is Q18750400(Eknath Shinde). I don't have the rights to edit this page. I kindly request you to update as you had updated this property last time Gopalkrishna92 (talk) 07:53, 3 January 2023 (UTC)

I updated it. But if you have a reference and a start date that would be optimal. BrokenSegue (talk) 08:03, 3 January 2023 (UTC)
As per this reference - https://en.wikipedia.org/wiki/List_of_chief_ministers_of_Maharashtra , date of appointment is 30th June, 2022 Gopalkrishna92 (talk) 08:20, 3 January 2023 (UTC)

LGBT-related as a film-genre ?

Whys is this used as a genre? This seams to be more like what IMDb label as plot keyword. I suggest that we drop using this as a genre. Ezzex (talk) 16:24, 31 December 2022 (UTC)

For reference:
-wd-Ryan (Talk/Edits) 18:50, 31 December 2022 (UTC)
personally I don't see what's wrong with having an expansive view of the idea of genre. BrokenSegue (talk) 18:53, 31 December 2022 (UTC)
I agree.
Amir (talk) 09:13, 17 June 2014 (UTC) ★ → Airon 90 10:29, 17 June 2014 (UTC) --Another Believer (talk) 15:00, 17 June 2014 (UTC) I am not terribly familiar with Wikidata, but offering my support! Gobōnobō + c 00:25, 18 June 2014 (UTC) OR drohowa (talk) 15:37, 18 June 2014 (UTC) Blue Rasberry (talk) 13:43, 20 June 2014 (UTC) SarahStierch (talk) 16:47, 30 June 2014 (UTC) (Been adding LGBT stuff on Wikidata for months, had no clue this existed!) MRG90 (talk) 10:15, 19 August 2014 (UTC) Ecritures (talk) 16:37, 17 July 2016 (UTC) Shikeishu (talk) 22:28, 17 September 2016 (UTC) OwenBlacker (talk) 20:34, 11 March 2017 (UTC) Ash Crow (talk) John Samuel 17:33, 2 July 2018 (UTC) SilanocSilanoc (talk) 12:18, 17 November 2018 (UTC) Daniel Mietchen (talk) 22:45, 2 February 2019 (UTC) Tdombos (talk) 22:14, 27 May 2019 (UTC) Mardetanha (talk) 17:01, 13 June 2019 (UTC) Theredproject (talk) 23:41, 20 June 2019 (UTC) Davidpar (talk) 20:30, 7 July 2019 (UTC) Gerarus (talk) 13:10, 1 August 2019 (UTC) Sweet kate (talk) 16:33, 4 August 2019 (UTC) Nattes à chat (talk) 22:45, 3 November 2019 (UTC) Lucas Werkmeister (talk) Hiplibrarianship (talk) 23:05, 8 March 2020 (UTC) Jamie7687 (talk) 22:27, 6 May 2020 (UTC) Nemo 16:25, 1 July 2020 (UTC) ViktorQT (talk) 14:25, 4 December 2020 (UTC) Christoph Jackel (WMDE) (talk) 13:27, 10 December 2020 (UTC) Mathieu Kappler (talk) 15:13, 12 December 2020 (UTC) Myohmy671 (talk) 14:10, 23 May 2021 (UTC) Ptolusque (.-- .. -.- ..) 23:32, 17 July 2021 (UTC) Zblace (talk) 07:21, 24 December 2021 (UTC) Clements.UWLib (talk) 01:55, 20 January 2022 (UTC) Lastchapter (talk) 22:42, 25 January 2022 (UTC) Idieh3 (talk) 14:28, 31 Januari 2022 (UTC) Koziarke (talk) 02:47, 26 June 2022 (UTC) Skimel (talk) 23:10, 18 July 2022 (UTC) MiguelAlanCS (talk) 19:02, 28 October 2022 (UTC) Rhagfyr (talk) 20:31, 26 December 2022 (UTC) -wd-Ryan (Talk/Edits) 18:59, 31 December 2022 (UTC) BlaueBlüte (talk) 05:35, 25 January 2023 (UTC) Léna (talk) 10:52, 24 March 2023 (UTC) Carlinmack (talk) 15:02, 15 May 2023 (UTC) Ha2772a (talk) 07:04, 18 May 2023 (UTC) La Grande Feutrelle (talk) 22:44, 23 May 2023 (UTC) StarTrekker (talk) 15:24, 19 August 2023 (UTC) Samthony (talk) 12:13, 25 August 2023 (UTC) Gufo46 (talk) 17:27, 30 September 2023 (UTC) Sir Morosus (talk) 06:03, 22 October 2023 (UTC) Cupkake4Yoshi (talk) Wallacegromit1
  Notified participants of WikiProject LGBT Thoughts? -wd-Ryan (Talk/Edits) 18:59, 31 December 2022 (UTC)
And what is the LGBT-stuff in this film [1]. I took the liberty to remove the element. I can't see anything on the article on English wikipedia either [2].--Ezzex (talk) 19:33, 31 December 2022 (UTC)
this is why [3] BrokenSegue (talk) 20:14, 31 December 2022 (UTC)
As with everything else on Wikidata, we should be reporting the claims of other sources. If there is a source that makes the claim that a film has a LGBT genre, then we should record that claim (with reference). In this case the claim was unreferenced but mentioned as an unsourced category on the Wikipedia page. I don't see anything wrong with removing the unsupported claim here. Though if there is a supporting reference, I see no issue with it being restored. From Hill To Shore (talk) 20:33, 31 December 2022 (UTC)
I wasn't saying removing it was wrong. I was explaining why it was tagged with that in the first place. BrokenSegue (talk) 20:40, 31 December 2022 (UTC)
And I never commented on what you were saying. There was a request above for more opinions on this and I provided an opinion. Nothing more. From Hill To Shore (talk) 22:26, 31 December 2022 (UTC)
The thing I see with IMDb keywords is that they can be numerous (379 in the linked example) and often highly specific, and most of them seem far less significant to a given work than its LGBT themes can be. "LGBT" could be a keyword or equivalent, but apparently so can minor features like "hit with a fire extinguisher", "blue skin", or "title spoken by character". A work can be of more than one genre, LGBT as a genre seems like a claim that is made in significant/notable ways outside of Wikidata, and we should be able to document that here. Jamie7687 (talk) 20:53, 31 December 2022 (UTC)
I think wee should do as Allmovie and create Themes. Themes are wider and more sufficient than plot-keyword [4]. I think its totally wrong to label LGBT-related as a film genre.--Ezzex (talk) 15:45, 2 January 2023 (UTC)
AllMovie's themes are an interesting example to take into account, but they still seem... smaller, somehow (e.g. "Hide the Dead Body" in your example link). And while AllMovie's genres and subgenres seem rigid and incomplete to me, they do still have a "Gay & Lesbian Films" subgenre, as well as a "Gay & Lesbian Show [TV]" subgenre. LGBT-related film (Q20442589) and LGBTI+ related TV series (Q85133165) are just broader versions of those, more aligned with how various streaming services, university libraries, and others label them as genres. Again, I think that we should be able to document these external claims here in Wikidata. Jamie7687 (talk) 18:35, 2 January 2023 (UTC)
Young should seriously consider doing something to clean up this mess. This ain't the only example (there are a lot of weird genres here). Wikidata is'nt exactly popular among the wikipediaes. Norwegian and Danish language versjons uses infoboxes with data generated from wikidata (mostly biographies), but that's about it. I don't think English wikipedia use data from Wikidata at all. I don't think german og italian use much either.--Ezzex (talk) 21:20, 2 January 2023 (UTC)
Many web-streaming services including Netflix maintain a genre called LGBT(Q+): https://www.netflix.com/fr/browse/genre/5977. Considering the (historically) low availability of films with LGBTQ+ characters and themes, having such a genre is helpful not just to (Wikidata) users/readers, but also the historians who want to comprehend the progress of LGBTQ+ rights around the world through different media. Please check Wikidata:WikiProject LGBT/Models and if you can help us improve these models, it will be quite helpful. John Samuel (talk) 18:29, 1 January 2023 (UTC)
  • Yes it is a genre I recognize that "LGBT+" does not fit the conceptual model of genre like action, horror, or comedy, but genre is about social convention and practice as much as it is art format. Agree with John Samuel - Netflix is well known for putting the LGBT+ genre as a prominent viewing option in hundreds of millions of households. In various languages of Wikipedia we have Category:LGBT arts (Q86045112), which sorts LGBT+ literature, music, and other art. In chain bookstores that sell paper books there are LGBT+ book sections, where the LGBT+ label is overriding of any other genre. There are lots of people who preferentially consume LGBT+ media, and this content is not just a switch from non-LGBT+ to LGBT+ characters. There are thematic fundamentals in this content. Bluerasberry (talk) 21:16, 1 January 2023 (UTC)
Yes, but it's still not a genre. Drama is, comedy is, horror is, crime is. It's not even a subgenre.--Ezzex (talk) 21:23, 1 January 2023 (UTC)
The main problem is that we never really figured what genre (P136) should be in the context of fiction, see also #Wikidata:Project_chat#Genres_on_Wikidata?. So right now it's all over the place. So maybe now it's the right time to start working on that? Quite a few genres seem to be an intersection between genre (P136) and instance of (P31)/main subject (P921).
Let's take a romance film (Q1054574). Most of them are probably heterosexual, but the other (gay, etc.) movies are still romance film (Q1054574), right? Why would this be a different genre (P136)? Isn't that just a different main subject (P921)? Multichill (talk) 21:54, 1 January 2023 (UTC)
You're very sure it isn't a genre but I don't know where that confidence comes from. The idea of genre is vague. Main subject and genre are different. An LGBT film isn't about LGBT things. It just includes LGBT elements. Likewise horror films aren't about horror. BrokenSegue (talk) 22:58, 1 January 2023 (UTC)
fwiw, Dartmouth Library lists LGBTQ as a genre, with sub-genres such as 'New Queer Cinema' - [5]. There seems to be an extensive literature on Queer Cinema. --Tagishsimon (talk) 01:26, 2 January 2023 (UTC)
Though they also have "genre" entries for James Bond films, Shakespeare on film, Star Wars films and streaming services. Seems like the site might have grown beyond just genres and added more sections/categories to be able to group resources on certain topics together.
new queer cinema (Q2724311) seems to be better suited as a value for movement (P135), similar to New Hollywood (Q377616), rather than a genre. --2A02:810B:580:11D4:D8E9:D54C:A781:5344 16:33, 3 January 2023 (UTC)
Wikidata have main subject [6], as used here.

Wikidata weekly summary #553

 
Here's your quick overview of what has been happening around Wikidata over the last week.
  • Events
    • Upcoming
      • The next Wikidata+Wikibase office hours will take place on Wednesday, 17:00 UTC on Wednesday, 18th January 2023 (18:00 Berlin time) in the Wikidata Telegram group. The Wikidata and Wikibase office hours are online events where the development team presents what they have been working on over the past quarter, and the community is welcome to ask questions and discuss important issues related to the development of Wikidata and Wikibase.
    • Ongoing
      • Weekly Lexemes Challenge #74, Urology
  • Tool of the week
    • OWL Map - a tool to help in linking Wikidata items with the matching object on OpenStreetMap.
  • Development
    • The development team took a break during the winter holidays and as a result, no development occurred.

You can see all open tickets related to Wikidata here. If you want to help, you can also have a look at the tasks needing a volunteer.

Belgian Legion Q7888577 in confusion

Belgian Legion (Q7888577) seems to be a bit of a mess. It it claimed to be a Wikimedia disambiguation page. However while the linked Dutch Wikipedia page is a disambiguation page those from other language Wikipedias are not. Please also be aware Belgian Legion (Q65065793) exists, which links to a English Wikipedia article. As I am indef blocked for competence is required I am not going to follow this up because of lack of access to that place. Please be aware I have switched the sitelink to enWikiquote from Q65065793 to Q7888577 as that seems the most reasonable thing to do but I'm happy for people to review that action. Thankyou. -- User:Djm-leighpark(a)talk 20:32, 3 January 2023 (UTC)

Hello from your new Community Relations Specialist

Hello everyone👋🏽

My name is Benedict Nnaemeka Udeh from Nigeria, a recently hired Community Relations Specialist (CRS) for Wikidata and Commons.

I'm a member of the Igbo Wikimedians User Group, the Founder of IG Wikidata Hub, and the Founder of Wiki Mentor Africa, a mentorship program that seeks to bring more African developers and technical writers into the technical space of Wikimedia. I'm passionate about community growth and capacity building. Like you, I'm incredibly spirited about the free knowledge movement.

In my new position as CRS, I will focus on collaboration and communication with you about Wikidata and Commons product development, the user-facing changes and roll-outs, and sometimes on projects with special needs for collaboration with Volunteers. In other words, I am here to serve you (the editors, contributors, and readers of Commons and Wikidata projects) and the working groups at the Foundation that support Commons and Wikidata. I've already started sharing some updates; you can find them on the Product and technical support for Commons Portal. Please feel free to use the discussion page to discuss these updates.

I look forward to working with and for you all and can't wait to see the fantastic things we will do together. Udehb-WMF (talk) 09:31, 4 January 2023 (UTC)

welcome. congrats. BrokenSegue (talk) 00:49, 5 January 2023 (UTC)

list of stadiums over 60000 places

list of stadiums over 60000 places 2A01:E0A:B5:DF10:694C:3E45:D1F:6050 19:59, 4 January 2023 (UTC)

Adapting something done a few years ago (and which is mainly designed to be displayed on a map, which the WDQS UI will allow you to select). I'd use the link shortener on this report, but of course the link shortener does not work for anything above fairly trivial report sizes, because WMDE prefer users to suffer, or something.
SELECT DISTINCT ?item ?itemLabel ?numericQuantity ?coord ?layer WHERE {            # put ?coord & ?layer in this select          
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE]". 
   }
  {
    SELECT DISTINCT ?item ?coord ?numericQuantity ?layer WHERE {  # put ?coord ?layer in this select
      OPTIONAL {?item wdt:P625 ?coord.}                           # get the coord
      ?item p:P31 ?statement0.
      ?statement0 (ps:P31/(wdt:P279*)) wd:Q483110.
      ?item p:P1083 ?statement1.
      ?statement1 (psv:P1083/wikibase:quantityAmount) ?numericQuantity.
      ?statement1 a wikibase:BestRank.
      FILTER(?numericQuantity > "59999"^^xsd:decimal)
      MINUS {
        ?item p:P3999 ?statement_3.
        ?statement_3 psv:P3999 ?statementValue_3.
        ?statementValue_3 wikibase:timeValue ?P3999_3.
      } 
      ?item p:P625 ?statement4.
      ?statement4 (ps:P625) _:anyValueP625.
      bind(if(?numericQuantity >100000,"100k+",                     # assign numeric ranges
        if(?numericQuantity >90000,"90k+",
        if(?numericQuantity >80000,"80k+",
        if(?numericQuantity >70000,"70k+",
        if(?numericQuantity >60000,"60k+",
        if(?numericQuantity >50000,"50k+",
        if(?numericQuantity >40000,"40k+",
        if(?numericQuantity >30000,"30k+",
        if(?numericQuantity >20000,"20k+",
        if(?numericQuantity >10000,"10k+",
        if(?numericQuantity >10000,"0k+",""))))))))))) as ?layer)
    }
  }
}
Try it!
--Tagishsimon (talk) 20:26, 4 January 2023 (UTC)

Nupur Sharma

I'm trying to get my head around the relationship and disambiguation around the following:

I think this place needs to improve its modelling of these items. Thankyou. -- User:Djm-leighpark(a)talk 04:26, 5 January 2023 (UTC)

I've got tow other unanswered questions above so I am minded I might have to have a first shot of resolving this myself. Perhaps in part as the result of my formulation of part of what I see as the problem above, I'm going to start by creating Nupur J. Sharma (Q116030561) and subsequently adding appropriate statements with references. To my best good faith knowledge she is not currently represented as I Wikidata Item and I don't want to try and re-leverage any of the items above as that might be a good mistake. I accept OpIndia is alleged to have made some controversial actions but I am neutral to that in any actions here. Thankyou. -- User:Djm-leighpark(a)talk 08:43, 5 January 2023 (UTC)

OK to summarise

All   Done - Basically self-resolved (hopefully) -- User:Djm-leighpark(a)talk 09:40, 5 January 2023 (UTC)

Mix-n-match new item creation

Hi, I am trying to import some data using mix-n-match and i find it imopssible to use the tool to create new items. When I click on "New Item" nothing happens, the line dims and it just stays that way. Am I doing something wrong? This is the catalog in question. It is safe to consider all the items in "unmatched" as candidates for new items, so please feel free to use it for tests. Cheers --Nikola Tulechki (talk) 07:37, 5 January 2023 (UTC)

Colby Walkmac

There are enough data on Colby Walkmac (Q116021066)? 2001:B07:6442:8903:D502:2199:2B18:5F7B 11:06, 5 January 2023 (UTC)

The problem with monolinguals

I've noticed operations on monolinguals in SPARQL is quite inefficient. A typical query will ask for labels in one language. This involves retrieving labels for all the items multiplied with all the languages the items have labels for. These labels are then filtered down to one language per item. The problem with this approach is that you fill up your server's caches with data which is then almost immediately discarded, pretty much guaranteeing the following operations will see cache-misses. AFAICT this happens because filtering by language tags isn't really optimized, whereas exact matches can use indices to make them fast.

Can we do better? I suspect so. If you designate a new prefix for language labels, and instead of storing labels as 

wd:Q12345 rdfs:label "Count von Count"@en

you store them as 

wd:Q12345 wdl:en "Count von Count"

you will avoid the problem of filling up the caches with unused data. You also avoid having to fetch the labels in other languages in the first place.

In other words, by introducing a change to how labels, aliases and descriptions are stored in the graph, you can get a free performance boost, a substantial one I suspect. Are there downsides? Yes, pretty much every old query will break.

I think this is worth an internal research project to try and quantify potential performance gains. When Wikidata has to move to a new triplestore/query engine, then the old way of storing labels etc. can be discarded in favor of the new way. This can be justified since there will be lots of breakages any way. Infrastruktur (talk) 18:13, 25 December 2022 (UTC)

You do realize that RDF and XML attributes is not something we made up here? In xml it's expressed like this: <rdfs:label xml:lang="en">Count von Count</schema:name>. Completely overhauling a data model making it no longer standard compliant for a performance problem seems not the best approach to me. Optimizer should just take care of this. Multichill (talk) 14:33, 26 December 2022 (UTC)
The standard also mentions RDF collections and containers. This isn't used on Wikidata either, for performance reasons. Instead you have series ordinals. Inferrence is turned off again for performance reasons. Instead constraints are used. Chosing not to use a part of the standard does not mean Wikidata is not "compliant". However in my proposal I didn't suggest eliminating monolinguals everywhere, because there are several reasons not to do this on statements. But it is a rather small change to the data model to do it to labels, descriptions and aliases.
The proposal should also work on most modern sparql engine/triple stores as they are typically implemented today. Some problems are irreducible within their constraints. In this case I doubt you can optimize monolinguals without coming up with a completely different way of storing the triples, or maybe a service of sorts. Blazegraph offers a service for full-text searches, since this is something that is impossible to do efficiently in plain SPARQL for fundamental design reasons.
Realistically there is only one point in time where it would be possible to change the data model. To not even look into this would be a giant wasted opportunity. Infrastruktur (talk) 12:30, 27 December 2022 (UTC)
To illustrate, grabbing english labels for 100 000 random items (https://w.wiki/6B4W), retrieves 4343 kiB worth of useful data, a total of 17077 kiB worth of data, giving an overhead of 12734 kiB worth of data (75%) that we don't need and that also flushes out useful data from the caches. This adds up quickly to degrade performance. Most items have english labels, if I had picked a different language these numbers would have been worse. Infrastruktur (talk) 17:49, 1 January 2023 (UTC)
For Bokmål (nb) i got 99% junk data. Just to show that the previous example was really the best possible case. For spanish and german i get 98%. Infrastruktur (talk) 18:01, 1 January 2023 (UTC)
It might be more efficient to rework the BlazeGraph index on language strings to take the language tag into account. Ideally it would optimize constructions such as 'filter(lang(?str)="en") …' to take advantage of the tag indexed-strings. author  TomT0m / talk page 16:42, 3 January 2023 (UTC)
I strongly agree with this. Instead of changing the data in Wikidata optimize the software that is used to access the data. Peter F. Patel-Schneider (talk) 17:45, 3 January 2023 (UTC)
If you're an academic, surely you can do better than a knee-jerk reaction? Infrastruktur (talk) 20:06, 3 January 2023 (UTC)
I suspect the reason this wasn't already optimized is that the makers of query engines had to weigh trade-offs against each other. The language-tag makes it easy to store a fourth bit of information in a triple, so it is a pro for storage size but a con for performance. But considering that if you didn't have language-tags you'd have to add more nodes to the graph to store the extra bit of data, and every time you follow an edge you take a performance hit, and so this had traditionally been considered a perfectly reasonable trade-off. You don't need extra nodes for labels, descriptions and aliases though, and this makes up a large percentage of all the data for every item. Labels are also typically requested in one language at least once per query. This makes it low-hanging fruit.
If you add another index for <triplestore>, that blows up the size of the database which is already unwieldily huge, this might negate any improvements from faster lookup. Hard to say without benchmarks. There might be papers on this, I haven't looked yet. Infrastruktur (talk) 19:09, 3 January 2023 (UTC)
Found an undergraduate thesis that discusses approaches to language filtering, pages 36-40. https://ad-publications.cs.uni-freiburg.de/theses/Bachelor_Johannes_Kalmbach_2018.pdf . Infrastruktur (talk) 19:50, 5 January 2023 (UTC)

Overhaul of P3012

I want to expand the Statistics Canada Geographic code (P3012) property to include other types of geographies. Is there a way to do that? As it stands right now, the only geographies that P3012 supports are province or territory of Canada (2-digit ID code), census division of Canada (4-digit code), and census subdivision (7-digit code). But I would like P3012 to also support 5-digit federal electoral district of Canada codes and 6-digit designated place of Canada codes. The problem is that Statistics Canada has different URL formats for each type of geography, and I do not know how P3012 could properly support such a schema:

Geography type Example P3012 value URL (P3012 value in bold type)
Province or territory Ontario (Q1904) 35 https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A000235
Census division Ottawa (Q1930) 3506 https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A00033506
Federal electoral district Ottawa Centre (Q3357743) 35075 https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2013A000435075
Designated place Chalk River (Q5068811) 350066 https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A0006350066
Census subdivision Ottawa (Q1930) 3506008 https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A00053506008

-- Denelson83 (talk) 23:47, 4 January 2023 (UTC)

if the URL formats are different we can either make a new property for each format or if the IDs are really unified we sometimes make a redirection service that does the right thing. I'm not following from your examples how the URL formats are different. Looks like they all have the same URL prefix? BrokenSegue (talk) 00:48, 5 January 2023 (UTC)
Not quite. look at the digit immediately before the bolded portion in each URL, which I have underlined. It is different for each type of geography. -- Denelson83 (talk) 01:28, 5 January 2023 (UTC)
We might be able to use the applies if regular expression matches (P8460) property as a qualifier to implement this schema. Maybe consider using these statements:
  • P1896: https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/index.cfm?Lang=F
    P407: French (Q150)
  • P1630: https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A0003$1
    P407: Q1860
    P8460: ^(1[0-3]|24|35|4[6-8]|59|6[0-2])\d{2}$
  • P1630: https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A0004$1
    P407: Q1860
    P8460: ^(1[0-3]|24|35|4[6-8]|59|6[0-2])\d{3}$
  • P1630: https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A0006$1
    P407: Q1860
    P8460: ^(1[0-3]|24|35|4[6-8]|59|6[0-2])\d{4}$
  • P1630: https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/details/page.cfm?Lang=E&DGUIDlist=2021A0005$1
    P407: Q1860
    P8460: ^(1[0-3]|24|35|4[6-8]|59|6[0-2])\d{5}$
Replace "Lang=E" in the P1630 value with "Lang=F" and replace the P407 value with Q150 to get the French-language counterparts for each of the last five statements.
-- Denelson83 (talk) 08:08, 5 January 2023 (UTC)
  • Surely the simple thing here is to include that extra prefix digit in the P3012 value (and maybe change the property label slightly to be clear it's not exactly the code but the code plus a prefix?) ArthurPSmith (talk) 12:32, 5 January 2023 (UTC)
Except that it's not one prefix digit, it's actually four according to Statistics Canada's own schema. And adding that extra prefix would make the code a bit more unreadable. I only want to use the portion of the code that starts with the regional indicator, which is a digit between one and six as outlined here. Now I have tried implementing this, as can be seen in the edit log for P3012, and Wikidata just seems to be ignoring the regex instruction and only looking at the first formatter URL entry. The example for the census subdivision is working, but the others are not. Is there something I am missing here? -- Denelson83 (talk) 16:41, 5 January 2023 (UTC)

Sargsyan items tangled

I believe there is some entanglement and confusion between Aram Sargsyan (Q501736) (AZS b. 1961) and Aram Gaspar Sargsyan (Q4408661) (AGS b. 1949). They seem to share Commons:Category:Aram Sargsyan. @Rubby, you seemed to do the sitelink Special:Diff/1552524456, can you review this? Please note I've raised [7]. Can some help fix/get everything in order please. Thankyou. -- User:Djm-leighpark(a)talk 21:49, 3 January 2023 (UTC)

  Done: [8] has fixed issues there (albeit with some winging). Its given me a bit of confidence to create Commons:Category:Aram Zaveni Sargsyan which has enabled AZG's image to be moved out of AGS's commons category. Crossed fingers this helps avoid future confusion. -- User:Djm-leighpark(a)talk 03:49, 6 January 2023 (UTC)

monosaccharide and glucose

Basic fact : glucose (Q37525) is kind of monosaccharide (Q133516).

I expect to see the fact in Wikidata, either in items or in query, but I didn't find it. Since my major is computer science, not chemistry, I need the expert in this chat room's help.

As much as I understand, it is usually in the hypernym (more specific) item, the relationship is made, via instance of (P31) or subclass of (P279), to connect the hyponym (more general). My questions are (1) which property is preferred, and (2) whether some qualifier is needed. Thanks. JuguangXiao (talk) 10:30, 6 January 2023 (UTC)

glucose (Q37525) is a subclass of aldohexose (Q1077067) which is a subclass of aldose (Q409079) and hexose (Q339725) which are both subclasses of monosaccharide (Q133516). There's no requirement for them to be directly linked. —Xezbeth (talk) 13:28, 6 January 2023 (UTC)
Thanks! Is there a way of query to find the elements among the path between such indirect subclass of (P279) connected items? JuguangXiao (talk) 13:47, 6 January 2023 (UTC)
I found a way to test connectivity of 2 items via a property indirectly: 
SELECT DISTINCT ?item ?itemLabel WHERE {
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE]". }
      wd:Q37525 wdt:P279* ?item. 
      ?item wdt:P279* wd:Q133516.
}
Try it!
Revised to show each step's subject-object, in graph view. 
#defaultView:Graph
SELECT DISTINCT ?x ?xLabel ?y ?yLabel WHERE {
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE]". }
      wd:Q37525 wdt:P279* ?x. 
      ?x wdt:P279+ wd:Q133516.
      ?x wdt:P279 ?y
}
Try it!
JuguangXiao (talk) 21:08, 6 January 2023 (UTC)
@JuguangXiao: Exactly that; see https://www.w3.org/TR/sparql11-query/#propertypaths --Tagishsimon (talk) 21:25, 6 January 2023 (UTC)
Love that spec link. Thanks! @Tagishsimon JuguangXiao (talk) 21:32, 6 January 2023 (UTC)

Property "entries"

Is there a property which holds the number of entries a dictionary for example has? Greetings, 109.42.241.192 23:29, 6 January 2023 (UTC)

I get the impression that number of records (P4876) is used; e.g. https://w.wiki/6CJE. --Tagishsimon (talk) 23:44, 6 January 2023 (UTC)

Restoring an item's original function VS creating a new item for it

Rosângela (Q38133486) was originally created for the female given name Rosângela. Then, less than a year ago someone changed it to Rosangela (without the diacritic). Noticing the absence here of an item for Rosângela, two days ago user:Minerva97 created a new item for it: Rosangela (Q116025262). Now we have one item for each form of the name. Is that OK as it is? Or should we restore Rosângela (Q38133486) to its original form Rosângela, merge the new item into it, and then create a brand new item for Rosangela? —capmo (talk) 13:41, 6 January 2023 (UTC)

If you take the view that Rosângela is distinct from Rosangela, then yes, revert the original item, merge the duplicate into it, and create a new item for the missing name. The meaning of items should not be changed, fullstop. --Tagishsimon (talk) 02:18, 7 January 2023 (UTC)
Well, I suppose they are different, just as Martin (Q18002399) and Martín (Q18002404), that have each an item. Thanks for your explanation. —capmo (talk) 21:09, 7 January 2023 (UTC)

highways A1 / A3 in Morrocco

Could somebody have a look at Rabat - Safi expressway (Q2126421): Expressway in Morocco? It seems to conflate two concepts (A1 and A3) but I’m not really sure how to disentangle this since I’m neither an expert on road modeling nor on Morocco. Thank you! --Emu (talk) 10:51, 7 January 2023 (UTC)

Tools for easily notifying/warning users?

Forgive me if this is obvious, but is there a way to easily welcome/warn/notify users for common issues (e.g. welcoming, vandalism, common mistakes, WD:N and other policy reminders), using a preset list of interlingual templates? Something akin (or identical) to Twinkle, to easily place templates such as Template:Welcome and those in Category:User warning templates on a user talk page? As Wikidata is a multilingual project, automated tools to facilitate communication regardless of native language are very important (and can help newcomers more quickly learn the ropes). For example, I'd like to easily remind users creating items to add references, external identifiers and/or links to other items so that the items can be clearly identified, distinguished, and used by others (if there's not a template for this, there absolutely should be). -Animalparty (talk) 19:13, 7 January 2023 (UTC)

I have a tool that adds a "warn this user" link to the list of links on the left side of the page. I forgot where I found it, but you can take a look at my .js page and copy the code there to import it. I think not all of the templates are translated, though, but that's not a failing of the tool but of the templates themselves. –FlyingAce✈hello 02:32, 8 January 2023 (UTC)

Volume name

What is the common practice to specify volume title of a book in case of a volume entity of multivolume work? title (P1476) and subtitle (P1680) are used to store information about multivolume work. volume (P478) usually stores volume number. For example, if volume is named as "Some book: some subtitle. Volume 1. Animals" what property should store "Animals" title? D6194c-1cc (talk) 09:45, 2 January 2023 (UTC)

May title (P1476) property (or any other) be used for volume name as a qualifier for the volume (P478) property? D6194c-1cc (talk) 10:55, 4 January 2023 (UTC)
I've started another discussion at Wikidata talk:WikiProject Books#Volume title and made some research. D6194c-1cc (talk) 12:57, 8 January 2023 (UTC)

Duplicate people and bogus items

I would like to inform you about a recent mass import of bogus data through improper OpenRefine usage by Allyshaleonard (talk • contribs • logs) from MoMA.

About 5.549 new items were created including a ton of duplicate people and many spurious "art institutions" like Q115634564.

What shall we do? She is not reacting to messages on her talk page. Lustiger Saunagast (talk) 12:34, 3 January 2023 (UTC)

It looks like all this user's edits were from December 8. I would allow a few more days for them to respond, but if no response within a reasonable time it might be a case for mass reversion, if there's not a more appropriate action. ArthurPSmith (talk) 18:08, 3 January 2023 (UTC)
I am currently trying to configure a mass reversion. These were not meant to be uploaded yet and there have been techincal difficulties in OpenRefine Allyshaleonard (talk) 14:50, 31 January 2023 (UTC)
(edit conflict) This is a problem. I guess warn, revert, and/or merge. A temporary block might be in order, although the user hasn't edited in about 3 weeks. Duplicates happen, and a fair amount should be expected when importing larger sets, but if the imports are done carelessly or recklessly (as seems to be the case here), the mess is magnified and harder to cleanup. When the same user creates multiple items for the person without a single external identifier or incoming link, the problem is magnified: For example 5 duplicates each for Will Rice Amon (1,2,3,4,5) Alexander Archer (1,2,3,4,5), or Terry Ashe (1,2,3,4,5). None have a Museum of Modern Art artist ID (P2174) value (ostensibly the reason for the mass import!), and only the last person has any additional unique info besides name (namely, a birthdate of 1942), to aid in identifying/disambiguating. Lord help us should dozens of John Smiths or Sarah Browns get imported with no other identifying features, and no one can tell whether all John Smiths refer to the same person or not. -Animalparty (talk) 18:23, 3 January 2023 (UTC)
I merged the 5 records for Will Rice Amon and completed it Jgalron (talk) 07:45, 8 January 2023 (UTC)

@ArthurPSmith, Animalparty: I would like to raise another issue. Would you mind if I'd move this post to the bottom of this page?--Bodhi Balboa (talk) 13:56, 6 January 2023 (UTC)

@Bodhi Balboa: Just start a new discussion and refer back to this one; we don't usually move discussions around on this page. ArthurPSmith (talk) 15:03, 6 January 2023 (UTC)

How do I remove an invalid 'sitelink redirect' icon from a Wikipedia link?

I recently updated the en-wiki link at Q18215011 which was previously a redirect, containing the sitelink-redirect icon ( ↳ ) to a full, mainspace article which should not contain this redirect icon. I don't see how to remove it, and when I click the icon, I get a drop-down which includes a lot of options, but not "remove" or "not a redirect" or some such. Mathglot (talk) 17:31, 8 January 2023 (UTC)

The UI is not v.intuitive here. Click on the sitelink-redirect icon, get a list of possible badges, click on the sitelink to redirect entry (which is somewhat highlighted) and you toggle it. Save. I've done it in this instance. --Tagishsimon (talk) 17:40, 8 January 2023 (UTC)
Correct. My sitelink badge bot also removes redirect badges that are no longer applicable; it runs once per week. —MisterSynergy (talk) 18:59, 8 January 2023 (UTC)

Merging of Cardamom Q14625808 and Q18360378?

The two items for Cardamom Q14625808 ("seed used in the kitchen") and Q18360378 ("group of plants, providing cardamom") appear to be very similar. While the titles suggest one is about the seed and other is about the plant family, there's no overlap between the different languages they are linked to on Wikipedia. I doubt the linked articles make this distinction. Salty-horse (talk) 20:19, 8 January 2023 (UTC)

Sure, but they're distinct concepts. One is a plant. The other, seeds. And then in particular their P31 values, rightly, differ. There may be issues as to which articles either item should point to, but there is no issue whatsoever but that WD should have two distinct items for these two distinct things. --Tagishsimon (talk) 20:38, 8 January 2023 (UTC)
Thanks! Where should I inquire about double-checking the Wikipedia associations? Salty-horse (talk) 20:58, 8 January 2023 (UTC)
I'm afaid it's a manual task of visiting each article, figuring out what it's about, and if necessary moving the sitelink from one item to another. Help:Handling sitelinks overlapping multiple items may also apply - the possibility is that redirect sitelinks might be necessary to handle so-called 'Bonnie and Clyde' articles (explained in that help document). There's nowhere to report it: it's just something that someone should do. Ideal task for you :) --Tagishsimon (talk) 21:03, 8 January 2023 (UTC)

author name string (P2093) and title (P1476) problems

I have come across in some scientific article items a non-standard way of writing non-ASCII characters like M Ga$zacute$dzicki (M Gaździcki) in Lambda production in central Pb $plus$ Pb collisions at CERN-SPS energies (Q57631268). It seems to appear in physics articles dated 2002ish ?

Would it be possible to assess the size of this problem ? Is is practicable to change by hand ? Or leave it to a bot run ?

Other instances of non-standard/erroneous spelling of author strings are for example instead of Małgorzata you can find Ma?gorzata, Ma Gorzata, Ma&lstrokgorzata and other variants. These problems stand in the way of author strings disambiguation into author (P50).
Kpjas (talk) 16:54, 6 January 2023 (UTC)

@Kpjas: This query finds 66 cases in articles in journals from that publisher. So it seems to me "by hand" should be fine. Of course there may be other publishers with this problem. ArthurPSmith (talk) 18:41, 9 January 2023 (UTC)

Wikidata weekly summary #554

 
Here's your quick overview of what has been happening around Wikidata over the last week.
  • Discussions
    • New requests for permissions/Bot:
      • Bean49Bot 2. Task/s: Adding statements to plwikisource items with no statements.
      • RPI2026F1Bot 4. Task/s: Reconstruct npm dependencies.
    • Closed request for permissions/Bot:
  • Events
    • Upcoming
      • Next Linked Data for Libraries LD4 Wikidata Affinity Group call January 10, 2023: Egon Willighagen will be introducing us to SARS-CoV-2 queries, a well-documented series of Wikidata queries around the SARS-CoV-2 virus and the pandemic. This work contributes to Wikiproject COVID-19 and documentation is currently available in Japanese, Dutch, Spanish, and Portuguese. Agenda
      • Talk to the Search Platform / Query Service Team—January 11, 2023. Time: 16:00-17:00 UTC / 08:00 PDT / 11:00 EDT / 17:00 CET
      • The full version of View it! Tool is out! The new user script uses structured data to display related media in any Wikimedia content pages—including Wikidata items—in an on-wiki image panel or gallery view. Please join the public launch and demo of the full version of this new tool on Thursday, January 12th, 5:00 UTC via Zoom (Meeting ID: 160 454 5329) to learn more and discuss forthcoming editing features.
      • The next Wikidata+Wikibase office hours will take place on Wednesday, 17:00 UTC on Wednesday, 18th January 2023 (18:00 Berlin time) in the Wikidata Telegram group. The Wikidata and Wikibase office hours are online events where the development team presents what they have been working on over the past quarter, and the community is welcome to ask questions and discuss important issues related to the development of Wikidata and Wikibase.
      • About 2 months left to register for the first digital@IANLS workshop: Learn the basics of Wikidata and how to use it in your Neo-Latin research. Event is online, free, and open to all. More info on event and registration.
    • Ongoing
      • Weekly Lexemes Challenge #75, Tools
  • Tool of the week
    • Web Hub allows users to navigate between origin and destination on the web using information from Wikidata, primarily on Wikimedia sites.
  • Development
    • REST API: Preparing the first release to test.wikidata.org on January 17th and to Wikidata on January 24th
    • Entity Schemas: Continuing to investigate and prototype technical approaches for version 2
    • Mismatch Finder: Working on giving mismatch providers access to the reviews of the mismatches they uploaded (phab:T304794)
    • Investigated how to best do language switching for the Query Builder and Mismatch Finder. So far they only support it by specifying the language via a URL parameter and not yet via the UI (phab:T324653)
    • Query Service:
      • Lexeme IDs now have tooltips so you can see the Lemma in the SPARQL code, similar as for Items and Properties (phab:T255245)
      • Fixed a bug where the y-axis label on a graph were misaligned and hard to read (phab:T325808)
    • Lexicographical data:
      • Statements linking to a Sense now also show the language of the Lexeme to make it easier to see for example which language a translation statement refers to (phab:T207392)
      • Fixed a bug in the language dropdown on Special:NewLexeme (phab:T324743)
      • The Lemma of a Lexeme is now shown in the title of the revision history of the Lexeme as well (phab:T312660)
    • Vector 2022 theme: worked on making the search on Wikidata work in this new theme. You can test it on test.wikidata.org now and next week on Wikidata. (phab:T316093)
    • mul language code:
      • Continued investigation around how to handle language fallbacks on mobile, which is a blocker for adding this language code (phab:T323098)
      • Preventing the addition of descriptions for mul (phab:T313027)
    • Constraint violations: Working on also showing them to non-logged in users (phab:T272132)

You can see all open tickets related to Wikidata here. If you want to help, you can also have a look at the tasks needing a volunteer.

Merge or such needed

I'm pretty certain Q116164699 is the same thing as Chinatown-International District (Q3153248). - Jmabel (talk) 20:25, 10 January 2023 (UTC)

It looks like the new one (Q116164699) was created to link a Commons Gallery page for images stored in the category of the old one (Q3153248). As there were no other statements in or links to the new item, I went ahead and merged them. From Hill To Shore (talk) 21:05, 10 January 2023 (UTC)
On a related subject, can someone take a look at Special:Contributions/218.102.0.44. This anonymous user has created a large number of empty items for Commons pages in the last 24 hours, many of which will have existing items, like the case above. However, there are many items for unusually specific concepts, such as Q116147564. I expect many of these items are in breach of Wikidata:Notability #4. Is this salvageable or should we consider a mass deletion? From Hill To Shore (talk) 21:15, 10 January 2023 (UTC)
I blocked them from the main namespace until they respond to comments. Hopefully they can explain their methodology and we can get to the bottom of this. If we can't then I support mass deletion. BrokenSegue (talk) 01:17, 11 January 2023 (UTC)

Can child's drawing (Q12407075) and child's drawing (Q50950479) be merged?

