Jump to content

Wikidata:Project chat: Difference between revisions

Shortcuts: WD:PC, WD:CHAT, WD:?
From Wikidata
Browse history interactively
← Older editNewer edit →
Content deleted Content added
VisualWikitext
M2k~dewiki (talk | contribs)
4,248,565 edits
(One intermediate revision by the same user not shown)
Line 336: Line 336:


Thanks a lot! [[User:M2k~dewiki|M2k~dewiki]] ([[User talk:M2k~dewiki|<span class="signature-talk">{{int:Talkpagelinktext}}</span>]]) 23:09, 1 August 2023 (UTC)
Thanks a lot! [[User:M2k~dewiki|M2k~dewiki]] ([[User talk:M2k~dewiki|<span class="signature-talk">{{int:Talkpagelinktext}}</span>]]) 23:09, 1 August 2023 (UTC)

:Hello, possible duplicates (by searching for ''template gmina'' + name of the gmina) can be found at
:* [[User:M2k~dewiki/Tools/possible duplicates templates gminas]]
:For example:
:* [[:d:Q22730319]]
:* [[:d:Q22802308]]
:* [[:d:Q22905161]]
:The current list with more than 1.400 possible duplicate items also includes some false positive matches, since it is simply based on string matching. [[User:M2k~dewiki|M2k~dewiki]] ([[User talk:M2k~dewiki|<span class="signature-talk">{{int:Talkpagelinktext}}</span>]]) 12:06, 3 August 2023 (UTC)


== Interwiki for external sites ==
== Interwiki for external sites ==

Revision as of 12:07, 3 August 2023

Edit


Property for linking vital records / civil registration to the institution they kept them

Hi, do we have or need the possibility to link online vital records or civil registrations to the institution (religious institution, state organization) that kept them? Something along the lines of:

This would probably help a lot if maintained consistently. A possible name could be “registration records available at“.

ChristianKl (talk) 15:11, 24 June 2017 (UTC) Melderick (talk) 12:22, 25 July 2017 (UTC) Richard Arthur Norton Jklamo (talk) 20:21, 14 October 2017 (UTC) Sam Wilson Gap9551 (talk) 18:41, 5 November 2017 (UTC) Jrm03063 (talk) 15:46, 22 May 2018 (UTC) Egbe Eugene (talk) Eugene233 (talk) 03:40, 19 June 2018 (UTC) Dcflyer (talk) 07:45, 9 September 2018 (UTC) Gamaliel (talk) 13:01, 12 July 2019 (UTC) Pablo Busatto (talk) 11:51, 24 August 2019 (UTC) Theklan (talk) 19:25, 20 December 2019 (UTC) SM5POR (talk) 20:17, 29 May 2020 (UTC) Pmt (talk) 23:22, 27 June 2020 (UTC) CarlJohanSveningsson (talk) 12:13, 30 July 2020 (UTC) Ayack (talk) 14:39, 12 October 2020 (UTC) EthanRobertLee (talk) 19:17, 20 December 2020 (UTC) -- Darwin Ahoy! 18:20, 25 December 2020 (UTC) Germartin1 (talk) 03:13, 30 December 2020 (UTC) Skim (talk) 00:13, 10 January 2021 (UTC) El Dubs (talk) 21:55, 29 April 2021 (UTC) CAFLibrarian (talk) 16:36, 30 September 2021 (UTC) Jheald (talk) 18:50, 23 December 2021 (UTC)[reply]

Notified participants of WikiProject Genealogy --Emu (talk) 12:04, 17 July 2023 (UTC) Emu (talk) 12:04, 17 July 2023 (UTC)[reply]

