Warning: file_put_contents(/opt/frankenphp/design.onmedianet.com/storage/proxy/cache/73f7f1bb79ae9ca75b6f1a8cb80adc77.html): Failed to open stream: No space left on device in /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php on line 36

Warning: http_response_code(): Cannot set response code - headers already sent (output started at /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php:36) in /opt/frankenphp/design.onmedianet.com/app/src/Models/Response.php on line 17

Warning: Cannot modify header information - headers already sent by (output started at /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php:36) in /opt/frankenphp/design.onmedianet.com/app/src/Models/Response.php on line 20
Wikidata:Project chat - Wikidata Jump to content

Wikidata:Project chat

Edit links
Add topic
Shortcuts: WD:PC, WD:CHAT, WD:?
From Wikidata
Latest comment: 2 hours ago by AnBuKu in topic Wikidata Game, several versions


Edit

Abu Obeida (Q12178106)

[edit]
Latest comment: 4 hours ago13 comments6 people in discussion

The item Abu Obeida (Q12178106) is fully-protected. Could someone add death date, death place, image and link to Commons Category:Abu Obaida (Hamas)? -- Vysotsky (talk) 23:09, 21 September 2025 (UTC)Reply

If you want information about date of death/death place added it would make sense to provide sources. ChristianKl ❪✉❫ 11:06, 22 September 2025 (UTC)Reply
https://www.bbc.com/news/articles/cm214r5rd29o.amp QwertyZ34 (talk) 10:01, 23 September 2025 (UTC)Reply
Also assassination of Abu Obaida (Q136030663). --Marsupium (talk) 17:00, 30 September 2025 (UTC)Reply
@Mohammed Qays: can you explain why you created the admin only protection / solve these issues? ChristianKl ❪✉❫ 21:09, 30 September 2025 (UTC)Reply
@ChristianKl The person has not been officially announced as deceased, and the item has experienced a lot of vandalism. I protected it for that reason, and it is proper that we wait for an official announcement from the organization they work with before confirming the death or that they are still alive. Mohammed Qays (talk) 07:33, 4 October 2025 (UTC)Reply
@Mohammed Qays Generally, semiprotecting items helps preventing real vandalism. Vandalism is not a word for "views that you disagree with".
"Only a official annoucement from the organization someone works for" seems to be a sourcing standard that's not our usual sourcing standard on Wikidata, so it sounds to me like you are trying to use your moderator rights to enforce a personal content preference without seeking consensus for it. ChristianKl ❪✉❫ 11:10, 4 October 2025 (UTC)Reply
@ChristianKl Thank you for the comment. I would like to clarify that the action is not related to any personal preference, but is based on clear reasons concerning the protection of the item from repeated vandalism and unstable edits, including those occasionally made by users with advanced rights.
According to established practice, the usual procedure in such cases is to restore the stable version prior to the vandalism, and then apply temporary protection until a community consensus is reached.
This is due to the existing disagreement regarding the death of the individual, and therefore, the purpose of the action is to maintain content stability and ensure data accuracy, with any proposed additions or modifications to be discussed on the talk page for further review.
In addition, other encyclopedic projects — including the Arabic Wikipedia — use Wikidata in their infoboxes, which makes the accuracy of this data directly affect the reliability of the content in those projects. Mohammed Qays (talk) 11:39, 4 October 2025 (UTC)Reply
The term vandalism means that a user intentionally crated a bad edit. It doesn't mean that the created an edit that you disagree with even if you would have good reasons for the disagreement. Which Wikidata users are you accusing of having created edits that they know to be bad in that item?
The way to handle disagreements in content is to have a discussion to find consensus about the disagreement. You seem didn't open any discussion on the talk page to find consensus on the question about how date of death should be modeled. Is there a discussion that you opened somewhere else to find consensus? ChristianKl ❪✉❫ 11:47, 4 October 2025 (UTC)Reply
@ChristianKl Thank you for your clarification. I appreciate your note, and I will make sure to open discussions on the talk page in the future whenever there is disagreement about content, in order to reach community consensus.
For clarification, the protection on the item was temporary and expired on 1 October 2025.
I would also like to point out that your comment included a degree of personalization, implying that I have a particular interest or personal views about this individual. In my opinion, this is not an appropriate approach for discussion and does not help in reaching constructive understanding.
Therefore, I prefer not to continue this type of exchange. Thank you for your understanding. Mohammed Qays (talk) 11:56, 4 October 2025 (UTC)Reply
Regarding the underlying content question here, it appears that someone (indeed many people) definitely died, but there remains uncertainty about whether the person who died is the same individual using this pseudonym.
A quick survey of the 26 Wikipedia projects with an article shows that (today) 13 state that he is dead, 4 report claims of his death with careful attribution, and 9 make no mention of it. Many in the last category are stub articles. This makes Arabic Wikipedia something of an outlier in providing substantial coverage yet with no mention of either death claims or identity claims.
Wikidata has to serve all client projects, even when they are inconsistent with each other. The appropriate response here is to report claims of the death with careful sourcing and use of statement supported by (P3680) and statement disputed by (P1310). There might also be a case for using a non-default rank, depending on the strength of sources.
Regarding the administrative actions, it does seem unusual to have used full protection in this instance, especially as it effectively supported the editorial position of one project among many. Raising questions about this is a normal and healthy part of community oversight and should not be taken either as personal criticism or as offence. Bovlb (talk) 21:19, 4 October 2025 (UTC)Reply
@Bovlb Thank you for your kind and precise reply. Best regards. Mohammed Qays (talk) 12:43, 5 October 2025 (UTC)Reply
@Mohammed Qays With power comes responsibility. If you use your admin power to silence estabilished users, looking at reasons why you do so is important. If your position is that you don't want to be held accountable for using admin powers like full protection, then don't use them. ChristianKl ❪✉❫ 12:34, 11 October 2025 (UTC)Reply

What's wrong with my Meetup.com Event ID proposal and how can I make it better.

[edit]
Latest comment: 6 days ago15 comments5 people in discussion

I'm confused about proposal here

I want to make it better. What can I do to get input on it? It's my goal to make Wikidata more friendly to major events that are organized online but exist in person. Things like Hackathons and Civic events.

Meetup.com Event ID authorities is just a taste of what I want. I'd love to have one page for something like an Anual Hackathon, https://www.wikidata.org/wiki/Q136147713

That has the meetup.com, eventbright, and devpost authorities.

It's hard to find prior examples, because Wikidata isn't currently very good at indexing events, but I want to help fix that.

As a side note, is there a way to run as an admin/moderator and get the ability to add/maintain properties myself (even if only just the ones I create), without going through this process? EvanCarroll (talk) 16:16, 25 September 2025 (UTC)Reply