Can child's drawing (Q12407075) and child's drawing (Q50950479) be merged? It looks like they are about the same subject. JopkeB (talk) 04:06, 16 January 2023 (UTC)

i think so BrokenSegue (talk) 04:15, 16 January 2023 (UTC)
@Daniel Mietchen would you please give your comment to this question, since you initiated child's drawing (Q50950479) in 2018, the one that was made after the other (2013), and therefor should be merged into the other? --JopkeB (talk) 13:34, 16 January 2023 (UTC)
Thanks for the checks and the ping. I merged them. Daniel Mietchen (talk) 14:53, 16 January 2023 (UTC)
Thanks for merging, Daniel Mietchen. --JopkeB (talk) 15:42, 16 January 2023 (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. --JopkeB (talk) 15:42, 16 January 2023 (UTC)

Freedom of panorama

Hello. Could someone please explain me what is the current policy regarding freedom of panorama in Wikidata. Commons follows both the US law and the law of the source country. What about Wikidata? For example the image at Q78221297 gets displayed in Slovenian Wikipedia via a template, which violates its local EDP. Should only the transcluded template be tweaked or should the image also be removed in Wikidata? --TadejM (talk) 11:12, 5 January 2023 (UTC)

I think that's indeed a general problem we have with Commons images used on Wikidata. There are quite a lot of files on Commons that simply can't be legally used in every jurisdiction. Simply because copyright rules can vary greatly between countries, and a work's copyright status in other countries can depend on lots of factors. Which results in a large number of cases where images are perfectly OK to use on Commons and Wikidata, but potentially violate copyrights when used on other projects.
So without an easy way to query a file's copyright status and automatically verify whether it's legal to use/display it in a specific project's jurisdiction, the current situation is bound to cause legal problems for users that aren't aware of this. --2A02:810B:580:11D4:A1:4540:326E:5D42 19:50, 5 January 2023 (UTC)
I do not think we have a policy, but I do not see how we could have a policy. We do not host images, we display images which are on Commons. Whereas we can in principle adopt a policy which would prohibit displaying images which are non-free in at least one jurisdiction, I do not see how this policy could be enforced. Ymblanter (talk) 20:09, 5 January 2023 (UTC)
Ymblanter: oh, I agree. Wikidata (and people editing our items) isn't really equipped to handle the complexities of international copyright law. But I think we do need a solution that at least allows users of our data to avoid images that aren't compatible with their local copyright laws. I'm not familiar enough with Structured Data on Commons to know whether Wikipedia templates could query a files' licenses/copyright status directly from Commons and then filter out problematic images. Which would still kinda leave the problem that users would first need to be aware of such potential traps, but I'm not sure what we could do about that. --2A02:810B:580:11D4:7D67:F45A:2EAC:762D 08:13, 7 January 2023 (UTC)
Put a disclaimer/warning on the front page. Indicate there might be images, Indicate there might be cultural references to how the data is presented to the end user that will upset one or couple of people that want to get receive attention. Pertaining to something. (Servers still need electricity..) And a proper check box for "I am an Adult and take responsibility for any laws I break." I am against burning the Library of Alexandria or Destroying any Artifacts or Data of Cultural significance. Omission of a thing is lie and deleting or censoring what other people see for what one person takes offense at isn't an excuse of that behavior.
This is a hosted site people who visit are guests.
If they pass a law tomorrow, somewhere else, that their religion doesn't believe in aggregating data that they aren't in charge of well.. what do we do? Delete the site? They don't like it they can leave. If they complain then block them. if the country complains full Geo IP block. Warn them, Give them Filters if needed. but don't kneel for attention seekers. 107.218.251.194 21:40, 11 January 2023 (UTC)
Wikidata doesn't host images itself but only links and loads images that are on commons as such we don't make additional requirements on images.
If you believe that there are currently legal problems with the status quo, it would make sense to speak with the Wikimedia legal team about whether they also think there are legal problems. If we create a solution based on a legal need it would be good to have an actual legal analysis to direct it.
We could have a system where we have a qualifier that could be used to prevent specific Wikipedia's from loading an image. Maybe also a system where the file is flagged on Commons (and then it might make sense to speak with the Structured Data on Commons team). ChristianKl ❪✉❫ 21:04, 5 January 2023 (UTC)
@Udehb-WMF: this seems like a WMF/Wikidata/Commons communication issue. ChristianKl ❪✉❫ 21:21, 5 January 2023 (UTC)
I'm a little confused by this thread. First, no-one has provided a link to the Commons page w:Commons:Freedom of panorama. Second, no-one has provided any details of a FoP policy problem between Commons and WD that needs to be solved. As far as I can see, Commons requires that images respect copyright restrictions - including the derivative image panorama issue - in the country of the object photographed (with the possible exception of images of object in country A photographed from country B - surely an edge case). Commons:FoP mandates that images not respecting the object country's laws be taken to Files for Deletion. If the Commons image on Q78221297 breaches that country's copyright law, then it should be nominated for deletion on Commons. Forgive me if I'm being a bit old-fashioned in wishing to see an evidence-based analysis of the supposed problem before jumping to 'solutions'; and it is possible that I've misunderstood Commons:FoP, in which case I hope to be corrected below. --Tagishsimon (talk) 00:57, 6 January 2023 (UTC)
Tagishsimon, TadejM seems to ask a pretty straight forward question. Files used as image (P18) values are from Commons, where they only have to be free in the US and their country of origin, but not necessarily in other countries. So since their project is using templates to populate infoboxes with Wikidata data, they are also getting images that they aren't legally allowed to use in their country. How can that be avoided without having to review all used images by hand? Is there an easy way to query the files' license/copyright status that can be used in templates to filter them out?
Because what good is image (P18) if we are exposing unaware data users from outside the US to potential copyright problems and don't even give them a way to avoid those (other than simply not using P18 data at all)? --2A02:810B:580:11D4:7D67:F45A:2EAC:762D 07:48, 7 January 2023 (UTC)
TadejM's pretty straightforward question asks about the FoP issue. I'm not seeing that there is an FoP issue which would cause an image legal in the US and the object's country, not to be legal in a third country. TadejM was not asking about the wider variable term before entering the public domain issue. So, again: is there an FoP issue we need to be concerned about? --Tagishsimon (talk) 09:59, 7 January 2023 (UTC)
FoP images show copyrighted buildings/artworks. If a third country doesn't have such FoP exemptions, but also grants foreign authors/copyright holders the same rights and protections as their own citizens, the author of the depicted work could sue the users for the unlicensed use of such images in that third country. --95.91.226.100 06:27, 10 January 2023 (UTC)
@TadejM: if a country doesn't have freedom of panorama, the image should be on Commons. What you should do is to just nominate it for deletion which I have done and commons:Commons:Deletion requests/File:Lj. Bezigrad, Holy Spirit parish church 01.jpg.
The input from the anonymous user is completely useless, incorrect and derailed the discussion. Just ignore whatever that person said.
@ChristianKl: no, we don't need to talk to the legal team. We just need to do what we always do when we encounter non-free material: Delete it
@Tagishsimon: you were right. Multichill (talk) 13:54, 7 January 2023 (UTC)
Hi Multichill. Thank you for the reply and for nominating the images at Commons. I guess this should resolve the issue. --TadejM (talk) 14:06, 7 January 2023 (UTC)
@TadejM: maybe you can update Holy Spirit Parish Church (Q78221297) with the architect and when it was build? Than we might know when to undelete it again in the (distant) future. Multichill (talk) 14:12, 7 January 2023 (UTC)
Excellent idea. I've added it. --TadejM (talk) 14:21, 7 January 2023 (UTC)
If it's an image/picture from a someones phone freely given or from the street view mapping vehicle I say leave it alone, don't make a note against it. If it's a part of their life, then they have no problem with attributing their name in submitting it and waving all copyrights. Rather then some unseen entity deleting a public site of religious significance (generating further conflict?). As far as the purpose of copyright? well, what The Art Museum Gift Shop charges for a postcard on behalf of the dead artist is absurd. Are there any verification sites that check for AI edited landmarks or cultural historic buildings? Maybe special case can be made for service that checks images. Upscaled photo might be okay. But replacing the Notre Dame Cathedral with Notre Dame Shopping Mall should be auto deleted, with a permanent note. ;) 107.218.251.194 22:04, 11 January 2023 (UTC)
The point here is that in countries that do not have freedom of panorama, the photographer is unable to waive all copyrights, because their image (of the church) is considered a derivative of the copyrighted architectural design of the church. Commons needs to deal with the law as it is, not as you'd like it to be. If you are not going to minimally educate yourself before contributing, your contribution will, like this one, tend to nothing. --Tagishsimon (talk) 22:19, 11 January 2023 (UTC)
Then let them come & sue me in the different country. I submitted it. I take responsibility for the photo I took on my Family vacation when I was 12, twenty years before that law was passed. If you flagged it for attribution and I say I take responsibility. "deal with the law as it is, not as you'd like it to be" No.no.no. That holds no water in any court other then one that holds animosity toward said landmark. International or otherwise. Grow a spine. And you replied to a different person, of course you probably think one building is another. 107.218.251.194 22:36, 11 January 2023 (UTC)
lol. No, you don't "take responsibility for it". Commons works to ensure that copyvio images are deleted, because Commons is on the hook if they are not deleted. No amount of ignorant blowhard "grow a spine" nonsense changes that. --Tagishsimon (talk) 22:47, 11 January 2023 (UTC)
"Slovenia
Article 55 of the Copyright and Related Rights Act of Slovenia states that "Works that are permanently situated in parks, streets, squares, or other generally accessible places shall be freely exploited," but this is prohibited if the intent of exploitation is for profit. In practice, however, this means that without permission from the author of the works, objects like buildings and statues whose copyrights have not yet expired can only be photographed for personal use, and publications of such images in a tourism portal or a newspaper are prohibited (since newspaper publishing is considered commercial)."
Clearly you want to sell post cards right? 107.218.251.194 22:54, 11 January 2023 (UTC)
No. Commons demands licences that do not have non-commercial restrictions, because the wikimedia community wishes its work to be reusable in commercial contexts. You may not like this, but that's the way it is. --Tagishsimon (talk) 23:00, 11 January 2023 (UTC)
For what it's worth. I understand that https://creativecommons.org/publicdomain/zero/1.0/ means it's open. but clearly No Copyright on Wikidata, and one simply can't indemnify everyone from selling data if it was given in good faith. Literally, turning around and selling an Actuary Table of Births and Deaths for Public Use (2020-2021) probably will get you gulag-ed in these times to be clear. If that is the case everything from countries, (not just churches) with living architects and that stipulation that contains any sky should be deleted OR put a disclaimer. and BTW, Maritime Law still applies especially on the seas. Again let them come, my Dad probably gave services there. Unless the person uses that picture and labels it a brothel. Then I'll be coming to them. That's all. I done. 107.218.251.194 23:07, 11 January 2023 (UTC)
No, my comment was not "completely useless, incorrect and derailed the discussion", just because in this very specific case the image happened to be a copyvio and can simply be deleted. But that's not always the case and that's a discussion that Wikidata has avoided for much too long. Because the carefree usage of Commons images on Wikidata will continue to lead to problems as long as we don't also supply data users with an easy way to query a files' license. Without the ability to verify whether they are even legally allowed to use these files, we expose data users to legal problems. --95.91.226.100 06:27, 10 January 2023 (UTC)
You might think about opening a thread on this issue so that it can be discussed. --Tagishsimon (talk) 22:23, 11 January 2023 (UTC)
There is a cat with 4 ears. One newspaper published the photos, talked about the family caring for the kitten and stated it's name the family gave it. Another newspaper much much bigger posted just the photos of this wild rare breed of the cat at significantly higher resolution. I took said photos, edit and directly edited the photo to include Subject: & Photographed By: from the original article on Social Media. Now there is not here. Let me ask you, If it was a public landmark site what kind of precedence would you declare? Anytime someone pressed the report button for any item be it image or paragraph to delete it? Allow spurious false copyright claims like Youtube and Google does any time someone fills out a roboform? Would you blind everyone to not see the image? Are you looking to sit on some sort of review and judgement committee like a Reddit art moderator? Do you have a image verification service you receive a commission for via recommendation? "we expose data users to legal problems" You are looking to lord over others. It should be said as kindly a possible then. Maybe you should go somewhere else. aka Get lost. 107.218.251.194 22:29, 11 January 2023 (UTC)
You seem to be annoyed because your self-declared copyvio of an image of a cat with four ears has been deleted. That'll probably be b/c Commons deletes copyvios. Probably not a thing to be annoyed about, when you think about it. What did you expect? --Tagishsimon (talk) 22:55, 11 January 2023 (UTC)
Holy $h1t. Do you even parse? And No it wasn't deleted. You didn't even read it correctly. and I stated There is Not Here. Four eared cat is outlier. I'll tell you what. I would happily tag a photographers name to any of their works. I would even send email them after looking at there profile to let them know I submitted it for preservation. You aren't even rational with any of your points now. 107.218.251.194 23:14, 11 January 2023 (UTC)

Merging of Q12206942 and Q51752

Q12206942 (Darth Vader) and Q51752 (Anakin Skywalker) should be merged as they are talking about the same character only one is the evil alter ego of the other. Jhowie Nitnek 19:19, 7 January 2023 (UTC)

Notwithstanding other considerations, some Wikipedias have distinct articles about Darth Vader and Anakin Skywalker. It’s not possible to merge those two items. --Emu (talk) 19:24, 7 January 2023 (UTC)
See WD:Bonnie and Clyde. Mathglot (talk) 17:37, 8 January 2023 (UTC)
And related Help:Modelling/Wikipedia_and_Wikimedia_concepts#Compound_Wikipedia_articles. - Jmabel (talk) 20:28, 10 January 2023 (UTC)
What does it mean? I don't really get it exactly Jhowie Nitnek 08:45, 11 January 2023 (UTC)
@Joeykentin: Just as two Wikipedia articles may cover essentially the same topic, thereby requiring a separate item for each article, a single WP article may also cover multiple topics. In order to let normal items associate with singular topics only, a compound Wikipedia article is treated somewhat like a disambiguation page, and the instructions at that Help page tell you what to do with compound articles.
See essential supremum and essential infimum (Q104804789) for an example (it doesn't matter what those two topics are; in our context the significant word in that label is "and").
So, if you were to find an article that covers both fictional characters Anakin Skywalker (Q51752) and Darth Vader (Q12206942) in equal measure as a cradle-to-grave kind of biography, and essentially only those two characters, the appropriate way to deal with the Wikidata item of that article would be to declare it a multi-topic item, listing those two topics, young Anakin + adult Lord Vader, even as they are temporal instances of the same individual "person". --SM5POR (talk) 19:35, 11 January 2023 (UTC)

Medical University of Lublin (Q74550808) and Medical University of Lublin (Q1916610)

Medical University of Lublin (Q74550808) vs Medical University of Lublin (Q1916610) isn't it the same thing ? Is there a reason to have two separate entities in this case? The former was created by Wikidata_talk:WikiProject_Open_Access#Other_sources_to_import and I can see another problematic National Institute of Geriatrics, Rheumatology and Rehabilitation and the Polish Rheumatological Society (Q74536848) which seems to be a conflation of National Institute of Geriatrics, Rheumatology and Rehabilitation (Q30261076) and Polish Rheumatological Society.
Kpjas (talk) 18:02, 10 January 2023 (UTC)

Someone needs to sort this one out, too

Amazon Headquarters Complex (Q116175358), linked to a French-language article and a Commons Category. No relations, not even instance of (P31). Given the naming, I'd guess it's supposed to be a category with something else as the main item, but I can't find that something else. Maybe badly named? @Arnaud 25: as creator. - Jmabel (talk) 23:20, 11 January 2023 (UTC)

Perhaps remove the Category: prefix and make it a main item. Ghouston (talk) 08:15, 12 January 2023 (UTC)
I have removed the prefix and added a few statements — Martin (MSGJ · talk) 11:10, 12 January 2023 (UTC)

"Move" command not working

Please be so kind to move the sitelink from Cinterion Wireless Modules (Q5121109) to Gemalto M2M (Q116180852).--Hart picken (talk) 12:44, 12 January 2023 (UTC)

Done — Martin (MSGJ · talk) 13:12, 12 January 2023 (UTC)

Populating new items with predefined values

I have about 500 heritage items I want to add to Wikidata. I don’t really want to go to the effort of doing a CSV export, but I would like to create a new item with predefined statements and values. Is this possible? - Chris.sherlock3 (talk) 12:03, 11 January 2023 (UTC)

See WD:QS. GZWDer (talk) 14:44, 11 January 2023 (UTC)
Or Wikidata:Tools/OpenRefine. Note that before importing it is important to match the dataset with existing items to avoid duplicates.--Jklamo (talk) 21:01, 12 January 2023 (UTC)

Mean anomaly

What is the difference between mean anomaly (P2325) and mean anomaly (Q216715) ?

- Io Herodotus (talk) 17:12, 12 January 2023 (UTC)

The first is a property and the second is an item - please see Wikidata:Glossary — Martin (MSGJ · talk) 17:44, 12 January 2023 (UTC)

QuickStatements

As a "new" editor (if you could call it that) I was wondering if there was a simple tutorial on how to use QuickStatements. I was confused by the info at WD:QS. Is there anything else that discusses the topic that could help me start using it? Lettler (talk) 17:35, 12 January 2023 (UTC)

Yes, the documentation is not the best but I think that is all we have. Feel free to ask questions here though. — Martin (MSGJ · talk) 17:43, 12 January 2023 (UTC)
@Lettler: Maybe a video? Several at https://www.youtube.com/results?search_query=quickstatements --Tagishsimon (talk) 19:25, 12 January 2023 (UTC)
One thing I'd caution against as a new user is doing a large import of new data that might not be notable. It's particularly ugly when admins end up deleting thousands of items a new user made with good intentions. Very large edit batches should be approved as bot actions. BrokenSegue (talk) 19:55, 12 January 2023 (UTC)

How do I change the title of a page?

The title of this item is completely inappropriate. Fontaine347 (talk) 20:04, 12 January 2023 (UTC)

It was vandalism. Find the edit in the edit hitory which caused it, and use the undo option. I've done so for this instance. --Tagishsimon (talk) 20:12, 12 January 2023 (UTC)

Additional information about a book

How to add additional information about a book in a book Wikidata item? I didn't find any suitable property for those purposes. This additional information is displayed in bibliographic entries (in my language). D6194c-1cc (talk) 09:35, 2 January 2023 (UTC)

There's no general property for freetext for additional information. You need to find the right property for each piece of information. ChristianKl ❪✉❫ 09:59, 2 January 2023 (UTC)
Well, it's like a second subtitle. It can have different information. For example, "textbook for students of all specialties", or "textbook for physical and mathematical specialties of universities" (translated into English), or "obligatory for all ministries, departments, enterprises and organizations, regardless of their legal form and form of ownership, as well as for individual entrepreneurs" (translated into English). This information is usually placed on the cover of a book. D6194c-1cc (talk) 12:02, 2 January 2023 (UTC)
You could use subtitle (P1680) multiple times with a qualifier of series ordinal (P1545). You would set subtitle (P1680) series ordinal (P1545) 1 for the first subtitle and subtitle (P1680) series ordinal (P1545) 2 for the second subtitle. From Hill To Shore (talk) 12:20, 2 January 2023 (UTC)
But in fact it's not a subtitle. That's why I asked this question. It's rather the description of a book title+subtitle. D6194c-1cc (talk) 12:30, 2 January 2023 (UTC)
if it's not a subtitle then figure out what it is and we can add a property for it. if it's just a generic description of the book then likely it doesn't belong anywhere. if it's a subtitle of a different sort put it under subtitle with a qualifier. BrokenSegue (talk) 17:45, 2 January 2023 (UTC)
When it comes to figuring out what kind of thing it is, see how libraries call it and whether they have mroe words than just "additional information" for it. ChristianKl ❪✉❫ 13:39, 3 January 2023 (UTC)
In official document (ГОСТ 7.1—2003) it is named as "information related to the title". It is defined as "title information that contains information that reveals and explains the title, including alternative title, information about the type, genre, purpose of the work, an indication that the document is a translation from another language, etc."
And it must be provided as stated in the source. So automatic detection of work type by instance of (P31) will not always work as it should. D6194c-1cc (talk) 14:50, 3 January 2023 (UTC)
That sounds like the field contains different kinds of information. In Wikidata we don't mix different kinds together into one property. If you want to add the genre you don't do it in some string based form but use genre (P136). Other information also need it's own property. ChristianKl ❪✉❫ 22:00, 3 January 2023 (UTC)
Well, it might be just property about bibliographic information of the edition as it's specified in the book. It can duplicate some information from other properties but it all depends whether the information is specified on the book cover or not (if not specified then square brackets are used for the information). It's the correct way to show bibliographic information using GOST standard in my county. Another way is to add property like object named as (P1932) for bibliographic information, but it would be more complicated. D6194c-1cc (talk) 09:37, 4 January 2023 (UTC)
Okay, the essential feature of what you are looking for is something like "additional information on the cover"? ChristianKl ❪✉❫ 11:12, 4 January 2023 (UTC)
This feature is described in 5.2.5 section of the standard. Here's a link via google translator: [9]. It doesn't say directly that information is from the cover, but it meant that. I think that something like "title related text" is much more informative. D6194c-1cc (talk) 19:32, 4 January 2023 (UTC)
The Dewey Decimal system and subsequent creation of the library cata(log/logue), 3 x 5 index card was a phenomenal effort of mind and concentration. and We would be wise to ignore critiques someone person sole effort and life work. Humbly I ask you to kindly observe this. From this site, "Documentation Made Easy - 3. Cataloguing or Bibliographic Description Section 10". It covers pen names, aliases, family names/heritages that in general, were as, the author self referenced. [10]https://www.nzdl.org/cgi-bin/library?e=d-00000-00---off-0hdl--00-0----0-10-0---0---0direct-10---4-------0-0l--11-en-50---20-about---00-0-1-00-0--4----0-0-11-10-0utfZz-8-10&cl=CL1.4&d=HASH01fd2e8f1c305b10d9bf3a7b.6.10&gt=2 It also DOES have a VERY SPECIFIC and Syntax Oriented Rules structure. Specifically when performing a cross-reference and self reference. "see", "see also", "see also reference". There was some ambiguity and allowing the librarian and/or filing clerk some say whether all of it should be included on the one forward facing card (or the back of the card could be used as well). Once the Head Librarian made the decision it was finalized for that Library.. til the end of time? I personally would be curious how the Library of Congress finalized their process when they went digital and most certainly they performed a knowledge capture before getting rid of the old cards. 107.218.251.194 21:02, 11 January 2023 (UTC)
It's very similar to the en:International Standard Bibliographic Description standard. The GOST standard that I cited before is very close to the ISBD standard. So some things can be aligned to that standard. D6194c-1cc (talk) 19:37, 13 January 2023 (UTC)
I probably found a way to handle this information in the items. I can use inscription (P1684) property with relative to (P2210) qualifier set to title (Q783521) and optional applies to part (P518) qualifier set to the position on the cover. It solves the problem as I think. D6194c-1cc (talk) 16:48, 6 January 2023 (UTC)
Any other ideas or suggestions? D6194c-1cc (talk) 17:19, 6 January 2023 (UTC)
For your example from above, the solution is intended public (P2360), e. g. Q98704357#P2360. I know this does not answer your initial question, but when possible, it is better to extract normalized information and use freetext only as a last resort. --Lockal (talk) 09:06, 8 January 2023 (UTC)
I know that. But according to the standard I need to display information as it present in the book. So I'll combine both approaches. Currently I use inscription (P1684) snaks filtered by subject has role (P2868) set to title related information (Q116158574) to display title-related information. Next stage I'll use identity of object in context (P4626) to associate inscription with certain terms and deduplicate automatically detected by instance of (P31) work types. So I'll have two groups of characteristics: automatically detected and manually written. Automatically detected items would be displayed in square brackets (as per standard). Manually specified would have text as in the book, but also they might have Wikidata entities associated with the text. D6194c-1cc (talk) 17:34, 10 January 2023 (UTC)
"see:", "see also:", "see also reference:". Would be good. Please do look at the link above, it includes Anthology Name (Trilogy/Epic Title - 'LotR, Chronicles of Narnia, His Dark Works'), Imprint Form (place, publisher, date of publish), its Collation ((Loose Leaf/Journal/Binder/String Bound/Stapled/Soft Cover/Hard Cover) w/ B&W,Sepia, Full Color & Illustrated, Photos or Maps.), Series (Non periodicals/magazines -'Annals of Medicine - Number 25 Volume 12 Supplement 5'). Many Thanks. 107.218.251.194 21:22, 11 January 2023 (UTC)

Proposal: More intuitive UX for complex properties

In my view adding some pieces of data to Wikidata is arbitrarily more difficult than it needs to be. For example, it's really intuitive to denote someone as a child (P40) but there is no equivalent way to do this for uncle (Q76557) (you need to add qualifiers). This is a problem more generally and makes wikidata inaccessible to newbies. But the solution is not to make a million different properties. Instead we can create a new kind of thing a "pseudo-property" or "Wikidata relationship".