There is also archives at (P485) and supporting values like collection (P195) and catalog (P972). Also noticing that most archives have their own identifier for å file or person i the archive like UK National Archives ID (P3029) and Swedish National Archive reference code (P5324) Pmt (talk) 14:06, 17 July 2023 (UTC)[reply]
When the subject item isn't identical to merely the registry itself (such as Saint Wolfgang Church (Q22041143) or Vienna (Q1741)), but is an entity with a purpose beyond being a registrar, the registry should preferrably have a Q-item of its own, and the appropriate URL property assigned to that item instead. This will allow a registry changing hands, splitting or merging it, and assigning various other existing generic properties to it, without the need to create a large number of new properties for the vital records only.
Applying the start time (P580) and end time (P582) qualifiers to the target URL would otherwise be ambiguous; say, would a start time of 2015 indicate that a new registry was established, or that an existing registry became available on the Internet that year? Merely providing a URL property for something that existed long before the Internet is insufficient; the URL should be optional.
And an archive is different from an actively maintained registry, which may later become archived, but I believe you know that already. --SM5POR (talk) 15:22, 23 July 2023 (UTC)[reply]
@SM5POR I think I understand your point but I‘m not sure about the specifics. In the case of Q116201377 we indeed have an organization (or at least a division within an organization) which is solely responsible for keeping life records. I’m not so sure about religious institution. Do you mean something along the lines of parish Vienna 03, Rennweg, Maria Geburt (Q105890802)? Although this is also kinda conflated because it mixes the institution of the local church parish with its district, i. e. its territorial (and personal) scope. Also, the parish has all sorts of other duties which are not limited to keeping life records. What would you propose in such (very common) cases?
I’m not sure about your concerns with start time (P580) and end time (P582). Wouldn’t this issue be solved with those values acting as qualifiers? Or do you want to point out that we use start time (P580) and end time (P582) for very different ontological concepts: The meaning of those qualifiers in Q156487#P39 is indeed very different to the meaning in Q48256#P227. It’s not as bad as described at User:Lucas Werkmeister/P642 considered harmful but it is an issue (CC Kolja21 because of a potential GND ID (P227) modeling issue)
Not so sure what you mean by your last sentence, could you maybe elaborate a bit? Thank you! --Emu (talk) 19:42, 23 July 2023 (UTC)[reply]
We have start of covered period (P7103) and end of covered period (P7104) which could be used here instead of start time (P580) and end time (P582). M2Ys4U (talk) 01:52, 24 July 2023 (UTC)[reply]
@M2Ys4U Thank you, I didn’t know those properties! That’s better, probably also for GND ID (P227), @Kolja21? --Emu (talk) 11:20, 24 July 2023 (UTC)[reply]
I don't see the need for it. If you look at Dnipro (Q48256): Q48256#P227
  • The years given are not about an archive the years are about the city.
GND + start time (P580) is used by 1.145 items. --Kolja21 (talk) 20:52, 24 July 2023 (UTC)[reply]
Yes, but GND 1269131966 was created on 2022-09-29 and that would probably be the “time an entity begins to exist or a statement starts being valid” in this case. --Emu (talk) 20:58, 24 July 2023 (UTC)[reply]
To clarify, Vienna (Q1741) is a city (the capital of Austria), not merely a vital records registrar. If you were to define a new property "death registry" with a URL datatype, pointing to a resource listing those who died in Vienna during a particular period in time, you are skipping a number of important links betwen the geographical location and the records you are referring to, which risks making the long-time maintenance (and use) of this property for multiple locations a nightmare, or at least an inconvenient burden. These are the links I'm thinking of, links that the users accessing the resource need to be aware if:
  1. The first entity has an arbitrary type, making its relation to the records unclear. In addition to the city of Vienna, you mention Saint Wolfgang Church (Q22041143) among your examples. Does this resource list residents of Vienna/members of that church who have died (somewhere), people who have died in Vienna/been buried at the church irrespective of their residency/membership, or people appearing in any death registry that happens to be maintained by the vital records registrar employed by the city/church? These are all different relations, and which records should be searched depends on what the user is looking for. Therefore I suggest one property should be "vital records registry", refering to the registry maintained by that corporate entity (city, church, whatever), and it must be a Q-item, not a URL. All cities may not have their own registries, but the registrar could be found on the provincial or national level.
  2. The registry, in turn, may have different parts, such as birth records, marriage records, or death records, organized differently by different registrars. If that registry has an official Internet presence, you could add a URL to their item. But are the vital records themselves available via that URL? Maybe, maybe not. If they are, you can of course add more specific URLs to the same item, using qualifiers to indicate what records they pertain to.
  3. The registrar should be in charge of the current registry to be able to update it, and may also retain records from long ago, never to be updated again (unless an error is discovered and corrected). But eventually a set of old records is transferred to an archive, which may be an entirely different organization, responsible for safekeeping and providing access to old records, but not updating them (this is what I referred to in my last sentence above). It would then need its own Q-item, and may be referred to from the registry item using archives at (P485). The archive could then have a set of URLs provding access to its holdings.
  4. In your examples, you mention Ancestry and FamilySearch. These are neither official vital records registries nor archives, but independent collections of records, often from multiple sources (typically archives, but also user-contributed data). Give them their own Q-items. Their holdings may be organized in entirely different ways, and the URLs listed for them should reflect that.
Adding a property to Vienna (Q1741) pointing directly at a FamilySearch URL providing records pertaining to Vienna would be inappropriate, in my opionon. A significant part of genealogy research is finding the best source of records, where researchers have to evaluate both reliability and relevance of individual pieces of information in the specific cases they are concerned about. Wikidata cannot do that job for them. At most, Wikidata could indicate which archives have been included in which collections (if that is even possible; limiting a search to a particular source of theirs or telling where their data comes from isn't very easy to do), but going beyond that and listing every single geographic location appearing in their collections (and adding corresponding URLs from each location to those collections) would be pretty pointless. You might just as well add pointers from year or name items to collections with records referring to them; it would serve no useful purpose.
The semantics of the start time (P580) and end time (P582) qualifiers would depend on the items and statements they are applied to. On the suggested "vital records registry" property, start time (P580) would refer to when the registry was established. On an archives at (P485) link, it would indicate the age of the oldest records from the subject item that exist at the archive; same thing with collection (P195). On a URL property if would indicate when that particular URL became valid. --SM5POR (talk) 09:47, 27 July 2023 (UTC)[reply]
It seems to be that we have very different use cases in mind. You wrote: A significant part of genealogy research is finding the best source of records, where researchers have to evaluate both reliability and relevance of individual pieces of information in the specific cases they are concerned about. Wikidata cannot do that job for them. How true, Wikidata will never be able to weigh different sources against each other. What Wikidata could do is solving rather trivial tasks that many historians, family history enthusiasts and/or Wikimedians face every day:
  1. “I know from later records that this person was baptised in church X, where can I find the records, preferably online?“
  2. “It is suspected that this person died around 1880 in the northern parts of Lower Austria. Which churches (that existent around the time of suspected death) are there in this region and where can I find their records, preferably online?“ In many cases, it’s simply trial and error.
There’s nothing improper about using third-party sources. Let me address some of your arguments:
  • Does this resource list residents of Vienna/members of that church who have died (somewhere), people who have died in Vienna/been buried at the church irrespective of their residency/membership, or people appearing in any death registry that happens to be maintained by the vital records registrar employed by the city/church? These are all different relations, and which records should be searched depends on what the user is looking for. Does it matter? In the case of Austrian Catholic death records this would often be quite fuzzy, something along the lines of “In principle, only Catholic people who lived and died in the parish, but sometimes dying there is enough and sometimes being buried here or somewhat nearby is enough and sometimes having some sort of religious ceremony is enough and sometimes it’s enough if the body was discovered within the parish or in the vincinity or indeed if somebody feels that this person or their family has a close connection to this parish even when another parish is officially responsible. Oh and sometimes there are records of Protestants, Muslims, Jews or atheists because maybe there was no appropriate registry here or Protestantism was still illegal or somebody just assumed they were Catholic or nobody could remember exactly what was happening.“ That’s all very interesting but you could never model that in Wikidata. It’s not really necessary anyway because in light of this fuzziness you often have to search in various records anyway.
  • Therefore I suggest one property should be "vital records registry", refering to the registry maintained by that corporate entity (city, church, whatever), and it must be a Q-item That doesn’t really answer my question from above. What is the “vital records registry“? The organization, the abstract concept of registry?
  • The registrar should be in charge of the current registry to be able to update it, and may also retain records from long ago, never to be updated again (unless an error is discovered and corrected). But eventually a set of old records is transferred to an archive, which may be an entirely different organization, responsible for safekeeping and providing access to old records, but not updating them (this is what I referred to in my last sentence above). That’s of course true but I’m not sure how this would be relevant here.
  • Ancestry and FamilySearch […] Give them their own Q-items. Each collection? I’m not sure that this would be a terribly good idea.
I’m not dead set on the URL datatype and I would be happy to accommodate any needs for a more complex modeling. However, I still don’t quite understand what you are proposing. Could you maybe give me a possible set of statements that could achieve to accommodate my use cases? Or is there a problem with my use cases? --Emu (talk) 11:45, 27 July 2023 (UTC)[reply]
no 5.21.241.21 11:23, 31 July 2023 (UTC)[reply]