Properties should ideally be relevant to existing Wikidata items. Your proposal only had one example and it appeared to be for a newly created item. It's not clear that the events you see this as a useful property for are likely to be added to Wikidata.
Wikidata has a property creator permission, but it, as with admin privileges, should only be used for properties you are not personally involved in. ArthurPSmith (talk) 17:15, 25 September 2025 (UTC)Reply
Please take a look at Wikidata:Property creation for the process we follow in property creation here. ArthurPSmith (talk) 19:22, 25 September 2025 (UTC)Reply
I'm asking for help. I did read that document. ;)
I'm not asking for the guidelines of notability to be changed here. Clearly annual events are _very_ often notable, and receive regional and national attention with major guests. For example, the Houston Hackathon is attended by the mayor. It was even announced by Mayor Parker's office in 2013.
https://web.archive.org/web/20130504231846/https://www.houstontx.gov/mayor/press/20130501.html
I don't really want to be put on the defense to defend the _content_ for notability. I just want to point out that if we don't add the plumbing for event authority we're missing out. I'll still create the events, even if I am the only one that does it. They're notable without establishing the platform that was used to facilitate them as an authority. EvanCarroll (talk) 20:30, 25 September 2025 (UTC)Reply
Can't you just add the two missing examples please? Trade (talk) 22:17, 26 September 2025 (UTC)Reply
Sure, I have to make something up. Because looking for them seems to be difficult. EvanCarroll (talk) 02:09, 28 September 2025 (UTC)Reply
If they are hard to find, what's your plan to provide extensive coverage on WD? Vicarage (talk) 00:37, 29 September 2025 (UTC)Reply
Build it and they will come. It's hard to find because Wikidata doesn't have very good support for events at all. I want to fix that. There are no authority ids for events currently. This is the first submission, but I have a few more lined up (facebook, and eventbrite should be next). EvanCarroll (talk) 17:58, 30 September 2025 (UTC)Reply
Done! I had to create events but that was always my plan. I just didn't see the point. It's all the same stuff. And it's not just us. There are tens of thousands of events worldwide created on Meetup.com. I thought the idea was demonstrated clearly with one example, but here are 2 more now. EvanCarroll (talk) 02:26, 28 September 2025 (UTC)Reply
@EvanCarroll: There are a lot of events on Meetup.com for which we likely don't want Wikidata items. This means that the sentence "There are tens of thousands of events worldwide created on Meetup.com" should make a reader suspicious about the proposal.
Generally, if you working on improving coverage of something like hackathon (Q46855) instances it's good to not only think about the new hackathons you want to add but also but those already in Wikidata. ChristianKl ❪✉❫ 18:02, 3 October 2025 (UTC)Reply
I've said many times I'm not looking to change the criteria for notability. This is clearly Wikidata property for an identifier that does not imply notability (Q62589320). EvanCarroll (talk) 19:54, 3 October 2025 (UTC)Reply
From that perspective the number of events at Meetup.com is irrelevant, yet you brought it up. The thing that matters is the number of events at Meetup.com that actually are notable for Wikidata. ChristianKl ❪✉❫ 01:05, 5 October 2025 (UTC)Reply
I brought it up to show the wide spread adoption of the platform and its potentially utility in whatever subset of it we wish to lean on. In the same way that one may bring up that OSM has a half-billion buildings without ever implying that they're all notable for Wikidata inclusion.
"The thing that matters is the number of events at Meetup.com that actually are notable for Wikidata", agreed fully! Are you asking that I audit Meetup.com to show that? I've already shown there are notable events on Meetup.com. I've also shown multiple examples. EvanCarroll (talk) 02:17, 5 October 2025 (UTC)Reply
Your proposal is not up to the standards of Wikidata in it's current form, so the system is designed to not let people create properties in that state. Your proposal has only one example. We generally want proposals to have at least three examples. As far as examples go, it's better to focus on examples on existing items that are already clearly notable than focusing examples on new items.
You filled in wrong information in the "Wikidata project" and "subject item" fields (taking a look at how other proposals fill out these kinds of fields is a good way to learn what they are for.
You just copied the label of the property into the description field instead of adding a description. For that look at how the description of other identifier (/ID) properties looks like and do something similar. Ideally, you add a few words to describe what meetup.com is as well in the description.
You capitalized "Event" in the label which deviates from the way most event IDs in Wikidata are labeled and did not provide a reasoning for why you deviated from the standard. In general, looking at the way that properties that are similar to the one you propose are structured is valuable to making a good proposal. ChristianKl ❪✉❫ 11:03, 26 September 2025 (UTC)Reply
I lowercased event in the label, and added the subsequent examples. I believe this fixes the problems? EvanCarroll (talk) 23:56, 28 September 2025 (UTC)Reply

Should I change the Wild Kratts episode numbering and order to that of the production codes?

[edit]
Latest comment: 5 days ago3 comments2 people in discussion

[continuing from my previous comment at archived discussion]

It is obvious to me that there’s no obviously correct order of Wild Kratts episodes, but the current numbering and order of them in Wikidata differs from both the production codes and the original air date order, which leaves me with the question: What system to adopt for Wikidata? Or should I remove the data altogether by setting series ordinal (P1545), follows (P155), followed by (P156) to unknown value?
— keepright! ler (talk) 10:34, 29 September 2025 (UTC)Reply

Production code would be a more appropriate value, as it is a better classification than publication dates QwertyZ34 (talk) 15:49, 30 September 2025 (UTC)Reply
Replying to prevent automatic archival … if no more options are suggested here until next weeks’s automatic archival, I’m going to change the Wild Kratts episode numbering and order so that it matches the production codes, as suggested by @QwertyZ34.
— keepright! ler (talk) 09:22, 6 October 2025 (UTC)Reply

How can Wikidata be useful IRL if it has less data than Wikipedia?

[edit]
Latest comment: 5 hours ago20 comments8 people in discussion

Currently, for many topics and applications, it would make more sense to query Wikipedia than to query Wikidata and to extract the data from there. This kind of negates the point of Wikidata.

Many people, software, projects, plants, diseases, films and organizations have more data in their infoboxes and categories than in their Wikidata item.

This could be images of the subject like a skin disease or the info which license a free software is under or which programming language it was programmed in.

Another more concrete example are spoken text audio (P989) – those are set in the articles of the Wikipedias but in Wikidata only English ones are widely set and even these are quite a way from being complete.

  • I think for Wikidata to be really useful and successful in the real world, such as for querying metadata about things like software or films, it needs to be AT LEAST as good as Wikipedia for the data and then move on from there to cover additional areas and contain additional data in more parseable formats.
  • One can also query Wikipedia in various ways so currently, it would often if not usually be more advisable to recommend or implement uses of APIs to query Wikipedia. I think striving to have all the data and cover the same applications when it comes to data as Wikipedia would be a great goal for Wikidata for now if it is to become successful in terms of people using it in practice (and by extension of many people being aware of it existing).

Currently, it seems like there only is one major tool to synchronize data from Wikipedia into Wikidata and from Wikidata into Wikipedia: Harvest Templates. That tool appears to be unmaintained, not used much, not known about much, there aren't many discussions or meta pages or coordination regarding it, and it has severe bugs and limitations. For example, it can't set a qualifier alongside a value and when trying to import spoken Wikipedia audios via the Wikipedia template for these it fails due to some bug after 30 seconds or so.

A nice thing about it is that people can share the configurations they use for importing data so other people could pick it up – https://pltools.toolforge.org/harvesttemplates/share.php Note that however, there are no indications for how much is still being done and whether the harvest is currently already regularly done or which language versions it could be used on too and so on. Here's the HT for Wikipedia audio versions that is also in that large list. You could use this for testing and to see the aforementioned bug. Those harvests are all specific to one language Wikipedia, one can't adjust it to import the narrated article audios from all the Wikipedias with a template for these so one would have to adjust the harvest for each and every of the Wikipedia of the hundreds of Wikipedias that have these...and then regularly update this manually every once in a while. It doesn't seem like anybody is doing or did this for most (well or all) Wikipedias. Likewise, Wikipedia has more images on skin diseases than the Wikidata items about the skin diseases and software items do not have the programming language info set despite that this metadata could be imported from for example GitHub and/or the Wikipedia categories.

I think being more useful than Wikipedia for some applications would be a second step after first outmatching Wikipedia on structured data.

This is not specific to this tool – it just seems like this would be the biggest way for equalizing data coverage of Wikidata and Wikipedia. Are there any further things or tools to know regarding this? How else than using and improving this tool could Wikidata improve on this and what could be done in regards to the former? In regards to the former, I made this wish in the Community Wishlist: Continue the development of the Harvest Templates tool that allows importing data to WData. Prototyperspective (talk) 00:15, 2 October 2025 (UTC)Reply

