Very bad news. Infopollution is problem everywhere. Firstly, lexeme stuff should be removed from Wikidata. Secondly, new users to not be allowed to create new items (e.g. in enwiki creation of articles is blocked for new-comers). Specially a lot of luck and success to Lydia Pintscher (WMDE). Hopefully, she and other technical specialists can save the sinking (?) ship Estopedist1 (talk) 07:33, 2 June 2025 (UTC)
Vandalism on Attack on Pearl Harbor (Q52418)
Latest comment: 1 month ago2 comments2 people in discussion
Latest comment: 1 month ago3 comments3 people in discussion
Here's your quick overview of what has been happening around Wikidata in the week leading up to 2025-06-02. Missed the previous one? See issue #681. Help with Translations.
Discussions
New requests for permissions/Bot: Wikidata Translation Bot - task/s: Automate translation of Item Labels and Descriptions across supported languages and submit them using the official Wikidata API.
New Linked Data for Libraries LD4 Wikidata Affinity Group project series! We have our next LD4 Wikidata Affinity Group event series on the Wikidata Graph Split project. Our first event will include guests from the Wikidata Search team to discuss the recent graph split project. Join us Tuesday, June 3, 2025 at 9am PT/ 12pm ET/ 16:00 UTC / 6pm CEST (Time zone converter). Please see our project page for more information and Zoom links.
Wikidata Qrank is a ranking signal for Wikidata entities. It gets computed by aggregating page view statistics for Wikipedia, Wikitravel, Wikibooks, Wikispecies and other Wikimedia projects. For example, according to the QRank signal, the fictional character Pippi Longstocking ranks lower than Harry Potter, but still much higher than the obscure Äffle & Pferdle.
Other Noteworthy Stuff
Should I watch this? - Enter a film title or IMDb ID to get a recommendation, uses data from Wikidata.
Job Openings - want to help shape the future of Wikidata or Wikibase?
Wikibase REST API: We are continuing the work on integrating simple search, specifically phrase matching (phab:T389011)
Query Service: We are working on an experiment to add a small dialog to inform people about alternative access methods for very simple queries that don't require SPARQL (phab:T391261)
It is a typical title for a software engineering position above senior status, without heading for a management role. Kind of a technical lead engineer. They also improve the software, but usually not anymore by cranking out much code or grinding user stories. —MisterSynergy (talk) 21:08, 2 June 2025 (UTC)
Broad or narrow external identifiers
Latest comment: 1 month ago9 comments6 people in discussion
@Jheald and I disagree over whether external identifers should have narrow or broad scope. Its in the context of introducing a new Wikidata:Property_proposal/trove.scot_ID to replace 4 Canmore (scottish heritage site being phased out) entries, as trove.scot promises to absorb many databases on diverse subjects across all of Scotland. I argue that people on WD want to know "What does trove.scot say on this subject", while he believes we should create multiple IDs based on the directory ontology they use, so a "place" has a different ID from an "object" or "planning decision". What do others think? Vicarage (talk) 20:03, 27 May 2025 (UTC)
It's also the ontologies *we* have for different sorts of items on wikidata, the different kinds of constraints we have for different kinds of items, and the different aims we may have for completion of different parts of their data.
A thesaurus is a database of a different kind of data, compared to a database of historic places.
Just because they are operated by the same organisation, that doesn't mean it makes sense to bundle them all up in the one property.
In the case of Trove one set of IDs are intended to be a direct continuation of the IDs for historic sites in the existing Canmore database (which is being retired). For these IDs it makes much more sense to simply change the URL formatter to point them to the entries in the new site, rather than muddling them together with entries for all sorts of other very different kinds of things. Jheald (talk) 22:23, 27 May 2025 (UTC)
You bring up a good point. Instead of one or many properties, one could group IDs belonging to similar classes in their own properties which might be the best option in many cases. Infrastruktur (talk) 18:06, 28 May 2025 (UTC)
How many different trove.scot "databases" are there? Historically we have supported both "broad" and "narrow" identifiers here; there are advantages to the narrow ones in regards to constraints etc. but I don't think there's really a strong preference on it, and I think a broad one is better than creating dozens of properties if that is the other choice. ArthurPSmith (talk) 16:05, 28 May 2025 (UTC)
We had a similar discussion last week at Wikidata:Property_proposal/Concertzender_ID − my opinion there leaned towards narrow identifiers. It depends also a bit how many properties we would be talking about should we choose to do narrow properties − the linked proposal could do a better job at explaining that IMO − reading through the discussion it would be 6? That sounds very reasonable (if you compare to, say, our 12 properties for MusicBrainz. Jean-Fred (talk) 17:30, 28 May 2025 (UTC)
We don't know the organisation's plans. They seem to be absorbing lots of different databases, but not preserving their names, just using simple keywords in URL sections, and we can see 6 at present. That doesn't help our plans to ensure coverage of Listings, Buildings at Risk, Dictionary of Scottish Architects etc, as we have no control of how they will fold databases into this new structure. I don't see why why our entries, with a very sophisticated ontology, also need a crude ontology for external site IDs, with each organisation having a different one, with our guessing what they mean by their subdivisions. If pre-existing database names are preserved on a site, we can respect that history in an id, but otherwise, we just know that they have a record of a concept. I'm glad I'm not writing queries to find out "what do music sites say about this". Vicarage (talk) 09:09, 29 May 2025 (UTC)
I (User:Jheald) asked User:Tagishsimon what his thoughts were on this, as he's done about as much work with the HES data as anyone (if not considerably more). He says that Sadly I'm still disinclined to contribute to / post on WD :( but thinks that Omnibus ID is a wretched idea, and sent me the following comments, that he said You're welcome to post into the discussion, if you wish:
" fwiw:
Trove currently hosts:
(A) Designations: corresponds with Historic Environment Scotland ID (P709) (listed buildings, scheduled monuments). 55k distinct IDs in WD, 62k records in Trove
(B) Places: corresponds with Canmore ID (P718) (canmore ID). 106k distinct IDs in WD, 330k records in Trove
(C) Objects: no current WD property. 2k records in Trove
I lean strongly to distinct properties for each of the distinct datasets WD will point to, irrespective of their all being hosted on a single website.
The chief reason for this is that 4 of the foreseeable datasets - (A) Designations, (B) Places, (C) DCA Building and (D) BaR Building all largely cover single real-world objects (a building or structure) which corresponds to a single WD item.
If a single property is used to represent all classes of Trove data for the building, then reporting becomes a right pain in the arse:
To ascertain which buildings are or are not on the Buildings at Risk register, or are or are not covered by DSA, the user will have to parse the contents of the single Trove property statement to ascertain whether or not the ID contains /BAR/ or /DCA/. That makes the SPARQL more unwieldy and the report slower.
To count the number of Listings, or Canmores, or BaRs, or DCAs, will also involve analysis of the property statement content, rather than a (COUNT(?Pxxx) as ?n) or a (COUNT(DISTINCT(?Pxxx)) as ?n)
The second reason is that all of the IDs have wide currency outside Trove and WD. To have to string-slice WD data to get an output from WD which corresponds with the real-world handing of these IDs (i.e. "place/12345" -> "12345") seems absurd. The ID is LB123 or (Canmore) 12345. The ID is not "designation/LB123" nor is it "places/12345". WD should store the ID, and not conflate the ID with the route to the record for the thing on Trove merely for the dubious convenience of having a single property in WD.
I don't see any real advantage at all in a single property.
So right now I'd be inclined to point Canmore ID (P718) at trove/places. When the HES portal is knocked on the head, point Historic Environment Scotland ID (P709) at trove/designations. Ditto for DCA and BAR properties when their time comes.
I don't foresee WD deciding to point at trove/objects, trove/images or trove/archives
I see value in pointing at the thesauruses, but if so would prefer a distinct property for each of the four distinct thesauruses." – The preceding unsigned comment was added byTagishsimon (talk • contribs) at 12:28 30 May 2025 (UTC).
IMO Tagishsimon makes a lot of sense in what he writes above. I'd also add (i) that if Trove can have up to 4 separate entries for the same building across its different databases, it's worth keeping those handled by different properties if we want constraint checking to be able to pick up potential duplicates; (ii) thesauruses apply to fundamentally different things than place registers (and are extremely valuable for ontology); (iii) it's much easier to work on completeness if small datasets (eg small thesauruses) are treated separately, and not subsumed in with the data from huge registers of places. Jheald (talk) 18:46, 30 May 2025 (UTC)
@Jheald @Vicarage Hi folks, just commenting to say that I'm mid-email in contacting HES about trove.scot, following a chat I had with a HES staff member at a conference yesterday. Been wondering how we can deal with this, glad to have found the conversation here. As ever please do drop me an email through my user page or leave me a talk page message if there's anything with which I can help. Sara Thomas (WMUK) (talk) 14:52, 3 June 2025 (UTC)
If the URL already contains a "?", then add "&action=purge"; otherwise add "?action=purge". There's a confirmation page. Bovlb (talk) 19:05, 2 June 2025 (UTC)
Many wikis have purge gadgets that add a button for purging a page. Search for "purge" on the gadget tab of your preferences on Commons or Wikidata. Samoasambia✎09:51, 3 June 2025 (UTC)
Unlinked ~ synonymous articles
Latest comment: 1 month ago5 comments3 people in discussion
de:Selbstbildung has as English synonym in the lead "self-education". On English Wikipedia, self-education redirects to en:Autodidacticism which also has that as a bold synonym in the lead. The subject of the articles also appears to be (more or less or entirely) the same. However, the English Wikipedia article does not link to the German article – only the other way around via an intentional redirect. Also, each have a mostly different set of interwikilinks.
This may be a discussion for German Wikipedia I thought but it's not just the DEWP<->ENWP link but also 5 other articles.
Moreover, this is probably not the only case. Some but not all of such cases could be identified via said to be the same as (P460)
Is there some dedicated place where such things are meant to be discussed?
While I wouldn't object to a merger, there may be a bit of a difference in connotation. In English, if someone is called an "autodidact" it tends to imply that they did not have significant formal education beyond (at most) secondary school and achieved, on their own, at least the equivalent of a bachelor's-level education (and typically even beyond that). I'm not sure if Selbstbildung has that connotation, and I'd be interested in hearing from a native German speaker whether it does. - Jmabel (talk) 23:31, 2 June 2025 (UTC)
I thought that may be the case but the article is about self-education in general. Self-education can also refer to that as alternative to formal education. Moreover, the article even makes it clear that it's not just that eg at This educative praxis (process) may involve, complement, or be an alternative to formal education.Prototyperspective (talk) 10:08, 3 June 2025 (UTC)
I'm not proposing merging as the option, I only thought and still think it would be among the options where the better one could be creating redirects and linking things accordingly
I saw that e.g. Polish Wikipedia has two separate super short stub articles. However, these could be merged.
Maybe the focus differs which is why I wrote (more or less or entirely) the same but possibly it only appears that way and those articles really are about the same thing. I thought it would be a different framing or focus where one puts the activity itself at the center (the process of self-education) while autodidacticism puts the concept / the literary and philosophical concept / the field into focus. But content-wise those don't really differ. Moreover, this applies to all sorts of articles which could be titled differently that would slightly change the framing or focus of the subject; when there are such articles separately then they are just merged. The only specialty here in regards to that I think is that both terms are about equally established/popular (here I think self-educations is a clearer better more general more descriptive term). For other articles where this is the case, it's often easier to select one of the terms and then put the alternatives into the lead and into redirects.