So we would create items that encapsulate these property usage patterns. So for example we would have an item to represent attaching a NYSE ticker symbol, assigning someone as an uncle, specifying a ship launch date or adding someone as a first author. (these are links to the test.wikidata where I've mocked up what this all might look like)

After making a ton of these property patterns items we would create gadgets that allow you to invoke them. So instead of typing relative (P1038) then the name of the relative then adding a qualifier kinship to subject (P1039) uncle (Q76557). You would just type "uncle" it would autocomplete to the pseudo-property and you would type the name. Then in the backend it would add a statement with the appropriate qualifier. This could be made to work for a lot of different usage patterns.

Thoughts? If there isn't interest I'll implement it privately in Wwwyzzerdd (Q108267084) but it seems like it would be cool to have this functionality public to any tool that wants to implement it. This require no changes to the core wikibase implementation. BrokenSegue (talk) 23:31, 9 January 2023 (UTC)

I like the idea for things like significant dates. When I asked some time about launch dates I was surprised there wasn't a keystroke configurable tool, but your usage patterns would achieve the same result Vicarage (talk) 23:54, 9 January 2023 (UTC)
@BrokenSegue, @Vicarage: I'm not sure I understand exactly what you are suggesting, but as long as you don't need to create additional properties or pseudo-items in Wikidata for this user interface to work, I'm all for it. As a way to determine what those property patterns are, I would recommend you to look into the concept of model items, which we are currently trying to promote better use of. If there is a model item (P5869) for the class of items you want to grow, you could use that as a template for properties and even default values of statements. Be careful though not to encourage sloppy copying of data from one item to another; the user must confirm that any data added is actually intended and correct.
As I hope in the long run there will be automated means of evaluating the structural consistency between different items using the model items as reference, it would be interesting to let your user interface operate in tandem with such automated mechanisms. Please participate in the discussions at Property talk:P5869 (or other relevant fora suggested there) and let us know what you need. --SM5POR (talk) 13:48, 11 January 2023 (UTC)
@SM5POR: Unfortunately this proposal does require new properties. And many of them at that. Take a look at the test wikidata examples to see the kinds of new properties we would need. And yes I am proposing making pseudo-items on Wikidata. The basic description of this proposal is that we capture patterns of adding data to wikidata into templates in wikidata itself and then let people use the templates as if they were just normal properties. BrokenSegue (talk) 17:48, 11 January 2023 (UTC)
@BrokenSegue: Well, that doesn't necessarily mean it will be incompatible with model items, but just that I'm less likely to endorse it as a good idea, since I think Wikidata has too many distinct properties of limited individual usefulness and want to encourage experimenting with what's already available in the form of generic properties in new combinations.
The property proposal and review process may seem like an obstacle that has to be challenged, but it exists for good reasons; without it Wikidata would probably be littered with millions of redundant or useless properties. I rather see the process as a gentle suggestion to take another route towards your goal.
Pseudo-items are a somewhat different matter since you can create them yourself; the number of items isn't a problem since we already have 100 million of them, but they still have to meet the notability criterion, so careful planning is important.
Now, besides the test implementation, do you have a written specification of the properties and pseudo-items you would need and how they are supposed to be used? I'm not suggesting you should change your approach on a whim, but I would like to take a look and see whether the problems could be solved also on a "shoestring" property budget, without compromising either functionality or compatibility with Wikidata as it exists today.
Given that Wikidata already contains detailed information on everything from subatomic particles to the entire universe, it seems implausible that you couldn't use it to describe also a user interface for editing Wikidata items. The value of a novel or a written work of non-fiction does not depend on the author using a large number of complicated words, but on the clever and innovative ways the words of an already familiar vocabulary have been combined. --SM5POR (talk) 18:50, 11 January 2023 (UTC)
@SM5POR: Per your advice I changed the design to require only a single new property. I made a small write-up for it at User:BrokenSegue/StatementPatterns. Tell me if this makes sense. BrokenSegue (talk) 17:16, 12 January 2023 (UTC)
There's also an alternate design that requires no new properties. I mocked that up here. BrokenSegue (talk) 17:19, 12 January 2023 (UTC)
It looks like you still need lots of new items though. My use case for ship launching would be me typing a keystring that auto-expanded to 'add statement|P354|Q229041|qualifier|P585' and left me in the input box ready to type my data. I would set up my most used sequences under different keyboard shortcuts or menu dropdowns. Javascript can monitor keystrokes and call functions so I guess it could do it, but it would need a more adept person than me. The Merge wizard pops up on Alt-M, could something have prefilled boxes for my common transactions and just leave me to add a date and submit for it to do its magic. Vicarage (talk) 22:28, 13 January 2023 (UTC)
@ Vicarage: yes this proposal is to make an item for each pattern we want to support. I think that's inevitable and acceptable. The only way to avoid that would be to push the data onto the property pages themselves. My backup plan is just to maintain this on github. BrokenSegue (talk) 23:06, 13 January 2023 (UTC)
That triggered me to have a hunt, and I found https://www.wikidata.org/wiki/User_talk:MichaelSchoenitzer/quickpresets which probably could be extended to have a second input box for qualifiers, which would cover my use case (perhaps a third box, as I often do "described by source" with both URL and Title. I'd need to be in ship mode a lot Vicarage (talk) 23:08, 13 January 2023 (UTC)
I like the general idea, and think it would be good to do it with a single new property. I created an example for adding someone as a first author works with one property.
Another way would be to store this relationship on the property itself. That way it would be easier for users to discover that it exists. ChristianKl ❪✉❫ 21:30, 13 January 2023 (UTC)
@ ChristianKl: see User:BrokenSegue/StatementPatterns and its talk page. I already revised the proposal to match your one new-property idea but SM5POR feels like the use of unknown values is improper. I'm still thinking about how else to do this. BrokenSegue (talk) 23:04, 13 January 2023 (UTC)

Allow draft articles to be linked to items (revisited)

See also: Wikidata:Project chat/Archive/2013/12#Draft name space and Wikidata:Project chat/Archive/2014/05#What do do with Wikipedia:Draft articles

Currently it is not possible to add a sitelink to an article in the Draft namespace in Wikipedia. I think this decision should be revisited, as there are numerous advantages of allowing such links, including:

  • Give easy access to data and sources to support creation of the article
  • Allow templates such as infoboxes and authority control to work correctly
  • Show links to other language versions, which can be used for translation
  • Identify existence of a draft article for a topic and help avoid different articles for the same topic being developed simultaneously

Caveats:

  • Sitelinks to drafts do not confer notability, so items should not be created on that basis alone
  • Interlanguage links to draft articles are not appropriate (but links from the draft to other language versions would certainly be useful)

I look forward to hearing your views! — Martin (MSGJ · talk) 13:30, 12 January 2023 (UTC)

There is usually a reason that a Wikipedia article is not yet in main namespace, so I am sceptical. We also need to consider that many projects do not have a dedicated Draft: namespace, but they manage drafts elsewhere (userspace, project space, etc.) so this could become quite messy. —MisterSynergy (talk) 13:53, 12 January 2023 (UTC)
Having links to draft items means that the Wiki that host those drafts will feel more need to delete offensive drafts. I don't think we should as Wikidata put that pressure on individual Wikipedia's without giving them the choice. From my perspective the decision about which links they want would ideally rest with individual Wikipedia's.
Similar to how we have our sitelink badges, we could add a "draft"-badge that allows to add sitelinks to drafts. The interlanguage tool could use the badge to not show articles tagged with it. ChristianKl ❪✉❫ 21:05, 13 January 2023 (UTC)

Widar -- no webservice

Cannot say that I have used Widar recently, though finding that it is offline with regard to its webservice, so unavailable for authentication for use of Petscan. Does anyone apart from MM have access to restart the service? Or offer an alternative? Thanks.  — billinghurst sDrewth 02:55, 14 January 2023 (UTC)

how to process ceb-language duplicate data

I think Q28795373 is the same with Xiamen (Q68744), and there's a disambiguation page, Xiamen Shi (Q29590641). Kethyga (talk) 11:36, 14 January 2023 (UTC)

Right of Survivorship (legal term)

The Right of Survivorship is defined in language that anyone can understand here: https://www.law.cornell.edu/wex/right_of_survivorship . However, when I do a search at WD using this term I am led to Concurrent estate (Q1939539), which is a much more general term describing the different forms of property ownership. I would like to start a new wd item (I think this is the correct terminology?) for The Right of Survivorship . How can I do this? Thanks in advance, Ottawahitech (talk) 15:16, 14 January 2023 (UTC)

The Special:NewItem page should allow you to create a new item. --M2Ys4U (talk) 22:09, 14 January 2023 (UTC)

How can you distinguish single-location bank from one operating many locations?

See Q9165022 vs Punjab National Bank (Q2743499) - how can I distinguish between them? I added brand (Q431289) to multilocation one but it feels like a hack. I do not really have exact count of bank branches, would want to just add that it has many Mateusz Konieczny (talk) 23:05, 13 January 2023 (UTC)

A discrete retail location of a bank should be coded as P31=bank branch (Q21073937), which is to say, there are distinct items for the bank business, and for bank branches. However, WD is a well of poor coding, so :( --Tagishsimon (talk) 23:10, 13 January 2023 (UTC)
@Tagishsimon: Well, Q9165022 is rather an independent bank operating a single location - not just a specific branch.
"WD is a well of poor coding" - oh, I know. I even have listing of detected cases where ontology is egregiously bad (User:Mateusz Konieczny/failing testcases) and I slowly report them at Wikidata talk:WikiProject Ontology (only few in both places but I can provide as many examples as people will fix) Mateusz Konieczny (talk) 08:18, 15 January 2023 (UTC)

Link elections items and legislative term items

How should we link items related to elections (for example 2022 French legislative election (Q106253677)) and the corresponding item for the legislature (for example 16th legislature of the Fifth French Republic (Q112567597))? Koxinga (talk) 10:59, 15 January 2023 (UTC)

Use elected in (P2715) on the legislature item? cf. https://www.wikidata.org/wiki/Q77685395#P2715 --Tagishsimon (talk) 11:38, 15 January 2023 (UTC)
It'd also make sense to qualify the election item's office contested (P541) with parliamentary term (P2937). --M2Ys4U (talk) 16:53, 15 January 2023 (UTC)
Cool thank you! Koxinga (talk)

A Voyage of Discovery and Research in the Southern and Antarctic Regions

The account of the Antarctic expedition by James Clark Ross, published in 1847 in two volumes, currently has six WD entries. There seems to be a work item for the whole work A Voyage of Discovery and Research in the Southern and Antarctic Regions, during the Years 1839-43 (Q56873383), work items for each of the two volumes, and edition items for each of the two volumes. (Plus a pair of items seem to be duplicates.) This organisation does not seem sensible, as subsequent editions of a work need not have the same number of volumes as the first edition. In this case there was only one edition published, although there are a couple of 20th century reprints or facsimile editions. I would like to tidy this up. Is the best option: work item - edition item (first edition) - edition item Vol 1 - edition item Vol 2; or should I just have the work and edition items with links to the two volumes in the edition item? I've seen it done both ways. Full text of the two volumes is available at BHL. Is there anything I have to watch out for to avoid breaking any links? Thanks Kognos (talk) 17:02, 15 January 2023 (UTC)

Probably best to ask this at Wikidata talk:WikiProject Books, or at least post a link there to the discussion here. From Hill To Shore (talk) 17:16, 15 January 2023 (UTC)

Category:Chicago Landmarks (Q105241578) and Category:Landmarks in Chicago (Q8582318)

No doubt there is some distinction, because each lines up to a category in en-wiki, but the difference is not very clear to me, and I suspect that [[:commons:Category:Chicago Landmarks}} associated with the latter should actually be associated with the former. I won't be the least surprised if the same is true for some of the various language Wikipedias, too. - Jmabel (talk) 00:48, 16 January 2023 (UTC)

I've amended the descriptions. Hinges on w:en:Commission on Chicago Landmarks. --Tagishsimon (talk) 00:52, 16 January 2023 (UTC)

Can "nonprofit organization (Q163740)" be considered a generic class of legal form?

Although this item does not represent an actual specific legal form in any jurisdiction, it has been commonly used in Wikidata as a generic class of legal form. I've often seen it used as a value in legal form (P1454) statements when no other more specific legal form item is available in the jurisdiction of a given organization (as in this example). Furthermore, many specific types of legal forms are stated to be a subclass of (P279)nonprofit organization (Q163740), for example in 501(c) organization (Q240625) and personne morale sans but lucratif (Q47091204).

Therefore, it would seem reasonable to state nonprofit organization (Q163740)subclass of (P279)legal form (Q10541491). However, this statement keeps being added and then undone from this item. Can we either state once and for all or else find an alternative item to designate as superclass of nonprofit legal forms? @BohemianRhapsody @Recherchedienst @Infovarius @Paulbe Please weigh in. Fjjulien (talk) 17:54, 10 January 2023 (UTC)

if i were being pedantic i would say it's not a legal form. BrokenSegue (talk) 18:59, 10 January 2023 (UTC)
The key issue here is expressed in the description of legal form (Q10541491) as "within certain legal system", i.e. I would say it depends on the jurisdiction. There may well be some jurisdictions that recognize "nonprofit organization" (or the corresponding term in the language of that jurisdiction) as a legal form of corporate body, but its not a universal property.
Speaking of "properties", even if it would be a legal form, subclass of (P279) would be the wrong property to use, as a subclass of legal form would not cover a limited number of organizations, but a limited number of legal forms (say, legal forms within a subset of all jurisdictions). The appropriate property to use would instead be "form of organization"instance of (P31)legal form (Q10541491).
There is a property applies to jurisdiction (P1001) which seems relevant here, and it could be used as a qualifier to some other statement, though I'm generally hesitant to apply qualifiers to core properties like instance of (P31) or subclass of (P279) and will not recommend that. Maybe you can find some other property to apply it to, such as nonprofit organization (Q163740)has characteristic (P1552)legal form (Q10541491)applies to jurisdiction (P1001)whatever jurisdiction you know this to be true in? --SM5POR (talk) 13:05, 11 January 2023 (UTC)
@Fjjulien: One more caveat though: I don't take lightly the task of interpreting the statutes of various foreign jurisdictions, and will especially discourage equating the Wikidata entity nonprofit organization (Q163740) with some legal form of organization in any specific jurisdiction, because equality is a transitive relation, and all non-profit legal forms in different jurisdictions certainly aren't equal. Instead, you already have the statements you referred to yourself above making those forms of organizations subclass of (P279) nonprofit organization (Q163740). I think that's actually enough. What additional information do you intend to express by your statement? --SM5POR (talk) 13:23, 11 January 2023 (UTC)
@SM5POR I'm only intending to document the legal of form of arts organizations with legal form (P1454) statements, but I face a few issues. Sometimes, there is evidence that an organization has some nonprofit status but the exact legal form (either under a provincial or federal jurisdiction) is not known. In such cases, a superclass denoting a generic nonprofit status would be better than. Other times, the exact legal form is known but there is no item for it in Wikidata. In those cases a generic nonprofit status would be a useful shortcut. Furthermore, a generic superclass would make it possible to query nonprofit organization regardless of the jurisdiction under which they are incorporated, which would be useful for statistical use cases. Fjjulien (talk) 02:42, 16 January 2023 (UTC)
For me, it is absolutely not correct to specify nonprofit organization (Q163740)subclass of (P279)legal form (Q10541491). I really do not see where the "therefore" jump from your first paragraph to your second paragraph comes from. I'm shocked by the logic of the first couple of sentences; let me rephrase them: the heaps of garbage data we have in WD justifies the creation of yet more garbage data. If you do not know the legal form of an organisation, it seems to me that you should not use legal form (P1454) to seek to convey that the organisation is a non-profit. --Tagishsimon (talk) 03:09, 16 January 2023 (UTC)

Problem of Party Number (Taiwan) (P5296): Taiwan Politic party ID url doesn't work anymore

The Party Number (Taiwan) (P5296) is still valid, but the url is not working anymore after the website is updated. How to remove the formatter URL? Supaplex (talk) 12:37, 16 January 2023 (UTC)

Country or country of origin for organisations

Should things in the organization (Q43229) tree have country (P17) or country of origin (P495)? Also websites. Generally we have country (P17) for places and country of origin (P495) says "creative work, food, phrase, product, legislation, etc.", which I suppose extends to websites as being creative, but I'm not sure for organisations. Vicarage (talk) 20:53, 14 January 2023 (UTC)

I'd have thought P17 for organisations and for legislation. On the latter, it's not as if legislation is exported to other countries. --Tagishsimon (talk) 21:06, 14 January 2023 (UTC)
Legislation is in the country of origin (P495) definition, I was wondering about websites. Vicarage (talk) 21:07, 14 January 2023 (UTC)
Just a comment: legislation is certainly exported to other countries. For example most of common law originally came from England, I think? Ottawahitech (talk) 21:22, 14 January 2023 (UTC)
For legislation, the definition is hopelessly hopelessly wrong: https://w.wiki/6Dxp ... P17 also seems to be preferred for websites: https://w.wiki/6Dxq
A statute passed by a legislature in one country does not somehow get exported to another country, Ottawahitech. Case law, and the adoption thereof, is not the same thing. --Tagishsimon (talk) 21:26, 14 January 2023 (UTC)
If you are saying that the word export does not mean adopt then I agree. Common law was adopted by many countries, not exported from the British Empire (Q8680). Ottawahitech (talk) 21:39, 14 January 2023 (UTC)
I'm not saying that, no. I'm saying that legislation is not the same as common law. The Sale of Goods Act 1979 has not and will not be exported to / adopted by any country other than the UK. Carlill v Carbolic Smoke Ball Co may, for all I know, be valid case law in any number of countries. The issue here was legislation, not case law. --Tagishsimon (talk) 21:51, 14 January 2023 (UTC)
We do have many cases where legislation is copied almost identically between legislatures. For example, Sale of Goods Act 1908 (Q22908459) from New Zealand is a near copy of Sale of Goods Act 1893 (Q7403813) from the United Kingdom. This isn't an isolated example. Now, how we model these in Wikidata is a separate question. Rather than relying on country of origin (P495) we could treat it like a derivative work and use based on (P144) and derivative work (P4969).
Another, but related, messy situation is when laws were applied to a territory that subsequently gained independence. Sale of Goods Act 1893 (Q7403813) applied to the United Kingdom but after Ireland gained independence in the 1920s, it remained part of Irish law. Now the country of origin (P495) is United Kingdom of Great Britain and Ireland (Q174193) but the applies to jurisdiction (P1001) should also include the post-independent Irish Free State (Q31747). A further complication is that once the law is incorporated into two different countries, it can take divergent paths. For example, Sale of Goods Act 1893 (Q7403813) was repealed in the UK by replacement laws in 1979 and 1981. The replacement laws do not apply to independent Ireland, so the legislation is either still in force there or has been repealed at a different time.
I'm not going to be getting myself too heavily involved in modelling legislation here (I have far too many other headaches at the moment) but I thought the additional context would be useful for the discussion. From Hill To Shore (talk) 14:49, 15 January 2023 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘

The initial question here made me think immediately of Yivo Institute for Jewish Research (Q259337), which has

. - Jmabel (talk) 00:54, 16 January 2023 (UTC)

I can see that knowing the original location (and hence country) for a now multi-national organisation is useful. I guess organisations that remain within a few countries can use country (P17) for their current activities, and you'd give up once the number became too big. Similarly a website has country (P17) for its hosting country until it becomes ubiquitous. Vicarage (talk) 18:36, 16 January 2023 (UTC)

Wikidata weekly summary #555

 
Here's your quick overview of what has been happening around Wikidata over the last week.
  • Development
    • Vector 2020 skin: the search that works properly with Wikidata has been deployed
    • Entity Schemas: Making progress on the proof of concept that will help us figure out the technical way forward
    • REST API: getting ready to deploy the first version to test.wikidata.org on February 17th
    • Lexicographical data: When linking to a Sense in a statement the language of the Lexeme for that Sense is now also shown. This makes it easier to understand translation statements for example (phab:T207392)
    • Constraints: Constraint violations are now also shown to logged out users (phab:T272132)
    • Fixing a regression where Wikipedia and co are no longer notified about a disconnected sitelink when the Item that contains the sitelink is deleted (phab:T326082)

You can see all open tickets related to Wikidata here. If you want to help, you can also have a look at the tasks needing a volunteer.

Joe Smith, Sr. versus Joe Smith Sr.

Comma or no comma, we are inconsistent, and English Wikipedia has standardized on no comma, for about 3 years. RAN (talk) 16:50, 16 January 2023 (UTC)

Wikisource has standardized on not including items like Sr. --EncycloPetey (talk) 17:27, 16 January 2023 (UTC)
Many things in Wikidata aren't consistent. To get consistency there needs to be a proposed standard either on a new page on as a section on Help:Label. ChristianKl ❪✉❫ 17:30, 16 January 2023 (UTC)

Interwiki links problem -

Hello Team WikiData,

Hope you are doing well, I am contacting you regarding a problem I am having in Wikipedia Spanish I would appreciate if you can please take a look and indicate whom shoul Id contact to fix it.

Thank you a LOT !!

Uruk from Spanish Wikipedia


Kanem Empire (Q1537016)/Kanem-Bornu Empire (Q1139762)

{{interwiki conflict |items = Q1537016/Q1139762

The problem happens within the said Qs in the realm of the Spanish Wikipedia. This problem has arisen during work we are performing within the completion of the 10,000 articles every Wikipedia should have.

Currently Q1139762 english language article link is to the "Empire Kanem-Bonu" article, and Q1537016 english language article link is to the "Empire Kanem" article. BUT in fact in english Wikipedia "Empire Kanem" is just a REDIRECTION to the article "Empire Kanem-Bonu"

The problem happens when I try to replicate the same model of links that work in English wikipedia within the Spanish wikipedia.


Currently Q1139762 spanish language article link is to the "Imperio Kanem-Bonu" article, and for Q1537016 there is no link provided to an Spanish language article. I have created a spanish article called "Imperio Kanem" (that is the official name of Q1537016), and in it just placed a REDIRECT to the Spanish article "Imperio Kanem-Bonu", the problem is that when I try to provide a link for Q1537016 to a spanish wikipedia article it would reject the insertion of the article "Imperio Kanem" !!

Would appreciate very much if you could process this request, and provide an exception; or guide me to someone that could fix the problem.