Dark mode

Can an interface admin please import the dark mode gadget on enwiki? This means importing the following:

The last page may need to be modified for Wikidata's special UI. Jasper Deng (talk) 10:07, 21 July 2023 (UTC)[reply]

@Jasper Deng: Imported. Do you want to have me activate it right away? --Matěj Suchánek (talk) 16:55, 28 July 2023 (UTC)[reply]
@Matěj Suchánek: Thanks for importing it. Yes, do activate it right away as a gadget.--Jasper Deng (talk) 17:36, 28 July 2023 (UTC)[reply]
✓ Done Should work now. --Matěj Suchánek (talk) 18:30, 28 July 2023 (UTC)[reply]
Nice, but this gadget seems to interact poorly with the ImageHeader gadget. The images at top right are shown with inverted colours, which is a little disconcerting for photographs. Bovlb (talk) 18:52, 28 July 2023 (UTC)[reply]

Catalonian buildings

These folks are working on something I can't quite figure out. The house in Barcelona has been demolished, apparently, but many of its properties are marked imported from Wikimedia project (P143): Catalan Wikipedia (Q199693), but what article? There has never been a linked article that describes this item only, as far as I can tell, so how do these quickstatements imports work?

Furthermore, there is a lot of deletion of instance of (P31): tourist attraction (Q570116) and unfinished building (Q1570262) from items. Sagrada Família is the sine qua non of unfinished buildings, as well as a verifiable tourist attraction! Why are properties such as this being removed en masse?

Catalan translation

Aquesta gent està treballant en una cosa que no acabo d'entendre. La casa de Barcelona ha estat enderrocada, pel que sembla, però moltes de les seves propietats estan marcades imported from Wikimedia project (P143): Catalan Wikipedia (Q199693), però quin article? Mai no hi ha hagut un article enllaçat que descrigui només aquest article, pel que puc dir, com funcionen aquestes importacions de declaracions ràpides?