Latest comment: 1 month ago3 comments2 people in discussion
I do not know who is personally responsible for this step, but I wanted to thank them for the final completion. The color set in the dark mode is pleasant, and editing is easy and unproblematic. No bugs found so far. Ymblanter (talk) 07:22, 30 May 2025 (UTC)
Maybe but not all properties use personal computer as a platform so i dont know if we want it to be universal Trade (talk) 20:45, 31 May 2025 (UTC)
Complex constraint cannot suppress a built-in one. We can only remove the constraint altogether, and "replace" it with a complex one. --Matěj Suchánek (talk) 11:39, 1 June 2025 (UTC)
For items, it's not that easy because the actual number is stored nowhere and you would need to compute it for all >117M items first. --Matěj Suchánek (talk) 12:53, 1 June 2025 (UTC)
@Matěj Suchánek, thanks, however, what about item with the most labels and/or descriptions?
What makes a statement unique? Suppose you mean the most properties, then GND ID (P227) has the most properties (32, [7]).
Computing this for items is, just like labels and descriptions, again unfeasible unless you can find a reasonable subset of items which definitely includes the wanted item. Maybe someone else has the necessary skills and resources to get the results for you. --Matěj Suchánek (talk) 07:51, 4 June 2025 (UTC)
What makes a statement unique? Suppose you mean the most properties
Latest comment: 1 month ago2 comments2 people in discussion
Freiheit (Q1168552) seems to be specific to Ernst Busch's 1938 recording of this 1936 song. We lack an item for the song itself, which is the topic of the linked English- and German-language Wikipedia articles (I didn't check the other two). My I presume we should have separate items for the song and for this particular recording? Also, I see that it is listed as an instance of single (Q134556), but I'm pretty certain that it was released only as part of an album of 78 RPM records (Six Songs for Democracy, of which I had a copy when I was a child; it had been my father's in the 1930s). - Jmabel (talk) 23:16, 2 June 2025 (UTC)
Based on the Wikipedia articles that were used for the original creation of the item, it's pretty clearly a musical work, not a particular performance, so I've removed the performer and changed the type from single to song. Tfmorris1 (talk) 16:51, 4 June 2025 (UTC)
Can’t add references
Latest comment: 1 month ago2 comments2 people in discussion
Please be more specific. Link to the "material" (is it an item?). What are you attempting to do? Which button(s) are you clicking? --Matěj Suchánek (talk) 16:14, 4 June 2025 (UTC)
Ongoing massive duplication of items that are instances of Scholarly article
Latest comment: 1 month ago7 comments4 people in discussion
I've noticed that we seem to have a high level of duplication in items that are instances of scholarly article (Q13442814) and thought I'd spend a little time looking into where they are mainly coming from, to which end I cobbled together this unlovely python script. It is driven by the recent changes database table, so examines item creations in the preceding 28 days (I gather).
It has written a table of results at User:William_Avery/ScholarlyDuplicates/202405. Unfortunately, the page is rather big, and not particularly suited to mobile devices; be patient while it loads.
I consider that an item is a duplicate creation if it has a value of DOI (P356) or PubMed publication ID (P698) that is currently found on a pre-existing item. If there were two such pre-existing items, two rows are output to the table, one for each item that was duplicated. There are sometimes more that two; for instance, Autophagy and apoptosis: parent-of-origin genome-dependent mechanisms of cellular self-destruction (Q134483339) duplicates unique identifiers found on seven pre-existing items. Other columns in the table show the user that created the item and the timestamp at creation, along with one of the properties that is a duplicated and the property's value.
It's clear from the output that most of the items involved are coming from bots, which are currently producing them at a rate that far outstrips the capability of human users to clean them up. William Avery (talk) 22:19, 30 May 2025 (UTC)
@William Avery, Epìdosis: This looks like a consequence of the WDQS graph split - I expect the bot is querying the main graph, not the scholarly graph. Also notice that the duplicated DOI is *not* showing up as a duplicate in the UI - i.e. the constraint check is failing. I count 9 duplicates of the mentioned article now - in regular search enter "haswbstatement:P356=10.1098/RSOB.140027" to see them. Do wikidata constraint checks depend on WDQS? @Lydia Pintscher (WMDE), DCausse (WMF), Sannita (WMF), GLederrey (WMF), ABaso (WMF): it looks like there's some more work needed here relative to the graph split! ArthurPSmith (talk) 16:13, 31 May 2025 (UTC)
Latest comment: 1 month ago3 comments3 people in discussion
Hello! I have created the item Q134711788, which is a registered SEO and digital marketing agency from Bosnia and Herzegovina. I would like to complete the necessary statements (instance of, official website, location, etc.) but currently don’t have enough editing privileges.
Could someone please help add the following properties, or guide me on how to speed up the approval process?
Ah, I see: basically I use this 'tool' to automatically redirect me to the URL, for example: hub.toolforge.org/Q25342?property=P1343.
I suppose in this case the property doesn't actually link anywhere, although I found that to be quite strange - like Wolfram has its own website and such, so I would expect a link to it like other Wolfram-related properties AddyLockPool (talk) 19:35, 4 June 2025 (UTC)
Latest comment: 1 month ago4 comments3 people in discussion
Hello,
I am new to this site and am poorly familiar with it. I am inquiring about posdibility of building a database of events, with each event having answers to "Who, where, why, when, what, how". For example:
who = mr donald trump
where = white house
what = inauguration
why = after election
when = jan 24, 2025
Noting that at Wikinews each news article has the 5Ws written clearly in each paragraph so I am wondering if it can be added to here.
A few questions:
Is it possible to do this here at Wikidata with making a new item for the event, which links them all together?
What to do with items which are text descriptions if they do not have wikilink? For example, "researchers found a new species when processing a park footage" does not have a clear item to link for "why".
Possibility to add one or more paragraph comments - I guess not possible
Examples, if this was already done before
Possibility of linking both to Wikipedia and Wikinews pages, if both exist
Whether ultra local events are in scope, such as local car crashes
Possibility to view recently added or recent dates current events and using a Wikifunction to generate a page about it with list of the 5Ws on it with links to Wikipedia.
That is too complicated. It says who, where, when, and a bunch of technical information. How does a newbie use this? Say, I wanted to add that "Mr Trump" (who) "dies" (what) "today" (when) in a "plane crash" (how) "because of unknown reason" (why) in "California" (where). Could you give me an example how this can be added? Thanks. :) Regards, -- Gryllida14:05, 5 June 2025 (UTC)
Vouch – Creative Freelance Gig Platform Powered by Trusted Referrals
Latest comment: 1 month ago4 comments3 people in discussion
I’m proposing a new Wikidata item for Vouch, a digital platform that facilitates creative freelance hiring through a trusted referral model. Vouch connects gig posters with creatives via introductions from vetted industry connectors, streamlining discovery and reducing hiring risk.
The item would include:
instance of (P31) → online marketplace (Q105470145)
main subject (P921) → creative freelancing (Q210167)
official website (P856) → https://vouch-app.com
headquarters location (P159) → New York City (Q60)
country (P17) → United States of America (Q30)
inception (P571) → 2023
developer (P178) → Vouch App Inc.
I’d appreciate input on appropriate item modeling, especially around referral-based platforms or marketplaces driven by social trust mechanisms.
Thanks for the feedback here, will circle back once we have a bit more public mentions in the coming weeks/months. Excited to bring visibility to a platform built for largely underserved audiences. Diminovakov (talk) 20:15, 5 June 2025 (UTC)
Latest comment: 1 month ago2 comments2 people in discussion
I want create wikidata item. My item about a musical artist. I have two news source and spotify,IMDb, Youtube artist Channel.My question is Can I create this wikidata item? Ranimita (talk) 18:13, 5 June 2025 (UTC)
Also please don't bombard random people and noticeboards across random projects with the same question about Wikidata. Bovlb (talk) 18:40, 5 June 2025 (UTC)
Notability of scientists?
Latest comment: 1 month ago28 comments11 people in discussion
definitely some discussions have already been taken place about this topic, but see Wikidata:Requests_for_deletions#Q134689749. In addition, enwiki has policy that the scientist must fulfill en:WP:SIGCOV. I also know that Wikispecies is pretty fuc*ed up ( :) ) because some articles about new species has 100+ authors, and Wikispecies tries to do entry any scientist who is described a species Estopedist1 (talk) 16:53, 4 June 2025 (UTC)
A scientist/academic who has authored/co-authored a scholarly article stored in Wikidata would clearly pass both:
WDN2: It refers to an instance of a clearly identifiable conceptual or material entity that can be described using serious and publicly available references.
WDN3: It fulfills a structural need, for example: it is needed to make statements made in other items more useful.