Kind regards, Uruk (talk) 21:21, 16 January 2023 (UTC)

When you add the sitelink you also need to add the badge "sitelink to redirect" or "intentional sitelink to redirect". Then it should work! — Martin (MSGJ · talk) 23:42, 16 January 2023 (UTC)

Elizabeth S. Greene

I have added some identifiers for Elizabeth S. Greene (Q116225257). Can anyone help to find more? Lights and freedom (talk) 01:13, 17 January 2023 (UTC)

Pesky symbols in identifiers

New property Longman Dictionary of Contemporary English Online entry (P11482) sometimes has as a value an identifier that includes the # symbol. These links are not working correctly. I know that there is a way to get them to work, so would someone please fix things so that URLs such as https://www.ldoceonline.com/dictionary/incorruptible#incorruptible__3 actually go to the correct place? There are two examples of identifiers containing the # on the property page: https://www.wikidata.org/wiki/Property:P11482

And the links for identifiers for property Government of Canada Core Subject Thesaurus ID (P9400) and Thésaurus des sujets de base du gouvernement du Canada ID (P9401) that have a + symbol in them are also not working. Can someone fix those too? For example, this URL works when entered in a browser, but doesn't work from the link in aviation accident (Q744913): https://canada.multites.net/cst/default.asp?lang=ENG&searchString=Aircraft+accidents&op=DISPTERM AdamSeattle (talk) 23:46, 12 January 2023 (UTC)

Handling earthworks (without a digger)

I have a problem with earthwork (Q11880367) and earthworks (Q1349587). The former is the more generic, the latter says its generic, but has an alias "Earthwork fortification" which could be my area, artillery forts. Each article has a English wikipedia entry, but the former's Earthworks (archaeology) has military use of earthworks up to motte and bailey castles, but the latter's Earthworks (engineering) mentions their military use throughout history. My inclination would be to have earthwork (Q11880367) be clearly labelled the military use, as archaeology can apply to WWI trench systems, and earthworks (Q1349587) be the more general civil engineering side. Perhaps the item titles should be "archaeological earthworks" and "earthworks". But each has has lots of Wikipedia articles I can't read, and while there is an obvious mistake putting the eo archaeology one in the engineering side, I don't know about the rest. Vicarage (talk) 17:52, 16 January 2023 (UTC)

Google translates helps with understanding Wikiarticles in other languages in a decent matter. You don't need to get everything to understand the topic and Google Translate is good enough for that. ChristianKl ❪✉❫ 18:31, 16 January 2023 (UTC)
Made the 2 things distinct and moved all fort stuff to one of them Vicarage (talk) 16:09, 17 January 2023 (UTC)

Merging Wikidata items and the impact on Structured data in Commons files

Yesterday, two Wikidata items were merged, see Can child's drawing (Q12407075) and child's drawing (Q50950479) be merged? Q50950479 was merged into the other. In Commons three files were linked to this Wikidata item, via Structured data. After the merge, they kept their old link to Q50950479, it has not been changed to Q12407075. In Wikidata item Q50950479 I could not find those files via "What links here". In Commons I did, so I changed them manually. But to me this does not look like a robust solution. Questions:

  1. In case of a merge of two Wikidata items, shouldn't the Structured data in Commons files be changed as well?
  2. If not necessary, will those files still be shown via View it!?

JopkeB (talk) 14:00, 17 January 2023 (UTC)

On Wikidata there's a bot that does the updates. If Commons currently doesn't have a bot that does that, it would make sense for someone to run a bot. ChristianKl ❪✉❫ 15:26, 17 January 2023 (UTC)

Voting now open on the revised Enforcement Guidelines for the Universal Code of Conduct

Hello all,

The voting period for the revised Universal Code of Conduct Enforcement Guidelines is now open! Voting will be open for two weeks and will close at 23:59 UTC on January 31, 2023. Please visit the voter information page on Meta-wiki for voter eligibility information and details on how to vote.

For more details on the Enforcement Guidelines and the voting process, see our previous message.

On behalf of the UCoC Project Team,

JPBeland-WMF (talk) 00:27, 18 January 2023 (UTC)

Personification (Q207174) statement

I think there should be a personification (Q207174) statement for example:

United Kingdom (Q145) -> Britannia (Q138396)

Berlin (Q64) -> Berolina (Q574396)

Scheldt (Q37620) -> Scaldis (Q116234451) Jhowie Nitnek 14:13, 17 January 2023 (UTC)

There is national personification (Q1142281) used with a qualifier. I think that works well enough, I don't think we need another property Vicarage (talk) 14:53, 17 January 2023 (UTC)
What qualifier exactly? Jhowie Nitnek 15:44, 17 January 2023 (UTC)
See how its used in Brittania Vicarage (talk) 16:07, 17 January 2023 (UTC)
Linking from the personification to the thing it presents like done in the Berolina item seems to make more sense. ChristianKl ❪✉❫ 15:02, 17 January 2023 (UTC)
Would a link from the thing to the personification not also be useful? GrandEscogriffe (talk) 20:07, 18 January 2023 (UTC)
Yes and no. Thing about WD is it is linked data. So long as the information is in a reasonable place - either in the subject or the object item, or an item linked to the subject or object - then we're good. Wanting all the information on Berlin to be in the Berlin item is very much WP thinking. WP is not WD. Items are not articles. --Tagishsimon (talk) 20:26, 18 January 2023 (UTC)
Its a strong WD principle that we avoid duplication of data, so we avoid inverse statements if possible. They can be inconsistent, and encourage queries that only get half the picture. So we say A is a member of B, but not also B has members including A. There are exceptions for convenience, but I don't think one is needed here. Vicarage (talk) 20:26, 18 January 2023 (UTC)

Birthhouses

We have a lot of birth house (Q19979289) such as Birthplace of Jiří Voskovec (Q33243704) which usually are not actual birth places but rather places where the person spent his/her early years. They're nevertheless still called birthhouses. What's a preferred way of modelling the relationship between these places and the person? Using instance of (P31) : birth house (Q19979289) / of (P642) : <<person>> is frequent but looks quite clumsy to me and the property of (P642) is now a little frowned upon. Would this deserve its own new property? Vojtěch Dostál (talk) 21:09, 17 January 2023 (UTC)

The individual can link to the house with residence (P551)object has role (P3831)birth house (Q19979289). If you have more information about when the person lived in the house start time (P580), end time (P582), and point in time (P585) can provide additional information.
In the other direction, named after (P138)subject has role (P2868)birth house (Q19979289) would work. ChristianKl ❪✉❫ 21:20, 17 January 2023 (UTC)
@ChristianKl Thanks! The idea with residence (P551) is good. Vojtěch Dostál (talk) 13:46, 18 January 2023 (UTC)

The Wikibase REST API is now available for testing on Test.Wikidata

Hello everyone,

As you may already know, we are developing a new REST API to improve programmatic access to data from Wikidata and other Wikibase instances. Many thanks to everyone who gave us the initial feedback on the proposed implementation plan of the REST API and subsequently tested the experimental phase of it on Beta Wikidata.

All the items listed in our previous communication as still needing to be completed for the initial release have now been completed and we are pleased to announce that the Wikibase REST API is planned to go live on Wikidata about a week later.

What is the current state capable of doing?

  • Retrieve the data of an Item with GET /entities/items/{item_id} and filter what fields (i.e. type, labels, descriptions, aliases, statements, sitelinks) are returned when reading the Item data
  • Retrieve all statements of an Item with GET /entities/items/{item_id}/statements
  • Retrieve the data of a single statement of an Item with GET /entities/items/{item_id}/statements/{statement_id} or GET /statements/{statement_id}
  • Conditionally request the data only if it has changed since the specified revision/timestamp (using If-None-Match, If-Modified-Since HTTP headers)
  • Create a statement on an Item with POST /entities/items/{item_id}/statements
  • Authenticate and authorize as a Beta Wikidata user when making edits using the API, as well as provide edit tags and edit comment, and mark an edit as one made by a bot.
  • Replacing the statement on an Item with PUT /entities/items/{item_id}/statements/{statement_id} or PUT /statements/{statement_id}
  • Removing the statement from an Item with DELETE /entities/items/{item_id}/statements/{statement_id} or DELETE /statements/{statement_id}
  • Automated edit summaries

The following functionality is still awaiting a final security review of a library and will be enabled once that is finished:

  • Editing a statement on an Item with PATCH /entities/items/{item_id}/statements/{statement_id}’ or ‘PATCH /statements/{statement_id}

Please note that the following items previously listed under "The following will likely not be available in the first version but follow later" have not yet been finished and will not be available in this initial release:

  • Creating or deleting an Item
  • Getting a statement from an Item based on the Property ID in the statement
  • Any operation (reading, adding, editing, removing) on sitelinks, labels, descriptions and aliases
  • Any operation on entity types other than Items (i.e. Properties, Lexemes, …)
  • Translated error messages

However, we will continue to work on these items in the future to provide a more comprehensive REST API. In order to prioritize our next steps, we would greatly appreciate feedback on which missing features you would find most useful to add next.

If you have any questions or want to provide feedback please let us know at Wikidata talk:REST API feedback round. As this new API is more tailored to the Wikibase data model than the Action API, we have outlined the differences in the JSON data format between the two for you for easy comparison.

Thank you for your patience and support as we continue to improve the Wikibase REST API.

Cheers, -Mohammed Sadat (WMDE) (talk) 18:55, 18 January 2023 (UTC)

remove multilingual wikispecies on Cryptolechia rectimarginalis (Q20686360)

Cryptolechia rectimarginalis (Q20686360) contains wikispecies link under multilingual sites, however, wikispecies is only in english language. how do we proceed? remove from multilingual sites or maintain status quo? <_>Jindam vani (talk) 02:31, 20 January 2023 (UTC)

Most of the business on Wikispecies is in English, but a lot of the content is in Latin and there are homepages in many languages. —Justin (koavf)❤T☮C☺M☯ 02:45, 20 January 2023 (UTC)
Having the link is important and the multilingual tab is currently the one that links to Wikispecies. I think you could argue that maybe a name like "Other Wikiprojects" would be better than "Multilingual sites" but that's not a discussion about individual items. ChristianKl ❪✉❫ 14:53, 20 January 2023 (UTC)


[Survey for Wikidata data Reusers] Help us Improve Wikidata by Sharing Your Experience on Ontology Issues and Data Reuse Impact

Hello,

We are conducting a survey to better understand the ontology issues within Wikidata and their impact on data reuse.

You may recall that in 2021 and 2022 we run the Data Quality Days, which generated a lot of very useful discussions on the processes around increasing/maintaining data quality and utility on Wikidata. We identified various types of ontology issues, and we would like to get input on which of these are the most problematic for you in your use of Wikidata's data.

This survey has two sections and will take 20–25 minutes to complete. Optional open-ended questions may lengthen the completion time.

In the first section, we will present descriptions of the ontology issues we have found. We will ask you to evaluate the impact of these issues on your work, and you are also invited to share any other ontology issues you have detected in case we missed them.

The second section focuses on how you use data from Wikidata. Most of the questions in this section are optional, meaning you do not have to share details of your work unless you find them relevant to the issues you would like to share with us. They are helpful to us in understanding the context of your issues better, however.

This survey is anonymous, but there is an optional email field at the end for follow-up questions. Providing an email will make your responses not completely anonymous. A summary of the results of the survey will be published as a whole at Wikidata:Ontology issues prioritization and will not include any identifying information.

If you would like to participate, please use this link (LamaPoll): https://wikimedia.sslsurvey.de/ontology-issues/

We kindly request your participation by Friday, February 17th at 23:59 UTC.

If you have any questions, please do not hesitate to let us know by replying directly to this message or leaving a note at Wikidata talk:Ontology issues prioritization.

Many thanks in advance for your participation.

Cheers, -Mohammed Sadat (WMDE) (talk) 16:50, 20 January 2023 (UTC)

Bot needed

Where we have instance=human and there is no description, can we have a bot fill in birth and death years in parenthesis as the description? I keep creating duplicate items, but if there was a minimal description, it would be obvious we have a variation of the name of that human already. RAN (talk) 20:48, 16 January 2023 (UTC)

doing this has caused controversy in the past. really this ought to be a feature of wikibase. BrokenSegue (talk) 21:26, 16 January 2023 (UTC)
It ought not to be controversial. If there is no description currently, then any accurate description is an improvement, no? — Martin (MSGJ · talk) 23:43, 16 January 2023 (UTC)
Some bots that currently add descriptions only add a description when the description field is empty. If you create a very minimal description than those bots would be stopped from adding a better description.
WikiFunctions moved in August into beta. I'm hoping that in the future it will be able to do the job of providing Wikidata descriptions. ChristianKl ❪✉❫ 16:09, 17 January 2023 (UTC)
'fraid I cannot buy this really this ought to be a feature of wikibase thing. First, it is perfect is the enemy of good. It'd seem to involve changes to the main UI, all APIs, the SPARQL engine and/or the triplestore; and I don't find it credible that anyone would prioritise that scale of change to do something that a pedestrian bot is capable of doing. Second, even were work to be done on this, I'm pessimistic anything would be delivered in a meaningful timescale. Third, it seems to have a perverse driver - the controversy over the pattern to be adopted for human item labels. Unclear why exactly the same controversy would not attach to it. It's depressing af that there's resistance on WD to the use of a standard pattern to include dob & dod where there is no ambiguity attached to either of those dates, but there we are; there's no accounting for folk. --Tagishsimon (talk) 23:23, 20 January 2023 (UTC)
I’m not sure about your third point: Are you suggesting that the debate would shift from “should manual descriptions contain life dates” to “should automatic descriptions contain life dates“? Because I don’t think so. --Emu (talk) 23:58, 20 January 2023 (UTC)
Weird that you oppose bots adding dob and dod, but would be alright with automatic descriptions including them. --Tagishsimon (talk) 00:14, 21 January 2023 (UTC)
the argument is that duplicating the DOB/DOD information leads to problems. it goes against the spirit of structuring the information. if the descriptions are automatic it's no problem. i'm honestly on the fence on this one. BrokenSegue (talk) 01:37, 21 January 2023 (UTC)
Exactly, thank you for summarizing the problem. --Emu (talk) 20:08, 21 January 2023 (UTC)

Automated way to upload a single entry in Findagrave?

I have creating entries in Wikidata for identified people at Commons in historic images. It would be great when I find the person at Findagrave, if I could just import that one entry into Wikidata. Do we have a mechanism in place already? RAN (talk) 23:28, 16 January 2023 (UTC)

@Richard Arthur Norton (1958- ) I suggest to create a corresponding Mix'n'Match catalogue. Vojtěch Dostál (talk) 09:35, 17 January 2023 (UTC)
New-Q5 tool is not fully automated, but it expedites the process of creating (or editing) items for humans. Pasting the birth/death dates and the URL from a Find a Grave profile (or WikiTree, or obituary, or any URL) reduces some of the tedious, repetitive steps. -Animalparty (talk) 18:25, 17 January 2023 (UTC)
@Animalparty Is there instructions or an example of this tool's use?
Thanks, -- Ooligan (talk) 20:28, 21 January 2023 (UTC)
Ooligan To my knowledge, no, aside from the GitHub documentation page. But it's fairly simple to play around with. Entering a name and (optionally) a date of birth and/or death, along with (optionally) a URL spits out a script that can be modified or run directly in QuickStatements. It sometimes automatically detects existing items with similar names (especially if the existing items lack birth/death dates), but not always: to add dates with nicely formatted reference to an existing item, I just replace "LAST" with the QID and remove "CREATE" before clicking Import QuickStatement (the script can be further modified as needed, e.g. removing the last line will prevent the addition of described at URL (P973), which is useful for reducing redundancy for for URLs that have corresponding external IDs like Find a Grave memorial ID (P535) and IMDb ID (P345)). And I always double check the item after creation to ensure things make sense (e.g. a woman's maiden surname may be recorded as a second given name, or the tool may choose the wrong name and infer the wrong gender based on homonymous names like José (Q66827836) vs. José (Q2190619)). Sometimes I just add a dummy date to get a nicely formatted, referenced statement that I can then use to source other properties (spouse, birthplace, etc.). -Animalparty (talk) 21:08, 21 January 2023 (UTC)
@Animalparty I understand the tool much better. Thank you for your very detailed response and your time. -- Ooligan (talk) 21:30, 21 January 2023 (UTC)

Number of viewers/listeners vs. Number of views/streams

Hello, I'm fairly new around here and I'd like to know if it would make sense to have a separate page from number of viewers/listeners (Q114771092) for number of views/listens or views/streams. If you go to Last.fm, for instance, artist pages have two separate numbers for their total amount of Listeners and Scrobbles, and currently I suppose there isn't a way to add 2 qualifiers for both numbers under Last.fm ID (P3192), only for one of them (Listeners). Moreover, for Spotify we might need to have a way to show monthly listeners. Of course, this type of data changes over time but so do other qualifiers such as number of subscribers (Q114771100), which is probably why users add point in time (P585). "Number of views" and "number of listens" being aliases seems misleading. Lily Allen having 720,956,190 views on YouTube doesn't (necessarily) mean that almost 721 million people watched her videos since one person (or bot) can watch multiple videos and/or watch the same video multiple times. We can easily realize this by looking at her Last.fm in which she has 2.6 million listeners vs. 80.9 million scrobbles. Guttitto (talk) 01:07, 18 January 2023 (UTC)

Numbers of views is something that consistently changes as such it doesn't work well for a non-changing external ID qualifier. ChristianKl ❪✉❫ 15:26, 19 January 2023 (UTC)
There is already a property for number of viewers/listeners (P5436) though, which I forgot to include in the original topic, and it is also variable. Guttitto (talk) 23:00, 21 January 2023 (UTC)
Same for number of subscribers (P3744) or social media followers (P8687). Guttitto (talk) 23:01, 21 January 2023 (UTC)
Also I didn't propose an external ID this is data type quantity. Guttitto (talk) 23:03, 21 January 2023 (UTC)


Merging corvée Q466259 and Corvée Q1369542

I think merging these two would be good. 178.26.170.21 12:28, 21 January 2023 (UTC)

A few of the language wikis appear to have distinct articles for both items, so you'd have to understand what's going on there; normally (despite the labels) there's a distinction to be drawn between the two item concepts; or else the wikis have duplicate articles, which ceb-wiki apart, is less likely. --Tagishsimon (talk) 13:07, 21 January 2023 (UTC)

Updating population of Hungarian municipalities

Please review my bot action to update the population of Hungarian municipalities, and if you are fine with that, please express your support on my request. Thank you. --Bean49 (talk) 16:48, 21 January 2023 (UTC)

Linking external images with an url

How do I do that? Like with external 360 deg pano images. I like to add images to venues. Because no text describes a location like an image. The offerings from wikimedia are sometimes ok, but often there is nothing. Or its dated. On the other hand there are always good high qulity pictures of venues by the venues which are made to be public and freely used. They are just not on wikimedia for different reasons. How to add them via urls to Wikidata? Thats something that would improve the description of a venue a lot. I'm looking for something like: 

image_url: url
external_image_url: url
image_resource: url
press_image: url

Sources would be something like Flickr or press photos offered by a venue on their website. Example: here from this page. I'm new. Am I missing something? There are a few ways to link external text with an url. But no way for images? Why is that? Thanx.

I think it is not suitable for wikidata for several reasons. Yet, it has been proposed before. --Christian140 (talk) 12:43, 17 January 2023 (UTC)
I understand and agree to some of this. What about restrictions? For examples images which are clearly described as for public consumption by the issuer. Like press images. What about a restriction to certain big image providers like flickr_image_url? They have their own spam protections. I think its bad policy to say: There will be problems with images, so we don't allow ANY urls, except if they are 360 degress. Which imho doesnt make that much sense. There is also some spam in panoramas. And what about a wikimedia image urls database? You have to first add the url to the "wikimedia image url database", and if it's kept there - with whatever restrictions und rules you like - you can use it in wikidata. if the url is deleted in wikimedia, its deleted in wikidata? Databob (talk) 13:51, 17 January 2023 (UTC)
@Databob: What's the motivation of not licensing press images under an open license? It's to shut down criticism that uses the press images. Press images for public consumption should be licensed under an open license. Our policy creates incentives for doing so.
If you see a press image that you want to include, you can sent the publisher an email: "Hey, I want to include your image on Wikidata and WikiCommons. For that your image would have to be published under an open license, can you please publish it under an open license?"
If the venue truly intends the images to be freely used, the might appreciate your efforts to help their images to be seen more broadly by licensing the image accordingly. If they don't want their images to be freely used, they won't. ChristianKl ❪✉❫ 14:15, 22 January 2023 (UTC)
@ChristianKl: The reasons I can think of: 1. They venue has no clue that their PR photos can not be used everywhere unless they upload it to Wikimedia. 2. The photographer - not the venue - of the fotos doesn't allow an open license for his work, even if if the venue payed for it. My first thought was yours: Email them and ask them to upload their PR photos to Wikimedia. My second thought was: All of them? There must be a better solution.
That the venues don't open the photos because they are afraid they might be used against them? I don't understand that.

Linking to Outreach wiki not an option

Linking to Outreach wiki is not an option among Multilingual sites of Wikidata items... Is there a (special) reason for this? Any other way to do it? --Zblace (talk) 18:40, 12 January 2023 (UTC)

I think the decision whether to promote a Wiki to be official in the sense that it's interlinked was historically done by the WMF. ChristianKl ❪✉❫ 20:44, 13 January 2023 (UTC)
phab:T171140.--GZWDer (talk) 14:53, 14 January 2023 (UTC)
That does seem like it's just an undone task. @Lydia Pintscher (WMDE):? ChristianKl ❪✉❫ 15:29, 17 January 2023 (UTC)
Mpfh. Indeed. Someone forgot to coordinate with us. I'll have a closer look what's missing. Lydia Pintscher (WMDE) (talk) 10:24, 22 January 2023 (UTC)

Scope of identifiers: must they be general and official?