A més, hi ha una gran quantitat de supressió de instance of (P31): tourist attraction (Q570116) i unfinished building (Q1570262) dels elements. La Sagrada Família és el sine qua non dels edificis inacabats, a més d'una atracció turística verificable! Per què s'eliminen en masse propietats com aquesta? Elizium23 (talk) 17:22, 23 July 2023 (UTC)[reply]

With instance of (P31) we try to avoid over use as so many properties can be described as an instance of something else. As all churches can be viewed as tourist attractions you don't want to label each as such, and state of conservation (P5816) would be better for buildings. Vicarage (talk) 05:59, 24 July 2023 (UTC)[reply]
As for the house, it was imported from the heritage register Inventari del Patrimoni Arquitectònic de Catalunya (Q1393661) to Cultural heritage monuments in Aiguafreda (Q11932939), and imported from the list to Wikidata. The article was merged with Habitatge a la carretera de Barcelona, 21 (Aiguafreda) (Q19255897). Once removed from the register, it has been removed from the list and marked as deprecated. You are right, these imports should have been reported more specifically. Vriullop (talk) 08:05, 31 July 2023 (UTC)[reply]

Diversity and bias in DBpedia and Wikidata as a challenge for text-analysis tools

Summary: Diversity Searcher is a tool originally developed to help analyse diversity in news media texts. It relies on automated content analysis and thus rests on prior assumptions and depends on certain design choices related to diversity. One such design choice is the external knowledge source(s) used. In the linked article, they discuss implications that these sources can have on the results of content analysis. They compare two data sources that Diversity Searcher has worked with – DBpedia and Wikidata – with respect to their ontological coverage and diversity, and describe implications for the resulting analyses of text corpora. They describe a case study of the relative over- or underrepresentation of Belgian political parties between 1990 and 2020. In particular, they found a staggering overrepresentation of the political right in the English-language DBpedia. --Jensbest (talk) 12:08, 27 July 2023 (UTC)[reply]

Was the representation on Wikidata better than DBpedia? I would expect it to be. BrokenSegue (talk) 00:55, 28 July 2023 (UTC)[reply]

How to depict freshmen

Any thoughts on what to do at freshman (Q60825090) to depict that a freshman is someone in their first year? {{u|Sdkb}} talk 15:08, 27 July 2023 (UTC)[reply]

First year of any educational institution? Including high school, university, primary school, etc.? Mateussf (talk) 23:41, 27 July 2023 (UTC)[reply]
Why do we need a separate item from freshman (Q1215869)? Elizium23 (talk) 00:42, 28 July 2023 (UTC)[reply]
@Elizium23, it looks like freshman (Q60825090) is specifically about the educational realm, whereas freshman (Q1215869) is a more generic version that could also apply to e.g. freshmen congresspeople. It seems a bit backwards that 1215869 is described as a subclass of (P279) of 60825090, but there may be non-English language stuff at play, so I'd be a little cautious before messing with it too much. {{u|Sdkb}} talk 02:39, 28 July 2023 (UTC)[reply]
There's also freshman (Q13883551) so some tidying up here is needed M2Ys4U (talk) 21:44, 28 July 2023 (UTC)[reply]
We seem to mix three different concepts into two items.
(1) There the generic concept.
(2) There the concept that's about being part in the educational institution (maybe there's an additional seperation here between people in the first year compared to the first semester)
(3) The status of being a nonregular member of a student association before that member becomes a full member which usually is after they were one year in the student organization. ChristianKl ❪✉❫ 11:22, 2 August 2023 (UTC)[reply]
@Mateussf, yep, that's how it's defined in the item. {{u|Sdkb}} talk 02:37, 28 July 2023 (UTC)[reply]

My colleagues have data on wikidata and this data appears on the Google search engine, but my problem is that my data does not appear on the Google search engine. Can you help me solve this problem, knowing that I waited more than two months for it to appear automatically, but to no avail. Zayn Hesham (talk) 16:56, 27 July 2023 (UTC)[reply]

@Zayn Hesham Hi! Thanks for following up.
First of all, you should know that we discourage editors from creating items about themselves.
Secondly, the relevant item appears to be Q119790860. There is a Google Knowledge Graph id which appears valid. When I search for "Zayn Hesham", Google appears to be using a knowledge graph entry, but everyone is going to see slightly different Google results.
What specifically are you doing, what is the result, and what were you expecting to happen instead? Bovlb (talk) 17:28, 27 July 2023 (UTC)[reply]
You may also want to join the discussion at Wikidata:Requests_for_deletions#Q119790860. Bovlb (talk) 19:14, 27 July 2023 (UTC)[reply]
Wikidata is not intended to improve the search results for a specific query. It's not associated with Google, so even if Wikidata items seem to effectively affect the Google search results, Wikidata isn't 'working' with the results directly. WD is just one of millions of websites indexed by Google (Yahoo, Yandex, whatever), and WD never guaranteed an assistance in a sort of promotion. Its main goal is structuring the data which could be used in Wikipedia and its sister projects.
Additionally, the search engine actually 'knows' you, and your name/alias is connected to a G- knowledge graph. --Wolverène (talk) 12:34, 28 July 2023 (UTC)[reply]