"Currently, for many topics and applications, it would make more sense to query Wikipedia than to query Wikidata and to extract the data from there. " There are hundreds if not thousands of infobox parameters which does not have an equivalent data property on Wikidata. That's the issue. There is also the issue that new data property proposals tends to face an heavy amount of scrutinization compared to other types of properties which can discourage users from trying to make new proposals. --Trade (talk) 00:30, 2 October 2025 (UTC)Reply
I am curious how to implement a query of Wikipedia, given the infoboxes are not structured : do you mean it's more useful to copy-paste article per article from a wikipedia than to run a wikidata query ? Bouzinac 💬●✒️●💛 08:57, 2 October 2025 (UTC)Reply
how to implement a query of Wikipedia, given the infoboxes are not structured since you got no reply and since Trade only mentioned this and I think only so in reference to what I said: Wikipedia can be queried using APIs and you can then extract the data in the infoboxes because they are structured by the predefined infobox parameters. Search the Web for example for wikipedia get data from infobox and you'll find many tutorials and premade tools for this using the Wikipedia API such as this post in 2012 with a premade tool to extract the infobox data. These are probably more viewed or popular on the Web than questions & tutorials about querying data from Wikidata. Additionally, there is a new dataset which has the data prestructured; see Matěj Suchánek's comment below. Just to clarify again the thread topic and intent: I think Wikidata would have to be at least as good as Wikipedia for structured data for this to change in the sense of (category&)infobox data there being all covered here. Prototyperspective (talk) 11:20, 5 October 2025 (UTC)Reply
Not sure what you mean by "data property proposals" vs other types of properties? External identifiers are the exception here, with slightly lower requirements for approval and creation.
More importantly we need more people regularly commenting on and supporting properties they want created; without clear support they won't be. See Wikidata:Property proposal/Overview for the current list, there are a lot still open with no comments in the last 3 months, many 6 months or more. ArthurPSmith (talk) 17:53, 2 October 2025 (UTC)Reply
I don't think the challenges or scrutiny faced by new data properties are a major factor – there already are very many properties but they aren't set on many items where Wikipedia has the data. The main example of one such was spoken text audio (P989) but this also applies to many other properties.
It may be an issue but it's not "the" issue and as far as I can see not among the main issues. Instead, people propose more and more properties but there is little work and consideration of how to get the data for these into Wikidata, not even when lots or all of that data is readily available in other Wikimedia projects. Prototyperspective (talk) 17:55, 2 October 2025 (UTC)Reply
"there is little work and consideration of how to get the data for these into Wikidata" There's plenty of people who do want the data into Wikidata but there are much less people willing to "lend" out their bots by giving it additional tasks Trade (talk) 22:05, 3 October 2025 (UTC)Reply
That doesn't or wouldn't change anything – so if you're arguing there is not little consideration but moderate or even much consideration of that, I couldn't see any evidence or indications of that and would welcome any info suggesting to the contrary. Moreover, these aren't two separate groups – if people want to have the data, they could for example get Harvest Template to work or ask about HT and at least document and discuss its issues or build a bot or ask for bots like that. Prototyperspective (talk) 22:42, 4 October 2025 (UTC)Reply
Combining data from multiple sources is a useful skill, more people should get into the habit of doing this. DBpedia might have more data imported from Wikipedia templates (e.g [1]), and thanks to sitelinks there should be no problem combining this with data from Wikidata in a query. Infrastruktur (talk) 10:21, 2 October 2025 (UTC)Reply
On a related note, Wikimedia Enterprise has recently started publishing datasets of Wikipedia's "parsed content", including infoboxes (see Tech News). I wonder if this is something where we could go for missing data and import them to Wikidata.
However, the habit of copying Wikipedia infoboxes to Wikidata has been somewhat controversial because of its error rate and lack of references to sources (or means to import them). On Wikidata, we can take a different approach and import directly from external sources (public registries, etc.), and we also indicate them as the source. Of course, this is possible on Wikipedia, too, but AFAIC imports like this are not done often. On Wikidata, this is "by design".
Also, queries over Wikipedia are more difficult, inefficient, or the available data varies by Wikipedia editions (e.g., categorization trees). Nothing comparable to WDQS is available for Wikipedia.
So to answer your question how can Wikidata be useful IRL if it has less data than Wikipedia: 1) by offering actual structured data, 2) by offering quality, 3) by offering efficient means for aggregating results over many data entries. --Matěj Suchánek (talk) 11:17, 2 October 2025 (UTC)Reply
  • publishing datasets of Wikipedia's "parsed content", including infoboxes (see Tech News). I wonder if this is something where we could go for missing data and import them to Wikidata. That's a great idea! Has somebody looked into this? This may enable importing more data more quickly more easily in comparison to Harvest Templates. Please also let me know if you know of any resources that are about importing data to Wikidata from these dumps.
  • Good point. 1. many people also edit items by hand (and without adding a source) 2. many values in Wikipedia do have sources which could be imported too 3. for many values like again spoken Wikipedia audios there is no need for a source as it would make no sense 4. if there are some flaws from time to time, these are quite rare on at least ENWP and other sources where data is being imported from can also be flawed; most importantly though if the value gets edited some tool could show a mismatch so that it can be fixed in Wikipedia or Wikidata wherever the error is.
  • There's the Wikipedia API and probably more questions, demos, and projects online to get data from Wikipedia infoboxes etc than from Wikidata even though querying Wikidata via sparql is often more convenient (despite that it can quickly run into timeouts and sparql is not sth many people know much about). That the available data varies per language is unimportant as, depending on the data sought, people can simply import from the largest Wikipedia, ENWP. They aren't necessarily more difficult or inefficient; people can use premade packages for making Wikipedia API calls and if you get less data with WD that is still less efficient even when WDQS is more designed for these kinds of things.
  • 1) Wikipedia also has structured data in infoboxes and categories. Here's just more and basically only structured data and things built with queryability in mind. That it has structured data doesn't answer the question as is; if the data is too incomplete it can be as structured as it can be but not enable any real-world uses 2) Wikipedia is checked by far more people and other databases have quality too so I wonder how that would answer the question – it's more like a goal or positive aspect; WD is not super reliable and how millions of for example book and study items with essentially no watcher can be all effectively protected from vandalism may be a subject for another day 3) appreciate the answer but here again what is missing is the real-world practical part – this may be great in theory but in practice even the best efficient means for aggregating results over many data entries are of no use if the data is very incomplete. The emphasis is on the use in practice, secondarily in specific uses that are of interest to many people and only third about Wikidata in terms of technical tool / potential aspects.
Prototyperspective (talk) 21:48, 2 October 2025 (UTC)Reply
"many people also edit items by hand (and without adding a source)" Perhaps Wikidata should make editing items by hand less cumbersome then? Trade (talk) 22:12, 3 October 2025 (UTC)Reply
You can't take out tiny parts completely out of context like that. I said this in the context of Suchánek's claim about error rate of data in English Wikipedia infoboxes. This doesn't address anything I said and Wikidata editing isn't cumbersome I think except that the add statement button shouldn't just be at a variable place at the bottom. Prototyperspective (talk) 22:38, 4 October 2025 (UTC)Reply
Well, since there appareantly isn't an issue with manual editing or adding sources being cumbersome i guess there's nothing for us to discuss then?
I guess the issues i run into must be a figment of my imagination and i'm just being lazy then Trade (talk) 00:51, 5 October 2025 (UTC)Reply
Your comments really aren't constructive. If you personally find there is nothing to discuss, I'd suggest to not comment. The comment by Suchánek whose thread you derailed by bringing up something entirely unrelated and unconnected to the thread topics was precisely ontopic, insightful and interesting. Prototyperspective (talk) 11:13, 5 October 2025 (UTC)Reply
it's more like a goal or positive aspect Yes, maybe this describes my view better.
are of no use if the data is very incomplete I agree. It's worth noting, though, that we can use it to measure (in)completeness, too.
about error rate of data in English Wikipedia infoboxes In infoboxes in general. Perhaps English Wikipedia has higher quality, yet it doesn't cover all items. If true, this would be a good motivation for trying to import these dumps.
let me know if you know of any resources Unfortunately, I'm not aware of any related resources or tools. OpenRefine? Mismatch Finder? --Matěj Suchánek (talk) 11:39, 5 October 2025 (UTC)Reply
Any effort to import data from infoboxes should be avoided unless the citations used in those infoboxes are imported as well. Too often I see large imports using imported from Wikimedia project (P143) as a reference. Citing the Wikipedia article is as bad as having nothing at all, not only because the data may change and we're left with an incongruity, and because those imports can sit for years without being properly reviewed and cited, but also for the same reason Wikipedia:Wikipedia is not a reliable source (Q22964187) exists on multiple sites. — Huntster (t @ c) 13:07, 5 October 2025 (UTC)Reply
that we can use it to measure (in)completeness, too If one could measure it that way by comparing both data then that would probably usually already 3/4th of importing the Wikipedia data into Wikidata. I guess except if one takes a small sample and compares the data for these. Perhaps English Wikipedia has higher quality, yet it doesn't cover all items It depends on the subject for many cases (and one could also limit the imports to these), it has like >95% of items across the Wikipedias. Basically one could limit infobox data imports to the global Wikipedia which is ENWP. Mismatch Finder? Interesting, didn't know about it and if that tool can't do this yet this seems like a perfect candidate tool for functionality to find mismatches and missing data in Wikidata when comparing the linked Wikipedia article(s). Prototyperspective (talk) 23:30, 5 October 2025 (UTC)Reply
My attempts to use wikipedia infoboxes generally fail as there is huge inconsistency in the way infoboxes are coded on each page, with attempts to parse fields scuppered by inconsistent units and endless inline templates. The raw wikicode is hard to parse, the formatted html is easier, but mostly other sources that just render databases are better still. And unlike here there is no attempt make consistent infoboxes across the whole of human knowledge, its very balkanised. Vicarage (talk) 18:16, 2 October 2025 (UTC)Reply
It would be very useful if you could document the difficulties you faces as well as the methods (including the tools) you used. In my limited experience templates simply had parameters that were set in the articles in consistent ways with no embedded templates other than refs and these kinds of [sup notes] which I think could be excluded via regexing away {{xyz}} basically. Interesting to hear from your experience with this.
I'll make a page about Wikipedia <-> Wikidata data. Prototyperspective (talk) 11:51, 11 October 2025 (UTC)Reply

area (P2046)