I guess that due to more and more complexity of science, massive co-authoring (25+ authors in one article) become usual, and then we have to resolve it in order to prevent that "random" scentists aren't included in Wikidata (not sure how Wikispecies will resolve it) Estopedist1 (talk) 19:18, 4 June 2025 (UTC)
Shouldn't all the authors of papers be stored in the recently hived off scholarly article Wikidata, with cross-links to entries in this WD if they are more generally notable? I would think there are more articles than authors, given the number of articles written in an academic's career and the typical number of collaborators, so they could strive for completeness there, but we would struggle here. Similarly for books. In a federated system a book WD could contain all books and authors, but this WD would only contain notable ones. This would involve what now seems an inevitable tightening of notability, given all the concerns in the Mass Edit RFC. At least with scientists we have the Science Citation Index and Impact Factors to define notability. Vicarage (talk) 19:44, 4 June 2025 (UTC)
Probably not if we don't have enough information to identify them and distinguish them from other authors. I don't think creation of empty or almost empty items with no sitelinks is a good idea, particularly if those items are about living people. Even with identfiers there is sometimes no public information available, or information has been removed from the source, or some identifiers mix information about different people with the same name. Peter James (talk) 20:43, 4 June 2025 (UTC)
I think the scholarly article items are still here, it's just that they are now separated from other items for the query service. Peter James (talk) 20:43, 4 June 2025 (UTC)
The "hundreds of authors" business is not relevant - essentially every one of those people has published at least one other paper where they are one of only a few authors, or has their name on hundreds of these "hundreds of authors" papers. The number of distinct identified researchers is always going to be significantly less than the number of articles. Scholia lists 45 million scholarly article items in Wikidata, while there are only about 6 million people (and there are probably about as many sportspeople here as researchers). Wikidata is missing a lot of researchers who really are notable in the sense of winning awards - I regularly enter people who have won research awards in physics and there is always a significant fraction missing. I agree we shouldn't be automatically creating items for these people without checking anything further, but identified people who are published researchers should not be deleted! ArthurPSmith (talk) 21:30, 4 June 2025 (UTC)
Perhaps the criteria should be that the individuals must be uniquely identifiable by more than just a name and the fact they are a researcher. External identifier, source providing employer, alma mater, DOB, etc. Otherwise we'll have a heap of items that cannot be uniquely identified and are useless outside of that one attached article they were part of. So long as that criteria is met, they should have an item. — Huntster (t@c)22:07, 4 June 2025 (UTC)
Yeah, to prevent duplication ideally we should not create items without IDs (such as Katherine S. Rocci (Q134609702)). However items with at least ORCID is fine since this already eliminated much possibility of duplication. GZWDer (talk) 03:05, 5 June 2025 (UTC)
This discussion surprises me. Things are of course different for vanity publishing but to my knowledge, the notability of all authors of “real” academic papers has never been seriously questioned. --Emu (talk) 10:53, 5 June 2025 (UTC)
@Emu: are we able to very roughly estimate how many real academic papers exist and how many authors? If a saw a news that some articles have 10k+ authors, I am bit worried ... And very well said @Peter James. If simple Google search doesn't give enough information to identify scientists and distinguish them from other authors, then we should delete such item and substitute it with author name string (P2093), see e.g. Q114066683Estopedist1 (talk) 16:29, 5 June 2025 (UTC)
So your position is that if we have two such authors, whose QIDs are different, and both have the name "John Doe", we should delete both QIDs from each of the items about the papers they authored, and replace them with the string "John Doe"? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits16:36, 5 June 2025 (UTC)
@Estopedist1 I get where you are coming from but your “Google search” heuristic seems highly dangerous: If information isn’t easily obtainable there’s actually more reason to keep in in Wikidata not less. Also scientists are generally relatively easy to distinguish (because they are linked to articles and those articles generally stay in a field that keeps getting more and more niche), at least compared to the people I generally work on. --Emu (talk) 17:06, 5 June 2025 (UTC)
"how many real academic papers exist and how many authors" - See Property_talk:P6366, the Open Academic Graph contains 254 million articles. This may likely cover 50%-100% papers ever published in the world. This means Wikidata covered 8%-18% of all papers ever published. The number of total different authors may be 200-500 million. GZWDer (talk) 06:24, 6 June 2025 (UTC)
You are certainly entitled to this opinion but it’s certainly not consensus and especially not what WD:N is saying. --Emu (talk) 06:13, 6 June 2025 (UTC)
Wikidata is not Wikipedia and items for researchers are useful for many purposes (e.g. they are linked to alma maters, employers, doctoral advisors; and used for generate Scholia profile). What is required is such item must be clearly distinguishable (not be conflated with any other people with same name). ORCID will serve this purpose. GZWDer (talk) 06:18, 6 June 2025 (UTC)
Agree. Also somebody/some org/team really should work on the scalability issues of Wikidata. It won't be very useful if it won't be able to get at least as large as other popular databases. Prototyperspective (talk) 12:06, 9 June 2025 (UTC)
So a scientist who does not meet en.Wikpedia's notability criteria, but who has articles on, say, the German, Arabic and Bengalli Wikipedias, should not have a Wikidata item? That's certainly an interesting position to take... Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits08:31, 6 June 2025 (UTC)
Latest comment: 1 month ago6 comments5 people in discussion
I currently experience a massive amount of simple queries they do not give complete or no results. Running the same request some seconds later gives correct results. Does anyone else have the same problem? A tool I run since two years never experienced such problems. GPSLeo (talk) 16:05, 5 June 2025 (UTC)
I had this problem yesterday for this query which shows a list of WikiProjects https://w.wiki/E8b3 It showed no results unless I removed line FILTER((LANG(?mainSubjectLabel)) = "en") (which results in the column having lots of text due to the labels for many languages). However, after retrying again some time later it worked. The listeria bot issue described in a section a bit above could also be related to this. A list it updated first was very incomplete but later the full list was added but it seems to cause some problems at other lists. Prototyperspective (talk) 16:39, 5 June 2025 (UTC)
I had a similar problem (June 4th) as well on a niche SPARQL query. Would you mind explaining what to look for on either of those links to check when when the "full reload" is finished (if that's indeed the solution)? Tæppa (talk) 21:35, 5 June 2025 (UTC)
Look for wdqs1022 on [9]. Compared to the other nodes, it currently holds ~50% of non-scholarly Wikidata. Based on the progress (+10% in 12 hours), it will probably take a few days to be fully reloaded.
This is just my hypothesis. But if a host is being reloaded (confirmed by [10]), then it's busy reloading and thus queries are more likely to timeout. And since it's being reloaded, it doesn't hold all data, thus the results may be incomplete. Users report either. I wonder why a reloaded node is reachable by users, though. --Matěj Suchánek (talk) 08:38, 6 June 2025 (UTC)
For https://query.wikidata.org/sparql API queries, I'm seeing 60s timeouts, then if I re-run seconds later, they complete in 30s. Perhaps I'm getting different nodes each time. Annoying as I'm restructuring my system to reduce query time at present. Vicarage (talk) 13:09, 6 June 2025 (UTC)
SSSIs in Scotland - when are the site & the SSSI different items?
Latest comment: 1 month ago5 comments3 people in discussion
Hi folks - hoping someone can advise. I'm doing some work around protected areas in Scotland, at the moment, I'm working specifically on SSSIs (notes on this are on my userpage). In importing the identifiers from NatureScot, I've come across a small number where I'm unsure if the SSSI and the site should be separate entitities. The person that I'd normally ask is no longer on Wikidata, I've had a look in WikiProjects but haven't been able to find the answer.
Example - Q41214 and Q134715256 - should these be two separate items, or merged, or linked? Is the mountain a separate entity to the SSSI?
I'd checked for dupes before importing, reconciling against both the NatureScot ID and label, and thought that I'd eliminated as much chance of duping as possible, but in checking the import I'm now finding a few that might be dupes, so I'd like to clean them up. However, I'm now wondering if they'd be better kept separate. Signposting to any guidance would be appreciated! Lirazelf (talk) 17:15, 5 June 2025 (UTC)
I would merge them. Nearly always on WD we have a named location that has SSSI properties. I'd ensure the merged item's area was given a caveat saying it was the SSSI's measure Vicarage (talk) 17:46, 5 June 2025 (UTC)
Ah, hadn’t clocked that, thanks. Similar issues with an island in Loch Lomond, one SSSI contains 2 islands (seems separate makes sense), one is mostly just the island but seems to also contain some of the water around it. Much case by case basis! Lirazelf (talk) 12:38, 6 June 2025 (UTC)
Tool to depicts on commons
Latest comment: 1 month ago2 comments2 people in discussion
Issues with setting docs pages & developer of Wikimedia tools
Latest comment: 1 month ago14 comments5 people in discussion
There is substantial data on Wikimedia software in Wikidata. There could be more of it and it could be of relatively high usefulness to aggregate such data on WD². See Wikidata:List of Wikimedia tools with Wikidata item. Next to the issue of a large gap of missing items, there are two issues when creating such items:
The developer usually is a Wikimedia user but one can't link to Wikimedia user pages or name such in developer (P178) and creating a new item for the person may not be the best approach or if it would be, it would make creating tool items a lot of hassle – asked about it at A way to just link a Wikimedia user page for W tools
Often, the docs are on a user subpage. Maybe something should be done about that e.g. by mass-moving such pages to a proper nonuser namespace. However, what about allowing user-subpages to be linked in the interwiki links of items if those are instances of Wikimedia tool or any subclass thereof? For example, graphDataImport (Q134713849) can't link to the docs page on Commons and the item for en:Wikipedia:reFill only links to the English and German Wikipedia docs pages because on other language version the docs are on some user page that can't be linked there.
² (this data could then for example be used by other sites that make it easy to find tools of interest, or lists by which one can find tools in a preferred programming language, or better enable people to locate docs) Prototyperspective (talk) 13:18, 5 June 2025 (UTC)
I think all significant Wikimedia tools are notable, especially when there are millions of items that, unlike items for Wikimedia tools, nobody uses/is likely to use. In any case, that page has 1. It contains at least one valid sitelink to a page on so that's part of the subject discussed here as the pages are often in userspace. Good question though. Prototyperspective (talk) 14:32, 5 June 2025 (UTC)
I mean, it’s technically just an example, but if we don’t stick to this requirement, “structural need” is devoid of any actual meaning, resulting in the de facto end of notability criteria. Which, if memory serves, seems to be pretty much your goal. --Emu (talk) 11:43, 6 June 2025 (UTC)
One could for example specify exceptions to that policy and/or modify that policy accordingly and/or specify that structural need is by default there for Wikimedia-related items; I think that which is the subject of this thread would be the better route for how these items are within scope: 1. It contains at least one valid sitelink to a page on… where one could specify that for Wikimedia tools, user-pages (and/or the source code?) would suffice or enable these to be linked as sitelinks as suggested here. Prototyperspective (talk) 11:53, 6 June 2025 (UTC)
Okay, which “structural need” are you thinking of anyway? Structure of what? Need what for? --Emu (talk) 13:41, 6 June 2025 (UTC)
A Wikimedia tool template that includes links in statements using user manual URL (P2078)/described at URL (P973) could be a work-around for the lack of sitelinks. As for developer (P178), I think creating an item for the developer(s) would be a good idea, but if you really don't want to create items for them how about using unknown value and then qualify the statement with a link to their user-page? M2Ys4U (talk) 16:33, 5 June 2025 (UTC)
Thanks, those are good ideas – which qualifier could be used for that? A Wikimedia tool template what do you mean by that? That it would suggest these two/either of the two properties to be filled when entering instance of: Wikimedia tool? Prototyperspective (talk) 16:36, 5 June 2025 (UTC)
Latest comment: 1 month ago4 comments3 people in discussion
What to do with claims that capture some important snippet of information in a clearly interpretable way, though not strictly conforming to the definition of the property that’s being used, like
(the specific apple tree genotype Idared has Jonathan as its direct ancestor) where child (P40) is defined as only for Homo sapiens? (People are deleting such claims as supposedly “obvious mistakes”.) Is there (shouldn’t there?) be a rule against deletions just for reasons of form? Is it an “obvious mistake” to enter such information before the specific case has been discussed? There is a lot of such improperly stored, but useful information on Wikidata that can be converted into a more suitable form more or less easily. It seems wrong to just delete it instead of requiring people to find a more productive solution. Cartoffel (talk) 10:44, 7 June 2025 (UTC)
Generally if you want to broaden the scope of a property you start a discussion on its Talk page, perhaps others can suggest a more suitable property that could be used/broadened. Or discuss on the relevant Project (though far too many are moribund). But its unwise to start adding data that has blatant constraint violations, and the relation between 2 species or cultivars is certainly not that of a child. Vicarage (talk) 11:10, 7 June 2025 (UTC)
One thing to remember is that "child" and "direct ancestor" may make (metaphorical) sense in English, a translation in another language may make the claim seem like gibberish.
The parent/child relationship is between 2 individuals. And its use is fine for racehorses like Shergar (Q2739031), so its not just homo sapiens. But that's very different from claims about classes like cultivars of apples, even not considering fiddly detail like diploid or tripolid cultivars. I'd ask Wikispecies for modelling advice. Vicarage (talk) 13:13, 9 June 2025 (UTC)
Other projects sidebar in userpages
Latest comment: 1 month ago3 comments2 people in discussion
Hi there!! I was wondering if there would be some way of adding the "In other projects" sidebar to my userpage so users can easily find me across projects. And in case this is currently not supported, where should I propose it's implementation? Thanks!! It's moon (talk) 13:56, 9 June 2025 (UTC)
As far as I'm aware this only works for Wikipedia sites. My request is about linking to other projects such as Commons or Meta. When adding [[commons:User:It's moon]] it behaves as a link (commons:User:It's moon) but doesn't add the "In other projects" sidebar, nor adds Commons as another language. It's moon (talk) 14:47, 9 June 2025 (UTC)
Did anyone notice that ListeriaBot is down?