@Буквы

  • Greetings Wikidata editors, I was examining some additions of YouTube video ID (P1651) and YouTube channel ID (P2397) (YouTube links) to many many items, such as Chernobyl disaster (Q486). The addition is here. However, it seems that most other Wikidata:Identifiers are for database entries that describe the whole item, such as a general encyclopedia entry. These YouTube links are sometimes to a specialized video, unofficial niche material, or to an unverified channel, and in general they seem to be more appropriate as references/reliable sources, if at all, than Identifiers in their own right. Perhaps I am wrong, and perhaps the usage of Identifiers is more general than I had assumed. What do you think?

Elizium23 (talk) 18:38, 21 January 2023 (UTC)

yeah I think the item should be official in some way. Maybe the official upload of the music video or the official youtube channel. Just a youtube channel about the subject or a random upload of a video of the subject are not correct uses for these properties. BrokenSegue (talk) 19:37, 21 January 2023 (UTC)
(change to using P and Q templates above) My preference is to add a new identifier for the YouTube channel, and then add the videos as described by source (P1343) with qualifier URL (P2699), see HMS Victory (Q213958) which promotes the excellent videos by Drachinifel (Q112737775). If useful I add a title qualifier too. That way the source can build up its own portfolio. Occasionally I use described at URL (P973). I think this is better than asserting a single YouTube video for the item as a whole. I certainly don't get hung on justifying the addition. I add things which I consider enhance the item, my standards are lower for more obscure items. Vicarage (talk) 19:45, 21 January 2023 (UTC)
Thank you both for this guidance.
On a related note, I've noticed "Described in source Obalky.cz" added to many items. Is this the right way to do that? Obalky doesn't seem to have any substantial information, and my translator informs me that this is a book catalog, not intended as a general encyclopedia or database.
@Vojtěch Dostál - thoughts? Elizium23 (talk) 02:53, 22 January 2023 (UTC)
This database provides pictures, often even in cases when Wikimedia Commons has none. Anyway, book catalogues are no less an integral part of Wikidata than general encyclopedias. Vojtěch Dostál (talk) 08:52, 22 January 2023 (UTC)

Watercourse distance

Most waterways distance are not referenced, which is not a problem since they can easily be checked on a map. Now, user:Fralambert is given the mission to depreciate their lengths if not referenced, which I consider a disorganization. Any advise to solve this ? Yanik B 13:20, 16 January 2023 (UTC)

Has this already been discussed somewhere? ---MisterSynergy (talk) 13:48, 16 January 2023 (UTC)
It is discussed at User_talk:Fralambert#Dépréciation. @Fralambert: is wrong to deprecate these statements. Deprection, per Help:Ranking is "used for statements that are known to include errors (i.e. data produced by flawed measurement processes, inaccurate statements) or that represent outdated knowledge (i.e. information that was never correct, but was at some point thought to be)." This is not the case with unreferenced river lengths. I don't think there is general support on WD for deprecating nor removing statements with no reference and I'm unaware of any specific support for this watercourse action. So "If there is no source for a length, I depreciate, period. No more complicated than that." is all of pointy, wrong and uncollegiate. --Tagishsimon (talk) 14:05, 16 January 2023 (UTC)
Should I delete them instead? Since they most of them seem to be original research. What I ask is for a source, nothing else. Just note that all these deletions [11], [12] or [13] come from @YanikB, not me. Fralambert (talk) 23:50, 16 January 2023 (UTC)
If you doubt them, it would sense to actually check whether the values are correct. If there's a pattern of most unsourced waterways distance being frequently wrong that might be ground to do something. In the absence of such a pattern being established the norms of Wikidata neither allow for decprication nor removal just because someone didn't provide a source. ChristianKl ❪✉❫ 22:41, 21 January 2023 (UTC)
Ok, I restore all the lenghts of Rat River (Q11984856) [14], the sourced ones, the unsurced ones, and even the one modified by @YanikB here. Just if you exclude the one of the total lenght of streems of the watersheed, their is a difference of 88 km between the smallest and the biggest values. That is a 67% of difference, that is a lot. And this is why you need reliable source. Fralambert (talk) 14:25, 22 January 2023 (UTC)
There are several possible lengths for a stream: the distance between the main source and the mouth; the distance between the farthest source and the watershed mouth; and sometimes the total length of the watershed’s streams. Ideally, the last two should be on the watershed element, however, if there is no watershed element, we can qualify the lengths and put a preferred rank on a length other than the one without qualifiers (the distance, following the course, between the main source identified on a map and the mouth). No need to devalue a volunteer’s work. --Yanik B 14:48, 22 January 2023 (UTC)
If the lenght are from your own calculation, it would be useful to use determination method (P459) as a qualifier and observed in (P6531) : Toporama (Q56419922) as a the source. Not sure this is the best solution, but at least this is a source. Fralambert (talk) 15:13, 22 January 2023 (UTC)
I don’t use the Toporama calculation tool, I rely on the layout provided by Canvec on OpenStreetMap. --Yanik B 16:08, 22 January 2023 (UTC)
Then cite that you use CanVec [15] (we don't have a item for it, we should probably create it.), OSM orCRHQ. It's not that hard [16]. Fralambert (talk) 16:22, 22 January 2023 (UTC)
Find. Now, how can we specify that a length applies to the coordinates given in coordinate location (P625) ? --Yanik B 17:36, 22 January 2023 (UTC)
If you want to calculate it from the other statements on Wikidata, inferred from (P3452) is there for that. ChristianKl ❪✉❫ 19:31, 22 January 2023 (UTC)

Wikidata weekly summary #556

 
Here's your quick overview of what has been happening around Wikidata over the last week.
  • Events
    • Past
    • Upcoming
      • Next Linked Data for Libraries LD4 Wikidata Affinity Group call January 24, 2023: We will be discussing plans for 2023, and requesting feedback and ideas for future programming. Based on a previous recommendation to host a discussion on advocacy for Wikidata within libraries, we will introduce and provide time to complete a short survey to foster conversation in a future Group Call. Please come ready to share your thoughts Agenda
      • Introductions to Wikidata, online meetings organized by Wikimedia Australia on January 25, 26 and 31
  • Development
    • Wikibase REST API: The new API is now available for testing on Test.Wikidata
    • Ontology issues: We finalized and published the survey to better understand which types of ontology issues are most problematic for reusers of our data.
    • Wikipedia and co: Fixing a regression where there is no entry in recent changes and watchlist when an Item is deleted that was connected to an article on that wiki (phab:T326082)

You can see all open tickets related to Wikidata here. If you want to help, you can also have a look at the tasks needing a volunteer.

View it! for adding missing P18 statements

 
Screenshot with View it! enabled.

We have shared about View it! before in the weekly summaries, but I wanted to post here to explicitly call out a new Wikidata-specific feature. View it! is a user script that will show you related images at the top of the page you are viewing—looking at a Wikidata item, it will show you any images in the Commons category if there is a Commons category (P373) statement and any images in a SDC depicts (P180) statement that use that Wikidata item.

Now there is a "+" button for every displayed image (see screenshot) that will allow you to easily add a missing image (P18) statement with a click—if the script is pulling up images but the item has no P18. We've been having some fun testing this feature. For example, here's a query of almost 500 Rijksmonument (Q916333) sites with no P18 statements, but with P373 statements, implying they have images available. Open them up in tabs and you can quickly click-to-add on them without ever leaving Wikidata (and faster even than Drag'n'drop). There is also a copy-to-clipboard button that copies the file name, in case you want to copy the name of an image you are seeing on the top bar in any other image property, like logo/flag/etc.

Please let us know if you have any feedback on using this tool, or would like to see it as a full gadget. Dominic (talk) 18:07, 23 January 2023 (UTC)

Nice idea, especially if gets more photos of my beloved forts. I sometimes go looking for aerial views, models and design plans of sites to add, so a possible extension would be a dropdown to choose which image type is to be populated. Vicarage (talk) 18:12, 23 January 2023 (UTC)
@Vicarage: We considered allowing other property types, but we didn't want to add too many clicks to the P18 experience or confuse users, if P18 is likely to be 95% of the usage. Maybe there's a different way to do it, though. Dominic (talk) 18:19, 23 January 2023 (UTC)
Had a go. The confirmation would start to irritate if I was doing it a lot, and I did wish that a tap on the photo would bring up a bigger image so I could judge it better, rather than going to the commons page. Vicarage (talk) 18:50, 23 January 2023 (UTC)

meta:Community Wishlist Survey 2023/Wikidata/Popup to link to or create a new Wikidata object after creating an article

Please see

Thanks! M2k~dewiki (talk) 18:54, 23 January 2023 (UTC)

Pubblication date

Hello to community!

In the Wikipedia pages, for call software version identifier (P348), is used {{#property:P348}}. What I should write, to call publication date (P577)?

Many thanks in advance!!! 2001:B07:6442:8903:4C3B:695B:B4BB:4B7C 11:35, 10 January 2023 (UTC)

Have you tried {{#property:P577}}? Vojtěch Dostál (talk) 18:18, 10 January 2023 (UTC)
@Vojtěch Dostál: yes, but do not work since publication date (P577) is a qualifier of software version identifier (P348), not a statement. --2001:B07:6442:8903:E58D:7411:B1F3:CFA9 08:52, 11 January 2023 (UTC)
You'll probably have to use a Lua module for that. Which Wikipedia language version are you working on? Vojtěch Dostál (talk) 11:51, 11 January 2023 (UTC)
@Vojtěch Dostál: japanese --93.34.224.108 14:08, 11 January 2023 (UTC)
Best to ask on Japanese Wikipedia, then. It is difficult for me to understand the templates in Japanese Wikipedia due to the language and script barrier. Sorry. Vojtěch Dostál (talk) 16:14, 16 January 2023 (UTC)
You can use Lua to get qualifiers. Here's the documentation page: [17]. D6194c-1cc (talk) 21:04, 23 January 2023 (UTC)
It looks like you have simpler option, use ja:Template:Wikidata. Infrastruktur (talk) 06:08, 24 January 2023 (UTC)

Request for assistance from anyone who speaks Czech and/or Polish

I'm trying to merge catechist (Q1735732) and catechist (Q11737267), because I believe they are both the same concept. I can't, because of conflicting Czech and Polish language links – cs:Katecheta vs cs:Katechista and pl:Katecheta vs pl:Katechista. Now, from what I understand, kaecheta in both Czech and Polish actually means catechist, so belongs as a link to the merged article; however, katechista in both means catechesis, and hence belongs on catechesis (Q1735729) instead. The problem is, the later is already linked to cs:Katecheze and pl:Katecheza. So, here's what I can't work out – in English, we have a distinction between catechesis (the field of Christian religious education) and catechist (a person who practices in that field). Whereas, in Czech and Polish, we seem to have not two words, but three: katecheta, katechista, and katecheze/katecheza – and I can't understand what the distinction between the three is. I take it, like English, one must be the field and the other a person who practices it, but what then is the third? Can anyone who understands Czech and/or Polish help here? SomethingForDeletion (talk) 21:59, 16 January 2023 (UTC)

In Czech (and probably also in Polish) catechist (Q11737267) is brader term and catechist (Q1735732) is only pedagogue. Cite from czech article: Technically, there is a difference between a catechist (Q1735732) in Czech - the one who prepares and leads catechesis, and a catechist (Q11737267) - the one who is legally authorized to grant baptism. JAn Dudík (talk) 21:01, 17 January 2023 (UTC)
Thanks User:JAn Dudík. But how to represent this in English–or most other languages–where the distinction does not exist? I am thinking the broader term (Q11737267 you say) is likely the equivalent in most other languages, and Q1735732 is some narrower concept which may only exist in Czech or Polish. What do you think? SomethingForDeletion (talk) 08:33, 24 January 2023 (UTC)

Invalid CSRF token

Dear devs,

after having troubles and needing 10x tries to get logged in I'm getting invalid CSRF token on many actions.

How long does it take until all my zombie sessions have timed out on the server? Zachary Zulock (talk) 21:03, 24 January 2023 (UTC)

there's an outage https://phabricator.wikimedia.org/T327815 BrokenSegue (talk) 21:09, 24 January 2023 (UTC)

Adding/Removing leading zeros for URL formatter

Hello, regarding this discussion: Topic:Xbcy58fywxh6shvz / Property:P381

Is it possible to add or remove leading zeros for parameters using the URL formatter?

For example, instead of

using

without leading zeros? (while keeping the same ID in the related object d:Q19362321)

Thanks a lot! M2k~dewiki (talk) 14:44, 25 January 2023 (UTC)

@M2k~dewiki: We can do some custom things with rewriting ID's for formatters with this toolforge service but it's probably better to follow whatever specific ID syntax the source website uses in cases like this... I'm not entirely clear on what you're asking for actually! ArthurPSmith (talk) 20:14, 25 January 2023 (UTC)
Hello @ArthurPSmith: what I am looking for, would be something like
for the URL formatter in Property:P381 (maybe some JavaScript could/should be used?)
On the other hand, I might be possible to remove the leading zeros in the script
M2k~dewiki (talk) 20:21, 25 January 2023 (UTC)
Oh, I see it's a toolforge script, yes it would be best to handle it there I think. ArthurPSmith (talk) 20:24, 25 January 2023 (UTC)
https://phabricator.wikimedia.org/T327956 M2k~dewiki (talk) 20:38, 25 January 2023 (UTC)

Using different URLs for the URL formatter, depending on a property?

Hello, regarding this discussion: Topic:Xbcy58fywxh6shvz / Property:P381

Property:P381 links to the Heritage-Tool-DB

For class/category A objects there would also be a official URL available:

The official URL is not available for class/category B objects:

Is it possible to use different URLs for the URL formatter, depending on a property? For example, depending on Property:P1435:

  • value Schweizer Kulturgut von nationaler Bedeutung -> official URL
  • value Schweizer Kulturgut von regionaler Bedeutung --> Heritage-Tool-DB

Thanks a lot! M2k~dewiki (talk) 14:45, 25 January 2023 (UTC)

The first version of the new Wikibase REST API is now available on Wikidata

Hi everyone,

We are working on making it easier to use Wikidata's data to build applications. A big part of that is a better API. Over the past months we have been developing the first version and gotten testing and feedback for it. Last week we made it available on test.wikidata.org. Today the first version of the new Wikibase REST API is available on Wikidata. I'd like to thank everyone who helped along the way by providing feedback and testing.

You can find more information on what you can already do with the new Wikibase REST API at Wikidata:REST API.

We will continue to build out its functionality and your input on what would be most useful for you will help prioritize the next steps. Please leave your thoughts on the feedback page.


Cheers Lydia Pintscher (WMDE) (talk) 15:29, 25 January 2023 (UTC)

Notability of the author of a Wikipedia source

Hi there,

Is the author of a Wikipedia source considered notable for Wikidata? This point is not clear for me.

Thanks in advance! 92.184.117.74 07:39, 24 January 2023 (UTC)

See Wikidata:Notability, especially Criteria 2 and 3. If by Wikipedia source you mean the author of a publication (book or journal article), then both the publication and the author would likely be notable: many authors can be clearly identified and described in catalogs like VIAF or ORCID, and a vast number of items for scientific journal articles and authors have already been mass imported (although the articles might not all be linked to their respective authors). If you mean the Wikipedian who created a Wikipedia article, then they are probably not notable enough for a Wikidata item. -Animalparty (talk) 08:36, 24 January 2023 (UTC)
Wikipedia uses a lot of blogs for sources. Blog authors mostly wouldn't be notable. --Christian140 (talk) 10:00, 24 January 2023 (UTC)
@Animalparty, Christian140: thank you! But what about such a source (which is not a scholarly article nor a blog: see CTHS person ID (P2383))? 92.184.107.173 11:06, 24 January 2023 (UTC)
Whether or not something is a blog isn't really the central criteria. We don't want that everyone who has a blog adds themselves for self-promotion purposes to Wikidata and thus don't treat that as being enough. The key question is whether there's a strutural need. In a case like your need I likely wouldn't create an item if there's just a single source for a certain person. If a person is however used frequently as a source there's a structural need to store that information in Wikidata. ChristianKl ❪✉❫ 15:18, 24 January 2023 (UTC)
In this specific case, the linked person has a French Wikipedia article, and existence of a Wikipedia article is automatic grounds for notability here. The entry can be found at Monique Kuntz (Q55758958). ChristianKl's explanation remains valid for similar cases lacking other grounds for notability. From Hill To Shore (talk) 09:00, 25 January 2023 (UTC)
@From Hill To Shore: I wasn't talking about Kuntz, but about the author of the entry dedicated to her on the CTHS directory.
@ChristianKl: he also authored other entries, such as this or this one.
92.184.105.6 14:46, 26 January 2023 (UTC)

Redirect

I need to redirect Q116444363 to Q14801279 because I didnt know the page already existed so I created a double page. But I don't know how to make the redirect. Thanks Coldbolt (talk) 09:25, 26 January 2023 (UTC)

  Done RVA2869 (talk) 13:08, 26 January 2023 (UTC)
@Coldbolt: Check out Help:Merge#Gadget. In short, a redirect is created when one item is merged with another. The merge gadget is the best tool for the job. --Tagishsimon (talk) 13:34, 26 January 2023 (UTC)

An ontology for a university faculty department

An ontology for a university faculty department might include classes such as "Department," "Faculty Member," "Course," and "Student." The Department class would have properties like "name" and "chairperson," and would be connected to the Faculty Member class through a property like "hasFacultyMember." The Faculty Member class would have properties like "name," "rank," and "office location," and would be connected to the Course class through a property like "teaches." The Course class would have properties like "name," "code," and "description," and would be connected to the Student class through a property like "enrolledIn." The Student class would have properties like "name," "studentID," and "major." Additionally, the ontology might include object properties such as "advisedBy" to connect a student to a faculty member, or "prerequisite" to connect one course to another. This ontology would allow for easy organization and querying of information about the department's faculty, courses, and students. Maishamidi (talk) 12:31, 26 January 2023 (UTC)

This certainly shouldn't be on wikidata. No reason to have items for classes or enrolled students here. Not notable for wikidata. --Christian140 (talk) 13:28, 26 January 2023 (UTC)
Unclear why @Maishamidi: is posting their ontological thoughts here. I've just deleted the Movies thread. I get the impression the contributions do not actually relate to WD; maybe to a course they're taking. Anyway: wrong forum, Maishamidi, I'm afraid. --Tagishsimon (talk) 13:36, 26 January 2023 (UTC)

coordinates from Google

Do we have any property for coordinates from Google? Do we have any plan to take coordinates from Google (perhaps Google Maps)? Geagea (talk) 09:38, 19 January 2023 (UTC)

No, and no. Such a thing as their copyright in their work & all. --Tagishsimon (talk) 11:35, 19 January 2023 (UTC)
Are coordinates copyrightable? Feels like they'd be some subset of PD-ineligible. DS (talk) 14:54, 19 January 2023 (UTC)
If they're just coordinates, WD has P625. If they're coordinates from google - i.e. the coordinate that google cites for an identifiable entity - then imo copyright exists in the arrangement of the entity with the specific coordinate. --Tagishsimon (talk) 18:19, 19 January 2023 (UTC)
The coordinates almost certainly aren't covered by copyright, however Google does have database rights over the entire collection of coordinates. Extracting coordinates from Google Maps would therefore likely be an infringement on these rights, at least in jurisdictions which recognise database rights either because they form part of domestic law or by mutual recognition of foreign "intellectual property" laws. meta:Wikilegal/Database Rights discusses this. --M2Ys4U (talk) 18:28, 19 January 2023 (UTC)
M2Ys4U thanks for the link. So, the database is protected as a compilation "the author made certain choices". The choice considered to be "creative expression", so it's part of copyright.
A. Google have coordinates for every location. Could this considered to be an "author made certain choice"?
B. Is there any database that is in the public domain?
Geagea (talk) 11:16, 20 January 2023 (UTC)
@Geagea I am not a lawyer but I don't think Google maps would be protected "because the author made creative choices" but because there are database rights attached to a mass use of any compilation, both in the US and EU. As for B, the closest thing to that would be any CC-0 database such as Wikidata. Vojtěch Dostál (talk) 11:56, 20 January 2023 (UTC)
@Vojtěch Dostál: Since when database rights exist in USA? But yes, despite CC0 usage of Wikidata in UK or EU has serious limitations due to basically ignoring database rights Mateusz Konieczny (talk) 13:04, 20 January 2023 (UTC)
@Mateusz Konieczny It's true that my layman knowledge of database rights is formed by the EU experience so I will not speculate any further on Google maps. Vojtěch Dostál (talk) 13:23, 20 January 2023 (UTC)
My understanding is that someone in USA can ignore database rights (sui generis database rights do not exist) - but such product may be illegal to use in EU or have severe limitations on its use (where database rights exist). This applies also to Wikidata, and its claimed CC0 license is quite misleading. Mateusz Konieczny (talk) 13:27, 20 January 2023 (UTC)
@Geagea: "Do we have any property for coordinates from Google?" - yes, many were indirectly harvested when copying Wikipedia data, where many coordinates were added by users copying them from Google Maps and asimilar sources. English Wikipedia openly encourages to copy coordinate data from Google Maps/OpenStreetMap/etc. which is perfectly legal in USA but database created from that is likely not legally usable in EU (warning: I am not a lawyer) Mateusz Konieczny (talk) 13:27, 20 January 2023 (UTC)
Usually, when a user manually ads a coordinate from Google Maps or OpenStreetMap they will chose the exact spot themselves. For cities that will usually mean that they are not choosing exactly the same point as Google Maps or OpenStreetMap. That's unprobleatic. It gets more complicated when you copy the exact location down to the meter and that choice is nontrivial (it's not simple the center). As far as I remember the question whether or not that's allowed is unsetteled. ChristianKl ❪✉❫ 01:52, 22 January 2023 (UTC)
@ChristianKl: people often copied this info semi-automatically or automatically, without choosing own spot. Also even if they selected exact spot in OpenStreetMap then they did it based on OpenStreetMap data only. And yes, that is likely not fully settled or not settled at all how much you can copy before database rights starts to apply (I am not a lawyer, none of my posts in this thread is an official opinion) Mateusz Konieczny (talk) 10:42, 22 January 2023 (UTC)
Thanks to all. Geagea (talk) 21:02, 26 January 2023 (UTC)

Searches involving "category:"

Anyone able to give some guidance on the best way to undertake category: searching for items? If I use search box, it works for type ahead, though actually doing a search does not bring back any results (for me). Doing a search without the word category, or the namespace is just not helpful with the results that I am getting. Working examples would be great. Thanks.  — billinghurst sDrewth 01:29, 27 January 2023 (UTC)

Add haswbstatement:P31=Q4167836 to your search query (example). It will break autocompletion, but search will work. Lockal (talk) 03:14, 27 January 2023 (UTC)

Is there a property (or best practice) for magazine archives?