[edit]
Latest comment: 6 days ago6 comments3 people in discussion

Q134166256#P2046

Which of these two statements am i supposed to mark as preferred?--Trade (talk) 00:24, 2 October 2025 (UTC)Reply

Generally, I would suggest that metric be preferred since that is so much more widely used. FYI, I corrected both figures as they were the length rather than area. — Huntster (t @ c) 02:19, 2 October 2025 (UTC)Reply
Why did you removed non-operating temperature (P5067)? Trade (talk) 03:09, 2 October 2025 (UTC)Reply
Sorry for not responding sooner. This was removed because the property is wrong for what the source is describing. The property is the bearable temperature when a device is turned off, but the source is describing its maximum operating temperature. — Huntster (t @ c) 22:45, 4 October 2025 (UTC)Reply
I've encountered a somewhat similar issue with dates: the choice between Julian and Gregorian calendars. An example is Euler's death date (see Q7604#P570): he died in Russia, so a Julian death date would be more logical. I believe the best approach would be to allow both variants and explicitly indicate their equivalence somehow, but I don't think that is currently possible. Difool (talk) 03:58, 2 October 2025 (UTC)Reply
What about the constraint error? Trade (talk) 06:23, 2 October 2025 (UTC)Reply

When child (P40) is a name

[edit]
Latest comment: 5 days ago11 comments5 people in discussion

170 people have child (P40) pointing to female given name (Q11879590). 

SELECT ?item ?itemLabel WHERE {
  ?item wdt:P40 ?child.
  ?child wdt:P31 wd:Q12308941
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],mul" }
}
Try it!

230 have a child who is a male given name (Q12308941). 

SELECT ?item ?itemLabel WHERE {
  ?item wdt:P40 ?child.
  ?child wdt:P31 wd:Q12308941
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],mul" }
}
Try it!

Obviously these statements are all false, although I can see that when used in infoboxes they might generate what looks like correct information. Would it be proper to delete all these statements, or could they be mined somehow (qualifiers? number of children (P1971)?) for information worth preserving? MartinPoulter (talk) 21:03, 3 October 2025 (UTC)Reply

Maybe we should just have a "child name string" property to when the child isn't notable? Trade (talk) 22:23, 3 October 2025 (UTC)Reply
Would something like child (P40)no valuegiven name (P735)Sue (Q16281769) work, perhaps? — Huntster (t @ c) 22:35, 3 October 2025 (UTC)Reply
No, that would imply they are childless Trade (talk) 23:15, 3 October 2025 (UTC)Reply
Also that wont show up in infoboxes anyways so what would the benefit even be? Trade (talk) 23:16, 3 October 2025 (UTC)Reply
1) That is true. Unfortunately our options are limited in that regard. 2) We aren't here to cater to infoboxes. Their code can be changed if they want to support a given value. 3) While a new property might have value, what would be more useful would be a fourth drop-down menu option in addition to 'no value', 'unknown value' and an item value: text string value. There are many times where simple text would suffice where an entirely new item would be unwarranted. — Huntster (t @ c) 23:51, 3 October 2025 (UTC)Reply
You can use object named as (P1932) as is done at Q117053000#P26 or Q124417003#P40. Difool (talk) 02:11, 5 October 2025 (UTC)Reply
That's not really a great solution. The statement implies that the spouse is unknown even though they are not Trade (talk) 05:13, 5 October 2025 (UTC)Reply
Wikidata:Data model#No value and unknown value mentions it can mean "the value is a known object, but there's no Wikidata item about the object (perhaps because it's not notable)." Difool (talk) 07:04, 5 October 2025 (UTC)Reply
What you've brought up seems to be an ideal solution. Correcting User:Trade, the statement implies that the spouse has no distinct Wikidata representation but that a fact is known about them. That's a true implication. MartinPoulter (talk) 10:06, 5 October 2025 (UTC)Reply
As Huntster said, but use 'unknown value' not 'no value' - this is exactly what 'unknown value' is for. ArthurPSmith (talk) 15:58, 6 October 2025 (UTC)Reply

Guillaume Guillaume V, Grand Duke of Luxembourg (Q161319)

[edit]
Latest comment: 7 days ago4 comments2 people in discussion

Should Guillaume V, Grand Duke of Luxembourg (Q161319) be described as Grand Duke of Luxembourg (Q3276071), Monarch of Luxembourg (Q113956811) or both? Particularly as the first should be a subclass of the second? 96.79.116.169 01:42, 4 October 2025 (UTC)Reply

Ok, I see now that "Grand Duke" is a "noble title" while "Monarch" is a "position held". The problem is that Grand Duke of Luxembourg (Q3276071) is showing "inverse constraint": "Henri, Grand Duke of Luxembourg should also have the inverse statement position held Grand Duke of Luxembourg." That goes for all of the Grand Dukes listed on that item. What should we do to get rid of the "inverse constraint" error? 96.79.116.169 01:49, 4 October 2025 (UTC)Reply
Perhaps don't double the information? We have queries for a reason. Sjoerd de Bruin (talk) 09:15, 4 October 2025 (UTC)Reply
@Sjoerddebruin I don't understand what it should be. I didn't set up the all the statements on Grand Duke of Luxembourg (Q3276071) say "inverse constraint" and that shouldn't be there because the issue should be fixed. 96.79.116.169 16:42, 4 October 2025 (UTC)Reply

instance of human and instance of racist

[edit]
Latest comment: 5 days ago6 comments5 people in discussion

https://www.wikidata.org/w/index.php?title=Q2902602&diff=prev&oldid=2380889023

Are there standard property constraints marking instance of role on human items? EyuOrg (talk) 10:19, 4 October 2025 (UTC)Reply

Its bad form to make a change and then raise the issue for discussion. Have some respect for the process by waiting until others have expressed their opinions. Vicarage (talk) 10:37, 4 October 2025 (UTC)Reply
Are you hunting me (see section "Can an organisation have a subclass statement")?
There is nothing bad in removing a false statement, especially if provably false and like here a statement on a human.
The issue raised here, is related to a missing constraint violation sign, not an issue regarding the statement itself. EyuOrg (talk) 10:45, 4 October 2025 (UTC)Reply
The statement in question seems clearly vandalism by an unregistered user. Removing it seems very sensible.
EyuOrg's question looks to me like it's asking about whether we can crate property constraints to prevent these kind of statements. The answer is that "if P31 Q5 -> no other P31" is not something that can be expressed in the current constraint system. ChristianKl ❪✉❫ 11:39, 4 October 2025 (UTC)Reply
The revert by EyuOrg was correct. On to the matter, I am not sure if we have a property for this.
We can also research what users have done with similar characteristics. For example, antisemite (Q15954665) is a subclass of racist (Q28528178), and there is a plenty of Qxxxoccupation (P106)antisemite (Q15954665) claims. But is it really an occupation?
--Matěj Suchánek (talk) 11:55, 5 October 2025 (UTC)Reply
It's a constraint violation, P1142=Q22649 should be used. There are already complex constraint violations on P31 for some things, I don't know if it catches these values. I've removed many invalid P31 values on Q5 items; there are some exceptions (instances of immortal (Q1207451) and one of legendary hero (Q12334344), and several values on Diva Depressão (Q104450771) which seems to be combine multiple things in one item). Peter James (talk) 19:18, 5 October 2025 (UTC)Reply

Which is the best way to edit disambiguous page?