Latest comment: 1 month ago15 comments6 people in discussion
I don't know why you thought this could possibly be related; it's only a proposal not describing something that is now newly getting implemented and about items and revision histories anyway, not the small number of lists that the listeria updates.
Edelseider thanks for asking. I did notice a new list I've created yesterday wasn't getting updated but that's relatively normal so I didn't yet recognize that the bot is down. I think the tool is a bit underused here, I mean it has more potential than what it's used for and most lists are hard to find and not about subjects of general Internet user or Wikidata contributor interest. Prototyperspective (talk) 11:44, 4 June 2025 (UTC)
Thank you! Magnus replied on his talk page that he restarted the bot, but it still doesn't update when prompted manually. It must be some reason beyond mere on/off. Edelseider (talk) 10:09, 5 June 2025 (UTC)
Well it now did update Wikidata:List of activities done as hobby so something is or was working again. I also tried updating another list which it didn't despite having status OK but that may be because nothing changed since the last update and this may also apply to some list that you tried it for maybe. A good opportunity to mention this new list which seems to miss many items about hobbies. Prototyperspective (talk) 11:15, 5 June 2025 (UTC)
It is working again, and most of the time it's working right, but what I see in cawiki is that today Listeriabot is making a lot of wrong edits, randomly deleting parts of lists. Most of these errors can be solved by clicking to update the list, but not all of them, as of now. Pere prlpz (talk) 15:33, 5 June 2025 (UTC)
I don't know how often the list is supposed to be auto-updated as I could not see this config there but the last update by the listeriabot there has been 16 June 2024.
Another thing: it works again but did remove items from some but not all lists it updated. Hopefully, it will soon be fully functional again so that it can retrieve all items so as to not remove items from lists that shouldn't be removed. Prototyperspective (talk) 12:05, 9 June 2025 (UTC)
About autoupdate: My experience in cawiki is that it doesn't autoupdate often, but usually it updates quite fast when you click "manually update list". However, I clicked it in your list and it hasn't still worked.
It stopped working here: [11]. It won't work without transcluding the template. I guess the reason was to prevent it showing the There is no consensus to use Template:Wikidata list in articles. error message in the only article using the template. The opening banner can be removed as it has no effect. --Matěj Suchánek (talk) 18:41, 9 June 2025 (UTC)
Help us test the new Search box: search by entity type
Latest comment: 1 month ago6 comments4 people in discussion
Currently, search-terms for a Lexeme, Property, or EntitySchema: need to be prefixed with L:, P:, and E: respectively and potential matching results are shown on the Special:Search page. This should improve the search function and save you some clicks.
What has changed
A new search dropdown menu, where an entity type can be selected.
Typeahead suggestions available for Item, Lexeme, Property and EntitySchemas: matching and partially-matching results are shown to you and can be clicked from the dropdown.
Will only be available in Vector 2022 skin.
What may break when deployed
Tools, gadgets, user-scripts or workflows that rely on the previous search behaviour may no longer function correctly.
If your search is modified by JavaScript or CSS.
Bugs, unusual behaviour, doesn’t work: Tell us!
We’d love to collect your feedback about this new feature before releasing it on Wikidata in approximately 1 week (11.06.2025). Testing can be done now on https://test.wikidata.org and https://wikidata.beta.wmflabs.org.
If you encounter any bugs or have technical questions, please reach out to us on the Phabricator ticket: phab:T321543.
Not sure where you are having general discussion, but I'd not want to have it in Phabricator. I'm not keen on having to click on the search icon for the entry box to appear. I'd rather have both visible. No problem having the dropdown appear rather Special:Search when the icon is clicked. How does this meet accessibility standards? Vicarage (talk) 20:40, 4 June 2025 (UTC)
I like the new search box, it makes sense to me. I'm not sure what Vicarage is talking about, the search entry box was there from the start on test.wikidata.org for me, and it defaulted to item entity type which is exactly what I would expect. ArthurPSmith (talk) 21:03, 4 June 2025 (UTC)
I was at a higher zoom level, where only the icon appears. It needs to have an entry box for high zoom levels. Zoomed out, It seems sized for the dropdown+box combination from the start, with a much bigger search box than before. Either it should always show the dropdown, or it should start smaller. I found it a distraction that I'd clicked in the box, it magically redrew with the actual input box jumping 2 inches to the right. Could the dropdown be positioned after the entry box to avoid this? Vicarage (talk) 10:43, 6 June 2025 (UTC)
Hello @Vicarage, could you please take a screenshot of your issue and upload to the Phabricator ticket? Is the issue that when the magnifying glass icon is clicked and the new search box is displayed, the dropdown box of entity types is 'off your screen' to the left? Thanks - Danny Benjafield (WMDE) (talk) 11:59, 10 June 2025 (UTC)
Latest comment: 1 month ago3 comments2 people in discussion
For example, it would be useful to have a column about number of subreddit subscribers and the number of pageviews the last x days (even better would be having a small graphic next to it showing the chart like in the Wikipedia app) in this list: Wikidata:List of Linux distributions.
Wikipedia Android App Top ReadOn the right is how the Wikipedia Android app's top read of the day looks like. There you can see a small graphic showing the pageviews trend as well as the pageviews number. It would be great if making pageviews queryable and rendering a chart image of them was made possible so that one could have a pageviews column in tables like the example one above. This would be really useful and would find many applications. Prototyperspective (talk) 13:55, 9 June 2025 (UTC)
The entity is too big. The maximum allowed entity size is 2.93 MB.
The entity is too big. The maximum allowed entity size is 2.93 MB.
How can this be solved? I think this data is more useful than lots of other things in that item. Maybe all those many values in child astronomical body (P398) could be moved out somehow, e.g. by moving them to a separate item or by just setting parent astronomical body (rather associated star or sth like it). I think also in general items would need to get a list of links to queries that are relevant for the item type, in this case e.g. a button that when clicked showed all the 'child astronomical bodies' of the item. Prototyperspective (talk) 16:59, 9 June 2025 (UTC)
It really needs an accessible button like "showed all 'child astronomical bodies' of this item" for items that are an instance of a subclass of star (Q523). Not all users know all relevant properties and how to quickly query for them for all types of items. Another example where this would be useful are items on TV series where instead of having the episode items on the series item, there is a button to view all episode items. Prototyperspective (talk) 19:35, 9 June 2025 (UTC)
For other readers of that thread: relateditems is a gadget that can be enabled in the preferences. Thanks, forgot how it was called and didn't use it in a while – I think it would best be
enabled by default (not hidden in a long list of preferences that the most active registered logged-in contributors change)
change the title or add sth to it because "show derived statements" is quite unclear
for every derived statements box enable the user to get the search link and/or sparql (some button for each box) to get these so that this can be used or adjusted as needed
some of the boxes can contain a large number of items; instead showing very many of these, cut the box after some and also show other types of derived statements – for example free software (Q341) shows only a long list of items that are an instance of it but not other kinds of relation such as items that have it as main subject (P921); if the relation is not on the first page one is out of luck and would have to think of some potential relation and then manually craft some query (e.g. finding forks of a free software using based on (P144))
Currently over 1.4 million gravitationally bound objects catalogued [12], almost all of them asteroids. 471 claims for child objects of the Sol with the sole classification of asteroid is actually misleading, which is an argument for getting rid of them. Infrastruktur (talk) 11:56, 10 June 2025 (UTC)
I'm puzzled why you want to add a 4 year old audio recording of a wikipedia article to anywhere, let alone WD. Surely there are modern text-to-speech systems that could trivially reproduce audio to the required quality level on demand, and always be current? Vicarage (talk) 21:51, 9 June 2025 (UTC)
Glad to see somebody sharing this view. The only difference is that 4 years is not that much to me, especially for an article that doesn't change that quickly in terms of new events or changes in humanity's understanding as the example article. As asked about here in regards to implementation, I think the qualifier about the version date of the audio's article should always be set for all spoken text audio items (via QuickStatements somehow). The problem for now is less 4 year old audios in Wikidata items to my mind though but 20 years old audios in Wikipedia articles. Also I started this new WikiProject to do more or less exactly what you described but it's stalled due to a lack of participants, if you'd like to make an impact, you could – millions of people listen to audio podcasts daily. Prototyperspective (talk) 22:01, 9 June 2025 (UTC)
And if, as I assume, you are merely providing a link to the original https://en.wikipedia.org/wiki/File:En-Sun.ogg stored somewhere in commons, and not uploading it anew, why is WD triggering a size warning? There doesn't seem to be a constraint for it, and why would WD care about the size of a file stored elsewhere. Vicarage (talk) 22:16, 9 June 2025 (UTC)
Latest comment: 1 month ago1 comment1 person in discussion
Here's your quick overview of what has been happening around Wikidata in the week leading up to 2025-06-10. Missed the previous one? See issue #682. Help with Translations.
Discussions
Open request for adminship: Coinhoe - RfP scheduled to end after 10 June 2025 23:49 (UTC)
Events
Upcoming events: New Linked Data for Libraries LD4 Wikidata Affinity Group project series! We have our next LD4 Wikidata Affinity Group event series on the Wikidata Graph Split project. Our second event will be a conversation with Daniel Mietchen and Lane Rasberry about Scholia, the Wikidata frontend which generates and presents scholarly profiles based on WikiCite content. They'll speak to Scholia's current state and roadmap, with consideration for the recent Wikidata graph split. Tuesday, June 10, 2025 at 9am PT/ 12pm ET/ 16:00 UTC / 6pm CEST. More info and Zoom links: project page.
GLAM:Memory of the World Report: Hannah Drummen at UNESCO, alongside data expert Martin, has completed a structured dataset of 496 International Register items, ready for bulk upload to Wikidata in June, with an aim to enhance accessibility and define best practices for future updates.
Wikidata QID updates to BHL catalogue: The BHL Lead Developer, Mike Lichtenberg, is ensuring periodic Wikidata Qid refreshes in the BHL Catalogue, with the working group advising a downloadable post-refresh report for OpenRefine integration, to be sent to the BHL Metacat group for reconciliation by Siobhan or other Wikidata editors.
Wikidata training & Datathon in Indonesia: Wikimedia Indonesia hosts WikiLatih Wikidata training to enhance skills in editing Indonesian cultural heritage data on Wikidata, while Datathon challenges participants to make the most edits on museum-related topics in Indonesia.
Papers
Wikidata for Botanists: Benefits of collaborating and sharing Linked Open Data By von Mering et al., (2025) - This paper explores Wikidata as a multilingual open knowledge base for botany, highlighting its role in connecting botanical information across sources, and calling on the botanical community to enhance its content.