Merge request (São Gonçalo do Sapucaí)

São Gonçalo do Sapucaí (Q22065156) should be merged with São Gonçalo do Sapucaí (Q952405) as both refer to the same city in Brazil. I don't understand why the former has two interwiki links. Mateussf (talk) 23:39, 27 July 2023 (UTC)[reply]

@Mateussf: This is another bot-created-article situation. The two WD items cannot be merged as there are Wikipedia articles at both ceb.wiki and sv.wiki. I'll mark it as a permanently duplicated item. — Huntster (t @ c) 13:31, 28 July 2023 (UTC)[reply]
For some reason, those two wikis have separate articles for the cities and the municipalities, even when the city is the whole municipality. I've seen it with many cities in my country as well. –FlyingAce✈hello 15:10, 31 July 2023 (UTC)[reply]
@FlyingAce: Yes, this is because those articles were created by a bot based on Geonames.org entries. It's frustrating, but this is the best way currently to deal with these situations. — Huntster (t @ c) 21:39, 31 July 2023 (UTC)[reply]

Baronets

See Jock Delves Broughton (Q120985038). I want to show the succession of Baronets, but looking at multiple entries, they are all set up differently. Should a baronetcy be set up as a position? I want to have an auto generate table that would list the succession, like we have for monarchs. RAN (talk) 16:31, 28 July 2023 (UTC)[reply]

@Richard Arthur Norton (1958- ) I think the general standard for hereditary titles like this is to have it as noble title (P97):baronetcy or the more specific noble title (P97):specific baronetcy. Someone would need to tweak the table script to be able to generate a list of holders, since I think it's only set up for P39, but in principle it should be a relatively small change. Andrew Gray (talk) 22:01, 29 July 2023 (UTC)[reply]
It looks like Charles III (Q43274) duplicated "Prince of Wales" as both a noble_title and a position, as a way to solve the puzzle, but we should only have to add the data once. --RAN (talk) 14:17, 2 August 2023 (UTC)[reply]

References

When I click to add a reference "reference url" is no longer popping up as a choice, what happened? How can we restore the choice? RAN (talk) 23:44, 28 July 2023 (UTC)[reply]

it's a choice for me if I type it out but I feel like it used to be suggested immediately. not sure what changed. it looks like hwbsgetsuggestions is returning reference url with lower probability. BrokenSegue (talk) 00:27, 29 July 2023 (UTC)[reply]
@Harej recently made an edit but that diff looks innocuous and unrelated to the autocomplete. Harej is a WMF employee, though, so perhaps changed something else behind the scenes? Elizium23 (talk) 00:41, 29 July 2023 (UTC)[reply]
I haven't been a Wikimedia Foundation employee in four years, and there is no way that edit would have caused it. Harej (talk) 18:04, 30 July 2023 (UTC)[reply]

Local notability policy

What is the "local notability policy", it keeps turning up in deletion arguments as a reason to remove people that aren't famous, but no one can define it, or point to some written policy guideline. It may be a concept imported from English Wikipedia. RAN (talk) 16:38, 29 July 2023 (UTC)[reply]

Do you mean a fragment of my reply on WD:RFD? I meant Wikidata:Notability. I didn't guess that the wording local notability policy may sound obscure for the English speakers, I was sure that local policy in this context means a policy within the project we are using at moment, here it's Wikidata. --Wolverène (talk) 18:38, 29 July 2023 (UTC)[reply]
Also, I never said that a person must be famous to be notable for Wikidata. --Wolverène (talk) 18:58, 29 July 2023 (UTC)[reply]
  • It sounds like someone who may only be famous locally, like in the town where I live where the reference material may be the local paper. You argument for deletion was: "Mr Archibald Olive just lived his life as a farmer, he wasn't either an election candidate, or a soldier, or a trade unionist.", which to me is more like the English Wikipedia policy of fame. --RAN (talk) 22:09, 29 July 2023 (UTC)[reply]
    If the local paper had anything substantial to say about Mr. Archibald Olive, we wouldn’t have this conversation, now would we? --Emu (talk) 22:51, 29 July 2023 (UTC)[reply]
  • Yes, we would still be here because the argument is that he is not an "election candidate, or a soldier, or a trade unionist." Of course if both of you just followed our actual policy where: "It refers to an instance of a clearly identifiable conceptual or material entity that can be described using serious and publicly available references", we would all be saved lots of time both at deletion debates and here at project chat. --RAN (talk) 23:37, 29 July 2023 (UTC)[reply]
    Of course, election candidates, soldiers, and trade unionists were examples of people who might be described by something more than US census data (mandatory existing for everyone by the law) or an obituary (roughly speaking, obituaries and written birthday greetings are equally useful in the matter of making subjects clearly identifiable). And by something more than a entry in a database which can be edited by anyone with any intentions. Also, AFAIK not every "Election candidate, or a soldier, or a trade unionist." is notable by the enWP policy... with which I probably familiar worse than you, but seems definitely not every of them. Should we really continue wasting time discussing the English Wikipedia?
    These debates are becoming more emotional than constructive, and it's not good. Is not it better to start a discussion which can help us all to understand if Find a Grave or similar websites suggest notability? I am not sure that only I doubt that they suggest. In partucular, I don't see Find a Grave memorial ID (P535) listed here.
    P.S. Just in case, I'd like to apologize for my words which might've sounded like I believe that farmers are not notable. Of course, some farmers are notable. Just like election candadates, some trade unionists and actors. --Wolverène (talk) 02:09, 30 July 2023 (UTC)[reply]
    @Richard, accusing two admins of not following notability policy is a pretty serious charge. Please at least try to understand that your reading of WD:N, while shared by a vocal minority, is far from general consensus. --Emu (talk) 08:21, 30 July 2023 (UTC)[reply]
  • If you can't win an argument by logic and persuasion, remember to threaten to use the Admin card. --RAN (talk) 15:38, 30 July 2023 (UTC)[reply]
    Again, it sure looks like you are accusing me of inappropriate behavior as admin (intimidating users). Per AGF and because English is not my first language, I will assume that this is not the case. Might I just remind you that I haven’t acted in my capacity as admin in this discussion? --Emu (talk) 17:08, 30 July 2023 (UTC)[reply]
  • Why would someone mention they were an Admin if it wasn't intended to intimidate? Why don't you and I just end it here, and let other people express their opinions. You made your point and I made my point, all the needless repetition is just wasting everyone's time. --RAN (talk) 22:38, 30 July 2023 (UTC)[reply]
    Because accusing a non-admin of not following notability policy is irrelevant because they don’t have to. Admins do. But sure, let’s end this discussion that you started at this point. --Emu (talk) 23:29, 30 July 2023 (UTC)[reply]