[edit]
Latest comment: 7 days ago2 comments2 people in discussion

According to the Vietnam Wikipedia, Thái Hòa (Q136434782) comprises of Taihe (Q417144), Q11555696 and so on. However when I did Special:Diff/2412510333, Wikidata noticed that I should use different from (P1889). There are lots of examples between alphabetic writing and Chinese characters. Please help. Thanks.--迴廊彼端 (talk) 10:35, 4 October 2025 (UTC)Reply

Actually maybe merge Thái Hòa (Q136434782) into Taihe (Q417144)? DinhHuy2010 (talk) 10:42, 4 October 2025 (UTC)Reply

Confusion regarding descriptions of US presidents

[edit]
Latest comment: 6 days ago5 comments3 people in discussion

US presidents has confusing descriptions, because in some of the descriptions values, the word "president" is either capitalized or lowercased:

Examples of presidents that has p lowercased: William McKinley (Q35041): president of the United States from 1897 to 1901; William Howard Taft (Q35648): president of the United States from 1909 to 1913; Ronald Reagan (Q9960): president of the United States from 1981 to 1989 and actor (1911–2004); Barack Obama (Q76): president of the United States from 2009 to 2017, etc.

Examples of presidents that has p capitalized: James K. Polk (Q11891): President of the United States from 1845 to 1849; Benjamin Harrison (Q35678): President of the United States, 1889-1893; Richard Nixon (Q9588): president of the United States from 1969 to 1974; George W. Bush (Q207): President of the United States from 2001 to 2009, etc.

Current US president Donald Trump (Q22686) has for description: "president of the United States (2017–2021, 2025–present)"; which is probably the favored format, should we change every US presidents en descriptions, like changing for Ronald Reagan (Q9960) "from 1981 to 1989" to "(1981–1989)"? Also, I believe that George Washington (Q23): Founding Father and first U.S. president (1789–1797) don't need to be changed, it's like Trump–give me your opinion, which is better?:

1. president of the United States (start time–end time)
2. President of the United States (start time–end time)
3. president of the United States from start time to end time
4. President of the United States from start time to end time
5. ranking President of the United States from start time to end time (ranking example: 26th)
6. ranking president of the United States from start time to end time
7. ranking President of the United States (start time–end time)
8. ranking president of the United States (start time–end time)
9. president of the United States from start time to end time (birth date–date death)
10. President of the United States from start time to end time (birth date–date death)

See query for US presidents with their corresponding (en) descriptions QwertyZ34 (talk) 13:39, 4 October 2025 (UTC)Reply

Isn't "President of the United States" a title that would be capitalized? ChristianKl ❪✉❫ 16:34, 4 October 2025 (UTC)Reply
Preferably, yes. President of the United States (Q11696) is a title that would be capitalized–most of sources recognizes this term as capitalized, so choices 2, 4, 5, 7, and 10 would be preferred as descriptions. QwertyZ34 (talk) 16:43, 4 October 2025 (UTC)Reply
Isn't that only if it is part of a name, like Sir David Attenborough and President of the United States Donald Trump, see this Chicago Manual of Style FAQ. I think Chicago Manual of Style recommends "Abraham Lincoln (1809–1865), president of the United States, 1861–1865" in running text, so () for life spans and no () for office terms. I am not sure how to put that in a description without the name. Difool (talk) 02:34, 5 October 2025 (UTC)Reply
For titles, "President of the United States" is capitalized, e.g. see https://digital.library.unt.edu/ark:/67531/metapth1136096/; for descriptive phrases, "president of the United States" is lowercased—see w:President of the Unites States which has for first sentence: "The president of the United States (POTUS)…", as descriptions are describing the item in general, and are descriptive, I believe that president should be lowercased in description, also like Abraham Lincoln (Q91): President of the United States from 1861 to 1865, it should:
be lowercased;
nothing to change for office term;
add life span (between brackets)
@Difool may tell us more and inform us more on this topic, I think "president of the United States from 1861 to 1865 (1809–1865)" is the best format. QwertyZ34 (talk) 11:48, 5 October 2025 (UTC)Reply

Names of European cities in different languages (C–D) (Q6961631)

[edit]
Latest comment: 5 days ago1 comment1 person in discussion

@Quesotiotyo: is restoring incorrect capitalised descriptive name for Names of European cities in different languages (C–D) (Q6961631) while it is not proper noun. Eurohunter (talk) 20:53, 5 October 2025 (UTC)Reply

Format for census population data

[edit]
Latest comment: 1 day ago4 comments3 people in discussion

Hello,

As my first substantial Wikidata contribution project, I had the idea of using QuickStatement to add population numbers for Bhutan, which are currently only sporadically present on Wikidata. I'm starting with the 20 districts of Bhutan and then will work on the lower-level subdivisions (towns and gewogs).

I've gathered 2005 and 2017 population data for all the 20 districts of Bhutan, including breakdowns by female vs. male and urban vs. rural. I've already figured out the syntax of QuickStatement. However, before I submit the batch, I'd like some clarifications on how to format the qualifiers and source information:

  • As a subpoint to this, Arunachal Pradesh (Q1162) uses an interesting variant of this approach. It provides a top-level urban population (P6343), but then has female population (P1539) as a qualifier to that, allowing one to see the detailed breakdown of urban & male, rural & female, etc. The Bhutanese census data is broken down similarly, so I could borrow this approach, but I'm not sure whether it's recommended.

So far, this is the basic format I have come up with:

Trashiyangtse District (Q600200)male population (P1540)8,861point in time (P585)2005determination method or standard (P459)Population & Housing Census of Bhutan 2005 (Q136443651)stated in (P248)Results of Population & Housing Census of Bhutan 2005 (Q136443667)

Thank you, Grešla (talk) 06:33, 6 October 2025 (UTC)Reply

For the second point, use the report item. Also don't forget to set retrieved (P813). --Matěj Suchánek (talk) 07:38, 6 October 2025 (UTC)Reply
@Grešla: Thanks for asking here. I think a more specific value for determination method or standard (P459) is preferred. I haven't seen the approach used on the Tibet Autonomous Region (Q17269) example for subsets of the population; I think you could go either way, but that approach makes sense to me and doesn't seem to violate any constraints. The "basic format" you list looks good. ArthurPSmith (talk) 16:05, 6 October 2025 (UTC)Reply
@Matěj Suchánek, ArthurPSmith: I submitted the batch based on your feedback. I hope it looks okay. I could also add literate population (P6499), illiterate population (P6498), and literacy rate (P6897) but I'm waiting for feedback on #Indicating an age range for a demographic property. Grešla (talk) 21:24, 9 October 2025 (UTC)Reply

Wikidata weekly summary #700

[edit]
Latest comment: 5 days ago1 comment1 person in discussion
Here's your quick overview of what has been happening around Wikidata in the
week leading up to 2025-10-06. Missed the previous one? See issue #699.

Events

  • Upcoming events:
    • WikidataCon 2025 - The first version of the conference schedule has been released! See the Event page for links to the schedule (and to register if you haven't already).
    • 13th Birthday Events - check the calendar to see which Birthday events are happening this week!
    • Webinar: Embedding Project - join us this week, October 9 1600 ~ 1700 UTC, to hear a presentation and then pose your questions regarding the Embedding Project
    • The next Wikidata+Wikibase office hours will take place on Wednesday, 18:00 CEST, 15th October 2025 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.
  • Past events:
    • Wikimedia CEE Meeting 2025 - Learning Day (September 25, at the Goethe Institut in Thessaloniki): the main topic was Wikidata and libraries; the program and the materials of the hands-on session are available in the Meta page, the slides of the presentations are available in the Commons category
    • Convegno NILDE (October 2-3, University of Genoa; website): in the first session, regarding Il futuro dei cataloghi [The future of the catalogs], many presentations (in Italian) have mentioned the use of Wikidata in collective library catalogs; the proceedings of the conference (in Italian) will be published in open access in early 2026:
      • Lianna D'Amato and Emanuela Secinaro, Cobis LOD: integrare e aumentare i cataloghi eterogenei delle biblioteche specialistiche [Cobis LOD: Integrating and Enhancing the Heterogeneous Catalogs of Specialized Libraries] (slides)
      • Stefano Bargioni, La catalogazione per entità, esperienze pratiche della rete URBE con Parsifal [Entity cataloging: practical experiences of the URBE network with Parsifal] (slides; Wikidata project page)
      • Elena Ravelli e Camillo Pellizzari, Authority file di SBN e Wikidata: un esempio di interoperabilità e riutilizzo dei dati [SBN Authority File and Wikidata: An Example of Data Interoperability and Reuse] (slides; Wikidata project page)

Press, articles, blog posts, videos

Tool of the week

  • Script newentity.js is a UserScript by GZWDer it helps to generates new items (including lexemes) from the given JSON object.