CS-KG 2.0: A Large-scale Knowledge Graph of Computer Science By Dessí et al., (2025) - This paper introduces CS-KG 2.0, an advanced AI-powered knowledge graph built from 15 million research papers, designed to enhance scientific exploration by structuring and interconnecting vast amounts of computer science literature.
Videos
Using the Wiki List tool - GoogleSheet with formulae for retrieving Wikidata values and writing QuickStatements commands.
Wikidata Phonemes This is the web application developed specifically for Wikidata IOLab. In here you can add phonemes to a whole bunch of languages, basing your work on the work that the brazilian students of their national olympiad did while editing Wikipedia.
Should I watch this? is a tool that helps users decide whether a movie or show is worth watching.
WikiProject highlights: Names/Belarusian - This WikiProject aims to add structured and linguistic data to Wikidata to enable the study of people's names across all time periods, regions, and languages.
Mobile editing of statements: We are doing initial development focusing on technical investigations and basic UI elements (phab:T394292, phab:T394886)
Lexemes: We are looking into a rare error when trying to do undo certain Lexeme edits (phab:T392372)
Watchlist/Recent changes on Wikipedia: We continued working on showing labels instead of IDs in the edit summaries of Wikidata changes that are shown in the watchlist and recent changes of Wikipedia and co (phab:T388685)
Wikibase REST API: Finishing touches on simple search (phab:T383126)
Query Service UI: Added experimental popup to point people running very simple queries to other available access methods (phab:T391264)
Latest comment: 1 month ago1 comment1 person in discussion
How do we mark the Honjibutsu of Shinto kami here? I have been using "said to be the same as" but is there a better one? Immanuelle (talk) 22:40, 10 June 2025 (UTC)
Bugs in filter-watchlist-languages.js
Latest comment: 1 month ago2 comments1 person in discussion
It allows you to filter out languages you don't understand from your Watchlist which makes it much more managable, overseeable, and time-efficient, saving lots of precious volunteer time you can use for actually contributing to Wikidata and making it less exhausting to quickly check the Watchlist.
It's also described as Filters your watchlist to show changes to labels, descriptions, and aliases that are in specific languages.
User:Lectrician1 seems more or less inactive and hasn't replied to the talk page issues. Another problem is that there is no documentation page for the tool, just the js page.
Could somebody please improve the script, mainly fix the two key issues of it hiding changes of one of the specified languages and the issue of it not hiding changes of languages not specified there?
I think this functionality is very valuable to nearly all Wikidata contributors and I even think it should be some native Watchlist functionality. Given that it at least currently isn't, I think it would be of high priority to fix those issues. Prototyperspective (talk) 12:07, 6 June 2025 (UTC)
Why are multiple undisclosed accounts not allowed on Wikidata?
Latest comment: 1 month ago3 comments3 people in discussion
According to Wikidata’s policy on multiple accounts, operating undisclosed multiple accounts are not allowed and are considered illegitimate if caught. But can you explain on why is this policy enforced on Wikidata? Why these accounts must be publicly declared, and the userpages of the alternate accounts must link back to the main account in an obvious fashion? What is the reason of disclosure though and above? Thank you for answering this. 2600:387:F:6118:0:0:0:623:55, 9 June 2025 (UTC)
What we call sock-puppetry is actually illegitimate use of alternate accounts (this is an important point). The thing is, if you do not declare alternative accounts upfront and what you are using them for we will absolutely assume that their use is nefarious, because you have given us no reason to assume otherwise. It is perfectly fine to have an alt-account, simply be upfront about it and its use. For example I have an account I have declared is used for testing, but that I might also use if I am logging in from a sketchy Mexican internet-cafe (apologies if you happen to be Mexican, this is besides the point). Infrastruktur (talk) 18:10, 11 June 2025 (UTC)
Help us improve the Wikidata editing experience on mobile devices
Latest comment: 29 days ago1 comment1 person in discussion
Hi everyone,
We’re excited to share a new prototype that brings statement editing on Items to mobile.
You have been asking for the ability to edit statements on Items via mobile devices. Until now, this hasn’t been possible in the mobile view, requiring editors to switch to desktop mode to make edits, resulting in a clunky and frustrating experience.
This prototype is designed to show how we are trying to make contributing to Wikidata easier and more accessible for editors using mobile devices. It has been updated based on a first round of user testing we conducted earlier in April. Now we’d like to get more feedback from you before we move forward.
@Vicarage: For permanent exhibits, they typically last many decades. I usually also photo-document them (but don't typically upload them). - Jmabel (talk) 04:43, 27 May 2025 (UTC)
So is the effective consensus really that the wall text at the Museum of Art Collections is unciteable, but some random Commons user is citeable? - Jmabel (talk) 18:13, 30 May 2025 (UTC)
Again: I'd really like either to have someone more active than I am in this site say, "Yes, that really is a consensus here" or "No, it's an accidental consequence of a series of more-or-less unrelated decisions." - Jmabel (talk) 04:06, 6 June 2025 (UTC)
@Jmabel: That is absolutely a citeable source that can and should be used, following the rules for "Adding a source to a statement" at Help:Sources#Adding_a_source_to_a_statement. The tricky bit is how to sensibly model the item for the source, which largely depends on how precise you want to get with it (e.g. is it sufficient to cite the exhibit as a whole, or do you want the source to be the specific label?) But as long as the reference provides enough information for someone else to theoretically verify that this source really does make the claim, then the specific modelling of the source item doesn't really matter that much right now: it can be extended or tidied up separately later. Oravrattas (talk) 08:40, 9 June 2025 (UTC)
Often these texts are used in this form or a revised one in the corresponding exhibition catalogue, so this could also be a possibility to cite something. --Dorades (talk) 15:45, 9 June 2025 (UTC)
I don't think there is a catalog (which would have been straightforward to cite). I think it's just wall text for the permanent exhibits. I took a photo of it, but that seems an odd thing to upload to Commons, we don't usually host those (even though it isn't a copyright issue, because it is below the threshold of originality). Again: I don't even know if they are right, but I would assume they are at least as worth citing as some random Commons user.
I take it from the above, though, that people don't think there is an established way to model this.- Jmabel (talk) 23:03, 9 June 2025 (UTC)
I've ended up citing a different source in this case. I still think we should establish a way to model museum wall texts as citations. - Jmabel (talk) 18:04, 12 June 2025 (UTC)
In any case, it may be useful to have an item for this in order to use it for SDC (Structured Data on Commons). I uploaded c:File:Rijksmuseum 25 October 2017 02.jpg this one years ago and was delighted to see it was finally updated back during the pandemic but I guess I never uploaded the updated one. Jane023 (talk) 06:13, 13 June 2025 (UTC)
Vote now in the 2025 U4C Election
Latest comment: 27 days ago1 comment1 person in discussion
OpenRefine: Can't login to Wikidata (login verification code)
Latest comment: 27 days ago3 comments2 people in discussion
Hello, I tried to login to Wikidata via OpenRefine (3.9.3), but it fails. At the same time I get emails with verification codes ("it looks like you are trying to log in from a new device on Wikidata." etc.) but I see no way to enter it in Openrefine (or anywhere else). Did someone run into the same problem? Have you found a solution? Thank you and kind regards, Valentina.Anitnelav (talk) 14:55, 14 June 2025 (UTC)
Latest comment: 26 days ago5 comments3 people in discussion
I am interested in using automated tools here on wikidata. I have written some programs to query wikidata, and I am interested in running something that scrapes geni and creates wikidata items or links the profiles on the pages. Is this kind of editing allowed without going through any kind of special approval process? Immanuelle (talk) 22:19, 10 June 2025 (UTC)
Geni.com contains a large amount of problematic data. Please do not mass import them without checking other sources. GZWDer (talk) 11:39, 15 June 2025 (UTC)
Importing data from categorization on wikis
Latest comment: 26 days ago3 comments1 person in discussion
Can we import Shrine categorizations from Japanese wikipedia as aspects of the wikidata items?
I am not quite sure what to make of it. But there are many different historical rankings of shrines that I think should be here
These are the categories I think should all be represented on the wikidata items
jawiki subcategories of this one ja:Category:府県社 Fuken-sha. Note this has a bit of complications because there are technically Ken-sha and Fu-sha which were legally combined.
Subcategories and problems
ja:Category:式内社 subcategories here are Shikinaisha by Province
I do not think we will want to duplicate this here, but add Shikinaisha categorization to all of these, all the ones in these subcategories (excluding the Myojin Taisha and Shikinai Taisha ones) are in the ranking of "Shikinai Shosha" Q9646930. This is the only category that is better done on other wikis than jawiki. But we would probably need some kind of query to get this done
The articles in this category ja:Category:式内社の一覧 are lists that we could get info from either indication of the shrines being Ronsha as marked by （論）
We could use some kind of script to import the other data from these pages. But it will be more difficult than just categorization
We could make wikidata items for all the Shikinaisha, or add the traditional pronounciations, or add some of the info about the ritual levels Immanuelle (talk) 08:24, 13 June 2025 (UTC)
I am doing this right now. Except for the table stuff because the table stuff is much more complicated and prone to error. Immanuelle (talk) 04:58, 15 June 2025 (UTC)
Yeah Shikinai Shosha are uncategorized on jawiki and only categorized on other wikis Q9646930. Chinese is most comprehensive but anything that is not a Shikinai Taisha or Myojin Taisha but is a Shikinaisha is by process of elimination a Shikinai Shosha.
I am thinking about making the Engishiki lists here too but that is for another day. Probably when I am doing the table things, since I think there is an ordering involved with them. For example this one Q11559808Immanuelle (talk) 05:44, 15 June 2025 (UTC)
Not import deprecated website?
Latest comment: 26 days ago4 comments2 people in discussion
I'm working on an infobox for Spanish Wikipedia. The website section imports two websites from Wikidata but one of them is deprecated, is there a way to ignore deprecated elements automatically? JaimeDes (talk) 17:56, 13 June 2025 (UTC)
Parser functions #property and #statements do this implicitly. You are probably using Lua. If it's a generic module, check its documentation. If you are coding it from scratch, select statements whose statement.rank~='deprecated'. --Matěj Suchánek (talk) 10:22, 14 June 2025 (UTC)
Thanks! Yes, I was working with Lua. And now that we are here, do you know why sometimes while using the normal parser, the template {{Property|P~}} doesn't import information from Wikidata but {{#invoke:|}} does? JaimeDes (talk) 21:08, 14 June 2025 (UTC)
Latest comment: 25 days ago6 comments3 people in discussion
I recently spent a little time looking at this little ontological chain while looking at something that happened on my wikicrowd tool on Commons.
In summary in this chain of edits the tool removed depicts road in favour of adding a specific depicts for the specific bridge. This happened due to the bridge itself being a toll bridge, which apparently is a subclass of road.
Anyone got any thoughts on this subclass chaining? Does it make sense? Should it be tweaked? Are toll bridges always toll roads? Anyone got any real world examples? ·addshore·talk to me!20:53, 14 June 2025 (UTC)
DecomissionedAircraftMap (see tool below) - The Decommissioned Aircraft Map project uses Wikidata to enhance its mapping of historic aircraft by pulling images from linked Wikidata entries. Users can contribute by adding or correcting Wikidata tags on OpenStreetMap, ensuring accurate representation of aircraft locations and visuals. By Watmildon.