Many magazines, newsletters, and other periodicals have curated, digitized archives of back issues. Sometimes back issues can be found in digital repositories like HathiTrust (Q3128305) or Internet Archive (Q461) or Issuu (Q1982075) (in which case HathiTrust ID (P1844) or Internet Archive ID (P724) or Issuu ID (P9921) can be used, respectively). But sometimes collected back issues may be found on subpages of official websites (if the publication is still extant), and/or on other third party websites: in the absence of dedicated properties, is there a generic property for essentially "back issues can be found here"? The properties full work available at URL (P953) or archives at (P485) seem a bit imprecise or inappropriate, and archive URL (P1065) seems intended for snapshots of individual URLs, not repositories. We do have newspaper archive URL (P7213), which is tailored to newspapers: should the scope of that be expanded to include all periodicals? Before I realized Internet Archive ID (P724) could include periodical collections (e.g. Life magazine) I would use archives at (P485)Internet Archive (Q461)URL (P2699)example.com or described at URL (P973)example.com, both of which seem redundant and/or suboptimal. Any tips for how best to link to back issues when no appropriate external identifiers exist? -Animalparty (talk) 03:17, 27 January 2023 (UTC)

Is ToolForge having problems or technical difficulties? Eugen Rochko (Q64876086) data Toolforge bypass

I had opened a discussion topic for Mastodon address (P4033) with proof of data not always going through Toolforge, ie. George Takei (Q110154) data(related to his Mastodon address (P4033)) does not go through Toolforge and resolves a 404 error when linking to his Mastodon address.

Eugen Rochko (Q64876086) the founder of Mastodon itself's Mastodon Address does not resolve on Wikidata. :) this is just too puzzling for me. I found this edit where the rank for "Formatter URL"/"URL Formatter" had been changed from preferred to normal in this edit for the property of Mastodon address (P4033).

Does changing "Formatter URL" rank priority somehow affect if some data of Mastodon address (P4033) will or will not go through Toolforge or does Toolforge have technical difficulties because there is also a very hard puzzle to crack where Taylor Lorenz (Q89135464) Mastodon Address resolves fine through Toolforge without a problem! This makes no sense to all. If an expert can explain to me what's going on I think I'm gonna learn something new today! Mastodeas (talk) 16:24, 27 January 2023 (UTC)

yes the formatter rank decides which formatter is used. it's probably mostly a caching issue and the priority. BrokenSegue (talk) 17:03, 27 January 2023 (UTC)

Fixing incorrect merge of publishing concepts

Previously we had the concept of "publication in German copyright law" under publication in German copyright law (Q15852766) and "publishing - process of production and dissemination of literature, music, or information" publishing (Q3972943). The former appears to be about a specific concept in German law for the very moment of publication. The latter was about the broader concept of publishing as it applies to the whole world. These were then merged on 6 November 2022 by User:DerMaxdorfer.[18] We are now left with an item that says the concept of publishing for the entire world is based solely on German law. This is clearly a mistake but I'd like a second opinion on how to fix it. Using the principles of managing conflations, if this was a recent merge we would reverse it and if it is a longer term merge we would have to manually split the concepts into two new items and deleted the conflated item. Do we think this is recent enough to just reverse the merge? From Hill To Shore (talk) 23:17, 22 January 2023 (UTC)

Based on this diff, I would restore the pre-merge versions of both items and then re-apply the handful of recent edits. PKM (talk) 00:41, 23 January 2023 (UTC)
I merged the two items because of the chaos in the linking of Wikipedia articles in different language versions and I did my best to connect always those articles that describe the same topic. To be honest, I didn't care enough about the statements already existing in the different Wikidata items, otherways I would have deleted some the statements dealing exclusively with the German copyright law. In my opinion the German term doesn't need an own item. The description as "publication in German copyright law" in publication in German copyright law (Q15852766) is a mistranslation/misinterpretation of the original German description added here. Originally, this German definition translates as "in the German copyright law: moment, in which a work is made available to the public" and should probably only make aware of the fact that the word could have slightly different definitions in other languages even if it means the same thing.
To make it more clear: There is no specific concept called "publication" in German copyright law. The German copyright law (§ 6.1) only has its own short paragraph defining from which day to count the copyright rules we all know from Commons or Wikisource. That one sentence translates as: "A work is (= counts as) published, when it is made available for the public with consent of the authorised entity." In my eyes this sentence doesn't create something different from the one thing that is internationally called "publication" – but that can of course be seen differently. Therefore I have two possible solutions:
1.) Leave everything as it is.
2.) Re-create the item about the defined instant of time at which a work of art, literature etc. is counted as published in the German copyright law. But that item can only refer to the already cited §6.1 UrhG, not to any Japanese or Dutch terms as previously mentioned in the Wikidata item.
Please let me know which solution you choose so that I can at least arrange the German Wikipedia articles according to the new arrangement of the Wikidata items. --DerMaxdorfer (talk) 06:17, 23 January 2023 (UTC)
@DerMaxdorfer: I've split the two items into publishing (Q3972943) and publication in German copyright law (Q15852766). To perform the split I had to move the dewiki link to publication in German copyright law (Q15852766). If you think it sits better under publishing (Q3972943) then you are free to move it, without merging. While merging may be the simplest way for you to get your site links in the right place, you need to consider carefully the concepts of the two Wikidata items first and decide if you are going to cause damage. In this case you transformed an item about the global concept of publishing into a German concept of publishing. When in doubt, ask at the Project Chat, especially when you are dealing with fundamental concepts linked from thousands of other Wikidata items. If you think publication in German copyright law (Q15852766) is not needed, you are welcome to request deletion of that item (It has no references and the only ID is a Google Knowledge Graph ID). From Hill To Shore (talk) 11:11, 28 January 2023 (UTC)
I know, I know. But most national copyright laws will have some short regulations concerning the publishing of a book. That's why I found it strange to handle the German law as a special case needing its own Wikidata item. However, that is not important enough for me to discuss it that painstaikingly, especially as I assume that you know the Wikidata rules better than I do... Thanks for your work and best regards, DerMaxdorfer (talk) 11:24, 28 January 2023 (UTC)

Linked Data Fragments predicate for labels, aliases, sitelinks?

I'm using the LDF endpoint (https://query.wikidata.org/bigdata/ldf) for querying backlinks, but I cannot find any references for how to query labels, aliases & sitelinks through LDF. Is it even possible? I would like to avoid SPARQL. The manual (https://www.mediawiki.org/wiki/Special:MyLanguage/Wikidata_Query_Service/User_Manual#Linked_Data_Fragments_endpoint) doesn't mention it. Whatever the answer, it would be great if the manual was updated to reflect it. ProbabilityCollapse (talk) 20:14, 26 January 2023 (UTC)

The predicate for label is rdfs:label, for description schema:description and for alias, skos:altLabel ... not sure if that helps or not. See https://www.mediawiki.org/wiki/Wikibase/Indexing/RDF_Dump_Format#Entity_representation --Tagishsimon (talk) 20:19, 26 January 2023 (UTC)
Thanks for the reply. It does certainly help with labels, descriptions and aliases. Still unsure about sitelinks, which was my main use case. There's an http://wikiba.se/ontology#sitelinks property, but that only seems to indicate the amount of sitelinks, not their actual values. ProbabilityCollapse (talk) 20:47, 26 January 2023 (UTC)
@ProbabilityCollapse: So here the ?article (URI of a wikipedia/wikimedia page) is the subject of a triple; predicate is schema:about and object is ?item. These are the three triples typically employed when dealing with sitelinks ... ?sitelink is the article title:
In more detail: https://www.mediawiki.org/wiki/Wikibase/Indexing/RDF_Dump_Format#Sitelinks --Tagishsimon (talk) 23:51, 26 January 2023 (UTC)
Thank you! Took a bit of playing around with the queries, but I think I get it now: Sitelinks are stored as subject=wikipedia-page and object=wikidata-entity. ProbabilityCollapse (talk) 09:46, 28 January 2023 (UTC)
Excellent; yes, you have it. --Tagishsimon (talk) 14:24, 28 January 2023 (UTC)

Time capsules

Hello! What are the best properties for date of creation, date to be opened and content/objects included in time capsules? Example: Philco time capsule (Q116447516). Thanks. Emijrp (talk) 21:04, 27 January 2023 (UTC)

Maybe significant event (P793) with an appropriate value and a point in time (P585) qualifier for the dates. contains (P4330) for the contents? --Tagishsimon (talk) 21:20, 27 January 2023 (UTC)
Nice, I created date of enclosure (Q116457511) and date of opening (Q116457521). Example. Emijrp (talk) 09:38, 28 January 2023 (UTC)
Good work, Emijrp; but I feel the organisers may have bought too strongly into the notion of the everlasting qualities of LaserDisc (Q273309). "Its content was created to last 500 years" versus w:en:LaserDisc#Laser_rot :(. --Tagishsimon (talk) 14:23, 28 January 2023 (UTC)
Yeah, maybe... Fortunately there are hundreds of different time capsules all over the world. I am using the same properties for messages in bottles too, like Message in a bottle by James Ritchie and John Grieve (Q116458197). Emijrp (talk) 15:05, 28 January 2023 (UTC)

Is no quorum valid reason for closing property proposals?

@DannyS712: Closed more than 50 property proposals as "no support for creation of this property". Some of them are reopened by either the proposer or other users. I do not want to argue those with still no consensus to create with reasonable amount of input (including those with valid objection); but for those received no comments with no one opposing, a property proposal at least indicates a need for a property for someone (the proposer), so I think a better solution is to accept them by default after one month without objection. Having a property is seldom positively harmful, especially identifier properties.--GZWDer (talk) 04:56, 28 January 2023 (UTC)

I think properties need community support, and while a external identifier might be harmless, we need a strong case for ontological ones, lest we get overlapping or too niche ones. One person can rarely see the implications of their idea. Vicarage (talk) 05:22, 28 January 2023 (UTC)
I strongly agree. General properties should not be created willy-nilly or else Wikidata becomes a unsustainable mess. I believe it's the property creator's job to think of reasons against the proposal and have them addressed. This is the reason the property creator role exists in the first place.
I think identifiers are more of a judgement call, as long as there are no unanswered objections, there are over ~100 entries and the creator believes it's useful you could still create it, as the proposer and creator have implicit votes too. I also think it's better to eventually close proposals rather than having them open indefinitely, Speaking of which, when do property creators usually decide to close a proposal? Infrastruktur (talk) 08:57, 28 January 2023 (UTC)
Many requests above receive no comments at all, let alone objections. For identifiers, I tend to accept them after one month. GZWDer (talk) 11:12, 28 January 2023 (UTC)
Most property proposals today are identifiers (as are most of closed one I concern). For those, the standard may be more lenient as redundancy is hardly a concern (though there are still concern like copyright, stability or spam). For properties in other datatypes (especially item), it may be beneficial to leave the request a bit longer to solicit enough inputs.--GZWDer (talk) 05:39, 28 January 2023 (UTC)
If there's been months with no comment then I don't know if we should create the property. I personally would leave them to sit but this does create a big backlog. I don't think the burden of relisting is that big a deal. That said if there's support (even minimal) I don't think we should close it as declined. Going over most of the closed proposals I'm mostly fine with them. Closing property requests is a thankless task so I don't want to even appear to be snapping at someone for doing a good deed. BrokenSegue (talk) 05:58, 28 January 2023 (UTC)
I'm not fine with closure of many requests like this one. (Many of) such stale proposals frustrate users who want to propose (and use) new properties.--GZWDer (talk) 14:33, 28 January 2023 (UTC)
again I agree that I wouldn't do this. but if your proposal doesn't get any comments in a couple months and you don't solicit comment from e.g. here then it doesn't seem super productive to leave it up. Maybe we should just decide that, say, 3 months is the point where if you don't have any comments it's considered dead. BrokenSegue (talk) 17:53, 28 January 2023 (UTC)
DannyS712 closed some proposals with that reason which were open for less than a month. I don't think lack of support is a valid reason for closing properties at that stage. I think it can make sense for property proposal older than six months.
I don't think there should be automatic approval for ID properties. ChristianKl ❪✉❫ 14:08, 28 January 2023 (UTC)
Currently property proposals are already so many that proposal pages hits template limit and can not transclude all proposals. (And this make relisting useless.) Many proposals are ill attended (reviewed by few users), and this make them worse.--GZWDer (talk) 14:26, 28 January 2023 (UTC)
Most property proposals are made with minimum effort. If someone doesn't get any attention for their external ID property, that's often a sign that the person making the proposal put in little effort to make the proposal accessible.
It's the role of the person making the proposal to explain why it's valuable to have that property and provide the relevant information to understand why it's important. If a property proposal doesn't do that for any reader of it that's a sign that it's lower quality than the average property we create. ChristianKl ❪✉❫ 15:40, 28 January 2023 (UTC)

Junk items. last report.

even though i dont give a fuck about junk on wikidata, i'll report this one last time. (previous reports: Wikidata:Project_chat/Archive/2022/12#IP_user_creating_lots_of_items_for_commons-only_cats Wikidata:Administrators'_noticeboard/Archive/2023/01#Report_concerning_User:G6zLZz2cEPKdEXB.)

these users, and possibly other accounts/ip that i didnt find out, have created hundreds of junk items ( https://xtools.wmflabs.org/pages/www.wikidata.org/G6zLZz2cEPKdEXB https://xtools.wmflabs.org/pages/www.wikidata.org/218.102.0.123 ) that link to granular commons cats that dont need an item, e.g. Q116298299 Q116298452. RZuo (talk) 12:41, 24 January 2023 (UTC)

  all deleted. Thanks for notifying @RZuo! Let we know if you discover another IPs/users. Estopedist1 (talk) 18:59, 24 January 2023 (UTC)
@Estopedist1: We have the same behaviour from the user at User talk:218.102.0.44 with large scale creation of bare items with only a Commons category site link. The pattern of edits (and similar ip) appears to indicate it is the same editor. From Hill To Shore (talk) 04:23, 25 January 2023 (UTC)
I spoke with them as shown in that talk. They were definitely operating a bot without approval but it seems they now understand not to do that. I'll go through and nuke the old items. BrokenSegue (talk) 05:42, 25 January 2023 (UTC)
also User:42.200.150.136 there's probably more too. BrokenSegue (talk) 06:05, 25 January 2023 (UTC)
@Estopedist1, From Hill To Shore, BrokenSegue:
another one here: Special:Contributions/219.78.39.49.
since on User talk:218.102.0.44 s/he claimed s/he would stop on 13 January, but 219.78.39.49 is obviously the same person doing the same thing on 25 jan, i suggest an immediate block from now on is proper. RZuo (talk) 08:30, 28 January 2023 (UTC)
also, c:Special:Contributions/Yrellag is the same user. RZuo (talk) 09:37, 28 January 2023 (UTC)
@RZuo wow, this is exceptionally productive user. Some items are good ones also, e.g. Q115954976 and Q115954956. Due to editing after warning, range block is set for 218.102.0.0/25 Estopedist1 (talk) 15:47, 28 January 2023 (UTC)
@Estopedist1: a handful of them are alright, but you have to make too much effort in order to sieve them out.
since s/he has moved on to new ip, maybe 219.78.39.49/25 should be also blocked? RZuo (talk) 15:59, 28 January 2023 (UTC)
if you make so much garbage you can't complain if we accidentally delete some good pages. I'm going to delete them all. BrokenSegue (talk) 17:23, 28 January 2023 (UTC)
Special:Contributions/182.239.87.228 3 more here. XD
bus route Q115972959 is ok. RZuo (talk) 19:02, 28 January 2023 (UTC)
also, i think, to prevent these kinds of spam happening, maybe a bot should be set up to detect "newly registered users / ip creating more than x items in y days"? like more than 20 items in 2 days? having a bot-generated report and spot checking it might be useful in detecting such spam. RZuo (talk) 16:02, 28 January 2023 (UTC)
@RZuo definitely a good idea, but our goal probably should be to restrict items creations and items merging by anonyms. E.g. in Commons, anonyms cannot upload files Estopedist1 (talk) 07:09, 29 January 2023 (UTC)

image

How will i add an image to an item Bright0708 (talk) 07:17, 29 January 2023 (UTC)

Bright0708 An appropriately licensed free image -- either in the public domain (no copyright whatsoever) or with a free license (such as CC-BY) -- needs to first be present on Wikimedia Commons. Some more info is at Wikidata:Tours/Images. Image or no image, strive to ensure that the item satisfies one or more of Wikidata's notability criteria. -Animalparty (talk) 07:34, 29 January 2023 (UTC)

AI authorship

If the subject of an item was generated by ChatGPT, Midjourney, Stable Diffusion or similar should the AI program be listed as the author? Or the human who entered the prompts? Trade (talk) 18:49, 29 January 2023 (UTC)

maybe both with a qualifier? though pretty sure that violates a constraint BrokenSegue (talk) 19:11, 29 January 2023 (UTC)
Rather than author (P50) I think fabrication method (P2079) might be more applicable in these cases. Ainali (talk) 19:18, 29 January 2023 (UTC)
What would the value of fabrication method (P2079) be? The item of the AI program used or something specific? Trade (talk) 21:47, 29 January 2023 (UTC)
What if an AI enters its own prompt into another AI? -wd-Ryan (Talk/Edits) 20:32, 29 January 2023 (UTC)
Why should we have such items, references... ? --Succu (talk) 20:54, 29 January 2023 (UTC)
then list both in the fabrication method? BrokenSegue (talk) 20:54, 29 January 2023 (UTC)

Two identical items, but "impossible" to merge

Q600310 and Q22064645 are about the same city in Brazil, Mogi-Mirim. However, it's "impossible" to merge them, because Q22064645 has two articles linked to it, in the Ceb and in the SV Wikipedias. The main problem is Q600310 also has articles for the same city in the Ceb and in the SV Wikipedia. As far as I can tell, these two Wikipedias have two different articles for the same city. The two articles in each of these two Wikipedias were created by a bot, Lsjbot. What we do? Merge the article in these WPs and then the items here or what? Kacamata (talk) 01:25, 29 January 2023 (UTC)

These two items look like they're about different - but related - things. Often there are separate items for a settlement and a municipality that covers the settlement. I don't know enough about Brazilian local government structure or the ceb and sv wikis' policies on these matters, but my first reaction would be that the items should not be merged but that the items should link to each other, probably by using located in the administrative territorial entity (P131) on the settlement pointing to the municipality. For a similar example from the UK, see Manchester (Q18125) (city) and Manchester (Q21525592) (district) --M2Ys4U (talk) 01:42, 29 January 2023 (UTC)
(EC)In this case, cebwiki is drawing a distinction between a city of 78k people - https://ceb.wikipedia.org/wiki/Mogi_Mirim_(lungsod_sa_Brasil) - and a municipality of 86k people - https://ceb.wikipedia.org/wiki/Mogi-Mirim . Cebwiki's city article is linked to Q600310, which for WD is the municipality item. Cenwiki's municipality article is linked to Q22064645 which for WD is a human settlement item.
So at the least, the sitelinks for Ceb (and probably SV) should be swapped around.
It's unclear how the Ceb and SV articles were put together, but on the face of it a source (reliable?) draws the city / municipality distinction. So the choices seem to be, fix the WD items (swap the sitelinks, tweak the P131 on the city item). Or redirect the Ceb and SV city articles into the Ceb and SV municipality articles, and merge on WD. There has been a very great deal of pruning of duplicate-ish articles on (iirc) SV, fwiw. Equally, there is some distinct content in the two Ceb articles which ideally would be properly merged, not lost.
tbh, sorting it out at the WD end looks to me like the simpler option. --Tagishsimon (talk) 01:53, 29 January 2023 (UTC)
Those items come from a bot that used the geonames database. See https://www.wikidata.org/wiki/Wikidata:WikiProject_Territorial_Entities/Geonames_and_CebWiki for more information. ChristianKl ❪✉❫ 03:39, 29 January 2023 (UTC)

natural locality vs administrative division

In this section I want to point out that natural localities, i.e. human settlement (such as city) and administrative divisions are two distinct things. For example:

Comparison of city and administrative division
Natural city Administrative division
Tokyo (Q7473516) - In Chinese Wikipedia, a (board-concept) article describing the natural city of Tokyo, settled since (at least) 1457 Tokyo (Q1490) - Tokyo Metropolis, found 1943
Paris (Q90) - a city populated since BC Departments of Paris (1968-2018), then a territorial collectivity
New York City (Q60) - settled 1624 City of Greater New York (Q5123708) - since 1898
Beijing - a city settled before 1000 BC Beijing (Q956) - direct-administrated municipality since 1949; In Chinese Wikipedia, the article is explicitly named after the administrative division
Lagos (Q8673) - a natural city Lagos city government dissolved in 1967. Currently there are no individual administrative division of any sort for Lagos.

They are distinct in many ways:

  1. Administrative divisions may be created or dissolved, but it make no effect on the actual city (unless it is abandoned). Obviously an administrative division (if exist) may be found much later than the actual city.
  2. The border of an administrative division is usually well defined, but the border of a city or town can have multiple definitions (contiguous urban area, metropolitan area, both of them are also not well defined).
  3. A locality can only be a small part of administrative division ("The municipality of Chongqing, China, whose administrative area is around the size of Austria, has the largest population for a city proper. However, more than 70% of its residents live in rural areas."), but can also extended beyond the administrative division.