Other Noteworthy Stuff

  • The Wikidata Team at Wikimedia Deutschland would like to talk to developers who use Wikidata’s data dumps, or have tried using them but ran into challenges. Your feedback will help us make data dumps easier and more reliable to use. If you’re open to chatting, we’ll schedule a 60-minute user interview. As a thank-you for your time, you’ll be eligible for compensation. Sign up here.
  • For anyone interested in open and interoperable data, the DCMI 2025 Conference might be interesting. More information and registration here.

Newest properties and property proposals to review

You can comment on all open property proposals!

Did you know?

Development

  • The {{USERLANGUAGE}} magic word is now enabled on Wikidata and Test Wikidata. It can be used to display templates in the user interface language, replacing the previous {{int:lang}} hack. (phab:T405830)

GraphQL: We have continued working on the functionality that provides labels of linked entities

  • Wikidata vector embedding and MCP: We have released the vector embedding and MCP
  • Wikidata integration in Wikipedia and co:
    • We are continuing the rollout of improvements to ow we track Wikidata changes in Wikipedia's watchlist and recent changes in order to reduce the size of the database table and reduce the noise on editors' watchlists. (phab:T401288, phab:T401290)
    • We are working on some improvements to the Databox module and template (phab:T400322)
  • Mobile statement editing:
    • We improved the display of qualifiers and made it possible to add qualifiers (editing will still come) as well as the value type (value/some value/no value) (phab:T404991, phab:T404956, phab:T402431, phab:T403246)
    • We are working on adding editing support for more datatypes (tabular data and geoshape) (phab:T403973)
    • We are working on making it possible to edit references (phab:T402433)

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.

Weekly Tasks

Request for help about query service

[edit]
Latest comment: 4 days ago4 comments3 people in discussion

I want to list all the universities in California using query service, but there is a problem I cannot resolve. Using instance of (P31) and located in the administrative territorial entity (P131) to filter, not all the universities are listed. For example, for University of California, Berkeley (Q168756):

If P31 refers to not university but another subclass like public university or public research university, query cannot be succesful. The same thing applies to P131. How can I solve this? 

SELECT ?item ?itemLabel WHERE {
  ?item wdt:P31 wd:Q3918;
        wdt:P131 wd:Q99.
  
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],tr,en". }
}
Try it!

Karacehennem (talk) 18:48, 6 October 2025 (UTC)Reply

This query returns the item. I tried with "wdt:P131+" instead of the manual three-level rollout, but it just timed out. Also, did you know about Wikidata:Request a query?
SELECT ?item ?itemLabel WHERE {
  ?item wdt:P31/wdt:P279* wd:Q3918;
        wdt:P131|wdt:P131/wdt:P131|wdt:P131/wdt:P131/wdt:P131 wd:Q99.
  
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],tr,en". }
}
Try it!

Bovlb (talk) 20:39, 6 October 2025 (UTC)Reply

(edit conflict) @Karacehennem: You can use an asterisk (*) to build a recursive property path (described in Wikidata:SPARQL tutorial). For example, ?item wdt:P131* wd:Q99 will select any item that is located in California, or located in an item located in California, or located in an item located in an item located in California, etc.
Similarly, ?item wdt:P31/wdt:P279* wd:Q3918 will select any item that is either an instance of a university, or an instance of a subclass of a university, etc. Thus:
SELECT ?item ?itemLabel WHERE {
  ?item wdt:P31/wdt:P279* wd:Q3918;
        wdt:P131* wd:Q99.
  
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],tr,en". }
}
Try it!
However, this is understandably more expensive. The query above times out. You must apply query optimization techniques to get it to work, e.g. reversing the order of the path:
SELECT ?item ?itemLabel WHERE {
  wd:Q3918 ^wdt:P279*/^wdt:P31 ?item.
  ?item wdt:P131* wd:Q99.
  
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],tr,en". }
}
Try it!

Grešla (talk) 20:43, 6 October 2025 (UTC)Reply

@Bovlb, @Grešla, thank you for your responses. I am just trying to learn how to write better queries, your answers helped me. Karacehennem (talk) 07:42, 7 October 2025 (UTC)Reply

Are they the same entities?

[edit]
Latest comment: 2 days ago4 comments2 people in discussion

Q131896022 and Q131640686 . Same name for companies, possibly different countries? I've pinged creators but they haven't replied to me.

Outside coutry not matching, there is no contradictory information. Could it be the same company registered in two countries? Piotrus (talk) 09:10, 7 October 2025 (UTC)Reply

u-blox AG (Q131896022) provides an official website (P856) and an EU Transparency Register ID (P2657).
U BLOX AG (Q131640686) provides an OpenCorporates ID (P1320) and a SIREN number (P1616).
All four of these sources list the address Zürcherstrasse 68, Thalwil, Switzerland, so I think it's safe to say that both entities represent the same company.
Do we need separate items for the company's branches in these different countries? I would say no. I'm not seeing any indication that u-blox's French operations are particularly distinguished from their headquarters in Switzerland, considering that they are still listed using their Swiss address and the Swiss business terminology (AG = Aktiengesellschaft), and fall under the French legal category of "Foreign company not registered with the RCS". Grešla (talk) 09:25, 7 October 2025 (UTC)Reply
@Grešla Thanks. Since one creator ignored my comment on their talk, and the other is inactive, I'd suggested merging them. Piotrus (talk) 16:15, 7 October 2025 (UTC)Reply
@Piotrus: There's also U-blox (Q2466238), which is about the same company.
Note: The Wikipedia article U-blox refers to U-blox Holding AG, which is the holding company for U-blox AG and thus technically a distinct entity, as discussed on page 41 of this PDF. We currently don't seem to have an entity for the holding company. Grešla (talk) 16:29, 9 October 2025 (UTC)Reply

Request for undeletion: Besart Haskaj

[edit]
Latest comment: 4 days ago3 comments3 people in discussion

Hello admins,

I would like to request the undeletion of the Wikidata item for Besart Haskaj, which was deleted earlier for notability reasons.

Besart Haskaj is a verified entrepreneur and co-founder of **AB Webstudios**, covered by multiple reliable and independent sources:

The subject meets Wikidata’s notability guidelines, and the links above provide verifiable coverage. Abdulsomodoladeji (talk) 10:47, 7 October 2025 (UTC)Reply

The fact that you use an advertisement (/ paid content) as independent sources sounds to me like either you don't know what you are doing or you purposefully try to mislead us. Therefore, I reject your request as I don't like to be lied to. ChristianKl ❪✉❫ 12:15, 7 October 2025 (UTC)Reply
Relevant items: Q136261787 (Abdulsomodoladeji). CC deleting admins @Fralambert
OP has contributed three other items, all German entrepreneurs, one being the other co-founder. Bovlb (talk) 14:05, 7 October 2025 (UTC)Reply

Notability of relatives

[edit]
Latest comment: 3 days ago5 comments3 people in discussion

Hi, I want advice on what makes relatives of notable people Wikidata:notable. Is being a child of a notable person a valid sitelink? The context is Edith Durham (Q287994)'s eight siblings, who are listed in Albania's Mountain Queen: Edith Durham and the Balkans (1st edition) (Q136439249). I found Herbert Durham (Q19654832), Frances Hermia Durham (Q18922236), Florence Margaret Durham (Q15046128) and their father Arthur Edward Durham (Q15637169) already had items and I have created items for Arthur Ellis Durham (Q136439309), Beatrice Bateson (Q136442645) and Frank Rogers Durham (Q136447490) based on obituaries. That leaves their mother, Mary daughter of William Ellis (Q8008557), who is mentioned in accounts of the daughters' early lives; Alice Lilian (Lilla) who appears in several books about the suffragette movement (and translated a book on Joseph Joachim (Q159976)); and Ellen Lucy (Nellie), who married a solicitor, who was one of the founders of King Alfred School (Q6411230)). TSventon (talk) 15:53, 7 October 2025 (UTC)Reply