DecomissionedAircraftMap (as a demonstration of the power of OpenStreetMap into Wikidata): pulls geodata for displayed aircraft from OpenStreetMap and generates thumbnails from linked Wikidata entries.
Query split tester (Beta): webtool to see the impact on the graph split on your SPARQL query.
Other Noteworthy Stuff
Nominations for the Coolest Tools Award 2025 are open. Nominate your favorite tool! Nominations are due by the 25th of this month already.
Newest WikiProjects: EMCO Wikidata CoP - EMCO promotes the discovery and use of the world’s knowledge by supporting metadata producers in library and other cultural heritage communities.
Nothing. Wikidata will not (re)order the data. It's up to the presenting tool (e.g., the template displaying the items) to do that. --Matěj Suchánek (talk) 16:12, 16 June 2025 (UTC)
Latest comment: 24 days ago2 comments2 people in discussion
Kasuga Gongen and Kasuga no Kami are pretty similar.
I think they should be merged or somehow consolidated, at least for the wikipedia articles. Is there a policy around this
Kasuga Gongen refers to a deity figure that existed prior to the Meiji-era policy of Shinbutsu Bunri (the separation of Shinto and Buddhism) and has since been abolished. The current enshrined deity is Kasuga no Kami, who was considered identical to Kasuga Gongen during the period of syncretism between Shinto and Buddhism. Unless the Japanese Wikipedia articles are merged, it is unlikely that they can be unified on Wikidata. Afaz (talk) 01:22, 17 June 2025 (UTC)
Wikimedia Foundation Board of Trustees 2025 - Call for Candidates
Latest comment: 24 days ago1 comment1 person in discussion
This year, the Wikimedia community will vote in late August through September 2025 to fill two (2) seats on the Foundation Board. Could you – or someone you know – be a good fit to join the Wikimedia Foundation's Board of Trustees? [3]
Learn more about what it takes to stand for these leadership positions and how to submit your candidacy on this Meta-wiki page or encourage someone else to run in this year's election.
Best regards,
Abhishek Suryawanshi
Chair of the Elections Committee
On behalf of the Elections Committee and Governance Committee
Latest comment: 22 days ago8 comments3 people in discussion
Items pertaining to parkrun (Q7138632) appear to be very haphazardly structured, so I'm wondering what the best structure would be before fixing.
A general description of parkrun and terms, in real life:
(Note: parkrun uses a lower case 'p' but obviously the items will be named as per Wikidata policies)
parkrun broadly describes the entire movement - the collection of organisations and events as described below.
parkrun Global Ltd is a charity based in the UK. it has sub-organisations in some countries in which it operates, e.g. parkrun Australia, parkrun Canada.
The events are known as parkrun (sometimes 5k parkrun to differentiate) and junior parkrun
There are over 2000 of these events across 23 countries (in parkrun terminology, event is the location of a weekly run. Think of event as a location and a run as an individual event).
Each event has a designated location and route, is organised by local volunteers, and has an official unique name, e.g. Montrose Foreshore parkrun, Coventry parkrun, Curro Mossel Bay School parkrun
Now, how does it look on Wikidata now? Very messy:
These would probably be best created via a script initially. There are maybe 1-3 new ones a week.
Do individual runs need their own entries?
No. There would be literally hundreds of thousands of them, with 2000+ new ones every week).
The only thing missing is parkrun at the very top - the overarching topic - it could be instance of (P31) of social movement (Q49773) perhaps, with participant (P710) including the global and national orgs. There shouldn't be a problem with this having the same name as the "aka 5k parkrun" object.
Another note - these aren't technically sporting events. They're more like community fitness programs. Just in case anyone is trying to work out what other properties would apply.
Sorry, this entry became waaay longer than I had hoped.
Is there anyone else, maybe more familiar with both Wikidata structures and parkrun, with further thoughts or ideas on this? -- Chuq(talk)06:43, 12 June 2025 (UTC)
Thanks @Vicarage, I could use some of the same properties, although it's a bit different - a single recurring annual event vs multiple recurring weekly events. Perhaps PAX (Q588487) would be somewhere in the middle - half a dozen or so different events, which have their own annual recurrences, but the annual events don't have their own. -- Chuq(talk)00:04, 13 June 2025 (UTC)
Could you expand on this parenthetical comment "Note: parkrun uses a lower case 'p' but obviously the items will be named as per Wikidata policies"? Bovlb (talk) 20:53, 12 June 2025 (UTC)
Basically pointing out my use of a lower-case "p" above is because that's how the organisation refers to itself and events - but English Wikipedia has a policy - w:MOS:TMLOWER. Only mentioned in case Wikidata used a similar style. -- Chuq(talk)00:56, 13 June 2025 (UTC)
Wikidata does not follow the English Wikipedia's MOS. The relevant guidance is Help:Label#Capitalization, which says "In the rare case that something intentionally breaks capitalization rules, the capitalization on Wikidata should reflect this, and not try and correct it." Bovlb (talk) 03:21, 13 June 2025 (UTC)
I wasn't sure, it was just a note to advise that any name examples I listed I hadn't yet crosschecked with any policy. (I prefer the Wikidata method, btw!) -- Chuq(talk)04:23, 13 June 2025 (UTC)
Mass romanizations of CJK characters and other scripts for multi language titles
Latest comment: 21 days ago17 comments7 people in discussion
I think a weakness of wikidata is it often lacks human readable names, particularly for CJK characters.
Have there been proposals to mass add romanizations of names in the multi language section?
CJK will be the most difficult because of differing japanese readings. But other phonetic scripts would be more easily romanized. Immanuelle (talk) 02:00, 16 June 2025 (UTC)
I find the translation tools problematic, as sometimes they transliterate, and other times they translate. You want Kamikaze, not Divine Wind. There are transliteration websites, but they are problematic for Cyrillic at least, as there are many standards. Not sure if the same problem exists for CJK. Vicarage (talk) 07:56, 16 June 2025 (UTC)
Here is my understanding:
Chinese mandarin transliteration is very predictable going into pinyin
Japanese is extremely difficult with kanji
But Kana transliterate pretty good. The main difference is just whether you use Hepburn or Nihon-shiki, which have slight differences.
Fortunately it seems a ton of wikidata entries exist in Japanese where the kana spelling is present, even though there is no english label, and it might even be trivially easy to make a bot to extrace the kana pronounciation from articles when they are present. Although there are a couple of things that really have no clear pronounciation because they are say one name in an ancient book or something. Immanuelle (talk) 08:28, 16 June 2025 (UTC)
I spend a lot of time romanizing Japanese by hand since kanji readings can be inconsistent. @Estopedist1 Please don't use Google translate for romanizing Japanese; someone was doing that a couple years ago and I asked them to stop because sometimes the romanization Google translate produced didn't match the kana transliteration on the person's Wikipedia page. In a couple instances Google translate applied the Chinese pinyin romanization, which doesn't work for Japanese at all.
A few years ago @Lockal added the "name in kana" property (P1814) to many items for Japanese people. I found this incredibly helpful, and pulled in all P31:Q5 with that property and date of birth/death into Openrefine, where I added romanizations and pushed them back out into the English labels. However, there are about 10,000 or so that I didn't do because they didn't have dates. Typically, modern Japanese people in Wikidata have their names listed in (Given name)(Family name) order, while historical figures (pre-1868) have their names recorded as (Family name)(Given name). I wanted to keep consistent with this practice, so since then I've been romanizing by hand. If we could have the "name in kana" property added to items annually, I think I'd be able to keep up with any items that have dates. Or if you could show me how you did it, I'm happy to add "name in kana" and romanizations quarterly when I import biographical items from Japanese Wikipedia.
As for Chinese transliterations, I think my question would be which romanization system one might use. Pinyin is very common in mainland China, but from what I understand Wade-Giles is more commonly used in Taiwan. When I romanize Chinese (rarely, since I'm more comfortable in Japanese) I always try to determine where the person is from so that I can use the romanization system that might be appropriate for them, and put the other romanization as an alias.
I have no opinions about Korean romanization, though I'd imagine it would have its own hurdles since the McCune-Reischauer system is still in use in places like the Library of Congress, while the South Korean government uses the Revised Romanization system. I've also heard that McCune-Reischauer is also the romanization system used in North Korea, so if that is the case then just like with Chinese romanization any automated system in Wikidata would have to consider country of citizenship (P27) when determining which romanization goes in the label and which goes in the alias.
Please let me know if you have any questions! I spend a lot of time thinking about how to add more romanizations to labels, but don't really have any ideas for how to do this automatically. Mcampany (talk) 15:14, 16 June 2025 (UTC)
And to be clear, I don't mind filling in romanizations manually! It's a good opportunity to fill in other properties that these items often lack. Mcampany (talk) 15:33, 16 June 2025 (UTC)
@Mcampany I wrote a script that takes from the listed kana and adds the Hepburn romanization as the main name and the Nihon-shiki as an alias. My most recent edits have this and I would be interested in filling out a bot request for this. Immanuelle (talk) 11:13, 17 June 2025 (UTC)
You're welcome to submit a bot request. How would you handle the question of name order that I raised above, though? We wouldn't want a shogun like Tokugawa Ieyasu (Q171977) to have their label be Ieyasu Tokugawa, nor would we necessarily want a modern figure like Haruki Murakami (Q134798) to be listed as Murakami Haruki. The latter case isn't the end of the world, but it would make any labels your bot romanizes out of sync with typical Wikidata convention, and would mean that the automatic duplicate prevention (which works by preventing labels and descriptions that are identical to another item from saving) wouldn't work. Mcampany (talk) 15:42, 17 June 2025 (UTC)
It parses the break between the first and last name, and if the person was born before the Meiji period then it goes Last name first name, and if the date of birth is absent or post 1868 then it does western order.
Date of birth being absent is a bit tricky but in what I have seen, the majority of the people with missing birth dates and kana readings and no english labels are in fact modern people.
Having tried this I think that doing it for people and place names is probably the most reliable. But other things will often require more nuance. Immanuelle (talk) 22:58, 17 June 2025 (UTC)
I'd support your bot if it worked only on items where there is a date of birth/death present. In this query of people with name in kana but no English label, I found a few instances of people without dates whose names shouldn't be automatically romanized, like Q60989675 and Q11385582. Personally, I still feel a little hesitant about automatically adding romanizations to items that don't have a date, but if you feel confident you certainly don't need my permission. ;)
Romanizing place names sounds helpful, though @Whym brought up some good questions below about how to transliterate them. I think I usually see just "Shirakawa" as the English label, but it would be ideal if your bot could add the other transliterations and translations they shared as aliases. Mcampany (talk) 15:01, 18 June 2025 (UTC)
I see from your talk page that you have already begun adding romanizations for place names. Maybe you should run some small batches, check carefully, and maybe ask @Mariobanana for help? I'm really not familiar with romanization conventions for place names in Wikidata, but it seems like there are plenty of other people who have thoughts about how to do this well. Mcampany (talk) 15:16, 18 June 2025 (UTC)
I have other work to do but I don't want to leave the incorrect labels, so I have been correcting them. And, there are still items that need to be corrected. Please get help a Japanese Speaker who is not me.
As for the current script, I have spent a lot of time and effort fixing the mistakes, so I think it would be better to input labels manually.