Immigrants

Do we mark people as subject_has_role=immigrant or significant_event=immigration, who migrate to other countries. It can be inferred from country_of_citizenship, and sometimes birth and death locations, but not all entries have that information complete. Not everyone that died in another country was an immigrant, they can die on vacation. RAN (talk) 23:47, 29 July 2023 (UTC)[reply]

I would be interested to see how consistently we use emigrant (Q4989857) and immigrant (Q12547146). Of course, most people are both, being reciprocal properties, but some events are more notable than others, such as Emigration from Nazi Germany 1933-1945 (Q110486791). English Wikipedia categories are a holy intractable mess, in this regard. Elizium23 (talk) 00:48, 30 July 2023 (UTC)[reply]
We have residence (P551) to show when someone picked up a country as a residence. I don't think it's good modeling ot use subject_has_role or significant_event here. ChristianKl ❪✉❫ 22:16, 1 August 2023 (UTC)[reply]

Authority Control on Neema Parvini´s Wikipedia page

Hello as he is on Wikidata and Wikidata now has references on his page to the British Library, I am wondering why Authority Control won´t show on his WP page, please correct me if my understanding is wrong, I would just like to know why? StrongALPHA (talk) 10:07, 30 July 2023 (UTC)[reply]

I think I've fixed it. Neema Parvini (Q120922120) RVA2869 (talk) 12:19, 30 July 2023 (UTC)[reply]

Still alive and well (centenarians)?

I own the book A Practical English Grammar (Q28132035) by Audrey Jean Thomson (Q105812645) and Agnes Wallace Martinet (Q112356982). Are they both still alive and well? I would be happy for them (they would be 103 and 102 years old, respectively)...-- Carnby (talk) 17:30, 30 July 2023 (UTC)[reply]

Mix'n'Match behaviour

Has Mix'n'Match changed its behaviour?

I'm using a new laptop (Windows 11, also new to me, less powerful and so slower than my usual Dell XPS, but with my usual browser, Firefox). When I confirm, or reject a match a new tab opens. The status bar of that tab shows progress, then the new tab self-closes.

This makes M'n'M much slower to use for multiple changes. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:38, 30 July 2023 (UTC)[reply]

Mark people as dead

Is it possible to mark a person a dead without knowing anything about its death except that it happened? --Jobu0101 (talk) 06:25, 31 July 2023 (UTC)[reply]

Ah, I guess with P570 and unknown value. --Jobu0101 (talk) 06:27, 31 July 2023 (UTC)[reply]
In all likelihood the source will have some clue as to when the death happened, you can specify a death happening some time in the 1800s for instance. This is preferable to using unknown value. You can also add the qualifiers earliest date (P1319) and latest date (P1326) and even sourcing circumstances (P1480). Example: Ramesses IV (Q1532). Infrastruktur (talk) 06:50, 31 July 2023 (UTC)[reply]
@Jobu0101: also see
M2k~dewiki (talk) 15:23, 31 July 2023 (UTC)[reply]