Generally, according to our notability policy children and parents of other notable people are notable because of the structural need provision. ChristianKl ❪✉❫ 16:24, 7 October 2025 (UTC)Reply
This seems a can of worms to me. Its not really remarkable that human beings all have parents and many have children, and if there was interest in someone notable breeding or not, then number of children (P1971) would satisfy the structural need if they weren't notable. We are constantly rejecting self-promoting people, do we really want their family trees if they get a toehold? Vicarage (talk) 16:56, 7 October 2025 (UTC)Reply
Presumably when to apply the structural need provision is a matter of judgement. Adding two less notable siblings to seven notable siblings is probably more useful than adding nine non notable children to a notable parent. TSventon (talk) 17:12, 7 October 2025 (UTC)Reply
@Vicarage what I'm saying matches how we interpreted the notability policy in the past. Generaological data has it's value and I don't remember it having causes any problems. (Expect maybe the peerage import, that I don't think was justified via structural needs but via peerage)
When it comes to living people, it can also make sense to not capture data on minors for privacy reasons.
Wikidata cares about linked data. Interlinked items don't stand on their own. ChristianKl ❪✉❫ 18:16, 7 October 2025 (UTC)Reply

hair color (P1884) and eye color (P1340) and shades of colors

[edit]
Latest comment: 5 hours ago4 comments2 people in discussion

Is there any guidelines for how specific the values are allowed to be? Are we allowed to use specific shades as values? Like if someone have hair or eyes that is Dodger blue (Q856862) can we use that as a value instead just simply blue hair (Q4930092)? Trade (talk) 01:44, 8 October 2025 (UTC)Reply

Both currently has a one-of-constraint that suggests it should be one of a list. eye color (P1340) has the constraint deprecated so that it's only supposed to provide suggestions, hair color (P1884) however currently doesn't, so has a suggestion that specific values should be used. From my perspective deprecating the constraint on hair color (P1884) would make sense, but it's not something I feel strongly about.
When it comes to more specific guidelines, we don't have that for a lot in Wikidata. One step to move in that direction in cases like that is to have the discussion about a property at the property talk page so that users can more easily find it in the future and the discussion becomes something like a guideline. ChristianKl ❪✉❫ 11:39, 8 October 2025 (UTC)Reply
Is there a compelling reason why we need to manually whitelist every eye or hair color in the constraints? Trade (talk) 19:43, 8 October 2025 (UTC)Reply
@Trade One reason it is done is so that you can automatic completions of eye/hair color when typing. One-of-constraints are as far as I remember currently the only way to give Wikidata a list of things to autocomplete. black (Q17244465) and black (Q23445) are two different items, so the current autocompletion helps with selecting the right item. ChristianKl ❪✉❫ 11:48, 11 October 2025 (UTC)Reply

Trade names as properties versus individual items

[edit]
Latest comment: 3 days ago2 comments1 person in discussion

I noticed an apparent inconsistency for some pharmaceuticals:

So what is the right approach here? As far as I can see, these four trade names should be treated in the same manner. --Njardarlogar (talk) 04:30, 8 October 2025 (UTC)Reply

Also, what is the distinction between trade name (Q1417728) and pharmaceutical product (Q28885102) in this context? Should e.g. all pharmaceutical product (Q28885102) also be instances of trade name (Q1417728) (using instance of (P31))? Cf. Jardiance (Q29006055), which seems to me like it should also be treated in the samme manner as the four trade names above. --Njardarlogar (talk) 04:39, 8 October 2025 (UTC)Reply

Change image for Q21010069

[edit]
Latest comment: 1 day ago3 comments3 people in discussion

Hello, I’m the person this article is about (Gaz / Gary Alazraki). I have uploaded a replacement image on Commons under CC BY-SA. Could someone with rights please update the image from File:Gary_Alazraki,_September_2017_(2322)_(cropped).jpg to File:Gaz_Alazraki_2025.jpg Thank you! Gazalazraki (talk) 05:57, 8 October 2025 (UTC)Reply

✓ Done at https://www.wikidata.org/w/index.php?title=Q21010069&oldid=2413960491 DinhHuy2010 (talk) 11:43, 8 October 2025 (UTC)Reply
I've reverted, because the image appears to be a copyvio, having been posted online in 2024. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:36, 9 October 2025 (UTC)Reply

Please log in to view other pages

[edit]
Latest comment: 2 days ago2 comments2 people in discussion

i am shocked to see this, when i search a word while logged out. since when did this become so user-unfriendly? RoyZuo (talk) 18:09, 8 October 2025 (UTC)Reply

So far as I'm aware, this message is only shown on MediaWiki installations that have special security enabled. This should not be the case here on Wikidata. Can you give more details about where you saw this? Bovlb (talk) 20:18, 8 October 2025 (UTC)Reply

Have your say: vote for the 2025 Board of Trustees

[edit]
Latest comment: 2 days ago1 comment1 person in discussion

Hello all,

The voting period for the 2025 Board of Trustees election is now open. Candidates are running for two (2) seats on the Board.

To check your voter eligibility, please visit the voter eligibility page.

Learn more about them by reading their application statements and watch their candidacy videos.

When you are ready, go to the SecurePoll voting page to vote.

The vote is open from October 8 at 00:00 UTC to October 22 at 23:59 UTC.

Best regards,

Abhishek Suryawanshi
Chair, Elections Committee

MediaWiki message delivery (talk) 04:47, 9 October 2025 (UTC)Reply

Indicating an age range for a demographic property

[edit]
Latest comment: 1 day ago2 comments2 people in discussion

Is there a standard qualifier to indicate the age range that a specific demographic property applies to?

To make this concrete, the 2017 Bhutanese census provides two different overall literacy rates: the proportion of people aged >= 6 who are literate, and the proportion for ages >= 15:

Bumthang District (Q463242)literacy rate (P6897)73.1 percentpoint in time (P585)2017 (for ages >= 6).

Bumthang District (Q463242)literacy rate (P6897)67.9 percentpoint in time (P585)2017 (for ages >= 15).

(It seems that the first value is higher because younger generations are receiving better education.)

I'm not sure which value to pick. The cutoff at age 15 is more standard, as that is the definition the UN uses to compare literacy rates between countries, but the cutoff at age 6 is actually more accurate in some sense because it takes more of the population into account. Ideally it would be possible to include both with appropriate qualifiers.

It might be possible to (ab)use minimum age (P2899) as a qualifier but I'm not sure that's recommended.

Grešla (talk) 19:05, 9 October 2025 (UTC)Reply

Maybe use the applies to people (P6001) qualifier with suitable (new?) items for those age ranges? Your other edits on this look good to me. ArthurPSmith (talk) 01:39, 10 October 2025 (UTC)Reply

Projekt Gutenberg IDs

[edit]
Latest comment: 1 day ago7 comments2 people in discussion

Currently, when an instance of literary work (Q7725634) is associated to a Project Gutenberg ebook ID (P2034), there appears an exclamation mark icon with a warning saying: Entities using the Project Gutenberg ebook ID property should be instances of version, edition or translation (or of a subclass of it). See for instance Faust (Q29478). But lots of other IDs are allowed, such as the Penguin Random House work ID (P9818). Is this intended, that is, do all the IDs which don't have this warning in fact relate to the work rather than to an edition or version?