I think there is often need for human judgment and a bit of research. If an entity is about a creative work, there is a good chance that the title is translated rather than transliterated from the original language into English and into other languages. The lack of an English label doesn't mean there is no existing translation in English and you are free to choose - it might just mean that no Wikidata editor entered it. People names are (much) less prone to this, but some people do decide to use a not-straightforward spelling in a foreign language intentionally. For place names, too, it's not straightforward to choose between Shirakawa Village (half translation) [16], Shirakawa-mura (transliteration) [17], Shirakawamura (another transliteration), and just Shirakawa. --whym (talk) 11:48, 17 June 2025 (UTC)
In the advent of language code mul, is this rule still true? Shouldn't it be reshaped to needs a label in specific lang OR mul? best --Herzi Pinki (talk) 13:46, 18 June 2025 (UTC)
Since it will only be used on items that have proper nouns, the the constraint should be only 'mul'. It would be counterproductive to ask for 'de' labels. For example if the property is for identifiers for biographical articles, then the 'label in language' constraint should be changed to 'mul'. 'label in language' constraints on things that are not proper nouns should remain as they are. Edit: I have a tool to check label coverage here. Infrastruktur (talk) 14:13, 18 June 2025 (UTC)
@Infrastruktur Not in all cases. e.g ČSFD film ID (P2529) is czech movie database. Most of movies in this database have czech label, so constraint have P2529 and label in cs is useful. But movies usually should not have mul label becasuse name differs by language. JAn Dudík (talk) 06:38, 19 June 2025 (UTC)
Correction. Use of 'mul' is still under discussion, but its current accepted use seems to be a lot more narrow than was originally intended. Help:Default_values_for_labels_and_aliases doesn't list street names, so the constraint should stay 'de'. But as I mentioned above properties for identifiers on people can be changed over to 'mul' as long as the property can only be used on humans. A possible exception here is for properties that is expected to be used on a lot of people whose names use a non-latin script. Infrastruktur (talk) 10:44, 20 June 2025 (UTC)
It feels rather subjective to me, as of course arms companies and shipyards profit from wars. Crimes require proof of a conviction of a organisation by the state, and I expect company law varies a lot by jurisdiction, and you need to be careful using properties related to people for companies. Vicarage (talk) 13:20, 21 June 2025 (UTC)
There were generally maximum profit levels set. Charging beyond these was in infraction, though it may have been dealt with by asking for a lower invoice, as morale was important. All the best: RichFarmbrough, 12:24, 22 June 2025 (UTC).
Either these 2 wikidata item merge or there should be 2 different English article. But the term are so close i think 2 different English article wont be possible.
So if these 2 merge together, what should be the best way to merge:
@Dark1618: The current situation is fine as it is and the items are working as intended, with ethnic group (Q41710) linking to a redirect on en.wiki. The two items cannot be merged because multiple other wikis have articles about both topics, and they are about two similar but separate concepts. If you wish to develop an en.wiki article for "ethnic group", that's an option, but it is not strictly necessary. — Huntster (t@c)01:16, 19 June 2025 (UTC)
Latest comment: 18 days ago5 comments5 people in discussion
The "search for pages containing" in the list still works, but the button just redirects to the main page, whether it's a name or ID of an item or anything else that is entered in the box. Peter James (talk) 15:17, 19 June 2025 (UTC)
A lifehack in the meanwhile: If I search for "Poortweg" I get indeed redirected to the main page; if I search for "Poortweg Delft" I get redirected to the search which shows me precisely the one item I need (note that there is no item with the name "Poortweg Delft"; only the one with the name "Poortweg" and the description "street in Delft, the Netherlands"). Ymblanter (talk) 11:07, 23 June 2025 (UTC)
Re: Search by entity type: help us test
Latest comment: 18 days ago4 comments3 people in discussion
The new search experience is now live on Wikidata!
You can now filter your searches by entity type (Item, Property, Lexeme, or EntitySchema) using the new dropdown menu in the search bar. Your search terms across each entity type will also have typeahead results, making it easier and faster to find exactly what you’re looking for.
Huge thanks to everyone who helped test the feature on test.wikidata.org and beta. Your feedback made this possible!
If you’ve got any questions or feedback please leave a comment on this ticket: phab:T321543.
I hope the issue (which I have as well) will be fixed soon, but I just want to mention that this is a great development. Thanks a lot to the developers team. Ymblanter (talk) 07:09, 23 June 2025 (UTC)
It may be a great development when it works, but I'm rather disappointed it was deployed without adequate testing — Martin (MSGJ · talk) 07:39, 23 June 2025 (UTC)
How to merge two items of a cross-border river with conflicting info?
Latest comment: 18 days ago3 comments3 people in discussion
Urpalanjoki (Q4417881) and Urpalanjoki (Q36228678) are about the same river that starts in Finland and flows to Russia ending in the Gulf of Finland. Normally I would just merge them but both items have a wikilink to cebuano wikipedia, likely bot-generated articles. They also have different geoname ids and gns unique feature ids. MKFI (talk) 13:21, 23 June 2025 (UTC)
The simpliest way is to change one of ceb articles to redirect (they are usually with same informations) and then merge. JAn Dudík (talk) 13:47, 23 June 2025 (UTC)
Is there anybody familiar how we handle items that are pseudonyms? I noticed for example that it also has a date of birth and death and wasn't sure if they belong to the pseudonym item. Would be happy to receive some help on this. Thank you! Fallen Sheep (talk) 21:06, 14 June 2025 (UTC)
@Fallen Sheep While Resource Description and Access (Q1519318) demands separate records for each pseudonym, this isn’t the case with Wikidata. We generally only differentiate between people and their pseudonyms in special cases:
I think I understand what Robert Galbraith (Q110929251) is trying to communicate, but I’m not sure if the modeling does a terribly good job at that (CC Daask as creator)
I usually don't create separate items for pseudonyms, and instead just add the qualifier subject named as (P1810) to every identifier which distinguishes names. In the case of Robert Galbraith (Q110929251), I opted to split the item because
There were a substantial number of authority control records for the pseudonym.
There were separate social media accounts for the pseudonym.
There is a separate website for the pseudonym.
Altogether, I thought a separate item was useful in that it might help users who search whatever computer system for Robert Galbraith to find the associated identifiers, social media accounts, and website without having them jumbled up with J. K. Rowling (Q34660). I'm welcome to being outvoted on this, and I am aware that it isn't the norm on Wikidata.
Another occassion where I would expect a separate Wikidata item for a pseudonym is whenever the pseudonym is also an instance of persona (Q1077857) or character (Q95074) with features described in some sources, eg. with a distinct birthplace or career history. I'm not sure to what extent this is true of Robert Galbraith (Q110929251)
There are often occasions where it seems an editorial decision whether to model items as separate or joint. When items can plausibly be modeled as a single item or two distinct items, I tend to add both of these claims:
Latest comment: 18 days ago5 comments3 people in discussion
Hi! I have a question — how many inclusions are generally enough for an identifier? If a new property has 100 references, is that sufficient? Or does it need 1000? 2000? Mitte27 (talk) 11:02, 23 June 2025 (UTC)
@Emu is there a specific procedure to use if you want to introduce the links to something before it gets formalized as an identifier? Immanuelle (talk) 12:48, 23 June 2025 (UTC)
Latest comment: 18 days ago1 comment1 person in discussion
Here's your quick overview of what has been happening around Wikidata in the week leading up to 2025-06-23. Missed the previous one? See issue #684.
Events
Join us for the third LD4 Wikidata Affinity Group event on the Wikidata Graph Split project, Tuesday, June 24 at 9am PT / 12pm ET / 16:00 UTC / 6pm CEST. We’ll share updates on the Query Service, tools, and SPARQL learning resources. Stanford researchers Shicheng Liu and Sina Jandaghi Semnani will present their Spinach Wikidata Agent, which translates complex questions into SPARQL queries Project page
Press, articles, blog posts, videos
Presentations
Slides of the panelMediaWiki-based tools and services in Digital Humanities workflows, part of the DARIAH-EU Annual Event 2025 in Göttingen: (the panel was about Wikibase instances, including Wikidata)
FEMIber: a digital humanities tool that uses structured data aligned with Wikidata principles to document and analyze how women are represented in medieval Iberian chronicles.
Newest WikiProjects: WikiProject Personal Collections - The WikiProject Personal Collections aims to develop standardized terms, multilingual ontologies, and best practices for consistently describing personal collections and archives on Wikidata.
Showcase Lexemes: газ (L183915) - Russian noun (ɡas) meaning "gas (physical state)", "natural gas", or "accelerator pedal"
Development
Query Service: It is now possible to download query results in Wikidata Query Service that include coordinates as GeoJSON (KML & GPX should also be available soon). Thanks to Atom.oil.2 for the patch. (phab:T216601)
Wikidata integration in Wikipedia and co: We are continuing the work on improving how Wikidata changes show up in watchlist and recent changes, specifically showing labels instead of IDs in the future. (phab:T388685)
Search: We enabled the new search box that makes it easier to search in Properties, Lexemes and EntitySchemas as well. We are working on minor fixes based on feedback.
Mobile statement editing: We are working on showing the first statements in the new way - mostly tech demo and nothing to see yet (phab:T394886)
Federation: We are looking into measuring and better understanding queries that use SPARQL federation.
REST API: We continued working on prefix search for Items in the API (phab:T388209)
Latest comment: 18 days ago2 comments2 people in discussion
Upminster Tithe Barn (Q17644980) is a listed building commonly known as Upminster Tithe Barn. The official listing name is "Medieval grange barn, 228m south-east of Upminster Court" and that is the name used here for the item, with Upminster Tithe Barn as an alias. Is that correct or should the common name be the main name for the item here? MRSC (talk) 15:52, 23 June 2025 (UTC)
Common names should be used for labels ,the formal name can be part of a property, often as a qualifier about the database listing. Vicarage (talk) 17:29, 23 June 2025 (UTC)
Rfd help
Latest comment: 17 days ago3 comments3 people in discussion
idk if this is the right place to ask this question, but is their any easier way of Rfd? It seems like you need to add topic this page for deletion req: Wikidata:Requests for deletions. This page is so long, it takes too long to load and somtime freezes mobile browser. Is this the only to do this? Can I request for deletion from item discusstion? Nimon didarul (talk) 07:22, 19 June 2025 (UTC)
Wikidata Generic Tree broken ? it's included in all (even inexistent) talkpages
Latest comment: 17 days ago1 comment1 person in discussion
On trees like https://wikidata-todo.toolforge.org/tree.html?q=23112&rp=131 the labels don't really seem to load anymore. It's a shame because the tool is useful and included in template such as {{Item documentation}} (as a link) which are shown in every talk pages (even if the page does not exist) by a mediawike header mechanism. It's probably a not so well known fact so I add another layer of information here.