Wikidata weekly summary #587

Here's your quick overview of what has been happening around Wikidata over the last week.
  • Tool of the week
    • osm.js will add a link to the navigation for items that have a Wikidata mapping on the OSM side.
  • Development
    • Wikibase REST API:
      • We released the following new stuff to Wikidata:
        • GET /entities/properties/{property_id}/statements (phab:T339356) (new route)
        • GET /entities/properties/{property_id}/statements/{statement_id} (phab:T338383) (new route)
        • GET /statements/{statement_id} (now supports statements on Properties)
      • We started work on a new route for replacing the data of a statement on a Property (phab:T340006)
    • EntitySchemas:
      • We worked on improving the termbox (the table of labels, descriptions and aliases) on EntitySchema pages to bring it more in line with the one on Items and Properties
      • We made sure that EntitySchemas are shown by their label instead of ID when available in Recent Changes and similar lists (phab:T339924)
    • Mismatch Finder: We added a language selector so you can switch the interface language (phab:T328149)
    • mul language code: We are working on showing more labels from fallback languags in the termbox (phab:T340832, phab:T338330)

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.

Individual Constituency Heves County (Hungary)

in Individual Constituency Heves County No. 3 (Q15891201) there is a lack of reference for eligible voters (P1867) and successful candidate (P991) seems to be wrong for an electoral district. I cannot read Hungarian, is there someone who can help me? Thanks.-- Carnby (talk) 21:40, 31 July 2023 (UTC)[reply]

Swedish warship prefix HMS or HSwMS depending on context

I raised this in https://www.wikidata.org/wiki/Wikidata_talk:WikiProject_Ships#HMS_or_HSwMS_for_Swedish_ships, but no-one else seems to contribute there. I think its worth getting views here, for the general case where the label an item has depends not merely on they language its used in, but the context too.

https://en.wikipedia.org/wiki/Talk:Swedish_Navy#HMS? describes the problem that Swedes prefix their naval ships with HMS, except when those ships are involved in international operations, when they use HSwMS, to avoid confusion with the Royal Navy. So what do we do for labels, when we don't know the context the names will be used in? Use en:HMS which will confuse many English speakers, use en-GB:HSwMS which is accurate, but who has en-GB as their language code, and only really benefits SPARQL queries where fallback codes can be defined, use en:HSwMS and expect Swedes to pick up sv:HMS. What about official name (P1448) where the official name is context sensitive based on operational use, mul:HMS or mul:HSwMS?

I personally think en-GB:HSwMS, en:HSwMS, sv:HMS, mul:HSwMS, as wikidata use is effectively "use in an international context". Do we have any precedents? English wikipedia usage is mostly HSwMS, but that may have been achieved using a template https://en.wikipedia.org/wiki/Template:HSwMS. Vicarage (talk) 10:21, 1 August 2023 (UTC)[reply]

Pronoun templates

I recently received a complaint about "misgendering" an editor whose pronouns I had no way of knowing. To remedy this, I have copied selected templates from ENWP into Category:Editor's pronouns templates. For example, by using {{they|ExampleUser}} instead of "they", it will automatically be substituted in line with the user's gender preference. Use {{ucfirst:{{they|ExampleUser}}}} for a capitalised version (e.g. "They"). Note that I have not copied over all related templates, just those that seemed most immediately useful. I recognise that this only resolves the problem for English, and this is an international project. Bovlb (talk) 18:04, 1 August 2023 (UTC)[reply]

You can also use the underlying GENDER magic word directly, which should work with any language: for instance, {{GENDER:Bovlb|He|She|They}} becomes He. ({{subst:GENDER:…}} also works, if you prefer that.) Lucas Werkmeister (talk) 08:54, 2 August 2023 (UTC)[reply]

cause of death (P509)

What is the correct way to show that someone died by being pushed in front of a train in the metro? Trade (talk) 20:32, 1 August 2023 (UTC)[reply]

Around these parts, that's typically known in health care as blunt trauma (Q770709). Elizium23 (talk) 02:40, 2 August 2023 (UTC)[reply]

Duplicates for navigation templates for gminas in Poland

Hello, there seem to exist some duplicates for navigation bars for gminas in Poland, for example

In same cases, there seem to exist three items per gmina (e.g. an extra item for french)

They can be found by searching for template gmina + name of the gmina, for example:

Is there a way to find and merge them per script, SPARQL, PetScan, Quickstatements, etc.?

In addition, property Property:P1423 could be added to the navigation template item and Property:P1424 could be added to the item for the gmina, so the naviation template item and the item for the gmina link to each other.

Thanks a lot! M2k~dewiki (talk) 23:09, 1 August 2023 (UTC)[reply]

Hello, possible duplicates (by searching for template gmina + name of the gmina) can be found at
For example:
The current list with more than 1.400 possible duplicate items also includes some false positive matches, since it is simply based on string matching. M2k~dewiki (talk) 12:06, 3 August 2023 (UTC)[reply]

Interwiki for external sites

Today I learned that we have interwiki capabilities for external sites, for example:

[[viaf:64009368|64009368]] instead of [https://viaf.org/viaf/64009368/ 64009368])

I have added this eample, for now, to VIAF cluster ID (P214) using interwiki prefix at Wikimedia (P6720).