Katdav-wd-lit (talk) 06:28, 10 October 2025 (UTC)Reply

There is an (annoying to me) distinction in WD for books with 3 aspects: the creative work, the edition of that work, a single physical book. The book people feel this is important, and that as a Gutenberg text came from a scanning a particular edition done by a publisher, it need to connect to a matching "3rd edition of Animal Farm" entry. Personally, tying Animal Farm (Q1396889) to the Gutenberg text seems more useful to me, without having to find out the edition information and have complicated queries. Of course a few books have multiple editions with significant differences between them, but the vast majority have a single edition and single text. I'd hope that entries with warnings would encourage the purists to add the intermediate steps rather than remove the information, Vicarage (talk) 07:46, 10 October 2025 (UTC)Reply
And all IDs that appear without warning have been decided NOT to refer to particular editions? Or is it just that no one has looked at whether these represent particular editions or not. Katdav-wd-lit (talk) 10:02, 10 October 2025 (UTC)Reply
I probably qualify as one of those book people... I agree that it is annoying to have to search for editions to find a gutenberg ID. But then, it is also annoying not to be able to filter literary works by the original works language, and for this and similar reasons (like, adding number of pages and page numbers to citations) it is rather horrible not to be able to link to specific editions. The cleanest approach would be, I think, to allow gutenberg IDs to be inherited upwards to the work, from any of its editions. I see no reason why not the work and the edition both should carry an ID -- currently, this makes a warning appear as well, though. The gutenberg ID is supposed to appear for only one entity. Katdav-wd-lit (talk) 10:12, 10 October 2025 (UTC)Reply
I see no reason either. But at the moment, Animal Farm, a book both out of copyright and with I assume lots of editions, just has the one entry here, for the creative work, and no-one has recorded that it can be read at https://gutenberg.net.au/ebooks01/0100011h.html Vicarage (talk) 10:25, 10 October 2025 (UTC)Reply
To me, as a book person, it is frustrating that when this ebook was scanned and uploaded, the edition and publisher of whatever was scanned wasn't recorded. Or did I just not find it?
But I suppose one could now record this gutenberg version as an edition in its own right. Katdav-wd-lit (talk) 10:41, 10 October 2025 (UTC)Reply
Best to make your suggestions at https://www.wikidata.org/wiki/Wikidata:WikiProject_Books, I'm just doing caves at the moment. Vicarage (talk) 11:21, 10 October 2025 (UTC)Reply

Sharing of bibliographical metadata

[edit]
Latest comment: 1 day ago2 comments2 people in discussion

We use wikidata to obtain metadata for literary work (Q7725634) in our research. But we also generate metadata, often about obscure literary works. Is sharing of metadata for literary works which do not have their own wikipedia page something that wikidata would welcome, or not? (I'm sorry if this is a naive question. It seems to me that some tutorials talk about bulk uploads of bibliographical data, and it was not suggested that the data must correspond to a specific wikipedia page. I also see lots of items on wikidata that don't have wikipedia pages, but of course they might end up existing because of an image, or because of other connections within wikidata itself, such as being an edition.) Katdav-wd-lit (talk) 06:45, 10 October 2025 (UTC)Reply

See https://www.wikidata.org/wiki/Wikidata:Notability. Items don't need wikipedia pages, but need to have solid references, using our existing properties if possible, or described by source (P1343) if not. I would always apply a de minimis filter to start off, so books, where we are missing a great number, are very welcome, but personal letters in a minor author's correspondence might not be. Works where the text is online seem more useful too. Remember to tie them into the rest of the project, not merely by quoting authors, but main subject (P921) too. Vicarage (talk) 07:38, 10 October 2025 (UTC)Reply

A Wikipedia article has no Wikidata entry or ID, what to do?

[edit]
Latest comment: 22 hours ago3 comments2 people in discussion

The Croatian Wikipedia article at the URL https://hr.wikipedia.org/wiki/O_djelima_kraljeva_Dalmacije_i_Hrvatske has no Wikidata ID and no Wikidata page. I do not understand what to do in such case, or where to look for instructions. I looked at ways to report Interwiki conflicts, but it does not seem to cover the case of missing Wikidata ID. I would like this page to have the Wikidata ID. Thanks for help! Tranquillus (talk) 10:16, 10 October 2025 (UTC)Reply

Hi, you may create a new Wikidata item at Special:NewItem :). Samoasambia ✎ 11:15, 10 October 2025 (UTC)Reply
Ah, so this is not different than an item which has no Wikipedia article? I thought there might be a difference. I guess then I add the Wikipedia page address in the "Wikipedia" segment of the Wikidata page. Thank you. Tranquillus (talk) 18:59, 10 October 2025 (UTC)Reply

Social media not a reliable source

[edit]
Latest comment: 1 day ago2 comments2 people in discussion

Hello, why isn’t social media a reliable source? What happens if you cite them? 2600:387:F:4B12:0:0:0:B 11:12, 10 October 2025 (UTC)Reply

Per Wikipedia's Reliable Sources guidance, social media is not a reliable source because it is self-published and user-generated. BTracy-WMF (talk) 14:56, 10 October 2025 (UTC)Reply

WMF board reform

[edit]
Latest comment: 1 day ago1 comment1 person in discussion

Since reform of the WMF Board of Trustees would affect more than just people from enwiki, active wikidata editors may be interested in the m:2025 WMF Board reform petition. Clovermoss (talk) 12:07, 10 October 2025 (UTC)Reply

Distinguishing spoken text audio of Wikipedia articles vs of other literary works

[edit]
Latest comment: 23 hours ago1 comment1 person in discussion

Currently, spoken text audio (P989) is apparently used for both audio versions of Wikipedia articles as well as of audio versions of texts which may or may not also have a Wikipedia article about it largely without specifying which type it is.

Most uses of that property are for spoken Wikipedia. However, there's also many items where this prop is used for the text itself. See for example A Calendar of Wisdom (Q2894412) A Midsummer Night's Dream (Q55873489) A Reminiscence of Dr. Samuel Johnson (Q4177057) A Wonderful Bird is the Pelican (Q19027266) All's Well That Ends Well (Q55964658) Article One of the United States Constitution (Q48416).

I found all these via this dynamic list (since they didn't have the recording date qualifier set): Wikidata:List of audio podcasts of Wikipedia articles in English. By the way, I also used this table by checking items with empty rec date for adding missing Commons categories like Spoken English Wikipedia and Audio versions of Wikipedia articles from 2005 to files. Note that the table is only meant to show audio versions of Wikipedia articles / Wikipedia articles with audio versions so all the other items should be removed from that table.

What would the solution to distinguishing between these two+ types of uses for the property?

  • If it's moving files to a new property, which would that be and if needed could you propose it (=splitting the prop)?
  • If it's setting a qualifier, could somebody set it at scale to the items with P989 set?
  • If you have any other idea(s) regarding this, please elaborate.

Prototyperspective (talk) 17:58, 10 October 2025 (UTC)Reply

Jeremy Allen White

[edit]
Latest comment: 9 hours ago2 comments2 people in discussion

Hello. I am not good at Wikidata but it would be very excellent if this new WikiPortrait File:Jeremy Allen White Springsteen-33 (cropped).jpg could be made the default image for Q1411012 - the existing image is 12 years old and ports to several automated infoboxen. Please and thank you. 166.199.151.38 22:35, 10 October 2025 (UTC)Reply

✓ Done

i made listeria error

[edit]
Latest comment: 17 hours ago1 comment1 person in discussion

i new with this software someone can explain what i need to fix? https://www.wikidata.org/wiki/Wikidata%3AWikiProject_Machinima%2Freports%2FFilms_made_with_Machinima Trollface 2006ALT (talk) 00:19, 11 October 2025 (UTC)Reply

Yūko Nakamura is part of WD disambiguation page

[edit]
Latest comment: 16 hours ago1 comment1 person in discussion

E.g. Yūko Nakamura (Q126947429) and the disambiguation page Q126894715 hasPart several humans.

How to change that without just removing it? The WD dab page could use main topic, or the human listedAtCatalogue or so?

https://w.wiki/Fdax - 78 cases of "human is part of WD disambiguation page" - which is wrong. OSMan2025 (talk) 00:54, 11 October 2025 (UTC)Reply

IUCN Red List

[edit]
Latest comment: 6 hours ago1 comment1 person in discussion

"The update was released at the IUCN World Conservation Congress in Abu Dhabi. The IUCN Red List now includes 172,620 species of which 48,646 are threatened with extinction."

So a bulk update of the Wikidata needs to be performed. @Succu: OJJ (talk) 11:18, 11 October 2025 (UTC)Reply

Wikidata Game, several versions

[edit]
Latest comment: 2 hours ago1 comment1 person in discussion

Hello

Please, be careful then using this tool and you have no knowledge of the circumstances or local conditions.

Example mismatch commons category with wikidata:

https://www.wikidata.org/w/index.php?title=Q41378342&oldid=2415084442

Thank you for your attention. AnBuKu (talk) 14:38, 11 October 2025 (UTC)Reply

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat&oldid=2415182760"