However, In most databases or sources (but not GeoNames), we does not differ these two. They should ideally be split, but splitting (most of) each city to two items will strongly confuse data users.--GZWDer (talk) 22:46, 29 January 2023 (UTC)

You've made this part of an existing topic, I think you wanted it to be a new topic. This area is a can of worms, as people think in terms of the natural names and boundaries of places, not the administrative boundaries (and clumsy names to combine 2 names in one organisation). But if you care about political hieracy, then you need items that make their purpose clear, so Orpington (Q123977) where I live was part of Bromley Rural District (Q4973696) until 1935, then Orpington Urban District (Q15264364), part of Kent (Q67479626) (but not Kent (Q23298)) until 1965, now part of London Borough of Bromley (Q208201), in Greater London (Q23306), London (Q23939248), which is not London (Q84). Clear as mud, but there is no way we want each settlement to record its political history, boundaries are fluid, and we don't record them anyway, so the only thing really worth recording is the current state of affairs, and then accept that people will say my house is in Orpington, not the Farnborough and Crofton (Q56295126) electoral ward which is the true bottom of the tree. I think the prospects of making a plan in this area, and then getting people to follow it are very remote. Vicarage (talk) 00:10, 30 January 2023 (UTC)
All Japanese city names in Japanese are modern administrative division names, so P571 is set after 1888. This often results in "Contemporary" violations of historical figures. Therefore, it seems necessary to have two entries, one for modern place names and one for historical place names. In English, both names would be the same. Afaz (talk) 02:20, 30 January 2023 (UTC)
Changing names are hard to handle, how would you say the Prince Regent visited Constantinople? When we say X is a Y, there is an implicit assumption that the author meant it for a range of dates, with the names as valid in the range. To extract the information we really need a query of the form "tell me about X's relation with Y at date Z", but not by coding dates in the relation, as some have by quoting country changes for border towns, but by extracting them from X and Y. This is a problem for ships, which could be solved by using official name (P1448) and dates, but coding the information and getting good queries is a big burden on the user. A possible solution would be for Wikibase to have date qualifiers for labels, and SPARQL to provide ?itemLabel(1925), but this would be a lot of work. And continuity is important, some argue a ship with a new name or purpose is a new ship, I'd argue that tracing the history of the hull that matters, and the same applies to places. Vicarage (talk) 09:54, 30 January 2023 (UTC)

I would need some help on test.wikidata.org to understand how Mastodon address (P4033) works or be allowed(limited time?) to experiment myself

Do you mind if I experiment some on test.wikidata.org in order to understand how and why Mastodon address (P4033) creates broken Mastodon links? Can you tell me who to get in touch with to ask for permissions for a few weeks to experiment there so that my edits don't just suddenly get wiped now as I'm editing? Who am I to tell if somebody decides that my edits are pointless and decide to wipe everything I've done so far? Though I have a genuine worry that this is way over my head but I won't know that until I've tried to the best of my ability.

Do I need to apply that I'm doing this as a 'little project' anywhere? test.wikidata.org does not have a project chat or community portal like Wikidata. So far I've been allowed to create new items and new properties so there is no real technical measure that is stopping me from experimenting.

Perhaps I don't need to ask anyone for permission and I can just experiment? I'm trying to copy how Mastodon address (P4033) works on test.wikidata.org and I'm saying this so that you can decide for yourselves if this is futile, if this is not what test.wikidata.org was designed for or some other insight I don't know about. That will probably save me a lot of time. Mastodeas (talk) 00:39, 30 January 2023 (UTC)

you don't need permission to test on test.wikidata.org. just play around there. BrokenSegue (talk) 07:09, 30 January 2023 (UTC)

Better explanation about "sitelink to redirect" needed

when adding a link manually, the given link Help:Sitelinks doesnt say anything about the difference between sitelink to redirect and intentional sitelink to redirect. something more accessible is needed. RZuo (talk) 17:06, 27 January 2023 (UTC)

@RZuo: It looks like the help page has been updated since your post here, does this address your concerns? See also Wikidata:Sitelinks to redirects which is linked from Help:Sitelinks and goes into more detail on your question. ArthurPSmith (talk) 17:09, 30 January 2023 (UTC)
thx. before your message i didnt find out Wikidata:Sitelinks to redirects at all. i only found the brief descriptions on items Q70893996 and Q70894304. RZuo (talk) 17:17, 30 January 2023 (UTC)

Rock that disappeared, p31=?

an item is about a big rock at a place. the rock has disappeared because of land reclamation. p31=? RZuo (talk) 09:11, 30 January 2023 (UTC)

To describe things in the past, create all the usual properties, but add dissolved, abolished or demolished date (P576) to indicate it has ceased to exist and a significant event (P793) to describe both the reason and date Vicarage (talk) 09:40, 30 January 2023 (UTC)
most precise date i can find is "19th century". how to input this for Hoi Yan Sek (Q11150367)? RZuo (talk) 13:45, 30 January 2023 (UTC)
You can just write "19th century" in all properties that are about dates. ChristianKl ❪✉❫ 13:50, 30 January 2023 (UTC)
@RZuo I've updated Hoi Yan Sek (Q11150367) to show an example of how to specify date precision of "19th century". Dhx1 (talk) 16:06, 30 January 2023 (UTC)
thx a lot. i wrote 19th century and it said "malformed". wikidata should be more clever to understand this common notation. RZuo (talk) 16:09, 30 January 2023 (UTC)

Wikidata weekly summary #557

 
Here's your quick overview of what has been happening around Wikidata over the last week.
  • Events
    • A citation hunt program was organized for secondary school students at IIS School in Dubai as part of the 74th Republic Day celebration of India. 16 students participated, with different mother tongues and from different states of India living as expatriates in Dubai. The goal was to increase references for Wikidata statements based on research. The students were introduced to the importance of protecting India's history on the internet, and were taught how to edit Wikipedia and Wikidata. They were given a special event page on Wikidata with instructions and a list of Wikidata items to add citations to, and they searched for references in their textbooks and search engines to add to the Wikidata items.More details are given here.
    • Next Linked Data for Libraries LD4 Wikidata Affinity Group Working Hour January 30, 2023: The #1Lib1Ref campaign is afoot, so we will be working on adding references to Wikidata statements. We'll have some data for you to work with, but you're welcome to bring your own. We'll be mainly working with adding webpages as sources, but if you want to get experience using books or articles as sources, we recommend bringing some to the session to use as references for Wikidata items you have identified. You are, as always, welcome to bring your own project to work on Event page
  • Tool of the week
    • theyrule.net is a website that helps to explore the boardroom connections of the largest US companies.
  • Development
    • Wikibase REST API: The first version of the new API is released on Wikidata now \o/ We'd love to hear your feedback on what to improve next in the API.
    • Property Suggester: We updated the data for the Property Suggester so suggestions for new statements to add to an Item should be more in line with current Property usage on other similar Items again.
    • Fixed an issue with Item deletions not showing up on Wikipedia and co if they affect an article on that wiki (phab:T326082)
    • Vector 2022: Made it possible to add a new sitelink on Wikipedia and co in the new Vector 2022 language selector (phab:T310259)
    • Lua: Working on two new Lua functions, getDescriptionByLang (phab:T230839) and getBadges (phab:T305378)

You can see all open tickets related to Wikidata here. If you want to help, you can also have a look at the tasks needing a volunteer.

Mixed mine and village

Drmno (Q3039540) not sure what is going on here Mateusz Konieczny (talk) 10:29, 31 January 2023 (UTC)

A user, Djordje110795, changed its meaning from village to coal mine in 2019. I've rolled it back. --Tagishsimon (talk) 10:40, 31 January 2023 (UTC)
Oh thanks - somehow I got confused and forgot that I should start from history checking Mateusz Konieczny (talk) 11:19, 31 January 2023 (UTC)

AbuseFilter/64 catching lots of cjk edits

briefly looking at Special:AbuseFilter/64, i guess it didnt consider cjk chars? might not be a big issue, but the filter should be improved. RZuo (talk) 11:43, 31 January 2023 (UTC)

Generic property-proposal page is overflowing

We currently have so many open Generic property proposals that not all of them are being transcluded into the overview page (because of some Mediawiki limit, I guess). A few of the proposals seem to be ready or withdrawn, though. What is the mechanism for moving those on/elsewhere? ―BlaueBlüte (talk) 22:38, 30 January 2023 (UTC)

If there are proposal that don't really belong there like authority control properties you can manually copy them over to the place where they belong. ChristianKl ❪✉❫ 23:14, 30 January 2023 (UTC)
  Done Moved some proposals; diffs: [19] [20] [21]
―BlaueBlüte (talk) 19:07, 31 January 2023 (UTC)

Allyshaleonard wrecking havoc (again)

Allyshaleonard (talk • contribs • logs) from MoMA is running OpenRefine again and not avoiding any possible duplicate. She is even creating items from the last run a 2nd time: Q115636727 Q116519325

The discussion from 3rd of January notes, that the doesn't reply to criticism on her talk page. Since then she hasn't answered any of the 4 questions.

The edits have the potential to fill Property talk:P214/Duplicates/humans and therefore better end quickly.-- Sharaban K (talk) 14:13, 31 January 2023 (UTC)

Hello Sharaban K.
I am a current fellow at MoMA and currently having technical difficulties with open refine. Things that are uploaded should not have been since the upload was incomplete.
I am currently working on resolving this issue.
I also see the note to change instances from individual to human and are doing so. My current leader at MoMA is still working to learn OpenRefine together and it has been a learning process. Any help you give is of course very welcome and I appreciate your patience as we resolve this issue Allyshaleonard (talk) 14:28, 31 January 2023 (UTC)
This is my first day back in office after a long break. I apologize for the occurrences and lack of response. Currently working to resolve all issues/comments and appreciate your patience in the meantime Allyshaleonard (talk) 14:32, 31 January 2023 (UTC)
To be honest, these were not even supposed to be uploaded to Wikibase yet as there were "technical difficulties" in OpenRefine, where the upload did not complete - hence the multiple uploads on the backend without knowing. I am currently resolving this issue and will be removing duplicates/merging. Allyshaleonard (talk) 14:37, 31 January 2023 (UTC)
UPDATE: all OpenRefine upload batches have been requested for reverse/deletion.
The instance of human has been added to the schema and everything will be reconciled again to avoid duplicates. Once all is deleted, we will give it another shot to meet all standards and merges.
I appreciate your patience. Allyshaleonard (talk) 15:31, 31 January 2023 (UTC)
same-topic discussion also at Wikidata:Administrators'_noticeboard#Please_Delete_Incorrect_Work_-_Deletion_Request_Made Estopedist1 (talk) 20:51, 31 January 2023 (UTC)

Collapsing property navboxes

See proposal in Wikidata talk:Property navboxes#Collapsing the navboxes. --Epìdosis 15:43, 31 January 2023 (UTC)

Has the Smithsonian online information removed DOB and Place of birth

I am trying to update the entry for https://www.wikidata.org/wiki/Q108744366 Kathy Caraccio in preparation of writing an article. It looks like SAAM, (as well as Nat Gallary) has removed DoB from their online information https://americanart.si.edu/artist/kathy-caraccio-6318 I noticed this change a few days ago, but attributed it to some front end updates. Does anyone on this notice board have any information about this? It looks like ti is just for living artists. Thanks for any insight. WomenArtistUpdates (talk) 17:05, 31 January 2023 (UTC)

After further investigation, it looks like I was working on a master printer where the data isn't in SAAM's database. Too obscure. Plenty of birth years on living artists. Sorry for the interruption. Best, WomenArtistUpdates (talk) 21:25, 31 January 2023 (UTC)

Last call to vote on revised UCoC enforcement guidelines!

Hi all,

A friendly and final reminder that the voting period for the revised Universal Code of Conduct Enforcement Guidelines closes tomorrow, Tuesday, 31 January at 23:59:59 UTC.

The UCoC supports Wikimedia’s equity objectives and commitment to ensuring a welcoming, diverse movement, and it applies to all members of our communities. Voting is an opportunity for you to be a part of deciding how we uphold this commitment to our community and each other!

To vote, visit the voter information page on Meta-wiki, which outlines how to participate using SecurePoll.

Many thanks for your interest and participation in the UCoC!

On behalf of the UCoC Project Team, JPBeland-WMF (talk) 21:32, 30 January 2023 (UTC)

Hello all,
The vote on the Universal Code of Conduct Enforcement Guidelines is now closed. The results will now be counted and scrutinized to ensure that only eligible votes are included. Results will be published on Meta and other movement forums as soon as they become available, as well as information on future steps. Thank you to all who participated in the voting process, and who have contributed to the drafting of Guidelines.
On behalf of the UCoC Project Team,
JPBeland-WMF (talk) 21:48, 1 February 2023 (UTC)

Should cites IMDb (Q103598310), cites Wikidata (Q103598309) and cites Wikipedia (Q103598308) exist?

According to my intuition all those aren't really grounds to deprecate the statement. Should we remove them and bring the related statement to normal rank? ChristianKl ❪✉❫ 18:44, 26 January 2023 (UTC)

Yes, they're very poor. The definition of deprecated rank is very simple, but again and again people misuse it as a bin in which to put things they don't like. Per Help:Ranking The deprecated rank is used for statements that are known to include errors (i.e. data produced by flawed measurement processes, inaccurate statements) or that represent outdated knowledge (i.e. information that was never correct, but was at some point thought to be). In the case of statements with poor referencing, the options are 1) to remove the statement 2) add a better reference 3) leave the damn thing alone. WD users must always act under the maxim caveat lector, let the reader beware; unreferenced and poorly referenced statements are suboptional but more often than not, still useful.
As none of these three are reasons for deprecation, I think the three items should be deleted and the rank on statements using them brought back to normal. --Tagishsimon (talk) 18:59, 26 January 2023 (UTC)
yeah they are bad BrokenSegue (talk) 19:12, 26 January 2023 (UTC)
There's a similar issue with uncitedness (Q2492572). ChristianKl ❪✉❫ 16:18, 27 January 2023 (UTC)
@ChristianKl: I agree those are insufficient as reasons for deprecation. They might serve a purpose as scoring criteria when evaluating relative "completeness" of items (say, to help identify model items), but we don't have such a system yet.
Now, is there a canonical list of valid reasons for deprecation somewhere, and is there a procedure for amending it? I have seen the list of a dozen "useful" examples at Help:Deprecation#Reason for deprecation (plus a query that might list 750 "reasons" currently used if it didn't time out), some 80+ reasons listed at has part(s) (P527) of list of Wikidata reasons for deprecation (Q52105174), and finally a table Help:Deprecation/List of Reasons for Deprecation with 319 items that are defined as instance of (P31) Wikibase reason for deprecated rank (Q27949697), whether right or wrong (the table includes those mentioned by you above). None of these lists look particularly authoritative to me. Can this situation be improved somehow? I have made a few suggestions at Help talk:Deprecation but I don't know if that is the appropriate forum. --SM5POR (talk) 21:17, 1 February 2023 (UTC)
@SM5POR: Here's a reworked version of the query - https://w.wiki/6HeA --Tagishsimon (talk) 21:23, 1 February 2023 (UTC)
@Tagishsimon: Thanks for the optimization, and it strengthens my point that this list of 751 reasons still can't be considered authoritative, because the most common reason used turns out to be marking inconsistent with the registered/official identifier/name (Q108180274), as it's used on 371171 different statements. That's apparently a robot running around deprecating the titles of scientific articles which it has itself set a few milliseconds earlier!
As the purpose of the deprecation mechanism is to help other editors find out what statements are invalid and should not be repeated, who is going to learn anything from this?
I may be wrong though; I'm just extrapolating from what I think I'm seeing. --SM5POR (talk) 23:18, 1 February 2023 (UTC)
@SM5POR:There is not an authoritative list. Users are, by & large, welcome to coin new reasons anytime they come across a circumstance not covered by any of the existing reasons. However it is certainly the case that a) many statements are deprecated for incorrect reasons and b) some 'reason for deprecated rank' values are inappropriate.. So wherever a deprecation, and/or a reason for deprecated rank, falls outside the deprecation policy set out in Help:Ranking, those would be errors that should be corrected. Then c) some items having a P31 of 'reason for deprecation', per the original post, should not exist since they are not valid reasons for deprecation. Meanwhile marking inconsistent with the registered/official identifier/name (Q108180274) looks to me like it's probably a valid reason for deprecation. I don't know the ins & outs, but a title is being deprecated and a different title enstated at normal rank ... I suspect from the examples I've looked at that there is an error in source A being depecated on the basis of a value in source B. --Tagishsimon (talk) 23:31, 1 February 2023 (UTC)
By design we don't have an authoritative list. Users should be able to express whatever reasons they have in mind. Using items in this way allows them to be translated and users of different language to understand the reason which wouldn't be the case if we would allow users to describe the reasons in free text. On the other hand, it's good to have processes to correct mistaken uses of deprecation.
There are two things to list. One is the reasons tagged with as reasons for deprecation and the other is the constraint violations that don't use any of them. I create the list for the items that are tagged that way so that it can be watched and when there are new items, those can be more immediately dealt with. ChristianKl ❪✉❫ 16:03, 2 February 2023 (UTC)
@ChristianKl: Letting users define additional reasons is fine, I think (and I just created one as I explain below); by "canonical list" I'm not asking for something only admins can edit, but a documented way of determining which reasons are currently identified. Since there are in effect four different lists of differing lengths, nobody can tell for sure how to go about registering one.
For the one I created, I made it an instance of (P31) Wikibase reason for deprecated rank (Q27949697), and also a subclass of (P279) obsolete (Q107356532) since I want to encourage maintaining some structure relating different reasons to each other, not just an unordered list of several hundred items.
Now, to satisfy the value constraint on reason for deprecated rank (P2241), it's sufficient to make the reason an instance of (P31) any subclass of Wikibase reason for deprecated rank (Q27949697), such as category mistake (Q1735821), which is claimed twice on edition number (P393) of Krishnamurti's Notebook (Q6437542) and an instance of (P31) informal fallacy (Q3312438). We can trace this item along a somewhat unexpected subclass path via
A few of the items in this path are also subclasses of error (Q29485), which is a true reason for deprecation, but that doesn't matter, since the constraint will not accept an instance of another instance. You could look at Talk:Q3312438, expand the class description and look for another path, but I can't find one. Of course, realizing that category mistake is a valid reason for deprecation just because it's some kind of medicine is the real giveaway here, and it only illustrates the problem many editors (myself included) have dealing with the fundamental concepts of instance and class.
In total, Wikibase reason for deprecated rank (Q27949697) has 11,414 subclasses, 365 of which have together yielded 9,308 instances, about two thirds of them being localized outbreaks of the COVID-19 pandemic. Only three of these instances have actually been used as stated reasons for deprecation; besides category mistake they are pseudepigraph (Q827597) and COVID-19 pandemic (Q81068910), so it's hardly a major problem.
It does however raise the issue of whether it's a good idea to let regular items that are constantly edited and (sort of) moved around in the class tree also serve part-time as reasons for deprecation. Maybe the latter should constitute a category of Wikidata internal items only, isolated from regular items (say, by a suitable constraint on subclass of (P279)? Or is it just the immediate subclasses of Wikibase reason for deprecated rank (Q27949697) that are the problem, while the hierarchy is preferably maintained as part of the main class tree? --SM5POR (talk) 23:56, 4 February 2023 (UTC)
@Kanashimi: can you tell us why your bot uses deprecation in this way? ChristianKl ❪✉❫ 15:54, 2 February 2023 (UTC)
These articles are non-English scientific papers. However, the source of the information I have obtained is in English only, not in the original language. So I think I should add a lower grade for them. Kanashimi (talk) 00:30, 3 February 2023 (UTC)
@Kanashimi: Thus the English-language references may be secondary sources, kind of. But deprecating the title is in effect the lowest possible grade you can give to a statement, short of deleting it. At the same time, Wikidata is full of completely non-sourced claims at normal rank. Surely a secondary/translated source is better than no source at all, don't you think?
If you have multiple sources to the same claim (or slightly different ones, such as one title in the original language and a translated one in square brackets), you could promote the primary/original source by giving that claim a preferred rank instead. --SM5POR (talk) 09:40, 3 February 2023 (UTC)
So it looks like I can just skip the step of lowering the ranking? Kanashimi (talk) 10:34, 3 February 2023 (UTC)
@Kanashimi: I think so, yes. Deprecation should be the exception, when something is actually wrong. --SM5POR (talk) 11:53, 3 February 2023 (UTC)
However, I see the example here. Maybe we need to modify this example as well? Kanashimi (talk) 12:02, 3 February 2023 (UTC)
@Kanashimi, @VIGNERON en résidence: I wasn't aware of that constraint. According to the edit history it was added in December 2018 among a number of format constraints, but many of them have since been deprecated, this one in September 2022. See Property talk:P1476#Constraint prohibiting '[' and ']' should be deprecated or modified for the specific reason why. And don't bother editing the deprecated constraint as it may confuse other editors as to why it was deprecated.
Also, the point of that constraint was to promote the title in its original language, but I don't recall seeing the Russian titles at all on the articles your robot was processing. In any case, I think deprecation is too severe as a response to a translated title (it may well be the title of a published translation of the work, unless that is supposed to have an item of its own). Better to give the original title preferred rank, if you can add that one. --SM5POR (talk) 14:46, 3 February 2023 (UTC)
I deprecated this constraint after this discussion [[Property_talk:P1476#Constraint_prohibiting_'['_and_']'_should_be_deprecated_or_modified]] (@Wostr:) as it wrongly assumed that a title beggining with [ is incorrect (it is rare but it's not). But indeed this constraint is also wrong about ranks (it's also no how it works). I'm wondering if we shouldn't just delete it. Cheers, VIGNERON en résidence (talk) 15:37, 3 February 2023 (UTC)
@VIGNERON en résidence: Considering all the work that seems to have gone into crafting those constraints, it would be pity to throw it all away (and then risk repeating it) just because we don't use them anymore. I wrote an incompatible with best current practice (Q116654895) reason for deprecation and added it to the constraint statement. It should be useful on other outdated constraint, usage example, model item or recommendation property statements as well, and will be easier to apply (especially when translated) than rewriting all those clarifications in multiple languages. Now, if we could have the editor interfaces make the reason for deprecated rank (P2241) stand out in bold or something to overshadow the deprecated text...
Progress is not solely based on an unbroken chain of recognized successful innovations, but even more so on uncountable (and indeed uncounted) mistakes, failures and utter disasters. "Those who forget history are doomed to repeat it." --SM5POR (talk) 17:24, 3 February 2023 (UTC)
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat/Archive/2023/01&oldid=1832235579"