I'm also told that using these puts less strain on the database than ordinary external links like the one in my example (see discussion at T343131).

A list of the existing interwiki shortcuts is at meta:Interwiki map; more are likely to be added soon, and they can be requested there.

I suggest we now need to consider:

  • changing constraints on P6720
  • how we could use this as a form of formatter URL (P1630), for internal use and to serve to, for example, {{Authority control}} on other projects (do we need a separate property?)
  • where else it might benefit us.

Thoughts? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:17, 2 August 2023 (UTC)[reply]

Wikifunctions

Wikifunctions, the latest Wikimedia project, has had a soft launch. It's probably too early to be looking at items for its entries, but we could already be linking to its meta pages, such as adding its version of this page to Project:Village pump (Q16503).

Is that in hand, or do we need to do something to get it moving? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:13, 2 August 2023 (UTC)[reply]

I don't think a project code has been implemented for multilingual sites. There is a phabricator ticket for it
Tracked in Phabricator
Task T342857
-- William Graham (talk) 21:56, 2 August 2023 (UTC)[reply]
The goal of this new wiki is not clear to me : is it there to simply help document each computer function (SQL function, Lua function, etc etc) or is there to store functions (module/templates) that could be used crosswiki into wikiarticles ? Bouzinac 💬●✒️●💛 07:50, 3 August 2023 (UTC)[reply]
Wikifunctions FAQ. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:00, 3 August 2023 (UTC)[reply]

Open Library <> Wikidata connection

Hello, I'm an volunteer for Open Library (Q1201876) and would like to have their system start pulling information from Wikidata for populating the author pages. For example, we could pull in date of birth/death and some identifiers. The tentative plan is to start by pulling the JSON for an author by wikidata ID (which is stored in Open Library) with a url like https://www.wikidata.org/wiki/Special:EntityData/Q110436.json when an author is edited (which doesn't happen too often). Then we'll start slowly integrating by doing thing like displaying the DOB and a little note that the data comes from Wikidata. From there we can build out more things like using identifiers, alternative names, names in other languages, etc. We're thoughtful about caching the results from Wikidata as to not hit the API more than needed. Doing a bulk import is also under consideration once we start using the data a bit.

I just wanted to put this out there and see if anyone had feedback.

PS: I'm just a volunteer so can't speak on behalf of OL but I have talked with their team and they're quite excited about the idea of pulling more from wikidata and directing users to contribute to wikidata for edits when possible. Thanks RayScript (talk) 21:56, 2 August 2023 (UTC)[reply]

With my Expounder sites (https://expounder.info) I started that way, but quickly switched to using SPARQL queries to generate custom JSON files for individual entries, and then csv files for blocks of entries. I found the custom output much easier to process, and the SPARQL support people here very helpful in designing queries. Vicarage (talk) 22:10, 2 August 2023 (UTC)[reply]
Https://fannish.expounder.info might be of interest as it has entries for authors with timelines and awards given, and links to other literary sites Vicarage (talk) 22:27, 2 August 2023 (UTC)[reply]
@RayScript: really cool idea. Are you aware of Open Library ID (P648), Open Library subject ID (P3847) which could be used for other integrations (or to bootstrap your efforts)? I might be a little cautious about SPARQL (though I'm not speaking with terribly much experience) because my guess would be that SPARQL is a lot more expensive for WMF than just hitting the EntityData endpoint. Or at least it can be. If it matters depends on the traffic you get (though honestly you'll want to cache even just for latency reasons, items can get big). They literally just finished making a new REST API for Wikidata that you would be a good test customer for. BrokenSegue (talk) 01:28, 3 August 2023 (UTC)[reply]
I found that a bulk query for 1000 authors was much better than 1000 individual requests for files. You can also tailor a SPARQL query to only ask for items that have been updated in a certain time window. OpenLibrary need to think whether to get updates when they edit an author, or poll to see if WD has updated them.
https://fannish.expounder.info/Q42 shows how I do things for Douglas Adams, with tabs for Wikipedia and Wikidata to allow easy updates.
And if you already store wikidata ID's for everyone, it would be very useful if OpenLibrary could provide an API to give access to their pages given the ID, so https://openlibrary.org/wd_api/Q42 or similar pointed to your page on Adams. Vicarage (talk) 08:25, 3 August 2023 (UTC)[reply]
What @BrokenSegue says is correct. We want to discourage use of the Query Service for access that doesn't actually require the graph. The REST API, action API and Special:EntityData are much preferred because scaling and caching for them is better. If you're interested in going with the new REST API that'd be great and I'd love to hear your feedback so we can continue adjusting it. Lydia Pintscher (WMDE) (talk) 08:42, 3 August 2023 (UTC)[reply]

Clumsily constructed Wikidata item

So, based on its history, Q1260524 ("time of the day") seems to have been created by a non-English speaker from the German-language Wikipedia article Tageszeit. Is there any way to correct this to match the standard English term "time of day" (in other words, drop the definite article)? -- DanielPenfield (talk) 11:59, 3 August 2023 (UTC)[reply]

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat&oldid=1945581958"