Inverses are generally avoided if possible because they cause data duplication with all negative effects. There are some exceptions for legacy and technical reasons. --Emu (talk) 06:01, 23 June 2025 (UTC)
Thanks @Emu, interesting, it's the first one I had come across. Is there any benefit in making all this automatic? e.g. any property with an inverse being added to an item causes the inverse property to be added automatically to the opposite item? I'm sure I'm not the first to notice inconsistences, example:
You can query to find these relations. The first example looks wrong to me and also you don't necessarily have to be employed by the organization you (co-)founded. Sjoerd de Bruin(talk)07:28, 24 June 2025 (UTC)
That's what I mean - two-fold: (1) The absence of the property causes people to work around it, using inconsistent methods (2) Could a query be used to automatically display the inverse property on the other item? -- Chuq(talk)14:01, 24 June 2025 (UTC)
After further thought and investigation, I'm starting to suspect that the shannonairportgroup URL was never correct--while I can't find it on archive.org to verify--I suspect it was the (possibly planned-but-not-published) official website for the company that runs the airport, but not really the website of/for the airport itself. (Currently, the company website appears to be https://www.snnairportgroup.ie/)
If that's the case, it would be appropriate to actually delete the URL, right?
Since it appears I don't have access to do that, do I post a request somewhere for someone to do it for me, or will this message suffice? Bsammon (talk) 08:36, 26 June 2025 (UTC)
Maybe, but I don't want to import by hand, so any advice importing that music data from external sources (e.g. MusicBrainz (Q14005), Discogs (Q504063), etc.) or maybe you (or anyone) help me to importing the data to here?
2. It refers to an instance of a clearly identifiable conceptual or material entity that can be described using serious and publicly available references.
Latest comment: 13 days ago5 comments3 people in discussion
Hi everyone,
One of my created items (Q134617360) was recently deleted via a mass deletion (tagged "Nuke").
As a new contributor, I’ve respectfully asked for clarification here: Talk:Q134617360.
I’m not challenging the decision, just hoping to understand what went wrong so I can learn.
I genuinely want to contribute more to Wikidata in the future, and understanding this case will help me avoid mistakes and follow community standards better.
Any feedback or guidance would be sincerely appreciated.
Latest comment: 14 days ago3 comments2 people in discussion
Hello, Where can I submit a request to hide my IP address? I accidentally edited an item using my IP instead of my account. Regards, Riad Salih (talk) 16:41, 27 June 2025 (UTC)
A vision of relevant, accessible, and impactful free knowledge has always guided the Wikimedia Movement. As the ecosystem of Wikimedia projects continues to evolve, it is crucial that we periodically review existing projects to ensure they still align with our goals and community capacity.
Despite their noble intent, some projects may no longer effectively serve their original purpose. Reviewing such projects is not about giving up – it's about responsible stewardship of shared resources. Volunteer time, staff support, infrastructure, and community attention are finite, and the non-technical costs tend to grow significantly as our ecosystem has entered a different age of the internet than the one we were founded in. Supporting inactive projects or projects that didn't meet our ambitions can unintentionally divert these resources from areas with more potential impact.
Moreover, maintaining projects that no longer reflect the quality and reliability of the Wikimedia name stands for, involves a reputational risk. An abandoned or less reliable project affects trust in the Wikimedia movement.
Lastly, failing to sunset or reimagine projects that are no longer working can make it much harder to start new ones. When the community feels bound to every past decision – no matter how outdated – we risk stagnation. A healthy ecosystem must allow for evolution, adaptation, and, when necessary, letting go. If we create the expectation that every project must exist indefinitely, we limit our ability to experiment and innovate.
Because of this, SPTF reviewed two requests concerning the lifecycle of the Sister Projects to work through and demonstrate the review process. We chose Wikispore as a case study for a possible new Sister Project opening and Wikinews as a case study for a review of an existing project. Preliminary findings were discussed with the CAC, and a community consultation on both proposals was recommended.
Wikispore
The application to consider Wikispore was submitted in 2019. SPTF decided to review this request in more depth because rather than being concentrated on a specific topic, as most of the proposals for the new Sister Projects are, Wikispore has the potential to nurture multiple start-up Sister Projects.
After careful consideration, the SPTF has decided not to recommend Wikispore as a Wikimedia Sister Project. Considering the current activity level, the current arrangement allows better flexibility and experimentation while WMF provides core infrastructural support.
We acknowledge the initiative's potential and seek community input on what would constitute a sufficient level of activity and engagement to reconsider its status in the future.
As part of the process, we shared the decision with the Wikispore community and invited one of its leaders, Pharos, to an SPTF meeting.
Currently, we especially invite feedback on measurable criteria indicating the project's readiness, such as contributor numbers, content volume, and sustained community support. This would clarify the criteria sufficient for opening a new Sister Project, including possible future Wikispore re-application. However, the numbers will always be a guide because any number can be gamed.
Wikinews
We chose to review Wikinews among existing Sister Projects because it is the one for which we have observed the highest level of concern in multiple ways.
Since the SPTF was convened in 2023, its members have asked for the community's opinions during conferences and community calls about Sister Projects that did not fulfil their promise in the Wikimedia movement.[1][2][3] Wikinews was the leading candidate for an evaluation because people from multiple language communities proposed it. Additionally, by most measures, it is the least active Sister Project, with the greatest drop in activity over the years.
While the Language Committee routinely opens and closes language versions of the Sister Projects in small languages, there has never been a valid proposal to close Wikipedia in major languages or any project in English. This is not true for Wikinews, where there was a proposal to close English Wikinews, which gained some traction but did not result in any action[4][5], see section 5 as well as a draft proposal to close all languages of Wikinews[6].
Initial metrics compiled by WMF staff also support the community's concerns about Wikinews.
Based on this report, SPTF recommends a community reevaluation of Wikinews. We conclude that its current structure and activity levels are the lowest among the existing sister projects. SPTF also recommends pausing the opening of new language editions while the consultation runs.
SPTF brings this analysis to a discussion and welcomes discussions of alternative outcomes, including potential restructuring efforts or integration with other Wikimedia initiatives.
Options mentioned so far (which might be applied to just low-activity languages or all languages) include but are not limited to:
Restructure how Wikinews works and is linked to other current events efforts on the projects,
Merge the content of Wikinews into the relevant language Wikipedias, possibly in a new namespace,
Merge content into compatibly licensed external projects,
Archive Wikinews projects.
Your insights and perspectives are invaluable in shaping the future of these projects. We encourage all interested community members to share their thoughts on the relevant discussion pages or through other designated feedback channels.
Feedback and next steps
We'd be grateful if you want to take part in a conversation on the future of these projects and the review process. We are setting up two different project pages: Public consultation about Wikispore and Public consultation about Wikinews. Please participate between 27 June 2025 and 27 July 2025, after which we will summarize the discussion to move forward. You can write in your own language.
I will also host a community conversation 16th July Wednesday 11.00 UTC and 17th July Thursday 17.00 UTC (call links to follow shortly) and will be around at Wikimania for more discussions.
The bot that compiles the constraint reports didn't consider OR-ed constraints last time I looked (IIRC it prevents that part of the report from being generated). The Mediawiki extension handles them just fine for live feedback. I haven't looked at the sources for my own Elasticsearch based label coverage tool since I wrote it, I'm guessing it will silently treat it as if you've AND-ed the constraints. Infrastruktur (talk) 18:08, 28 June 2025 (UTC)
Property for former names?
Latest comment: 12 days ago2 comments2 people in discussion
However, that would cause a loss of precision, since the action (felling an Acacia tree) is specifically done by an African elephant. Also, I don’t quite get why taxons aren’t considered classes of living things. — keepright! ler (talk) 17:04, 29 June 2025 (UTC)
Requests for comment notification
Latest comment: 12 days ago1 comment1 person in discussion
Painting on long term loan to a gallery, but not owned by that gallery; owner anonymous
Latest comment: 10 days ago2 comments2 people in discussion
I'm starting to make manual entries for paintings. One is on long term loan to a UK national art gallery. The name of the formal owner is unknown. I keep getting error messages as I try to configure the listing. I would be grateful for advice about the correct way to model this relationship.
https://www.wikidata.org/wiki/Q135153181S.M.NGL E9 (talk) 10:21, 30 June 2025 (UTC)
Latest comment: 11 days ago1 comment1 person in discussion
The Wikidata flagship event of the year, Wikidatacon 2025 is now open to receive your proposals!
This year we are celebrating the connections Wikidata promotes and creates, whether between Wikimedia Projects, WikiProjects, institutions and data reusers, across languages and communities. This year, we have 6 programme tracks waiting to receive your proposals.
The call for proposals is open until September 1stAoE. The 6 Tracks are:
Latest comment: 11 days ago2 comments2 people in discussion
I seem to run into rate-limiting for creating items quite a bit ("As an anti-abuse measure, you are limited from performing this action too many times in a short space of time, and you have exceeded this limit. Please try again in a few minutes."). It happens the most when I try to use Mix'n'Match, but recently I've even been running into it when I've not been using MnM. With MnM I will regularly have edits take a long time or even never succeed, and if I open my browser console, I can see the rate-limiting messages popping up over and over. I thought this was just an unavoidable MnM issue, but when I mentioned it on Discord, User:AntiCompositeNumber said the rate limit is 90 edits per 60 seconds and that I really shouldn't be hitting it. You can look at my contributions – I did do a batch import on June 21, but since then my edit speed has been very human.
Wikidata for Jewish Studies, a hands-on Wikidata workshop at the Akademie der Wissenschaften und der Literatur, Mainz from September 8-9, 2025. (Registration by July 18)
Visualizing Knowledge - how Indonesians turn Wikidata into stories that matter. This article details the efforts and winners of a WM Indonesia organised 30-day competition to create cool visualizations powered through Wikidata.
User:Jon Harald Søby/senseItemLabel.js What it does is that when you add a P5137 statement in a lexeme, it will let you add the lemma as a label to the item in one click, if the item doesn't have a label in that language yet, or if the label is different (very useful for case differences, for instance)
Other Noteworthy Stuff
The Wikidata Ontology Course has finished its course sessions. The slides for all sessions of the course are available on the course page. Part of participating in the course was to set up a project related to the Wikidata Ontology. Current information on the projects, many of which are continuing after the end of the course is now available. If you are interested in one or more of the projects please follow the project, add something to its discussion page, or contribute to it.
PhD Position in Digital History at the University of Luxembourg is seeking someone with experience in Wikibase and Linked Open Data.
defining Prolog formula (Prolog rule expressing the intended logical behavior of a property, for use in ontology formalization and rule-based reasoning.)
Newest WikiProjects: WikiProject Personal Collections - The WikiProject Personal Collections aims to develop structured, multilingual standards and best practices for describing personal collections and archives on Wikidata.
Wikibase REST API: We continued working on Item prefix search (phab:T388209) and started working on stemming support (phab:T397605)
Integration in Wikipedia and co: The improved edit summaries are now live on the first wikis (ca, uk, he). We are now showing the labels for linked entities instead of just their IDs (phab:T388685)
Mobile statement editing: We are continuing working on the basics of showing statements in the new way - nothing useful to see yet (phab:T394886)
Resolved Fallen Sheep (talk) 18:56, 2 June 2025 (UTC)
Comment https://www.nature.com/articles/nature.2015.17567 has 5000+ authors. Per e.g. user:Piecesofuk we should auto-create 5000 Q-items!!?? Estopedist1 (talk) 20:06, 4 June 2025 (UTC)
Done 
Merged --Emu (talk) 07:08, 15 June 2025 (UTC)
